ABSTRACT
Title of dissertation: IDENTIFICATION AND
CHARACTERIZATION OF NOVEL
PLANT IMMUNE COMPONENTS
USING THE ARABIDOPSIS-
POWDERY MILDEW PATHOSYSTEM
Qiong Zhang
Doctor of Philosophy, 2018
Dissertation directed by: Professor Shunyuan Xiao
Department of Plant Science
& Landscape Architecture/
Institute for Bioscience
& Biotechnology Research
Despite tremendous progress in the field of plant immunity in the past two
decades, how plants mount spatiotemporally appropriate defenses against pathogens
is still not well characterized. My thesis project took both forward and reverse ge-
netic approaches to uncover novel mechanisms used by plants to fight against fungal
pathogens, or exploited by fungi to adapt to host plants using the Arabidopsis-
powdery mildew pathosystem. Through a reverse genetics approach, I found that
two phospholipase D (PLD) genes PLDα1 and PLDδ play opposing roles in modu-
lating basal, post-penetration resistance against mildew through a novel, yet-to-be
characterized mechanism that is independent of EDS1/PAD4 (key immune com-
ponents), salicylic acid (SA), and jasmonic acid (JA). Inspired by this finding, I
designed and performed a large-scale forward genetic screen in the background of a
super-susceptible Arabidopsis eds1-2pad4-1sid2-2 (eps) triple mutant. By screening
EMS-mutagenized eps plants using powdery mildew species with different levels of
adaptation on Arabidopsis, 5 susceptible to non-adapted PM (snap) and 18 compro-
mised immunity yet poor infection (cipi) mutants have been identified. So far, this
has led to the characterization of the MAP KINASE PHOSPHATASE1 (MKP1)
gene, which is a negative regulator of PAMP-triggered immunity, and the MILDEW
RESISTANCE LOCUS 2 (MLO2) gene, a susceptibility factor of powdery mildew,
both of which act independently of EDS1, PAD4, and SA. Together, results from
this work should contribute to a better understanding of the multi-layered plant
immune system and powdery mildew’s host adaptation mechanisms.
IDENTIFICATION AND CHARACTERIZATION OF NOVEL
PLANT IMMUNE COMPONENTS USING THE
ARABIDOPSIS-POWDERY MILDEW PATHOSYSTEM
by
Qiong Zhang
Dissertation submitted to the Faculty of the Graduate School of the
University of Maryland, College Park in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
2018
Advisory Committee:
Professor Shunyuan Xiao, Chair/Advisor
Associate Professor Daniel C. Nelson
Professor Caren Chang
Associate Professor Stephen M. Mount
Associate Professor Louisa Wu
Assistant Professor Yiping Qi
©c Copyright by
Qiong Zhang
2018
Acknowledgments
I owe my gratitude to all the people who have made my PhD journey an
invaluable experience that I will forever cherish.
First and foremost, I would like to thank my advisor, Dr. Shunyuan Xiao for
giving me the opportunity to work on these extremely interesting projects studying
plant immunity that always fascinates me. I would not have become the scientist I
am today without his guidance, encouragement, inspiration, and support, and the
stimulating environment he’s created. His dedication to science will always be an
inspiration to me. I’m very grateful.
I would also like to thank my advisory committee, Drs. Daniel Nelson, Caren
Chang, Louisa Wu, Stephen Mount, Yiping Qi, and Liqing Yu for their kind support,
encouragement, and constructive advice throughout the years.
I want to thank all the collaborators who have helped me with my research –
Drs. Xuemin Wang, Joshua J. Blakeslee, Jinshan Lin, Teun Munnik.
I also want to thank my neighbors at IBBR, especially Dr. Shuwei Li, the
Nelson lab and the Roy Mariuzza lab members, for sharing their lab equipments,
reagents, and lab supplies, and for their friendship.
Thanks to all my lab mates, past and present, for their assistance, accompany,
and friendship. They have made everyday lab life enjoyable and productive. Special
thanks to Bob for initiating the phospholipase project and many others, and training
me when I first started; Harley for being the big brother helping me since day one,
from whom I can always find my sanity when overwhelmed; Yi for his kindness and
ii
generosity in helping me in many ways; Xianfeng for isolating the tobacco powdery
mildew, which greatly helped to advance my PLD project; Ying and Wanpeng
for analyzing the sequencing data; Bruce for all the fun conversations on science
and non-science; my high school interns Sharon and Anna for helping with the
phospholipase project and others; Ouyang, Lili, and Yuheng, and my undergraduate
interns Anya, Robin and Jessica for helping with the mutant screening project;
and our greenhouse manager, Frank for maintaining a functional and clean growth
facility for plants.
I’m grateful for the four-year stipend financial support from China Scholarship
Council, 2016 Summer Research Fellowship from the Graduate School, and 2017
Summer Research Fellowship from CBMG.
I’m very lucky to have my husband, my best friend, Wanpeng, to keep me
company during this journey with his love and support, who has already and will
continue to enrich my life.
I owe my deepest thanks to my family, my mother and father for their uncon-
ditional love, support, encouragement, and believing in me.
iii
Table of Contents
Acknowledgements ii
List of Tables vii
List of Figures viii
List of Abbreviations x
1 Introduction and Literature Review 1
1.1 The Plant Immune System . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.1 PTI and ETI – detection of pathogens . . . . . . . . . . . . . 3
1.1.2 Plant defense signaling network . . . . . . . . . . . . . . . . . 9
1.1.2.1 The MAPK signaling cascade . . . . . . . . . . . . . 11
1.1.2.2 Plant hormone signaling . . . . . . . . . . . . . . . . 12
1.1.2.3 The EDS1 signaling node . . . . . . . . . . . . . . . 15
1.2 Plant–Powdery Mildew Interaction . . . . . . . . . . . . . . . . . . . 15
1.2.1 The Extra-Haustorial Membrane . . . . . . . . . . . . . . . . 16
1.2.2 Pre- and Post-invasion Resistance . . . . . . . . . . . . . . . . 17
1.2.3 The Arabidopsis-Golovinomyces cichoracearum Pathosystem . 19
1.2.4 RPW8, a Unique Powdery Mildew Resistance Locus . . . . . . 21
1.2.5 MLO, Another Unique Powdery Mildew Resistance Locus . . 24
1.3 Phospholipase D and Phosphatidic Acid in Plant Immunity . . . . . . 27
1.3.1 PLD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.3.2 Signal-induced PA Production . . . . . . . . . . . . . . . . . . 29
1.3.3 PLD and PLD-derived PA in Plant Immunity . . . . . . . . . 31
1.4 Conception and Significance of this Study . . . . . . . . . . . . . . . 32
2 Arabidopsis Phospholipase Dα1 and Phsopholipase Dδ oppositely modulate
basal resistance against powdery
mildew independent of EDS1/PAD4 and salicylic acid 35
2.1 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
iv
2.3.1 PLDα1 and PLDδ play opposing roles in post-penetration re-
sistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.3.2 Loss of PLDα1 or PLDδ affects basal resistance against an
oomycete but not ETI . . . . . . . . . . . . . . . . . . . . . . 48
2.3.3 PLDδ is dispensable for RPW8-mediated resistance . . . . . . 52
2.3.4 PLDα1 and PLDδ have distinct subcellular localizations . . . 54
2.3.5 PLDδ contributes to resistance independent of EDS1/PAD4,
SA, and JA signaling pathways . . . . . . . . . . . . . . . . . 58
2.3.6 Loss of PLDα1 and/or PLDδ has no significant impact on SA,
JA, and ABA levels and signaling . . . . . . . . . . . . . . . . 64
2.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.4.1 PLDδ and PLDα1 modulate post-penetration resistance against
powdery mildew . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.4.2 PLDα1 and PLDδ may modulate defense via a potentially
novel pathway . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.4.3 PLDα1 may repress PLDδ-mediated defense signaling . . . . . 72
2.5 Material and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 74
2.5.1 Plant lines and growth conditions . . . . . . . . . . . . . . . . 74
2.5.2 DNA Constructs, Plant Transformation and Microscopy . . . 75
2.5.3 Pathogen Infection, Disease Phenotyping, and Quantification . 77
2.5.4 In situ Detection of H2O2 Accumulation and Callose Deposition 78
2.5.5 Determination of Endogenous SA, JA, and ABA Concentrations 79
2.5.6 qRT-PCR Analysis . . . . . . . . . . . . . . . . . . . . . . . . 81
2.5.7 JA sensitivity assay . . . . . . . . . . . . . . . . . . . . . . . . 81
3 Mutant Screens to Identify Novel Immune Components against Powdery
Mildew in Arabidopsis 83
3.1 Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
3.3.1 The design and implementation of a sensitive mutant screen . 88
3.3.2 Infection phenotypes of the mutants to two powdery mildews
and an oomycete . . . . . . . . . . . . . . . . . . . . . . . . . 90
3.3.3 Mapping and identifying causal mutations by whole-genome
sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.3.4 CIPI1 encodes MAPK Phosphatase 1, a negative regulator of
plant immunity . . . . . . . . . . . . . . . . . . . . . . . . . . 96
3.3.4.1 Identification of the causal mutation in cipi1 . . . . . 96
3.3.4.2 Verification of the role of MKP1 in immunity by
CRISPR/Cas9-induced targeted mutagenesis . . . . 98
3.3.5 Five cipi mutants contain mutations in MLO2 . . . . . . . . . 101
3.3.5.1 Isolation of five cipi mutants with better resistance
to Gc UCSC1 than to Gc UMSG1 . . . . . . . . . . 101
3.3.5.2 Identification of five independent mlo2 alleles . . . . 104
3.3.5.3 Knocking out MLO6 and MLO12 in cipi3 . . . . . . 107
v
3.3.5.4 MLO2-GFP exhibits focal accumulation at the PM
penetration site . . . . . . . . . . . . . . . . . . . . . 109
3.3.6 snap1 is susceptible to a more distant, non-adapted PM Po-
dosphaera aphanis . . . . . . . . . . . . . . . . . . . . . . . . . 111
3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
3.4.1 The eps-Gc UMSG1 artificial pathosystem enables a sensitive
genetic screen . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
3.4.2 Plant defense mechanisms independent of EDS1/PAD4 and SA116
3.4.2.1 Negative regulation of PTI signaling by MKP1 . . . 116
3.4.2.2 Possible mechanisms of mlo-based powdery mildew
resistance . . . . . . . . . . . . . . . . . . . . . . . . 118
3.4.3 Towards better understanding multi-layered non-host resistance121
3.5 Material and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 123
3.5.1 Plant lines and growth conditions . . . . . . . . . . . . . . . . 123
3.5.2 Pathogen Infection, Disease Phenotyping, and Quantification . 123
3.5.3 Plant Transformation and Microscopy . . . . . . . . . . . . . . 123
3.5.4 EMS mutagenesis . . . . . . . . . . . . . . . . . . . . . . . . . 123
3.5.5 Preparation of CRISPR DNA constructs . . . . . . . . . . . . 124
3.5.6 Genotyping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3.5.7 Genome Sequencing and Mapping . . . . . . . . . . . . . . . . 126
3.5.8 Accession Numbers . . . . . . . . . . . . . . . . . . . . . . . . 127
4 Conclusions and Future Directions 129
A Co-authored Non-thesis Manuscripts Published or In-preparation 134
A.1 Published Manuscripts . . . . . . . . . . . . . . . . . . . . . . . . . . 134
A.1.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
A.1.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
A.1.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
A.1.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
A.1.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
A.2 Manuscripts In Preparation . . . . . . . . . . . . . . . . . . . . . . . 143
A.2.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
A.2.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
A.2.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
Bibliography 149
vi
List of Tables
2.1 Arabidopsis T-DNA insertion mutants screened in Chapter 2 . . . . . 41
2.2 Primers used in Chapter 2 . . . . . . . . . . . . . . . . . . . . . . . . 82
3.1 Summary of cipi and snap mutants subjected to WGS analysis . . . . 94
3.2 Primers used in Chapter 3 . . . . . . . . . . . . . . . . . . . . . . . . 128
vii
List of Figures
1.1 Schematic of the plant immune system. . . . . . . . . . . . . . . . . . 5
1.2 Strategies for NLR-mediated detection of pathogens . . . . . . . . . . 7
1.3 Plant defense signaling network . . . . . . . . . . . . . . . . . . . . . 10
1.4 RPW8.2 Is Targeted to the EHM . . . . . . . . . . . . . . . . . . . . 22
1.5 Phylogenetic analysis of selected MLO proteins . . . . . . . . . . . . 25
1.6 Arabidopsis PLD domain structures and biochemical properties . . . 28
1.7 Formation and attenuation of phosphatidic acid . . . . . . . . . . . . 30
2.1 Disease reaction phenotypes of T-DNA insertion mutants . . . . . . . 43
2.2 PLDα1 and PLDδ play opposing roles in post-penetration resistance
against Gc UCSC1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.3 PLDα1 and PLDδ can complement pldα1 and pldδ . . . . . . . . . . 46
2.4 pldα1 and pldδ infected with Gc UMSG1 . . . . . . . . . . . . . . . . 47
2.5 H2O2 production and callose deposition in pldα1 and pldδ . . . . . . . 49
2.6 pldα1 and pldδ infected with oomycete pathogens . . . . . . . . . . . 50
2.7 Loss of PLDα1 and/or PLDδ does not affect ETI against bacterial
pathogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
2.8 PLDα1 and PLDδ are dispensable for RPW8-mediated resistance . . 53
2.9 The PLDδ-eGFP and PLDα1-eGFP fusion proteins are functional . . 55
2.10 Subcellular localizations of PLDα1 and PLDδ . . . . . . . . . . . . . 56
2.11 pldδ-containing mutants infected with Gc UCSC1 . . . . . . . . . . . 59
2.12 pldδ-containing mutants infected with Gc UMSG3 . . . . . . . . . . . 61
2.13 pldα1-containing mutants infected with Gc UCSC1 . . . . . . . . . . 63
2.14 Impact of the pldα1 and pldδ mutations on the levels and signaling
of SA, JA and ABA . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.15 Working model for PLDα1 and PLDδ in plant immunity . . . . . . . 68
3.1 Infection phenotypes of eps to different powdery mildews . . . . . . . 87
3.2 Schematic of the sensitive mutant screen . . . . . . . . . . . . . . . . 89
3.3 Mapping approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
3.4 Identification of MKP1 as a negative regulator of plant immunity . . 97
3.5 Knocking out MKP1 by CRISPR/Cas9 results in leaf necrosis and
PM resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
viii
3.6 Infection Phenotypes of the five cipi mutants containing mutations
in MLO2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.7 Identification of the causal mol2 mutations in the five cipi mutants . . 105
3.8 Indel mutations induced by CRISPR/Cas9 in MLO6 and MLO12 in
cipi3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
3.9 MLO2-GFP is functional and exhibits focal accumulation around PM
penetration sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
3.10 Mutant snap1 is susceptible to strawberry mildew Pa UMSG1 . . . . 112
ix
List of Abbreviations
Avr avirulence factor
Bgh Blumeria graminis f.sp. hordei
BSA bulked segregant analysis
CC-NB-LRRs coiled-coil–nucleotide-binding site–leucine-rich repeat
CRCK3 CALMODULINBINDING
RECEPTORLIKE CYTOPLASMIC KINASE 3
DAB 3,3’-diaminobenzidine
DAG diacylglycerol
DAMP damage-associated molecular pattern
CIPI compromised immunity yet poor infection
DGK diacylglycerol kinase
DGPP diacylglycerol pyrophosphate
dpi days post inoculation
DR disease reaction
EDS1 ENHANCED DISEASE SUSCEPTIBILITY 1
EF-Tu elongation factor thermo unstable
EFR EF-Tu Receptor
EHM extra-haustorial membrane
ET ethylene
ETI effector-triggered immunity
ETS effector-triggered susceptibility
FLS2 Flagellin-sensitive 2
Gc Golovinomyces cichoracearum
Hpa Hyaloperonospora arabidopsidis
hpi hours post inoculation
HR hypersensitive response
JA jasmonic acid
LPP lipid phosphate phosphatase
x
MAPK mitogen-activated protein kinase
MAPKK MAPK kinase
MAPKKK MAPK kinase kinase
MKP1 MAP kinase phosphatase 1
MLO Mildew Resistance Locus O
NB-LRR nucleotide-binding site–leucine-rich-repeat
NDR1 NON-RACE-SPECIFIC DISEASE RESISTANCE 1
NPR1 NONEXPRESSER OF PR GENES 1
PA phosphatidic acid
Pa Podosphaera aphanis
PAD4 PHYTOALEXIN-DEFICIENT 4
PAK PA kinase
PAMP pathogen-associated molecular pattern
PCD programmed cell death
PEN1 PENETRATION 1
PEN2 PENETRATION 2
PEN3 PENETRATION 3
PIP5K phosphatidylinositol 4-phosphate 5-kinase
PLC phospholipase C
PLD phospholipase D
PM plasma membrane (Chapter 2)
PM powdery mildew (Chapter 3)
Pma Pseudomonas syringae pv. maculicola
pPLA patatin-related phospholipase
PR pathogenesis-related
PRR pattern recognition receptor
PtdIns(4,5)P2 phosphatidylinositol-(4,5)-bisphosphate
PTI PAMP-triggered immunity
PTP1 PROTEIN TYROSINE PHOSPHATASE 1
RLK receptor-like kinase
RLP receptor-like protein
SA salicylic acid
SAG101 SENESCENCE-ASSOCIATED GENE 101
SAR systemic acquired resistance
SID2 SALICYLIC ACID INDUCTION DEFICIENT 2
xi
SNAP susceptible to non-adapted pathogens
SNC1 SUPPRESSOR OF npr1-1, CONSTITUTIVE
TIR-NB-LRRs Toll-interleukin 1 receptor–NB-LRRs
UBC9 ubiquitin conjugating enzyme 9
WGS whole-genome sequencing
xii
Chapter 1: Introduction and Literature Review
Plants are nutritious food for humans, animals, and not surprisingly, mi-
crobes. Epidemics of plant diseases caused by pathogenic fungi, oomycetes, bac-
teria, and viruses may lead to devastating consequences for agriculture at a local or
regional scale. For example, the notorious potato late blight caused by the oomycete
pathogen Phytophthora infestans, led to the Irish Famine in the 1840s and still trou-
bles potato production this day (Saville et al., 2016); wheat stem, stripe, and leaf
rusts caused by the obligate fungal pathogens Puccinia graminis, P. striiformis, and
P. triticina, respectively, threat wheat production on a global scale (Figueroa et al.,
2018); and citrus greening caused by the bacteria Candidatus liberbacter asiaticus is
devastating millions of acres of citrus crops throughout the world (Wang and Trivedi,
2013). It has been estimated that 15% of the global crop production is lost due to
preharvest plant diseases every year (Dangl et al., 2013). These plant pathogens
pose a constant threat to food security especially in the face of an ever-growing
world population and climate change.
The past three decades have witnessed not only great advances made towards
understanding of the molecular mechanisms of the plant immune system, but also
potential/successful application of new knowledge in breeding disease resistant crops
1
(Dangl et al., 2013, Kreuze and Valkonen, 2017). To date, numerous disease resis-
tance (R) genes have been identified from various plant species and their major
defense pathways have been characterized. Conventional breeding for disease resis-
tance has been greatly facilitated by marker-assisted R gene selection and R gene
pyramiding. The advent of plant transgenic technologies has further expanded our
ability to create crop cultivars with novel disease resistance. For instance, the Hawaii
papaya industry was saved from the devastating Papaya ringspot virus (PRSV) by
deployment of transgenic papaya plants expressing the coat protein of PRSV (Gon-
salves and Ferreira, 2003). Similarly, transgenic squash expressing viral coat proteins
have shown to display complete resistance to multiple viruses (Tricoll et al., 1995).
Many other transgenic crop plants with engineered disease resistance have been un-
der field trials and may be deployed in agriculture in the near future (Dangl et al.,
2013).
Despite tremendous progress mentioned above, many fundamental questions
concerning plant-pathogen interaction still remain to be answered. For example, the
ways plants mount spatiotemporally appropriate defenses against different pathogens
with distinct modes of parasitism, and how fungal and oomycete pathogens subvert
host immunity and acquires nutrients from host cells are still not well character-
ized. Moreover, both the host and the pathogen are constantly co-evolving or even
engaged in an ever-escalating arms-race especially in agricultural settings. Thus,
understanding the molecular, genetic and evolutionary principles underlying the
dynamic plant-pathogen interaction remains to be a continuous challenge for the
field of plant pathology and crop protection.
2
1.1 The Plant Immune System
Unable to move around, plants have to directly face various challenges from
the environment including all kinds of pathogens. Moreover, unlike animals, plants
lack mobile immune cells and a somatic adaptive immune system. Thus, plants have
evolved preformed physical and chemical barriers (such as epicuticular layers and
constitutive production of certain secondary metabolite of antimicrobial activities)
and a sophisticated inducible innate immune system to protect themselves from
various (potential) pathogens. The molecular basis of the latter has been the subject
of extensive study in the past three decades (Jones and Dangl, 2006) and will be
briefly reviewed below.
1.1.1 PTI and ETI – detection of pathogens
The plant innate immune system consists of two evolutionarily inter-related
branches. Activation of the plant immune system and the response strength are
mostly determined by modes of pathogen detection.
The first branch uses cell surface-localized pattern recognition receptors (PRRs)
to detect pathogen- or microbial-associated molecular patterns (PAMPs/MAMPs)
and damage-associated molecular patterns (DAMPs). PRRs are receptor proteins
containing a ligand-binding ectodomain, a single-pass transmemebrane domain, and
depending on the presence/absence of an internal C-terminal kinase domain, are
grouped into either receptor-like kinases (RLKs) with an intracellular kinase domain,
or receptor-like proteins (RLPs) lacking any obvious internal signaling domains
3
(Boutrot and Zipfel, 2017). Thus, for RLPs to transduce a signal, they often work
together with RLKs (Boutrot and Zipfel, 2017). PAMPs/MAMPs from pathogens
are mostly highly conserved molecules that are both unique and essential to mi-
crobes and are exposed to PRRs during initial contact with plants. Well-studied
examples of plant PRRs and the PAMPs they recognize include FLAGELLIN SEN-
SITIVE2 (FLS2), an RLK that detects bacterial PAMP flagellin (Gómez-Gómez and
Boller, 2000, Hann and Rathjen, 2007, Robatzek et al., 2007, Takai et al., 2008, Trdá
et al., 2014), EF-TU RECEPTOR (EFR), another RLK that senses bacterial EF-Tu
(Zipfel et al., 2006), and OsCEBiP, a rice RLP that perceives fungal chitin (Kaku
et al., 2006). DAMPs, on the other hand, are molecules derived from damaged
host tissues upon pathogen invasion. Though induced by pathogens, the produc-
tion of DAMPs is less pathogen-specific. This enables plants to detect a broader
range of pathogens. Extracellular ATP is such a DAMP that is detected by Ara-
bidopsis DORN1 (Choi et al., 2014). Perception of these non-self and damaged-self
molecules by PRRs results in an immune response termed PRR-triggered immunity,
or PAMP-/pattern-triggered immunity (PTI) (Fig. 1.1).
However, as an outcome of a long iterative host-adaptation process of a par-
ticular microbe, PTI can be suppressed by effector proteins delivered into the plant
cell by an invading microbe, resulting in effector-triggered susceptibility (ETS) and
the establishment of a host-pathogen relationship (ETS, Fig. 1.1).
To counter effector-mediated suppression of PTI, plants have evolved intra-
cellular immune receptors to recognize the presence and/or virulence activity of
specific effectors and activate a stronger and longer-lasting defense response termed
4
Figure 1.1: Schematic of the plant immune system. In this
scheme, the ultimate amplitude of disease resistance or susceptibil-
ity is proportional to [PTI - ETS + ETI]. In phase 1, plants detect
microbial/pathogen-associated molecular patterns (MAMPs/ PAMPs,
red diamonds) via PRRs to trigger PAMP-triggered immunity (PTI). In
phase 2, successful pathogens deliver effectors that interfere with PTI, or
otherwise enable pathogen nutrition and dispersal, resulting in effector-
triggered susceptibility (ETS). In phase 3, one effector (indicated in red)
is recognized by an NB-LRR protein, activating effector-triggered immu-
nity (ETI), an amplified version of PTI that often passes a threshold for
induction of hypersensitive cell death (HR). In phase 4, pathogen iso-
lates are selected that have lost the red effector, and perhaps gained new
effectors through horizontal gene flow (in blue)these can help pathogens
to suppress ETI. Selection favours new plant NB-LRR alleles that can
recognize one of the newly acquired effectors, resulting again in ETI.
Figure from (Jones and Dangl, 2006)
5
effector-triggered immunity (ETI). This is the second branch of the immune system
(Fig. 1.1). As in animals, these plant intracellular immune receptors belong to the
nucleotide-binding, leucine-rich repeat (NLR) superfamily. Specific NLR-effector
recognition occurs at the very first step of, and is essential for ETI. Both direct and
indirect recognitions of effectors by NLRs haven been observed, and four modes of
NLR sensing mechanisms are illustrated in Fig. 1.2 (Jones et al., 2016).
The “direct” mode suggests a direct physical interaction between an NLR and
the cognate effector, which is an extrapolation from the “gene-for-gene” hypothesis
by Flor in the 1940s (Flor, 1971), as demonstrated by direct and specific interactions
between the NLRs encoded by the R genes at the AL5/6/7 loci of flax and the
effectors (also termed avirulence factors) encoded by the cognate AvrL5/6/7 genes
from the flax rust pathogen (Ellis et al., 2007). However, such direction interactions
are not observed for most of the NLR-effector pairs. Indirect interactions are more
prevalent. In this case, NLRs (guards) monitor/guard the integrity of the host
proteins (guardees) targeted by effectors. These guards activate immunity once
modifications of the guardees as a result of virulence activities from the effectors are
detected. This is described as the “guard hypothesis” by Dangl & Jones in 2001
(Dangl and Jones, 2001). For example, RIN4 is a host target of multiple pathogen
effectors guarded by NLRs RPM1 and RPS2. It can be phosphorylated by effectors
AvrB or AvrRPM1 and cleaved by effector AvrRpt2, which can activate RPM1- and
RPS2-mediated resistance, respectively (Chung et al., 2011, Axtell and Staskawicz,
2003). As mentioned earlier, the primary purpose of pathogen effectors is to suppress
immunity and promote pathogenesis. Conceivably, effector manipulation of such
6
Figure 1.2: Strategies for NLR-mediated detection of pathogens.
Four conceptually distinct strategies are illustrated. The details of how
each strategy is implemented for a specific NLR example may vary. The
guard and decoy strategies are analogous: In both cases, the guardee
or decoy proteins are involved in maintaining the NLR in an inhibited
state, and in both cases, the inhibition is relieved upon effector-mediated
modification of the guardee or decoy. Guardees are distinguished from
decoys by having an additional and separate function in host defense,
whereas decoys are merely mimics of host defense proteins. Guardees are
thus the intended targets of effectors, whereas decoys are inadvertently
targeted by effectors. Figure from (Jones et al., 2016)
7
host target proteins can contribute to ETS in the absence of guard NLRs, which
imposes selective pressure on the hosts to mutate target proteins to evade pathogen
manipulation. Indeed, paralogs of some guardees have been found to have mutations
such that they are still targets of effectors but do not contribute to host susceptibility.
This has been described as the “decoy” model (van der Hoorn and Kamoun, 2008).
In addition, to avoid possible separation of decoys from the guard NLRs due to
recombination, direct incorporation of the decoy domain into an NLR has been
observed in some cases (Bernoux et al., 2014), which may be evolutionarily more
favorable for plants. Similarly, for NLRs working in pairs , genetic linkage between
the two could also reduce the risk of inappropriate allelic combinations to avoid
unwanted consequences such as autoimmune responses. The Arabidopsis RPS4 and
RRS1 are a pair of such NLRs that form an immune complex for the recognition of
bacterial effectors, AvrRps4 and PopP2 (Narusaka et al., 2009, Huh et al., 2017).
This is described as an “integrated decoy” model, which is evolutionarily more stable
(Fig. 1.2, Jones et al., 2016).
Based on the features of the N-terminal domains, NLRs are divided into two
major classes, Toll-interleukin 1 receptor (TIR)-NLRs and coiled-coil (CC)-NLRs.
While characterized TIR-NLRs require the nucleocytoplasmic lipase-like protein EN-
HANCED DISEASE SUSCEPTIBILITY 1 (EDS1) for signal transduction, most
CC-NLRs engage the plasma membrane-anchored integrin-like protein NON-RACE-
SPECIFIC DISEASE RESISTANCE 1 (NDR1) for signaling (Aarts et al., 1998, Cui
et al., 2015).
Essentially, the plant immune system performs surveillance of non-self, damaged-
8
self and altered-self controlled by PRRs at the cell surface or NLRs within the cell.
In general, PTI contributes to basal and broad-spectrum resistance against non-
adapted as well as adapted pathogens and ETI is more race-specific with much
stronger and longer-lasting responses (Cui et al., 2015). While the defense outcome
strength is very different, PTI and ETI are believed to be evolutionarily inter-
related and mechanistically interconnected, as both share an overlapping array of
downstream defense responses including PR gene expression, reactive oxygen species
(ROS) production and callose deposition via conserved interwoven signaling path-
ways that are regulated by salicylic acid (SA), jasmonic acid (JA), and ethylene (ET,
C2H4) (Bari and Jones, 2009, Pieterse et al., 2012). In addition, ETI is thought to
be an amplified PTI, in which case PTI is derepressed by activated NLRs (Cui et al.,
2015).
1.1.2 Plant defense signaling network
Following pathogen detection in either PTI or ETI, despite their differences in
modes of pathogen detection and outcome strength, an overlapping array of down-
stream defense responses are evoked including activation of mitogen-activated pro-
tein kinase (MAPK) signaling, ROS production, Pathogenesis-related (PR) gene ex-
pression, phytohormone elevation and rapid programmed cell death (PCD) known
as the hypersensitive response (HR) at the site of infection to restrict the spreading
of the pathogen (Thomma et al., 2011, Cui et al., 2015).
9
Any pathogen Biotrophs & hemi-biotrophs Necrotrophs
& insects
RR
s
P
ETI
PTI CC-NLR TIR-NLR
NDR1 EDS1/PAD4
MEKK1 MAPKKK3/5
MKK1/2 MKK4/5 SID2
MPK4 MPK3/6 JA/ ET
SA
MKP1
PTP1 NPR1
AP2C1
Resistance Resistance & HR Resistance
Figure 1.3: Plant defense signaling network. Schematic illustration
of major components in plant defense signaling.
10
1.1.2.1 The MAPK signaling cascade
MAPK signaling cascades are highly conserved signaling modules in eukaryotes
that mediate intracellular signaling transduction from external stimuli. It is one of
the earliest signaling events activated following the detection of pathogens in both
PTI and ETI (Meng and Zhang, 2013, He et al., 2018). A typical MAPK cascade
consists of three consecutive kinases to relay signals from MAPK kinase kinases
(MAPKKKs) to MAPK kinases (MAPKKs) and to MAPKs.
Perception of PAMPs by PRRs can trigger two distinct MAPK cascades in
Arabidopsis (Fig. 1.3). One consists of MAPKKK3/MAPKKK5, MKK4/MKK5,
and MPK3/MPK6, which is activated by RLCK VII-4 acting downstream of at least
FLS2, EFR, LYK5 and PEPR (Bi et al., 2018, Asai et al., 2002). The other one is
composed of MEKK1 (a MAPKKK), MKK1/MKK2 (two redundant MAPKKs),
and MPK4, which is found to act downstream of FLS2 (Ichimura et al., 2006,
Nakagami et al., 2006, Qiu et al., 2008, Gao et al., 2008, Suarez-Rodriguez et al.,
2007).
In ETI, MPK3/MPK6 shows prolonged activation to induced SA-responsive
genes in the CC-NLR RPS2-mediated ETI (Tsuda et al., 2013). The MPK4 cas-
cade is guarded by another CC-NLR SUMM2 (suppressor of mkk1 mkk2) through
monitoring the phosphorylation state of a MPK4 substrate, CALMODULINBIND-
ING RECEPTORLIKE CYTOPLASMIC KINASE 3 (CRCK3). Disruption of the
MPK4 cascade by a bacterial effector HopAI1 from Pseudomonas syringae could
reduce the phosphorylation of CRCK3, resulting in SUMM2-mediated immune re-
11
sponses (Zhang et al., 2012, 2007, 2016b).
Proper control of the magnitude and duration of the MAPK signaling is essen-
tial for the host cell to mount appropriate defense responses, while avoiding poten-
tially detrimental effects. Protein phosphatases are known negative regulators that
dephosphorylate and inactivate MAPKs (Bartels et al., 2010). In Arabidopsis, phos-
phatases of three major types have all been shown to negatively regulated pathogen-
induced MAPK cascades, including AP2C1 (a PP2C (Ser/Thr protein phosphatases
of type 2C)-type phosphatase), PTP1 (PROTEIN TYROSINE PHOSPHATASE 1,
a Tyr-specific phosphatase), and MKP1 (MAP KINASE PHOSPHATASE1, a Thr
and Tyr dual specificity phosphatase) (Fig. 1.3, Shubchynskyy et al., 2017, Bartels
et al., 2009).
1.1.2.2 Plant hormone signaling
The plant immune network is regulated by various plant hormones, among
which SA, JA and ET are the major defense hormones during PTI or ETI activation
(Fig. 1.3, Bari and Jones, 2009, Pieterse et al., 2012).
SA plays a major role in mediating localized PCD as well as systemic acquired
resistance (SAR) against biotrophic and hemi-biothrophic pathogens that rely on
living host cells for establishing infection (Vlot et al., 2009). Cellular SA accumula-
tion constitutes an early signaling event during ETI. This step requires components
including EDS1, and its homologous & interacting partners PHYTOALEXIN DEFI-
CIENT 4 (PAD4) and SENESCENCE-ASSOCIATED GENE 101 (SAG101) (posi-
12
tive regulators of SA signaling) (Falk et al., 1999, Jirage et al., 1999, Feys et al., 2001,
Wagner et al., 2013), as well as SALICYLIC ACID INDUCTION DEFICIENT 2
(SID2) (required for 90% stress-induced SA biosynthesis) (Wildermuth et al., 2001)
and EDS5 (required for SA transport from the chloroplast to the cytoplasm) (Ser-
rano et al., 2013). Elevation of SA level has been shown to change the cellular
redox states, leading to monomerization and translocation to the nucleus of the
cytoplasm localized NPR1 oligomer, the master regulator of SA (Spoel and Dong,
2012). Monomeric NPR1 acts as a co-activator of transcription factors to induce
expression of antimicrobial proteins encoded by pathogenesis-related (PR) genes.
Meanwhile, NPR1 accumulates in neighboring cells to promote cell survival and SA-
mediated resistance relating to promote SAR (Fu and Dong, 2013, Yan and Dong,
2014). Defenses activated in the infected cell also generate mobile immune signals
such as azelaic acid (AzA) (Jung et al., 2009), glycerol-3-phosphate (G3P) (Chanda
et al., 2011), methyl salicylic acid (MeSA) (Park et al., 2007), abietane diterpenoid
dehydroabietinal (DA) (Chaturvedi et al., 2012) and N-hydroxylpipecolic acid (Hart-
mann et al., 2018). These molecules can travel to distal uninfected plant tissue to
activate defense against secondary infection, known as SAR, a broad-spectrum and
long-lasting immune response. In addition, SA signaling engages a feedback circuit
to amplify defense responses (Wiermer et al., 2005), which is negatively regulated
by EDR1, a Raf-like MAPK kinase kinase (Frye et al., 2001, Xiao et al., 2005).
How SA is perceived in the cell has long been an interest of study. While two
separate studies show that NPR1 also serves as the cytoplasmic receptor of SA (Wu
et al., 2012, Manohar et al., 2015), another study identifies NPR3 and NPR4, two
13
paralogs of NPR1, but not NPR1 to be SA receptors (Fu et al., 2012). This remained
controversial until a recent report demonstrated that both NPR1 and NPR3/4 are
bona fide SA receptors with NPR1 being the positive transcriptional regulator and
NPR3/4 acting independently as the co-repressors of defense-related gene expression
(Ding et al., 2018).
On the other hand, JA and ET are critical in immunity against necrotrophs
that aim to kill and feed on dead host cells (Pieterse et al., 2012). In many cases,
the JA/ET pathways work antagonistically with the SA pathway (Glazebrook, 2005,
Pieterse et al., 2012, Zheng et al., 2012, Van der Does et al., 2013). However, by
examining the contributions of SA, JA, ET, and PAD4–four major defense signal-
ing sectors–to plant immunity systematically, it has been revealed that these four
sectors could have synergistic relationships in PTI triggered by flg22 (Tsuda et al.,
2009, Kim et al., 2014), as well as compensatory relationships in ETI induced by
AvrRpt2 (Tsuda et al., 2009). Whether it’s antagonism or synergism or compensa-
tion between these four signaling sectors or even other hormonal pathways largely
depends on the context of specific plant-pathogen interactions (Robert-Seilaniantz
et al., 2011). The sophisticated control of the relationships between these four sig-
naling sectors enables a resilient immune network that is robust enough to fight
against fast-evolving pathogens while maintaining tunability to optimize immune
responses and minimize fitness cost (Tsuda et al., 2009, Kim et al., 2014, Hillmer
et al., 2017).
14
1.1.2.3 The EDS1 signaling node
EDS1 is a lipase-like protein with no lipase activity detected so far for regulat-
ing plant defense signaling, and is considered a lipase-like molecular switch (Wagner
et al., 2013, Cui et al., 2015). Its homologous signaling partners PAD4 and SAG101
compete for binding to EDS1 to form soluble nucleocytoplasmic (EDS1-PAD4) and
nuclear (EDS1-SAG101) complexes (Feys et al., 2005, Rietz et al., 2011, Wagner
et al., 2013). Balanced partitioning of EDS1 in the nuclear and cytoplasm is impor-
tant for appropriate defense output (Garćıa et al., 2010). EDS1 is not only required
for basal and all tested TIR-NLR resistance together with PAD4 to induce SA ac-
cumulation through SID2 activity, but also functions in parallel with SA in basal,
TIR-NLR, and CC-NLR resistance, which makes the plant immune system resilient
to perturbation or inhibition of SA signaling from pathogens (Fig. 1.3, Cui et al.,
2017).
1.2 Plant–Powdery Mildew Interaction
Powdery mildew is a fungal disease affecting about 10,000 plant species from
a wide range of monocotyledonous and dicotyledonous (angiosperms) plants includ-
ing many agriculturally and economically important crops (wheat, barley, cucumber,
grapevine, etc.) and ornamental plants (rose, magnolia, azalea, etc) (Takamatsu,
2004). It is ranked among the most important plant diseases. The causal agents are
obligate biotrophic fungi that require living host cells to survive and reproduce be-
longing to Ascomycetes in the order of Erysiphales. The fungal mycelia (filamentous
15
vegetative structures) and conidia (asexual spores) can grow superficially on any of
the green tissue, and thus, infected plants display whitish to greyish dusty powder
on the above ground part, most prominently on both the adaxial and abaxial sides
of the leaves. Defoliation, cosmetic damage of ornamentals, reduced yields, and
lowered quality are the common damages caused by powdery mildew.
1.2.1 The Extra-Haustorial Membrane
Haustoria are feeding organs of oomycetes and several fungi including powdery
mildew. These specialized intracellular hyphae, in the case of powdery mildew
contain a spherical main body with small lobes (as in most dicot mildew) or long
finger-like membrane structures emanating from the main body. Upon penetration
of the host epidermal cell wall by the appressorium, a sporeling of powdery mildew
differentiates a haustorium from the tip of the penetration peg inside the host cell.
Early electron microscopic studies suggest that the haustorium is separated from the
host cytoplasm by an interfacial membrane named the extrahaustorial membrane
(EHM) (Gil and Gay, 1977, Roberts et al., 1993). Conceivably, the EHM is the
major host-pathogen battleground where pathogens send effectors across to suppress
host defense and manipulate nutrient uptake, and the host plants may activate their
haustorium-targeted defenses there. However, the EHM is poorly characterized, and
even less is known about the molecular interaction between the host and pathogen
at the EHM. Using the Arabidopsis thaliana-G. cichoracearum (Gc UCSC1, which
causes powdery mildew on many dicot plants including Arabidopsis) pathosystem,
16
Koh et al in 2005 showed that eight host plasma membrane proteins were absent
from the EHM, suggesting that the EHM is distinct from the plasma membrane
Koh et al., 2005. Indeed, Berkey et al recently demonstrated that EHM is physically
separable from the PM and most likely is de novo synthesized during the biogenesis
of the haustorium(Berkey et al., 2016).
1.2.2 Pre- and Post-invasion Resistance
While the same two-branched central immune system is utilized by plants to
fight against a broad range of pathogens including bacteria and fungi, plant defense
against different types of pathogens exhibits distinct spatiotemporal characteristics.
Plant resistance against powdery mildew can be conveniently divided into pre- and
post-invasion stages, also known as penetration and post-penetration resistance.
Penetration resistance is cell-wall based and often manifested by papilla formation,
which is a thickening of the cell-wall with deposition of callose (1,3-beta glucan,
contributing to the formation of papillae) and other defense chemicals at the site of
penetration. Previous studies have shown that there are at least two independent
mechanisms contributing to penetration resistance in Arabidopsis. One involves a
focal exocytosis of antimicrobial materials controlled by PENETRATION1 (PEN1),
a syntaxin, and its SNARE partners (Collins et al., 2003, Kwon et al., 2008); the
other engages PEN2 myrosinase to produce an antifungal glucosinonate product,
which is then transported by the PEN3 ATP-binding cassette transporter (Lipka
et al., 2005, Stein et al., 2006, Bednarek et al., 2009). Both mechanisms are likely
17
activated upon recognition of PAMPs by PRRs, and thus may be part of or con-
nected to PTI (Jones and Dangl, 2006, Hückelhoven and Panstruga, 2011).
While most nonhost powdery mildew pathogens can be stopped from enter-
ing the plant cell by penetration resistance, some can still overcome this layer of
defense. Sow thistle powdery mildew Golovinomyces cichoracearum (Gc) UMSG1
is such an example (Wen et al., 2011). Despite successful penetration, its further
growth and development is inhibited by stage I post-penetration resistance, which
involves encasement of the haustorial complex by a callose-rich cell wall-like struc-
ture thought to be derived from the papilla by “rim growth” and HR cell death.
Both SA-dependent and SA-independent mechanisms have been shown to account
for this stage I post-penetration resistance, part of which may stem from PTI (Wen
et al., 2011). Since Gc UMSG1 fails to sporulate (i.e. cannot complete its life
cyle) in all 25 Arabidopsis accessions tested, by definition it is still a non-adapted
pathogen of Arabidopsis. Therefore, post-penetration resistance also contribute to
plant non-host resistance against powdery mildew (Lipka et al., 2005, Wen et al.,
2011).
Once stage I post-penetration resistance is suppressed by effector proteins
secreted from a better-adapted pathogen, the fungus can develop functional haus-
toria to extract nutrients from plant cells, leading to successful reproduction and
colonization. To resist haustorium-based infection, plants have evolved stage II
post-penetration resistance to defeat such better-adapted pathogens to presumably
kill haustortia or constrain their function. Resistance at this stage often occurs
through recognition of the fungal effectors by NLRs to initiate ETI, which is often
18
accompanied with HR at the infection sites. As demonstrated in cereals, barly Mla
and wheat Pm3, Pm21, and Pm60 all encode CC-NLRs that confer resistance to
their respective powdery mildew pathogens (Zhou et al., 2001, Srichumpa et al.,
2005, Bhullar et al., 2010, Xing et al., 2018, Zou et al., 2018). Genes from wild
North American grapevine species Muscadinia rotundifolia MrRUN1 and MrRPV1
encodes two TIR-NLR to confer resistance to powdery mildew and downy mildew,
respectively (Feechan et al., 2013). The defense triggered by these NLRs is often
race specific, and only effective against the powdery mildew isolate that contains
the cognate effector.
In summary, non-host resistance consists of both pre- and stage I post-penetration
resistance that is primarily PTI, while host-resistance depends mostly on post-
penetration resistance, which requires both PTI and ETI.
1.2.3 The Arabidopsis-Golovinomyces cichoracearum Pathosystem
The establishment of Arabidopsis-powdery mildew pathosystem has facilitated
the research on the interactions between host plants and powdery mildew fungi.
Three Golovinomyces cichoracearum (Gc) isolates exhibiting different levels of adap-
tation on the Arabidopsis Col-0 accession are have been maintained in the Xiao lab
at University of Maryland. Gc UCSC1 is a well-adapted host pathogen on Ara-
bidopsis Col-0 wild type plants (Adam and Somerville, 1996). Spores of Gc UCSC1
germinate in 6–10 hours post inocualtion (hpi) and differentiate haustoria at 12–
24 hpi. Within 4 days post inoculation (dpi), a healthy sporeling forms an epiphytic
19
fungal network consisting of mycelia and initial conidiophores. Conidiophores are
specialized hyphae that produce asexual conidia, which when matured fall on host
surface to start a new life cycle. In about 10 days, the fungal colony expands to pro-
duce a massive network with hundreds of thousands of conidiophores with mature
conidia visible as whitish powder to the naked eye.
Gc UMSG1 is an isolate that is infectious to sow thistle (Wen et al., 2011). It
has largely overcome penetration resistance of 25 Arabidopsis accessions examined
and is capable of forming initial haustoria and secodnary hyphae. However, further
development of the initial haustoria is arrested by stage I post-penetration resistance
in Arabidopsis. Consequently, Gc UMSG1 has very limited hyphal growth on the
leaf surface and cannot complete its life cycle. It thus still should be considered
a non-adapted pathogen of Arabidopsis. Interestingly, the eds1 mutant defective
in SA accumulation and signaling can support the fungus to complete its life cycle
with weak sporulation, suggesting that SA-dependent defense contributes to stage
I post-penetration resistance against non-adapted powdery mildew pathogens (Wen
et al., 2011).
Gc UMSG3 is a powdery mildew isolate infectious on tobacco plants. Exhibit-
ing limited sporulation on Arabidopsis Col-0 wild-type, Gc UMSG3 is a weakly-
adapted powdery mildew isolate of Arabidopsis. However, removal of EDS1 or
PAD4 in Arabidopsis allows profuse growth and sporulation of Gc UMSG3 (results
from this dissertation project), further supporting the notion that SA-signaling plays
a critical role in stage I post-penetration resistance.
Exploitation of these three G. cichoracerum isolates enables fine dissection of
20
different layers of plant resistance against powdery mildew invasion.
1.2.4 RPW8, a Unique Powdery Mildew Resistance Locus
Arabidopsis gene locus RESISTANCE TO POWDERY MILDEW8 (RPW8)
identified in the Ms-0 accession contains two homologous R genes, RPW8.1 and
RPW8.2 (Xiao et al., 2001). These two R genes are unique in several respects.
First, unlike most characterized R genes that confer race-specific resistance, RPW8.1
and RPW8.2 (thereafter referred to as RPW8 unless otherwise indicated) activate
broad-spectrum resistance to all powdery mildew pathogens tested including Gc
UCSC1. However, like other R-mediated resistance, genetic analyses reveal that
RPW8-mediated resistance engages SA signaling and requires defense signaling com-
ponents such as EDS1 and PAD4 (Xiao et al., 2003, 2005). Second, these two RPW8
genes encode atypical R proteins with putative N-terminal transmembrane domains
and coiled-coil (CC) domains (Xiao et al., 2001, 2004b). Interestingly, both RPW8
proteins show significant homology (∼25% sequence similarity) to the N-terminal
domain of an ancient clade of NLRs, and thus this N-terminal domain is designated
as the RPW8 domain. Intriguingly, such RPW8-NLRs constitute a distinct NLR
clade sister to the TIR-NLRs and CC-NLRs (Collier et al., 2011, Shao et al., 2014,
Zhang et al., 2016a, Zhong and Cheng, 2016, Qian et al., 2017, Shao et al., 2016)
Later, in an effort to understand how RPW8 proteins confer broad-spectrum
resistance to powdery mildew, RPW8.1 was found to localize to membranous struc-
tures around the chloroplasts in mesophyll cells beneath a haustorium-invaded epi-
21
Figure 1.4: RPW8.2 Is Targeted to the EHM. Leaves of Col-0 or
Col-0 transgenic for NP:RPW8.2-YFP were inoculated with Gc UCSC1
and used for haustorium isolation at 2 days post inoculation. Isolated
haustoria were stained with PI and imaged by confocal. Bars, 5 mm. (A)
A single optical section of an isolated haustorium complex (HC) from
Col-0. Note the weakly propidium iodide (PI)-stained EHM (arrowhead)
and numerous lobes emanating from the main body of the haustorium.
(B) A single optical section of an isolated HC showing precise colocal-
ization of RPW8.2-YFP with the PI-stained EHM. (C) An isolated HC
showing localization of RPW8.2-YFP to the outer surface of the HC.
Note the lightly PI-positive encasement of the HC (EHC). Figure from
(Wang et al., 2009)
22
dermal cell (Wang et al., 2007, Ma et al., 2014). Strikingly, RPW8.2 was found
to be specifically induced by powdery mildew infection and is precisely targeted to
the EHM encasing the haustorium in the leaf epidermal cells (Fig. 1.4, Wang et al.,
2009).
RPW8.2-targeted haustoria often display thick callose-enriched encasement,
interface-focused H2O2 accumulation, and even distortion, suggesting that RPW8
activate defense to constrain the development or function of the haustorium. Given
that two basic residue-enriched motifs in RPW8.2 are critical for EHM targeting
(Wang et al., 2013) and a dominant negative RPW8.2 mutant compromises EHM-
localization and defense function of the wild-type RPW8.2 protein (Zhang et al.,
2015), it appears that a specific protein trafficking pathway is engaged to target
RPW8.2 to the EHM. However, how RPW8.2 achieves EHM-specific localization
to activate SA-dependent, haustorium-targeted defense remains to be determined.
Because phosphoinositides are known to bind target proteins and regulate their
subcellular localization, a tempting speculation is that RPW8.2 may interact with
a particular phospholipid to realize its specific targeting (Di Paolo and De Camilli,
2006, Behnia and Munro, 2005, Pendaries et al., 2005).
Both RPW8.1 and RPW8.2 have shown potential for application. Ectopic ex-
pression of RPW8.1 boosts PTI to fight off powdery mildew and downy mildew in
Arabidopsis (Wang et al., 2007), and against the blast fungus Pyricularia oryzae and
bacterial pathogen Xanthomonas oryzae pv. oryzae in rice (Li et al., 2017). Simi-
larly, extopic expression of RPW8.2 restricts powdery mildew growth in grapevine
(Hu et al., 2018).
23
1.2.5 MLO, Another Unique Powdery Mildew Resistance Locus
The broad-spectrum resistance conferred by RPW8 is reminiscent of that medi-
ated by the loss-of-function of the barley Mildew resistance locus o (Mlo). However,
these two types of resistance are genetically and mechanistically distinct. While
RPW8 is a dominant R gene that activates haustorium-targeted, post-penetration
resistance via an EDS1- and SA-dependent pathway, mlo confers recessively inher-
ited non-race specific durable resistance at the penetration stage to prevent host en-
try of the powdery mildew pathogens via an as-yet unknown mechanism (Jørgensen,
1992, Büschges et al., 1997). The mlo-mediated resistance was first discovered in
barley in the 1930s and 1940s, and has been used in agriculture since the late 1970s
and early 1980s (Jørgensen, 1992). Resistance conferred by mlo is remarkable as it
stops the pathogen at the penetration stage of infection with callose deposition, ac-
cumulation of the phenol conjugate p-coumaroyl-hydroxyagmatine as well as ROS
production in the infection sites, turning the host into almost a non-host (Skou,
1982, Skou et al., 1984, von Röpenack et al., 1998, Hückelhoven et al., 2000, Pif-
fanelli et al., 2002). The gene was not cloned until the 1990s and was then shown
to encode a membrane protein with seven transmembrane domains that has an ex-
tracellular N-terminal domain and an intracellular carboxyl tail (Büschges et al.,
1997, Devoto et al., 1999). Subsequent studies found that mlo-mediated resistance
is highly conserved and effective in many agriculturally and economically impor-
tant monocotyledonou and dicotyledonous plants including Arabidopsis (Kusch and
Panstruga, 2017).
24
Figure 1.5: Phylogenetic analysis of selected MLO protein. Phy-
logenetic relationship of selected MLO (Mildew Resistance Locus O) pro-
teins. Phylogenetic (neighbourjoining) tree of Arabidopsis (AtMLO1–
AtMLO15), barley (HvMLO), tomato (SlMLO1), Medicago truncatula
(MtMLO), Lotus japonicus (LjMLO1) and pea (PsMLO1) MLO pro-
teins. The clade harbouring presumptive (co)orthologues is highlighted
in grey. Proteins known to play an important role in conferring pow-
dery mildew susceptibility are highlighted in bold; the three AtMLO
co-orthologs of barley HvMLO are circled in blue. The numbers at the
edges designate the bootstrap support based on 1000 replicates. The
scale bar indicates the number of amino acid exchanges per site. Figure
adapted from (Humphry et al., 2011)
25
Though the availability of Arabidopsis mlo mutants have made it easier to
study mlo-mediated resistance, the mechanism is till enigmatic. Of the 15 MLO
genes in the Arabidopsis genome, AtMLO2, AtMLO6 and AtMLO12 appear to be
functional co-orthologs of barley Mlo, and they are grouped into the same phy-
logenetic clade (Fig. 1.5). Interestingly, they contribute unequally to powdery
mildew susceptibility with AtMLO2 being the major one and the other two the
minor ones. Specifically, while the Atmlo2 mutant exhibits clear resistance against
adapted powdery mildew fungi, the Atmol6, Atmol12 single, and Atmol6mlo12 dou-
ble mutants showed wild type-like susceptibility (Consonni et al., 2006). However,
the Atmlo2mlo6mo12 triple mutant displays complete resistance to powdery mildew
to a similar resistant degree seen in barley mlo (Consonni et al., 2006).
Extensive genetic analyses of the Atmlo mutations in various genetic mutant
backgrounds showed that although Atmlo2-based resistance can be partially sup-
pressed by several mutations (e.g. pen1, pen2), none of them affects powdery mildew
resistance of the Atmlo2/6/12 triple mutant (Consonni et al., 2006, Kuhn et al.,
2017). These results indicate that mlo-based resistance is independent of all known
defense mechanisms including the SA-, JA-, PAD4- and PEN1/2/3-dependent path-
ways (Kuhn et al., 2017). Despite being a research focus for a long time, the molec-
ular basis of mlo-based powdery mildew resistance remains a mystery. Elucidation
of mlo-mediated resistance will certainly broaden our knowledge of plant defense
mechanisms and how to better apply this powerful resistance gene in agriculture.
26
1.3 Phospholipase D and Phosphatidic Acid in Plant Immunity
While protein components are essential for plant immunity and have been
extensively studied, important roles for signaling lipids and their corresponding me-
tabolizing enzymes in plant immunity have also been observed but are relatively
understudied. Phospholipase D (PLD) is a family of enzymes that catalyzes the
hydrolysis of structural phospholipids, such as phosphatidylcholine (PC) and phos-
phatidylethanolamine (PE), to produce phophatidic acid (PA) and the respective
head groups (Wang, 2004). In plants, PLD and PA have been shown to be involved in
multiple cellular processes, such as cytoskeleton rearrangement, vesicular trafficking,
membrane remodeling and degradation to modulate plant growth and development,
and signaling events in response to both abiotic and biotic stress (Wang, 2004).
1.3.1 PLD
Compared to animals (which have two PLD genes) or yeast (which has a single
PLD gene), plants have a family of PLD genes with varied numbers (Bargmann and
Munnik, 2006). The Arabidopsis genome contains 12 identified PLD genes and the
encoded 12 isoforms can be classified into six types, PLDα (3), PLDβ (2), PLDγ
(3), PLDδ (1), PLD (1), and PLDζ (2), base on sequence similarities, domain
structures and biochemical properties (Fig. 1.6, Wang et al., 2006). All these PLDs
have an active catalytic site consisting of two conserved duplicate interacting HKD
(HxKxxxD/E) domains (Qin and Wang, 2002). Besides, while α, β, γ, δ, and
 isoforms contain a C2 domain (Ca2+-dependent phospholipid binding domain)
27
Figure 1.6: Arabidopsis PLD domain structures and biochemical
properties. PC, phosphatidylcholine; PE, phosphatidylethanolamine;
and PIP2, phosphatidylinositol 4,5-bisphosphate.. Figure adapted from
(Hong et al., 2016)
28
near the N-terminus, the two ζ isoforms, similar to animal and fungal PLDs, have
pleckstrin homology (PH) and phox homology (PX) domains instead (Eliáš et al.,
2002). C2 domain is unique to plant PLDs. It is required for enzyme activation
and determines µM to mM Ca2+ requirements for the activity of different PLD
isoforms (Qin and Wang, 2002, Zheng et al., 2000, Hong et al., 2016). The PH and
PX domains are typical phosphoinositides binding domains with distinct affinities
(Lemmon, 2003). In addition, there are other domains present in some of the PLDs
for binding of PI(4,5)P2 (phosphatidylinositol 4,5-bisphosphate), or actin, or oleate
(Zheng et al., 2002, Kusner et al., 2002, Wang and Wang, 2001).
These structural and biochemical properties make the PLDs diverse in sub-
strate preference, subcellular localization, and tissue distribution, accounting for
production of distinct pools of PA in a spatiotemporal manner, and thus their ver-
satile cellular and physiological roles (Hong et al., 2016).
1.3.2 Signal-induced PA Production
Being the simplest phospholipid class, PA is not only a central intermediate in
glycerolipid biosynthesis but also a signaling molecule involved in regulating cellular
processes such as lipid metabolism, signal transduction, cytoskeletal rearrangements,
and vesicular trafficking.
As a signaling molecule, the concentration of PA is normally very low in
plant tissues and can be induced rapidly by various stimuli. Signal-induced PA
is mainly produced via two distinct enzymatic pathways: (i) direct hydrolysis of
29
Figure 1.7: Formation and attenuation of phosphatidic acid
(PA). Phospholipases are highlighted in yellow, lipid kinases in blue
and lipid phosphatases in green. Abbreviations: DAG, diacylglycerol;
DGPP, DAG pyrophosphate; DPP, DGPP phosphatase; P, phosphate;
PAK, PA kinase; PAP, PA phosphatase; PC, phosphatidylcholine; PE,
phosphatidylethanolamine; PtdIns(4,5)P2, phosphatidylinositol-(4,5)-
bisphosphate. Figure from (Testerink and Munnik, 2005)
30
structural phospholipids (PC, PE) by a PLD; and (ii) a two-step enzymatic process
that involves generation of diacylglycerol (DAG) from inositol phospholipids such as
PI(4,5)P2 by PLC, followed by DGK-catalyzed phosphorylation of DAG to produce
PA. PA may be rapidly phosphorylated to diacylglycerol pyrophosphate (DGPP)
by PA kinase (PAK), or dephosphorylated to DAG by lipid phosphate phosphatase
(LPP) and PA hydrolase (PAH). DGPP can also be converted back to PA by LPP
(Fig. 1.7, Testerink and Munnik, 2011).
1.3.3 PLD and PLD-derived PA in Plant Immunity
The enzymatic activities or expression levels of different isoforms of PLD
can be induced by various elicitors including PAMPs/MAMPs, pathogens, and SA
(Young et al., 1996, de Torres Zabela et al., 2002, Krinke et al., 2009), indicating
differential participation of the PLD isoforms in plant immunity. Since the enzy-
matic activity of PLD produces PA, elevated levels of PA derived from PLD have
also been observed in response to pathogen challenges or SA treatment (Tetiana
et al., 2013, Rodas-Junco et al., 2015, Saha et al., 2016). Subsequent genetic or bio-
chemical studies provided more definitive evidence to support differential or even
opposing roles of PLD and PLD-derived PA in plant defense response under different
pathocontexts.
The Arabidopsis PLDδ and PLDδ-derived PA have been shown to positively
contribute to penetration resistance against two non-adapted powdery mildew fungi,
barley mildew Blumeriagraminis f. sp. hordei (Bgh) and pea mildew Erysiphe pisi
31
(Pinosa et al., 2013). The negative role of PLDβ and PLD-derived PA in plant im-
munity is revealed in studies showing that genetic depletion of specific PLD isoforms
in tomato, rice and Arabidopsis resulted in elevated defense responses (Bargmann
et al., 2006, Yamaguchi et al., 2009, Zhao et al., 2013). Moreover, a recent study
shows that the grapevine VvPLD genes can be differentially expressed in Botrytis
cinerea infected-grape cells with isoforms VvPLDβ1, VvPLDβ2, VvPLDδ2, VvPLDρ
and VvPLDζ being upregulated, and VvPLDα and VvPLDδ being down-regulated
(Yu et al., 2018).
The mechanism of how PLD and PLD-derived PA contribute to plant immu-
nity is still unknown. It is conceivable that PLD regulates plant innate immunity
through PA-binding proteins, or changing membrane curvature and integrity, or
producing abundant PA for generation of other lipid messengers, such as lysoPA,
free fatty acids, and oxylipins (Testerink and Munnik, 2011, Wang, 2004).
1.4 Conception and Significance of this Study
Fast-evolving plant pathogens pose a constant threat to global food security.
Plant pathogens like rust, oomycetes and powdery mildew deploy a similar infection
strategy, which is to form a feeding structure–haustorium–in close contact with the
host cytoplasm, to suppress host immunity and take up nutrients. The resistance
protein RPW8.2 from Arabidopsis is specifically targeted to the plant-haustorial
interface (i.e. the extrahaustorial membrane; EHM) to activate defenses to constrain
the haustorium. However, how RPW8.2 achieves such a high specificity in targeting
32
to the EHM is still not well understood. To investigate if phospholipids regulate
specific targeting of RPW8.2 to the EHM to activate on-site defenses, I screened a
panel of Arabidopsis T-DNA mutants defective in phospholipases or kinases involved
in lipid metabolism and/or signaling. While neither gene was found to be required
for RPW8-mediated resistance, the study did reveal that PLDα1 and PLDδ play
opposing roles in basal, post-penetration resistance against powdery mildew through
a novel, yet-to-be characterized mechanism that is independent of EDS1/PAD4, SA,
and JA signaling (Chapter 2). Inspired by this finding, I designed and carried out a
sensitive genetic screen in the background of an immunocompromised eds1-2pad4-
1sid2-2 (eps) mutant, hoping to identify novel defense pathways where the two
PLDs or other novel immune components may function (Chapter 3). In the EMS-
mutagenized eps mutant background, I isolated 5 susceptible to non-adapted powdery
mildew (snap) and 18 compromised immunity yet poor infection (cipi) mutants upon
powdery mildew infection. Whole-genome sequencing-assisted gene mapping has
identified one nonsynonymous mutation in the MAP KINASE PHOSPHATASE1
(MKP1) gene to be the causal mutation in cipi1, and five independent mutations
in the MILDEW RESISTANCE LOCUS O2 (MLO2) gene the causal mutations in
five cipis. While MKP1 is a negative regulator of plant immunity, MLO2 is more
likely to be a host susceptibility factor used by powdery mildew fungi for host cell
entry. Despite the difference in mechanism, the resistance mediated by mkp1 and
mlo2 is independent of EDS1/PAD4 and SA signaling. Gene identification with the
remaining snap and cipi mutants are in progress. Together, results from my work
not only will contribute to a better understanding of the multi-layered plant immune
33
system and host adaptation of powdery mildew, but also should provide invaluable
starting materials for future investigation and exploitation of novel mechanisms of
planting immunity to help plants fight against fungal pathogens.
34
Chapter 2: Arabidopsis Phospholipase Dα1 and Phsopholipase Dδ
oppositely modulate basal resistance against powdery
mildew independent of EDS1/PAD4 and salicylic acid
This chapter is published as:
Zhang, Q., Berkey, R., Blakeslee, J.J., Lin, J., Ma, X., King, H., Liddle, A., Guo,
L., Munnik, T., Wang, X. and Xiao, S., 2018. Arabidopsis phospholipase Dα1 and
Dδ oppositely modulate EDS1-and SA-independent basal resistance against adapted
powdery mildew. Journal of Experimental Botany, 69(15): pp.3675-3688.
35
2.1 Abstract
Plants use a tightly regulated immune system to fight off various pathogens.
Phospholipase D (PLD) and its product, phosphatidic acid, have been shown to
influence plant immunity; however, the underlying mechanisms remain unclear.
Here, we show that the Arabidopsis mutants pldα1 and pldδ, respectively, ex-
hibited enhanced resistance and enhanced susceptibility to both well-adapted and
poorly adapted powdery mildew pathogens, and a virulent oomycete pathogen, in-
dicating that PLDα1 negatively while PLDδ positively modulates post-penetration
resistance. The pldα1δ double mutant showed a similar infection phenotype to
pldα1, genetically placing PLDα1 downstream of PLDδ. Detailed genetic analy-
ses of pldδ with mutations in genes for salicylic acid (SA) synthesis (SID2) and/or
signaling (EDS1 and PAD4), measurement of SA and jasmonic acid (JA) levels,
and expression of their respective reporter genes indicate that PLDδ contributes to
basal resistance independent of EDS1/PAD4, SA, and JA signaling. Interestingly,
while PLDα1-enhanced green fluorescent protein (eGFP) was mainly found in the
tonoplast before and after haustorium invasion, PLDδ-eGFPs focal accumulation
to the plasma membrane around the fungal penetration site appeared to be sup-
pressed by adapted powdery mildew. Together, our results demonstrate that PLDα1
and PLDδ oppositely modulate basal, post-penetration resistance against powdery
mildew through a non-canonical mechanism that is independent of EDS1/PAD4,
SA, and JA.
36
2.2 Introduction
Many fungal and oomycete pathogens penetrate the plant cell wall and extract
nutrients from host cells by a similar feeding structure called the haustorium. Plant
defense against these haustorium-forming pathogens occurs at both penetration and
post-penetration stages. Penetration resistance is usually sufficient to prevent non-
adapted pathogens from entering the host cell by forming a papilla, which is cell
wall thickening with deposition of callose (1,3-β-glucan) and other defense chem-
icals at the penetration site. This process is likely activated upon recognition of
conserved pathogen-associated molecular patterns (PAMPs) by cell-surface pattern
recognition receptors (PRRs), and thus may be part of PRR-triggered immunity,
or PAMP-/pattern-triggered immunity (PTI) (Jones and Dangl, 2006, Hückelhoven
and Panstruga, 2011). Pathogens that can breach penetration resistance face post-
penetration resistance. Stage I post-penetration resistance stops the pathogen from
finishing its life cycle by terminating early stage haustorial development (Wen et al.,
2011), which may continue to engage PTI. Adapted-pathogens secret effector pro-
teins to suppress this layer of defense and establish successful colonization. Stage II
post-penetration resistance is activated through the action of plant resistance (R)
proteins by detecting the presence or activity of cognate effector proteins termed
avirulence factors (Avrs), which in most cases is equivalent to effector-triggered im-
munity (ETI). ETI often exhibits race specificity and features with rapid cell death
at the infection site, namely the hypersensitive response (HR) (Jones and Dangl,
2006). Most characterized R proteins are nucleotide-binding, leucine-rich repeat
37
(NLR) intracellular immune receptors. Based on the N-terminal domains, NB-
LRRs are divided into two major classes, Toll-interleukin 1 receptor (TIR)-NLRs
and coiled-coil (CC)-NLRs. While characterized TIR-NLRs require the nucleocyto-
plasmic lipase-like protein ENHANCED DISEASE SUSCEPTIBILITY 1 (EDS1) for
signal transduction, most CC-NLRs engage the plasma membrane (PM)-anchored
integrin-like protein NON-RACE-SPECIFIC DISEASE RESISTANCE 1 (NDR1)
for signaling (Cui et al., 2015).
Detection of pathogens triggers a conserved signaling network orchestrated
by salicylic acid (SA), jasmonic acid (JA), and ethylene (ET), resulting in the
activation of defense responses including pathogenesis-related (PR) gene expres-
sion, reactive oxygen species (ROS) production, and callose deposition (Bari and
Jones, 2009, Pieterse et al., 2012). SA plays a critical role in activating local as
well as systemic acquired resistance (SAR) to fight against biotrophic and hemi-
biotrophic pathogens. Depending on the context of specific plant–pathogen interac-
tions, the SA pathway could act antagonistically or synergistically with the JA/ET
pathways, which are mainly effective against necrotrophic pathogens (Glazebrook,
2005, Robert-Seilaniantz et al., 2011). EDS1 and its interacting homologous part-
ner PHYTOALEXIN-DEFICIENT 4 (PAD4) are both required for adequate SA
synthesis and signaling, and play a role in the antagonism between SA and JA/ET
pathways (Zhou et al., 1998, Falk et al., 1999, Feys et al., 2001, Wiermer et al., 2005).
Furthermore, EDS1 and PAD4 have also been shown to regulate SA-independent
defense responses (Feys et al., 2005, Venugopal et al., 2009, Zhu et al., 2011, Wagner
et al., 2013, Cui et al., 2017).
38
Two non-NB-LRR Arabidopsis R proteins, RPW8.1 and RPW8.2, confer
broad-spectrum resistance to powdery mildew fungi (Xiao et al., 2001), which re-
quires EDS1, PAD4, and SA signaling (Xiao et al., 2003, 2005). RPW8.2 is specif-
ically targeted to the host-derived haustorium–host cytoplasm interface, the extra-
haustorial membrane (EHM), to activate on-site defenses including callose-enriched
haustorial encasement and interface-focused H2O2 production to constrain the haus-
torium (Wang et al., 2009, Berkey et al., 2016). Previous studies suggest that a
specific protein trafficking pathway is engaged for targeting RPW8.2 to the EHM
(Wang et al., 2013, Zhang et al., 2015). However, how RPW8.2 achieves haustorium-
targeted defense remains to be determined. A tempting speculation is that RPW8.2
may interact with a signaling lipid(s) to realize its specific targeting. In an effort to
test this speculation, we instead found that two phospholipase D (PLD) enzymes
play opposing roles in plant defense against powdery mildew fungi, but neither of
them seems to be required for RPW8-mediated resistance.
Both positive and negative roles of PLD in plant defense have been described
depending on the PLD isoforms involved and/or pathosystems examined (Zhao,
2015, Zhang and Xiao, 2015, Hong et al., 2016). In one study, Zhao et al. demon-
strated a negative role for PLDβ1 in the SA signaling pathway (Zhao et al., 2013),
while Pinosa et al. reported a positive role for PLDδ in penetration resistance
in Arabidopsis (Pinosa et al., 2013). This is not surprising since Arabidopsis has
12 PLD isoforms (Bargmann and Munnik, 2006). These PLD family members in-
evitably exhibit functional redundancy in plant defense. For example, repression
of PLD-produced PA by n-butanol in Arabidopsis strongly inhibited ETI, yet not
39
a single PLD gene was found to be responsible for this (Johansson et al., 2014).
Together, these studies suggest that PLDs play important roles in plant defenses
with functional redundancy among family members. However, whether and how
PLDs (or PLD-derived PA)-mediated signaling intersects with the well-defined SA
and/or JA/ET signaling pathways is poorly understood (Zhao, 2015, Zhang and
Xiao, 2015, Hong et al., 2016).
Here, we screened a panel of Arabidopsis mutants with T-DNA insertions in
PLD, pPLA (patatin-related phospholipase), PLC, DGK, and PIP5K (phosphatidyli-
nositol 4-phosphate 5-kinase) genes for altered infection phenotypes to adapted pow-
dery mildew fungi. We found that while PLDδ knockout plants showed enhanced
susceptibility, PLDα1 knockout plants displayed enhanced resistance, suggesting
that PLDα1 and PLDδ play opposing roles in post-penetration resistance against
powdery mildew. We thus conducted a detailed analysis to determine the genetic
relationships between these two PLD genes, their possible involvement in PRW8.2s
localization and function, and the defense pathways they might modulate.
2.3 Results
2.3.1 PLDα1 and PLDδ play opposing roles in post-penetration re-
sistance
To investigate whether PLDs or PLD-derived PA play a role in basal resistance
to powdery mildew, we had tested a panel of T-DNA insertion lines (Table 2.1)
including six PLD knockout mutants (pldα1, pldδ, pldβ, pldα1δ, pldα1δα3, and
40
Table 2.1: Arabidopsis T-DNA insertion mutants screened in Chapter 2
Mutant name Gene Locus T-DNA line
pldα1(Zhang et al., 2004) At3g15730 SALK 053785
pldβ1(Zhao et al., 2013) At2g42010 SALK 079133
pldδ(Pinosa et al., 2013) At4g35790 SALK 023247
pldα1δ At3g15730/At4g35790 see above
pPLAIIIδ-knockout (ko) At3g63200 SALK 029470
pPLAIIIγ-ko At4g29800 SALK 088404
pPLAIIIβ-ko At3g54950 SALK 057212
pPLAIIIα-ko At2g39220 SALK 040363
pPLAIIδ-ko At4g37060 SALK 090933
pPLAIIβ-ko At4g37050 SALK 142351
pPLAIIα At2g26560 SALK 059119
pPLAI-ko At1g61850 SALK 087152
pip5k7-1 At1g10900 SALK 151429c
plc3/5 At4g38530/At5g58690 SALK 037453/SALK 144469
plc5/7 At5g58690/At3g55940 SALK 144469/SALK 030333
plc3/6/9 At4g38530/At2g40116/At3g47220 SALK 037453/SALK 090508/SALK 025949
dgk1/2 At5g07920/At5g63770 SALK 053412/SAIL 718 G03
dgk3-1 At2g18730 SALK 028600
dgk4-2 At5g57690 SALK 069158
dgk5-1 At2g20900 SAIL 1212 E10
dgk6-1 At4g28130 SALK 016285
dgk7-1 At4g30340 SALK 51 E04
pldα1δα3 At3g15730/At4g35790/At5g25370 SALK 067533 / SALK 023247 / SALK 122059
pldα1δ At3g15730/At4g35790/At1g55180 SALK 067533 / SALK 023247 /KONCZ68434
pldα1δ) with Golovinomyces cichoracearum (Gc) UCSC1, a well-adapted powdery
mildew isolate. Interestingly, we found that the pldδ mutant with compromised pen-
etration resistance (Pinosa et al., 2013) showed clear enhanced disease susceptibility
(‘eds’) while pldα1 defective in abscisic acid (ABA) signaling (Zhang et al., 2004)
and pldα1-containing mutants (pldα1δ, pldα1δα3, and pldα1δ) exhibited enhanced
disease resistance (‘edr’) to Gc UCSC1 (Fig. 2.1; Fig. 2.2A, B). The ‘edr’ pheno-
type of pldα1δ led us to speculate that PLDα1 may act genetically downstream of
PLDδ to modulate plant immunity negatively. Visual scoring of fungal mass on
the leaf surface at 12 day post-inoculation (dpi) and quantification of fungal spore
production showed that the level of the ‘eds’ of pldδ was almost comparable with
that of Col-nahG, a Col-0 transgenic line defective in SA signaling due to conversion
of SA to catechol by the bacterial SA hydrolase encoded by nahG as a transgene
41
(Fig. 2.2A, B). All other mutants tested exhibited levels of disease susceptibility
similar to those of the Col-0 wild type (Fig. 2.1; Fig. 2.2A, B). Consistent with the
results at 12 dpi, pldδ supported significantly more conidiophores per colony while
pldα1 and pldα1δ had fewer conidiophores per colony than Col-0 during early in-
fection stage at 4 dpi when the fungus begins asexual reproduction (Fig. 2.2C, D).
Interestingly, Col-nahG supported a similar amount of conidiophores to Col-0 at
4 dpi (Fig. 2.2D), suggesting that PLDδ-mediated defense against Gc UCSC1 prob-
ably occurs earlier than SA-mediated defense. This raises an intriguing question as
to whether PLDδ (and PLDα1) functions in a signaling pathway distinct from the
SA-dependent pathway.
To test whether the ‘edr’ phenotype of pldα1 and the ‘eds’ phenotype of
pldδ are indeed due to the loss of PLDα1 and PLDδ, respectively, multiple pldα1 and
pldδ lines expressing the respective wild-type genes were generated and tested with
Gc UCSC1. These lines displayed similar disease phenotypes to Col-0 (Fig. 2.3), in-
dicating genetic complementation of these two genetic mutations by their respective
wild-type genes. Thus, our genetic data established a positive role for PLDδ and
a negative role for PLDα1 in basal, stage II post-penetration resistance against
well-adapted powdery mildew in Arabidopsis.
To test if the PLD genes are also involved in stage I post-penetration resistance,
we inoculated the pld mutants with Gc UMSG1. Gc UMSG1 is a powdery mildew
fungus infectious on sow thistle. It has largely overcome penetration resistance of
25 Arabidopsis accessions examined and is capable of forming initial haustoria but
arrested before sporulation by stage I post-penetration resistance in Arabidopsis
42
A Col-0 pPLAIIIδ-ko pPLAIIIα-ko pPLAIIα-ko pPLAI-ko
DR=3 DR=3~4 DR=3~3.5 DR=3 DR=3
Col-nahG pldα1 pldα1δ pldβ1 pldδ
DR=4 DR=1 DR=1 DR=2~3 DR=4
B Col-0 pldα1 pldδ pldα1δε pldα1δα3 eds1-2
DR=3 DR=1~2 DR=4 DR=1~2 DR=1~2 DR=4~5
dgk1dgk2 dgk3-1 dgk4-2 dgk5-1 dgk6-1 dgk7-1
DR=3 DR=3 DR=3 DR=3 DR=3 DR=3
pld3plc5 pld5plc7 pld3plc6plc9 pip5k7-1
DR=3 DR=3 DR=3 DR=3
Figure 2.1: Disease reaction phenotypes of pPLA, PLD, PLC,
DGK, and PIP5K T-DNA insertion mutants infected with Gc
UCSC1. Representative plants of indicated genotypes infected with Gc
UCSC1 at 8 dpi (A) or 11 dpi (B). Note that pldβ1 showed slight ‘edr’ to
Gc UCSC1, and pldα1δα3 and pldα1δ triple mutants displayed similar
level of ‘edr’ to pldα1. The disease reaction (DR) scores (0, resistant; 1
to 2, intermediate; 2 to 3 or 3 or 3 to 4, susceptible; 4 to 5, ‘eds’ ; Xiao
et al., 2005) are shown above the photos.
43
A Col-0 Col-nahG pPLAIIIδ-ko B
c c
200
150 a a
100
pldα1 pldδ pldα1δ
50 b b
0
0 o 1 δ δ
Co
l- G
-na
h δ-k ldα pld α1
l IIIA po L p
ld
C
C D p
P
Col-0 Col-nahG pldα1 c
50 a
a
40
30 b b
20
pldδ pldα1δ
10
0
ol-
0 hG α1 ldδ 1δ
C l-n
a pld p ldαp
Co
Figure 2.2: Arabidopsis PLDα1 negatively while PLDδ positively
modulates post-penetration resistance against well-adapted
powdery mildew Gc UCSC1. (A) Representative images of Ara-
bidopsis leaves of indicated genotypes infected with Gc UCSC1 at 12 dpi.
Note, pldα1 and pldα1δ were less susceptible while pldδ was more sus-
ceptible than Col-0. (B) Quantification of spore production in indicated
genotypes at 10 dpi normalized to leaf fresh weight (FW). Data repre-
sent mean ± SEM of three samples (n = 3, 4 leaves each) from one
experiment, which was repeated three times with similar results. (C)
Representative microscopic images of single colonies of Gc UCSC1 on
leaves of indicated genotypes at 4 dpi. Fungal structures were stained
by trypan blue. Bars, 200µm. (D) Total number of conidiophores per
colony on leaves of indicated genotypes at 4 dpi. The boxplot shows
combined data from three independent experiments (at least 20 colonies
were counted for each genotype per experiment). The bold line within
box represents the median. The bottom and top edge of box represent
the first and third quartile, respectively. Ends of whiskers represent the
minimum and maximum of data points. Grey dots represent outliers.
Different lowercase letters indicate statistically different groups (P <
0.01) as determined by multiple comparisons using one-way ANOVA,
followed by Tukey-HSD.
44
Conidiophores/colony Spores/mg FW (x100)
(Wen et al., 2011). We assessed the growth of Gc UMSG1 on the pld mutants by
measuring the total hyphal length of each microcolony at 5 dpi. Not surprisingly,
pldδ supported significantly more hyphal growth than Col-0 (Fig. 2.4B), which is
similar to eds1-2 (in Col-0; Bartsch et al., 2006), known to support better growth
of Gc UMSG1 (Wen et al., 2011). However, while limited sporulation ofGc UMSG1
can occasionally be seen on eds1-2, indicating breakdown of nonhost resistance, it
has never been observed on pldδ, suggesting that PLDδ acts differently from EDS1
and is not as critical as EDS1 in stage I post-penetration resistance defined by this
pathosystem. However, hyphal growth in pldα1 and pldα1δ showed no significant
difference from that in Col-0 (Fig. 2.4).
To investigate how defense responses at the subcellular level may be affected in
the pld single and double mutants, we examined powdery mildew-induced H2O2 pro-
duction and callose deposition in infected epidermal cells. Because Gc UCSC1 can
largely suppress the production of H2O2 in Col-0 (Xiao et al., 2005), the non-adapted
isolate Gc UMSG1 was used to challenge the plants and in situ H2O2 production
was visualized by 3,3’-diaminobenzidine (DAB) staining (Thordal-Christensen et al.,
1997). We divided the haustorium–epidermal cell interaction in terms of H2O2 pro-
duction into three types: (i) H2O2 is undetectable; (ii) H2O2 accumulates in the
haustorial complex; and (iii) H2O2 is found in both the haustorial complex and
the whole cell (Fig. 2.5A). Of >750 interaction sites evaluated in Col-0, 39.5, 25.7,
and 34.7% were (i), (ii), and (iii), respectively, and the pld mutants showed a simi-
lar frequency distribution for the three interaction types (Fig. 2.5B). This suggests
that H2O2 production induced by haustorium invasion is not affected due to loss of
45
A Col-0 pldα1 p35S:PLDa1/pldα1 B Col-0 pldδ p35:PLDδ/pldδ
C 120 D b
a 200
a
90
150 a
a
60 b
100
30
50
0
l-0 1o 1
/ 0
C pld
α α /
PL
D 0 δ
: Co
l- pld LD
δ
5S S:
P
p3 α1 35
pld p δpld
Figure 2.3: Genetic complementation of the pldα1 and pldδ mu-
tants by their respective wild-type genes. (A and B) Represen-
tative images of leaves of indicated genotypes infected with Gc UCSC1
at 13 and 10 dpi, respectively. (C and D) Quantification of spore pro-
duction in leaves of indicated genotypes from (A and B) normalized to
leaf fresh weight (FW). Data represent mean ± SEM of four samples in
(A) and three samples in (B) (4 leaves each sample), from one experi-
ment, which was repeated twice with similar results. Different lowercase
letters indicate statistically different groups as determined by multiple
comparisons using one-way ANOVA, followed by Tukey-HSD (P < 0.01).
46
Spores /mg FW (x100)
Spores /mg FW (x100)
A Col-0 eds1-2
pldα1 pldδ
pldα1δ B b
5 b
4
3 a
2 a a
1
0
0 1 δ δ 2
Co
l- -
pld
α pld α1d s1pl ed
Figure 2.4: PLDδ in Arabidopsis contributes to post-penetration
resistance against a poorly adapted powdery mildew Gc
UMSG1. (A) Representative microscopic images of typical Gc UMSG1
fungal microcolonies grown on leaves of indicated genotypes at 5 dpi.
Bars, 100µm. (B) Total hyphal length per microcolony of indicated
genotypes at 5 dpi. The boxplot shows combined data from three in-
dependent experiments (n > 60). Different lowercase letters indicate
statistically different groups as determined by multiple comparisons us-
ing one-way ANOVA, followed by Tukey-HSD (P < 0.01).
47
Total hyphal length per 
microcolony (mm)
PLDα1 or PLDδ or both. Next, we examined callose deposition at the fungal pen-
etration sites (i.e. papillae) or around the haustorium (i.e. haustorial encasement)
by aniline blue staining after Gc UCSC1 inoculation. Again, callose deposition was
grossly unaffected in the pld mutants compared with that in Col-0 based on visual
scoring (Fig. 2.5C). These suggest that the ‘eds’ phenotype of pldδ and the ‘edr’
phenotype of pldα1 are not apparently associated with these two typical subcellular
defense responses.
2.3.2 Loss of PLDα1 or PLDδ affects basal resistance against an
oomycete but not ETI
Hyaloperonospora arabidopsidis (Hpa) is a fungus-like oomycete pathogen of
Arabidopsis. To test if post-penetration resistance to Hpa is also altered in the pld
mutants, we inoculated 10-day-old seedlings of Col-0, pldα1, pldδ, pldα1δ, and two
known ‘eds’ mutant lines, eds1-2 and pad4-1sid2-2, with Hpa isolate Noco2 (viru-
lent on Col-0). While pldα1 and pldα1δ were significantly less susceptible, pldδ was
significantly more susceptible (albeit not as susceptible as eds1-2 and pad4-1sid2-
2) to this pathogen than Col-0 (P < 0.01) (Fig. 2.6B, upper panel). These fur-
ther support the distinct roles of PLDα1 and PLDδ in post-penetration resistance
against haustorium-forming pathogens, despite that the phenotypic alterations in
these mutants were relatively less pronounced compared to those derived from pow-
dery mildew infection.
To test if loss of PLDα1 or PLDδ impacts ETI, we tested the mutants with an
48
A B
(i) No detectable H O (ii) H2O2 in (iii) H2O2 in haustorial 
1.00
2 2 haustorial complex complex and whole cell
0.75 (i)
0.50
H (ii)H
H
0.25 (iii)
0.00
l-0 α1 Dδ
δ
o D L Dα
1
C PL P PL
C Col-0 eds1-2 plda1 pldd plda1δ
Figure 2.5: Loss of PLDα1 or PLDδ or both do not impact H2O2
production and callose deposition in the haustorium-invaded
epidermal cells. (A) Representative images of three types of H2O2 pro-
duction in haustorium-epidermal cell interaction site: (a) H2O2 is not de-
tectable; (b) H2O2 accumulates in the haustorial complex; and (c) H2O2
is found in both haustorial complex and the whole cell. Leaf samples
were inoculated with Gc UMSG1 and stained by 3,3’-diaminobenzidine
(DAB) at 3 dpi. (B) Frequencies of the three types of H2O2 produc-
tion shown in (A) in each of the indicated genotypes. Total of between
750 to 1300 interaction sites combined from three independent exper-
iments were evaluated for each genotype. (C) Representative images
showing callose formation in the indicated genotypes. Leaf samples were
inoculated with Gc UCSC1 and stained blue by aniline blue at 3 dpi. Ar-
rowheads indicate three types of callose deposition: encasement of the
haustorium, half encasement of the haustorium, and callose is restricted
to the penetration site. Bars, 50µM. BF, bright field.
49
Merge Aniline Blue BF
Proportion of host cell−haustorium 
interaction site/H2O2 Class
A B
Disease Class Hpa Noco2
90
60 **
** **
30
0
Hpa Emwa1 (RPP4)
90
60
30
0
0
ol- ldα
1
pld
δ 1δ
C dα s1
-2 2-2 ldδ
p pl d i
d 2pe
4-1
s 1-
ad ed
s
p
Figure 2.6: Loss of PLDα1 and/or PLDδ affects basal resistance
against oomycetes but not ETI mediated by RPP4. (A) Repre-
sentative cotyledons showing disease phenotypes of the indicated disease
classes at 7 dpi. Ten-day-old seedlings were inoculated with virulent
Hyaloperonospora arabidopsidis (Hpa) isolate Noco2 or avirulent isolate
Emwa1. Sporangiophores (Sp) per cotyledon were assessed at 7 dpi, and
categorized into 5 classes as indicated by the corresponding figure keys.
(B) Quantification of the number of cotyledons (n > 100 for each of
the indicated genotypes) per class of the indicated genotypes infected
with Hpa isolate Noco2 (upper panel) or avirulent isolate Emwa1 (lower
panel) based on categorization of leaf infection defined in (A). χ2-test
was used to test statistical significance for disease degree between Col-0
and the indicated mutant lines at 7 dpi (**P < 0.01).
50
>15 Sp 11-15 Sp 6-10 Sp 1-5 Sp 0 Sp
Number of Cotyledons/Class
A B
Pma ES4326 0 dpi 3 dpi Pma ΔhrcC 0 dpi 3 dpi
6 6
4 4
2 2
0 0
l-0 -2 α1 dδ 1δ l-0 2 1 δo 1 ld pl α o 1
-
s ldα pld α1
δ
C ed
s p pld C ed p pld
C D
Pma avrRpm1 0 dpi 3 dpi Pma avrRps4 0 dpi 3 dpi
6 **6
4 4
2 2
0 0
l-0o s1
-2 1 δ
ldα pld α1
δ ol-
0
C s1
-2 α1ld pld
δ 1δ
C ed p pld ed p pld
α
Figure 2.7: Loss of PLDα1 and/or PLDδ does not affect ETI
against bacterial pathogens. Fifteen-day-old seedlings were dip-
inoculated with Pma ES4326 (A), Pma ∆hrcC (B), Pma avrRpm1 (C),
and Pma avrRps4 (D). Seedling samples were collected at 0 dpi (1 hour
post inoculation) and 3 dpi, and bacterial growth was quantified. Data
represent mean ± SEM (n = 4). One-way ANOVA followed by Tukey-
HSD was conducted to evaluate whether there was any significant dif-
ference in bacterial growth between Col-0 and the indicated genotypes
(**P < 0.01, P > 0.05 for the remaining). FW, fresh weight.
51
Log cfu/mg FW Log cfu/mg FW
Log cfu/mg FW Log cfu/mg FW
avirulent oomycete strain Hpa Emwa1 (recognized by RPP4, a TIR-NB-LRR; von
Malek et al., 2002), and Pseudomonas syringae pv. maculicola (Pma) ES4326 strains
expressing either AvrRpm1 (recognized by RPM1, a CC-NB-LRR; Grant et al.,
1995) or AvrRps4 (recognized by RPS4/RRS1, a pair of TIR-NB-LRR immune
receptors; Narusaka et al., 2009), since no NB-LRR-mediated resistance against
powdery mildew has been defined in Arabidopsis. While eds1-2 and pad4-1sid2-
2 were compromised in resistance against Hpa Emwa1, the pld mutants displayed
similar levels of resistance to that seen in Col-0 (Fig. 2.6), indicating that loss of
PLDα1 and/or PLDδ does not seem to affect RPP4-dependent ETI. Similarly, no
significant difference was detected between pldα1, pldδ, pldα1δ, and Col-0 (Fig. 2.7C,
D) in defense against Pma, further supporting that PLDα1 or PLDδ individually or
together do not play a significant role in ETI. In addition, the pld mutants remained
resistant like Col-0 to Pma ∆hrcC, which is unable to inject type III effectors to
suppress PTI, implying that the PTI against bacterial pathogens is not affected by
the loss of PLDα1 and/or PLDδ (Fig. 2.7B). This could be due to functional redun-
dancy among the PLD enzymes in defense against bacterial pathogens as suggested
in an earlier study since there are 12 PLD isoforms in Arabidopsis (Johansson et al.,
2014).
2.3.3 PLDδ is dispensable for RPW8-mediated resistance
RPW8.1 and RPW8.2 (referred to as RPW8 in later text unless otherwise in-
dicated) confer post-penetration, haustorium-targeted resistance to powdery mildew
52
A RPW8-RFP RPW8-RFP B S5 S5/pldδ
/Col-0 /pldδ
20 µm 100 µm H O 20 µm 20 µm2 2 H2O2
C Col-0 Col-nahG pldδ S5 S5/pldδ
DR = 3.0 DR = 4.0 DR = 4.0 DR = 1.0 DR = 1.0
D Col-0 pldα1 S5 S5/pldα1 
DR = 3.0 DR = 2.0 DR = 1.0 DR = 1.0
Figure 2.8: PLDα1 and PLDδ are dispensable for RPW8-
mediated resistance to Gc UCSC1. (A) Subcellular localization of
RPW8-RFP in Col-0 and pldδ in Gc UCSC1 haustorium-invaded cells.
The confocal images shown are Z-stack projections of 15 optical sections
taken at 2 dpi. Note that RPW8-RFP localization in pldδ mutant was
not affected. (B) RPW8-triggered H2O2 accumulation in haustorium-
invaded epidermal cells of S5 (Col-0 expressing RPW8) and S5/pldδ was
visualized by 3,3’-diaminobenzidine (DAB) staining at 3 dpi with Gc
UCSC1. Haustoria are indicated by arrows. Bars, 20µm. (C and D)
Representative plants of indicated genotypes infected with Gc UCSC1 at
13 dpi (C) and 14 dpi (D). The disease reaction (DR) scores (0, resistant;
1 to 2, intermediate; 2 to 3 or 3 or 3 to 4, susceptible; 4 to 5, eds ; Xiao
et al., 2005) are shown above the images.
53
(Xiao et al., 2001, Wang et al., 2009). To examine whether PLDα1 and/or PLDδ con-
tribute to RPW8-mediated resistance, we first stably expressed the RPW8.2-RFP
(red fluorescent protein) transgene from the native RPW8.2 promoter in pldα1 and
pldδ. Confocal microscopy showed that the localization of RPW8.2-RFP to the
EHM was unchanged in pldα1 or pldδ (as represented by RPW8.2-RFPs localiza-
tion in pldδ; Fig. 2.8A), indicating that neither PLDα1 nor PLDδ is required for
precise EHM-targeting of RPW8.2 (Wang et al., 2009). Next, we individually in-
troduced these two mutations into S5 (a Col-gl line expressing RPW8, Xiao et al.,
2005). Both S5/pldα1 and S5/pldδ displayed the same levels of resistance to Gc
UCSC1 (Fig. 2.8C, D) and H2O2 production as S5 in haustorium-invaded cells (as
represented by H2O2 production in S5/pldδ; Fig. 2.8B). Given that RPW8s defense
function requires SA signaling (Xiao et al., 2005), these results support that the
PLDα1/PLDδ pair most likely function via an SA-independent signaling pathway.
2.3.4 PLDα1 and PLDδ have distinct subcellular localizations
Since there is active membrane trafficking and biogenesis (of the EHM) in
haustorium-invaded cells (Berkey et al., 2016), we wondered whether the contrasting
defense responses of pldα1 and pldδ to adapted powdery mildew are due to possible
differential subcellular enzymatic activities of PLDα1 and PLDδ in haustorium-
invaded cells. To test this, we fused eGFP to the C-termini of the genomic DNA of
the two PLD genes and expressed the fusion constructs from the 35S plus the native
promoter (for PLDα1-eGFP) or the 35S promoter only (for PLDδ-eGFP) in pldα1 or
54
A Col-0 pldδ p35S:PLDδ-eGFP/pldδ
p35S-pPLDα1:PLDα1-eGFP
B Col-0 pldα1 /pldα1
Figure 2.9: The PLDδ-eGFP and PLDα1-eGFP fusion proteins
are functional. Representative plants of the indicated genotypes in-
fected with Gc UCSC1 at 12 dpi. While the transgene 35S:PLDδ-eGFP
could fully rescue the ‘eds’ phenotype of pldδ (A), p35S-pPLDα1:PLDα1-
eGFP could partially restore the ‘edr’ phenotype of pldα1 (B). Leaves
marked with red arrowheads display typical disease phenotypes of indi-
cated genotypes.
55
A PLDδ-eGFP C PLDδc-eGFP E PLDδc-eGFP G PLDα1-eGFP
H PM T H
P P
H PM
H
Gc UCSC1
Single plane Plasmolysis Single plane
B PLDδ-eGFP D PLDδc-eGFP F PLDα1-eGFP H PLDα1-eGFP
H T
P H
PM H PM H
Plasmolysis
Figure 2.10: Differential subcellular localization of PLDα1 and
PLDδ in powdery mildew-infected epidermal cells. Stable trans-
genic lines were inoculated with Gc UCSC1 or Gc UMSG1. At 2 dpi,
sections of infected leaves were stained with propidium iodide (PI, 0.5%
aqueous solution) for 40–60 min for staining haustoria (H, red) and
mycelia (red) before confocal imaging. All representative images shown
are merged (GFP, PI, and bright field) Z-stack projections of 15 to
20 optical sections unless otherwise indicated. (A, B) Localization of
PLDδ-eGFP (from the PLDδ genomic sequence translationally fused
with eGFP) in a Gc UMSG1-invaded cell (A) or a Gc UCSC1-invaded
cell (B). Arrows, concentric ring and dots. (C–E) Localization of PLDδc-
eGFP (derived from the PLDδ full-length coding sequence translation-
ally fused with eGFP; Pinosa et al., 2013) in a Gc UMSG1-invaded
cell before (C; arrows, peri-haustorial membrane) or after plasmolysis
(E; 0.5 M NaCl for 20 min; arrows, dots and membrane retained around
the haustorium), or a Gc UCSC1-invaded cell (D). (F–H) Localization of
PLDα1-eGFP in a Gc UMSG1-invaded cell (G), or a Gc UCSC1-invaded
cell before (H) or after plasmolysis (F). Scale bars=10µm. PM, plasma
membrane; P, penetration site; T, tonoplast.
56
Gc UCSC1 Gc UMSG1
pldδ respectively, since the GFP signal from the native promoter-driven PLDδ cDNA
(PLDδc) in fusion with eGFP was reported to be too weak for imaging (Pinosa et al.,
2013). PLDδ-eGFP could fully, while PLDα1-eGFP could partially, rescue the re-
spective mutant phenotypes (Fig. 2.9), indicating that these fusion transgenes are
(partially) functional. We then used leaves of the respective transgenic lines in-
fected with Gc UMSG1 or Gc UCSC1 at 2 dpi for subcellular localization analysis
using confocal microscopy. When examining leaves infected with Gc UMSG1, we
detected PLDδ-eGFP in the plasma membrane (PM) of all epidermal cells and in
two or more concentric rings around the penetration site forming the bulls eye do-
main (Assaad et al., 2004, Koh et al., 2005) often with small dots or bulbs within
or nearby (Fig. 2.10A). However, it was rarely seen in the Gc UCSC1 penetration
site (Fig. 2.10B), implying that the adapted pathogen suppresses the recruitment of
PLDδ-eGFP to the probably perturbed PM around the papilla. PLDδc-eGFP was
reported to exhibit focal accumulation around the Blumeria graminis f. sp. hordei
(Bgh) penetration site in Arabidopsis epidermal cells (Pinosa et al., 2013). We thus
examined the subcellular localization of the PLDδc-eGFP expressed from 35S in our
pathosystems. In the case of Gc UMSG1, PLDδc-eGFP was often more preferen-
tially detected in the bulls eye domain (Fig. 2.10A) or in an EHM-like membrane
surrounding the constrained haustorium than PLDδ-eGFP (Fig. 2.10C). After plas-
molysis (0.5M NaCl for 20 min), GFP signal was retained around the haustorium
in small dots or bulbs (Fig. 2.10E), similar to those in the papilla at the penetra-
tion site (Fig. 2.10A), indicating that PLDδc-eGFP is not at the EHM because the
EHM largely remains intact within 30 min of such plasmolysis treatment (Berkey
57
et al., 2016). In the case of Gc UCSC1, the PLDδc-eGFP signal was much weaker
at the penetration site (Fig. 2.10D), suggesting that recruitment of PLDδc-eGFP
to the penetration site is also similarly suppressed by the adapted powdery mildew
pathogen. The slight discrepancy in localization between PLDδ-eGFP and PLDδc-
eGFP may be attributable to alternative splicing of PLDδ (Wang and Wang, 2001)
which is pertinent to the PLDδ-eGFP construct but irrelevant to the PLDδc-eGFP
construct for which a full-length PLDδ cDNA was used (Pinosa et al., 2013).
A strong fluorescence signal of PLDα1-eGFP was found in a peri-haustorium
membrane similar to the EHM (Fig. 2.10G, H), which could be completely separated
from the haustorium after plasmolysis (Fig. 2.10F). This indicates that PLDα1-
eGFP is not localized to the EHM but rather it may be in the tonoplast that tightly
wraps around the haustorium.
These results in general agree with the subcellular localizations of PLDα1 and
PLDδ inferred by protein localization and fractionation analyses in earlier studies
(Pinosa et al., 2013, Wang and Wang, 2001, Wang, 2000). The distinct localization
patterns of these two PLDs may in part contribute to their opposing roles in post-
penetration resistance against powdery mildew pathogens.
2.3.5 PLDδ contributes to resistance independent of EDS1/PAD4,
SA, and JA signaling pathways
Our earlier results (Fig. 2.2C, D, 2.5; Fig. 2.8) suggest that PLDδ and per-
haps PLDα1 may function through an SA-independent pathway. To define this
58
A Col-0 eds1-2 pad4-1 pad4-1sid2-2 eds1-2pad4-1
eds1-2pad4-1sid2-2 B   ** c n.s.
400 n.s. n.s. de ce
de de de dede
300 bd bd
200 a
100
0
C Col-0 dde2-2
D
300 b
a a
200
a
100
0
ol-
0 2-2 2-2 ldδe id p
sid2-2 pldδ C dd s
Figure 2.11: Gc UCSC1 infection phenotypes of pldδ-containing
double and triple mutants and relevant controls. (A) Represen-
tative leaves of indicated genotypes (defined by name IDs from both X
and Y axises) infected with Gc UCSC1 at 10 dpi. (B) Quantification
of spore production in indicated genotypes at 10 dpi normalized to leaf
fresh weight (FW). (C) Plants of indicated genotypes infected with Gc
UCSC1 at 10 dpi. (D) Quantification of spore production of plants in
(C). Bars represent mean ± SEM of four samples (n = 4, 4 leaves each)
from one experiment, which was repeated three times with similar re-
sults. Different lowercase letters indicate statistically different groups
as determined by multiple comparisons using one-way ANOVA, followed
by Tukey-HSD (**P < 0.01). Note that no significant (n.s.) difference
(P > 0.05) was found in three of the four indicated pair of genotypes in
(B).
59
pldδ PLDδ
Spores/mg FW (x100)
Col-0
pldδ
eds1-2
eds1-2pldδ
pad4-1
Spores/mg FW (x100) pad4-1pldδ
pad4-1sid2-
p 2ad4-1sid2-2pld
e δds1-2pad4
e -1ds1-2pad4-1
e pd ls d1 δ-2pad4-1sid2-2
pathway further, we made double and triple mutants by crossing pldα1 or pldδ to
well-characterized SA-dependent (sid2-2) (Wildermuth et al., 2001, Dewdney et al.,
2000) or both SA-dependent and -independent signaling (eds1-2 and pad4-1) mu-
tants (Bartsch et al., 2006, Venugopal et al., 2009).
We first examined if pldδ-mediated ‘eds’ phenotype is additive to the ‘eds’ phe-
notypes of eds1-2 or pad4-1 in response to the well-adapted Gc UCSC1 isolate and
found that eds1-2pldδ and pad4-1pldδ were not statistically more susceptible than
the single mutants (Fig. 2.11A, B). We then made pad4-1sid2-2pldδ, eds1-2pad4-
1pldδ, and eds1-2pad4-1sid2-2 triple mutants, and compared the disease phenotypes
between these and the two double mutants. No significant differences were detected
between the mutants except pad4-1sid2-2pldδ versus pad4-1sid2-2 (Fig. 2.11A, B),
suggesting that either PLDδ somehow acts in the SA pathway or the pldδ-mediated
‘eds’ phenotype may be masked in the various double or triple mutants because
Gc UCSC1 is too aggressive on these mutants to allow reliable detection of any
phenotypic differences.
To test the latter possibility, we used Gc UMSG3, a powdery mildew isolate
from tobacco which can only weakly sporulate on Col-0, to resolve subtle infection
phenotypic differences between different genotypes. Sporulation of Gc UMSG3 was
found to be very weak on both Col-0 and pldδ; however, a whitish fungal mass was
more easily discernible on pldδ at 11 dpi (Fig. 2.12A, B). Interestingly, eds1-2, pad4-
1, eds1-2pad4-1, and pad4-1sid2-2 all supported profuse sporulation (Fig. 2.12A),
suggesting that EDS1 and/or PAD4 make a major contribution to stage II post-
penetration resistance to Gc UMSG3 probably via both SA-dependent and SA-
60
A Col-0 eds1-2 pad4-1 eds1-2pad4-1 pad4-1sid2-2 C eds1-2pad4-1sid2-2
B 300 ** n.s. D 400fg fcfg **
cg ***
** bc 300
200 de be de
d 200
100
100
a a
0 0
Figure 2.12: PLDδ in Arabidopsis contributes to post-
penetration resistance via an SA- and EDS1/PAD4-
independent pathway. (A, C) Representative leaves of indicated
genotypes (defined by name IDs from both X and Y axises) infected
with Gc UMSG3 at 11 dpi. Note that fungal mass is more noticeable
on most of the leaves especially the mid-vein area (arrowheads) from
the middle panel when carefully compared with the corresponding
leaves from the top panel. (B, D) Quantification of spore production in
indicated genotypes in (A, C), respectively, at 11 dpi normalized to leaf
fresh weight (FW). Data represent mean ± SEM of four samples (n = 4,
4∼5 leaves each) from one experiment, which was repeated three times
with similar results. Different lowercase letters indicate statistically
different groups as determined by multiple comparisons using one-way
ANOVA, followed by Tukey-HSD (B, **P < 0.01), or by Students t-test
(D, ***P < 0.001). n.s., not significant.
61
eds1-2pad4-1sid2-2 pldδ PLDδ
Spores/mg FW (x100)
Col-0
pldδ
eds1-2
eds1-2pldδ
pad4-1
pad4-1p
p lda δd4-1
p sa idd 24 -- 21sid2
e -2d ps1 ld- δ2
e pd as d1 4- -2 1p
e ad ds 41 -1-2 pp lda δd4-1sid2-2
Spores/mg FW (x100) pldδ PLDδ
eds1-2pad4-1sid
e 2d -s 2
s 1id -22 p-2 ap dl 4d -δ 1
independent mechanisms.
Notably, eds1-2pldδ and pad4-1pldδ supported significantly more fungal growth
(white powder around the mid-vein in particular) than eds1-2 and pad4-1 visually
(Fig. 2.12A) and quantitatively (Fig. 2.12B), indicating that PLDδ contributes to
resistance against Gc UMSG3 through a mechanism(s) that is at least partially
EDS1 or PAD4 independent. Interestingly, pad4-1sid2-2 was as susceptible as pad4-
1pldδ (Fig. 2.12B), which seemingly implies that PLDδ and SID2 may act in the same
signaling pathway. Yet, pad4-1sid2-2pldδ was significantly more susceptible than
pad4-1pldδto Gc UMSG1 (Fig. 2.12A, B) and pad4-1sid2-2 to Gc UCSC1 (Fig. 2.11A,
B). Similarly, eds1-2pad4-1pldδ exhibited an even higher level of susceptibility than
eds1-2pad4-1 and pad4-1pldδ (Fig. 2.12A, B). Finally, eds1-2pad4-1sid2-2pldδ ex-
hibited significantly higher susceptibility to Gc UMSG3 than eds1-2pad4-1sid2-2
(Fig. 2.12C, D). These observations together support that PLDδ acts through a yet
to be characterized pathway to limit fungal infection at the post-penetration stage.
It is worth pointing out that eds1-2pldδ showed a similar level of susceptibility to
eds1-2pad4-1pldδ(Fig. 2.12A, B), implying that EDS1 and PAD4 are both required
for resistance against Gc UMSG3. Supporting this inference, eds1-2pad4-1 was not
statistically more susceptible than eds1-2 or pad4-1 (Fig. 2.12A, B).
To assess if PLDδ functions through the JA pathway, the Gc UCSC1 infection
phenotype of pldδ was compared with that of dde2-2, which is impaired in JA
biosynthesis (von Malek et al., 2002). The susceptibility of dde2-2 was similar to
that of Col-0 (Fig. 2.11C, D), consistent with our earlier finding that the JA signaling
receptor mutant coi1 did not show ‘eds’ to Gc UCSC1 (Xiao et al., 2005), suggesting
62
A Col-0 eds1-2 pad4-1 sid2-2 pad4-1sid2-2
B n.s. n.s.
200 n.s.
150 n.s.
100
**
50
0
Figure 2.13: The ‘edr’ phenotype of pldα1 to Gc UCSC1 is sup-
pressed by the eds1-2, sid2-2 and/or pad4-1 mutations. (A)
Representative leaves of indicated genotypes (defined by name IDs from
both X and Y axises) infected with Gc UCSC1 at 10 dpi. (B) Quantifi-
cation of spore production in indicated genotypes at 10 dpi normalized
to leaf fresh weight (FW). Data represent mean ± SEM of four samples
(n = 4, 4 leaves each) from one experiment, which was repeated three
times with similar results. No significant (n.s.) difference (P > 0.05,
Student t-test) was detected in all the pairs indicated, except for the
Col-0, pldα1 pair (**P < 0.01).
63
pldα1 PLDα1
Spores/mg FW (x100)
Col-0
pldα1
eds1-2
eds1-2pldα1
sid2-2
sid2-2pldα1
pad4-1
pad4-1pldα1
pad4-1sid2-2
pad4-1sid2-2pldα1
that the JA pathway has little or very limited contribution to defense against Gc
UCSC1. Taken together, PLDδ is unlikely act through the JA pathway.
Next, we investigated if the ‘edr’ phenotype of the pldα1 mutant is affected by
the sid2-2, eds1-2, or pad4-1 mutations by first crossing pldα1 to the three single and
pad4-1sid2-2 double mutants and then testing their infection phenotypes. Intrigu-
ingly, eds1-2pldα1, pad4-1pldα1, sid2-2pldα1, and pad4-1sid2-2pldα1 all displayed
similar ‘eds’ to Gc UCSC1 to the respective single or double mutants with wild-type
PLDα1 (Fig. 2.13). This suggests that pldα1-mediated edr is completely neutral-
ized/suppressed when the SA- and/or EDS1/PAD4-mediated signaling is defective,
genetically placing PLD1 upstream of EDS1, PAD4, and SID2, which is in sharp
contrast to the epistatic effect of pldα1-mediated ‘edr’ over pldδ-caused ‘eds’. A
mechanistic model is proposed to explain the distinct yet related roles of PLDδ and
PLDα1 (see the Discussion; Fig. 2.15).
2.3.6 Loss of PLDα1 and/or PLDδ has no significant impact on SA,
JA, and ABA levels and signaling
To investigate if PLDα1- and/or PLDδ-mediated defense mechanisms are con-
nected with defense-related phytohormones, we first measured levels of endogenous
SA, ABA, and JA in pldα1, pldδ, and pldα1δ along with Col-0 and eds1-2 prior to
and at 5 dpi with Gc UCSC1 using LC-MS/MS. Compared with näıve plants, SA
levels increased by 5 to 16 fold in mildew-infected Col-0 and pld mutants, but re-
mained low in eds1-2 (Fig. 2.14A), indicating that pathogen-induced SA biosynthesis
64
A UCSC1 Inoculation E F Col-0 dde2-2 coi1-1 pldα1 pldδ pldα1δ
1.5 Col-0
a ad a
plda1 8 adeab
pldd ab de
1 eplda1d ab 6
eds1.2 4 c
0.5 c
b b b b 2 bc
bc
b b b
0 0
0 dpi 5 dpi 0 dpi 5 dpi MeJA 0 µM
B a G
300 Col−0 dde2−2 pldd
coi1−1 plda1 plda1d
200
b c
100 b b b b b b b b
0 b
a a a
0 dpi 5 dpi 4
C a
6
d MeJA 5 µM
4 3
c
2
0 a
0 dpi 5 dpi 2 c
D a a c
15 ab db bd e
10 1c db
5 b b c
0 0 10 20 30 40 50 MeJA 25 µM
0 dpi 5 dpi MeJA Concentration (μM)
Figure 2.14: Impact of the pldα1 and pldδ single and double mu-
tations on the levels and signaling of SA and JA before and
after powdery mildew infection. (A-C) Plant hormones SA (A), JA
(B) and ABA (C) levels were measured by LC-MS/MS in leaves of six-
week-old plants of indicated genotypes prior to (0 dpi) and post (5 dpi)
Gc UCSC1 inoculation. Notably, before inoculation, the JA level of
pldα1δ was higher than that of the two single mutants and was reduced
by ∼4 fold at 5 dpi. Bars represent mean ± SEM of three independent
experiments combined (n = 3 for each experiment). (D, E) Log2 fold
changes of PR1 (D) or PDF1.2 (E) relative to UBC9 encoding ubiquitin
conjugating enzyme 9. Bars represent mean ± SEM of three biological
replicates. (F) Representative pictures showing 10-day-old seedlings of
indicated genotypes grown on MS-agar medium without or with 5µM
and 25µM Methyl JA (MeJA). (G) Doseresponse curve of root growth
of indicated genotypes upon MeJA treatment. Root length of 10-day-
old seedlings growing on MS-agar medium supplemented with exogenous
MeJA at 0, 5, 10, 25 or 50µM were measured and presented as mean ±
SEM at each MeJA dosage. The line graph shows combined data from
two independent experiments (n > 15 for each experiment). Different
lowercase letters indicate statistically different groups (P < 0.05) as de-
termined by multiple comparisons using one-way ANOVA, followed by
Tukey-HSD.
65
PR1/UBC9 (log ) ABA (ng/g FW) JA (ng/g FW)2 SA (µg/g FW)
Root length (cm)
PDF1.2/UBC9 (log2)
is intact in the pld mutants. To see if SA-signaling is affected in the pld mutants,
the expression of the marker gene PR1 (Wiermer et al., 2005) was measured and
found to be induced to a level similar to that in Col-0, suggesting that SA-signaling
was not affected by any of the pld mutations (Fig. 2.14D). These results support
the inference from our genetic data that PLDα1 and PLDδ oppositely modulate
post-penetration resistance via an SA-independent pathway. No significant changes
in ABA levels were observed in Col-0 and the pld mutants before and after powdery
mildew infection (Fig. 2.14C).
Surprisingly, the JA level in uninfected pldα1δ was higher (3- to 6-fold) than
that in all other genotypes (Fig. 2.14B), and the expression of its marker gene
PDF1.2 was significantly higher in unchallenged pldα1 and pldα1δ compared with
that in Col-0 (Fig. 2.14E), suggesting that PLDα1 and PLDδ may act together
to repress JA production/signaling in the absence of pathogens. At 5 dpi with Gc
UCSC1, JA in pldα1δ was reduced to a level that is only slightly higher (∼2 fold)
than that in other plants (Fig. 2.14B), which is probably caused by an antagonistic
effect from enhanced SA biosynthesis and signaling in the mildew-infected plants.
However, despite a slight decrease in JA levels in all the genotypes at 5 dpi, expres-
sion levels of PDF1.2 showed a similar increase (2.5- to 12-fold) in all the plants, with
no significant difference between the pld mutants and Col-0 (Fig. 2.14E). Together
these results indicate that (i) although well-adapted powdery mildew infection does
not induce JA biosynthesis, it can still induce JA signaling; (ii) the altered defense
phenotypes in pldα1 and pldα1δ do not correlate with the changes in JA levels
and/or JA signaling.
66
It is known that high JA levels inhibit root growth (Staswick et al., 1992).
To test the results concerning the endogenous JA levels further, we examined root
growth of pldα1δ along with Col-0, pldα1, pldδ, and two JA mutants, dde2-2 (de-
fective in JA synthesis; von Malek et al., 2002) and coi1-1 (insensitive to JA; Xie
et al., 1998) in Murashige and Skoog (MS)-agar medium without or with supple-
ment of exogenous methyl jasmonate (MeJA). Consistent with the results from the
JA level measurements, only roots of pldα1δ grown in MeJA-free MS-agar medium
were significantly shorter (∼76.9% of Col-0) (Fig. 2.14F, G). Roots of all geno-
types, except those of coi1-1, showed similar rates of growth inhibition in MS-agar
medium supplemented with different concentrations of MeJA (5, 10, 25, and 50µM)
(Fig. 2.14G). This indicates that JA signaling in the pld mutants is not affected.
Taken together, our results further demonstrate that PLDα1 and PLDδ oppositely
modulate defense in an SA-independent manner but may act together to curb JA
accumulation in näıve plants.
2.4 Discussion
In this study, we collected genetic evidence to demonstrate that Arabidopsis
PLDα1 and PLDδ oppositely modulate basal, post-penetration resistance against
powdery mildew, and oomycete pathogens via an EDS1/PAD4-, SA-, and JA-
independent pathway.
67
No pathogen Pathogen
PLD PLD
riu
m
o
us
t
a
H
EDS1 EDS1
PAD4 PAD4
PLD 1 PLD 1
SID2
O OH
OH
OH
OH
SA
Defense Chemicals 
(basal level) Defense Chemicals
Defense
Figure 2.15: A working model for the roles of PLDα1 and
PLDδ in plant immunity. In this model, PLDδ positively whereas
PLDα1 negatively modulates plant basal resistance against powdery
mildew with PLDα1 acting downstream of PLDδ. We hypothesize
that upon perception of pathogen invasion, plasma membrane-associated
PLDδ is activated and functions through a novel, SA-independent, sig-
naling pathway(s), which is also distinct from, but possibly overlapping
with, the EDS1/PAD4-dependent pathway(s) (indicated by a dashed
line). By contrast, intracellular PLDα1 is involved in removal of defense
chemicals produced from basal activities of PLDδ- and EDS1/PAD4-
dependent pathways, thereby preventing inappropriate activation of de-
fenses in the absence of pathogens. However, in the presence of powdery
mildew or oomycete pathogens, PLDδ is activated, repressing PLDα1 ac-
tivity, which leads to accumulation of defense chemicals, resulting in ac-
tivation of defense responses. This model also implies that PA pools
produced in different subcellular compartments have distinct roles in
regulation of plant defense responses.
68
Vacuole
Vacuole
2.4.1 PLDδ and PLDα1 modulate post-penetration resistance against
powdery mildew
Pinosa et al. previously reported that the loss-of-function pldδ mutant is
compromised in penetration resistance against the non-adapted barley mildew Bgh
(Pinosa et al., 2013). Here, we show that the same pldδ mutant exhibited ‘eds’ to
a well-adapted powdery mildew isolate Gc UCSC1 (Fig. 2.2) and supported more
hyphal growth of the non-adapted powdery mildew isolate Gc UMSG1 that has
overcome penetration resistance (Fig. 2.4, Wen et al., 2011). This implies that the
PLDδ-based defense mechanism operates throughout the entire infection cycle of
powdery mildew and apparently has not been (fully) suppressed by even aggressive
powdery mildew pathogens such as Gc UCSC1. To determine if PLDδ-mediated
defense is effective against other pathogens, we tested pldδ with the fungus-like
oomycete Hpa Noco2 that also employs a haustorium-based nutrient acquisition
strategy. Notably, pldδ was significantly more susceptible than Col-0 but not as
susceptible as eds1-2 or pad4-1sid2-2 to Hpa Noco2 (Fig. 2.6B). Given that powdery
mildew fungi only invade host epidermal cells while oomycete pathogens invade both
epidermal and mesophyll cells (Takemoto et al., 2003), it is possible that PLDδ-
mediated defense is more effective in epidermal cells compared with mesophyll cells.
It is also possible that oomycete pathogens may be able to suppress PLDδ-mediated
defense more effectively than powdery mildew. In addition, PLDδ-mediated defense
may be attenuated under higher humidity (>90%) conditions necessary for infection
of Hpa Noco2. High humidity-caused suppression of resistance has been reported
69
for several different defense mechanisms (Xiao et al., 2003, Zhou et al., 2004, Wang
et al., 2007). Similar to what was reported earlier (Johansson et al., 2014), we did
not observe any difference in growth of virulent bacteria between Col-0 and pldδ,
suggesting that PLDδ is specifically involved in defense against cell wall-breaching
pathogens. Notably, among all reported genes involved in penetration and post-
penetration resistance, PLDδ is unique in that it contributes to both penetration and
post-penetration resistance against powdery mildew fungi. In contrast to pldδ, both
the pldα1 single and the pldα1δ double mutant exhibited ‘edr’ to virulent powdery
mildew and oomycete pathogens (Figs 2.2, 2.6). This suggests that genetically
PLDα1 and PLDδ function oppositely in the same pathway with PLDα1 acting
downstream of PLDδ. We reported earlier that loss of PLDα1 and PLDδβ1 resulted
in ‘edr’ to virulent bacterial pathogens and ‘eds’ to a necrotrophic fungal pathogen
Botrytis cinerea (Zhao et al., 2013), suggesting a positive role for PLD1 in the
JA pathway and a negative role in the SA pathway. We found in this study that
pldβ1 showed slight ‘edr’ to Gc UCSC1 based on our visual scoring of the infection
phenotypes (Fig. 2.1A), supporting a role for PLDα1 and PLDδβ1 in modulating
SA-JA signaling. Whether PLDα1 and PLDα1 and PLDδβ1 share similar regulatory
mechanisms and/or have overlapping function remains to be determined.
70
2.4.2 PLDα1 and PLDδ may modulate defense via a potentially novel
pathway
Three lines of genetic evidence collectively support our conclusion that PLDδ func-
tions through an SA-independent pathway. First, RPW8-mediated resistance, which
is known to engage SA signaling, is intact in pldδ (Fig. 2.8C); secondly, adding the
pldδ mutation to the SA signaling mutants eds1-2 and pad4-1 , or the SA biosyn-
thesis mutant sid2-2, resulted in increased ‘eds’ to the poorly-adapted isolate Gc
UMSG3 (Fig. 2.12); lastly, pldδ showed similar elevation of SA levels and induction
of PR1 expressions to Col-0 upon powdery mildew infection (Fig. 2.14A, D).
Because EDS1 and PAD4 are believed to function upstream of SA and modu-
late defense via both SA-dependent and SA-independent pathways (Bartsch et al.,
2006, Venugopal et al., 2009), the increased ‘eds’ of eds1-2pldδ, pad4-1pldδ, eds1-
2pad4-1pldδ, and eds1-2pad4-1sid2-2pldδ to Gc UMSG3 (Fig. 2.12) also provide
clear genetic evidence to support a role for PLDδ in defense through an EDS1
and/or PAD4-independent pathway. However, based on our genetic analyses alone
we could not exclude the possibility that PLDδ also contributes to EDS1/PAD4-
dependent resistance. It is possible that the defense pathways mediated by EDS1,
PAD4, and PLDδ may be interconnected or partially overlapping, since the pheno-
typic differences among the single and double mutants concerning these three genes
were largely diminished when they were tested with the aggressive isolate Gc UCSC1
(Fig. 2.11).
We also evaluated whether PLDα1 and PLDδ function via the JA-pathway.
71
Our results from genetic analysis (Fig. 2.11C, D, Xiao et al., 2005), measurements of
JA levels (Fig. 2.14B), and PDF1.2 expression (Fig. 2.14E) showed that the altered
defense phenotypes of the pld mutants could be uncoupled from the changes in the JA
levels and signaling, thus excluding the possibility that PLDα1 and PLDδ modulate
defense through the JA-pathway.
Taken together, our results indicate that PLDα1 and PLDδ play opposing
roles in modulating resistance against powdery mildew via a pathway that is inde-
pendent of the EDS1/PAD4, SA, and JA pathways. Notebaly, mlo-based durable
and broad-spectrum resistance against powdery mildew has recently been shown to
be independent of all the known defense pathways (Kuhn et al., 2017). Therefore,
it will be interesting for future studies to determine if PLDα1 and PLDδ have a
mechanistic connection with MLO or other known defense pathways such the ET
and mitogen-activated protein (MAP) kinase signaling pathways (Kuhn et al., 2017,
Tsuda et al., 2013, Kim et al., 2014, Hillmer et al., 2017).
2.4.3 PLDα1 may repress PLDδ-mediated defense signaling
We previously reported that PLDα1 promotes H2O2 production whereas PLDδ fa-
cilitates downstream H2O2 signaling in guard cells to regulate stomatal closure pos-
itively during drought stress (Zhang et al., 2009, Guo et al., 2012). However, our
genetic data from this study position PLDα1 as a negative regulator downstream of
PLDδ-mediated defense. Consistent with this, powdery mildew haustorium-induced
H2O2 production was not affected in any of the pld mutants (Fig. 2.5A, B). Given
72
that drought response relies on the movement of guard cells, whereas plant defense
against powdery mildew pathogens mostly occurs in the leaf pavement cells, it is
possible that the proteins interacting with these two PLDs and/or their substrates
during drought stress and pathogen infection are different. Hence, it is conceivable
that PLDα1 and PLDδ probably participate in distinct signaling networks in these
two different types of cells in response to abiotic and biotic stresses.
It is unclear to us how PLDδ positively modulates while PLDα1 negatively
modulates post-penetration resistance against powdery mildew pathogens. One pos-
sible mechanism is that PLDα1 and PLDδ exert their opposing roles in defense by
producing distinct pools of PA to modulate distinct cellular processes by targeting
spatiotemporally-restricted proteins at different subcellular localizations (Fig. 2.15).
Our confocal microscopy show that while PLDδ-eGFP is localized at the PM, around
the penetration site and peri-haustorium, PLDα1-eGFP is most likely to be asso-
ciated with the tonoplast and other intracellular membranes (Fig. 2.10), which are
compatible with results previously reported (Pinosa et al., 2013, Wang and Wang,
2001, Wang, 2000). Notably, the eGFP signal of PLDδ-eGFP was the strongest
around the penetration site of non-host barley mildew (Pinosa et al., 2013), weaker
around the penetration site and/or the haustorial complex of the non-adapted Gc
UMSG1, and almost undetectable in such subcellular compartments induced by the
well-adapted Gc UCSC1 (Fig. 2.10A–D). This suggests that PLDδ is recruited to
the PM around the penetration site to produce PA to (in)activate target proteins
locally, and adapted powdery mildew pathogens may suppress this recruitment to
interfere with PLDδs role in defense activation. As for PLDα1, its tonoplast lo-
73
calization may be related to vacuole-based removal of defense molecules to prevent
inappropriate activation of defense in the absence of pathogens. However, its sup-
pression is relieved by PLDδ-triggered signaling once pathogens invade. Future work
will be directed to identifying relevant immunity proteins that are modulated by the
two functionally distinct PLDs.
2.5 Material and Methods
2.5.1 Plant lines and growth conditions
All mutants used in this study were in the Arabidopsis thaliana accession
Col-0 background. Sequence data of the genes in this chapter can be found in
the Arabidopsis Genome Initiative or GenBank/EMBL databases. The accession
numbers of all genes used in this study are listed in Table 2.1. Mutants sid2-2
(Wildermuth et al., 2001), eds1-2 (Bartsch et al., 2006), pad4-1 (Jirage et al., 1999),
dde2-2 (von Malek et al., 2002), coi1-1 (Xie et al., 1998), pad4-1sid2-2 (Tsuda et al.,
2009) and eds1-2pad4-1 (Kim et al., 2014) have been described preciously. The
phospholipase-related mutants used for initial infection tests with Golovinomyces
cichoracearum (Gc) UCSC1 are listed in Table 2.1. The homozygous double (sid2-
2pldα1, eds1-2pldα1, pad4-1pldα1,sid2-2pldδ, eds1-2pldδ, and pad4-1pldδ), triple
(pad4-1sid2-2pldα1, pad4-1sid2-2pldδ, eds1-2pad4-1pldδ, and eds1-2pad4-1sid2-2),
and quadruple (eds1-2pad4-1sid2-2pldδ) mutants were generated by genetic crosses
and identified by PCR genotyping. S5 is a Col-gl line transgenic for a single copy of
a genomic fragment that contains RPW8.1 and RPW8.2 with their respective native
74
promoters (i.e. the RPW8 transgene) (Xiao et al., 2005). The selection marker for
the transgene RPW8 in S5 is BAR gene, which confers resistance to herbicide Basta.
To identify S5/pldα1 and S5/pldδ homozygous plants, the RPW8 transgene was first
selected by treating 7-day-old F2 seedlings with Basta. Of the Basta resistant plants,
homozygous pldα1- and pldδ-containing plants were identified by PCR genotyping,
which were passed down to F3 generation. Spray Basta again on F3 progenies to
search for the RPW8 homozygous lines which no longer segregated. The RPW8
transgene was further confirmed by PCR. Primers and restriction enzymes used for
genotyping of the mutants are listed in Table 2.2.
Seeds were sown in Sunshine Mix #1 (Maryland Plant and Suppliers) and cold
treated (4 ◦C for 2 days) before moving to growth chambers. Seedlings were kept
under 22 ◦C, 75% relative humidity, short day (8 h light at ∼125µmol m−2 s−1, 16 h
dark) conditions for experiments.
2.5.2 DNA Constructs, Plant Transformation and Microscopy
For genetic complementation, the genomic sequences of PLDα1 and PLDδ were
amplified by PLDα1-F/PLDα1-R2 and PLDδ-F/PLDδ-R primers (Table 2.2), re-
spectively, using Q5 DNA polymerase (New England Biolabs, M0491L), cloned into
pCX-SN (Chen et al., 2009) containing the 35S promoter, and introduced into
pldα1 and pldδ, respectively, via Agrobacterium-mediated transformation using the
A. tumefaciens strain GV3101 (Clough and Bent, 1998).
For determining subcellular localizations of PLDα1 and PLDδ, the p35S-
75
pPLDα1:PLDα1-enhanced green fluorescent protein (eGFP), p35S:PLDδ-eGFP, and
pPLDδ:PLDδ-eGFP constructs were made according to a previous report (Pinosa
et al., 2013). Briefly, to make the p35S-pPLDα1:PLDα1-eGFP construct, the ge-
nomic sequence and a ∼2.0 kb untranslated sequence immediately upstream of the
start codon ATG of PLDα1 was amplified from wild-type genomic DNA using
PLDα1-pF/PLDα1-R1 (Table 2.2) and cloned into pK7FWG2 (Karimi et al., 2002)
for C-terminal translational fusion with eGFP under the 35S promoter through
Gateway cloning (Invitrogen). Similarly, to make the p35S:PLDδ-eGFP construct,
the genomic sequence of PLDδ was amplified by PLDδ-F/PLDδ-R2 (Table 2.2) and
cloned into pK7FWG2. To the make the pPLDδ:PLDδ-eGFP construct, a ∼1.4 kb
sequence upstream of the ATG start codon of PLDδ was first amplified by H-PLDδ-
pF/S-PLDδ-pR (Table 2.2), digested by HindIII and SpeI and replaced the 35S
promoter of pK7FWG2. Then the PLDδ genomic sequence without stop codon was
cloned into pK7FWG2(pPLDδ). Construct p35S-pPLDα1:PLDα1-eGFP was intro-
duced into pldα1 and Col-0, while constructs p35S:PLDδ-eGFP and pPLDδ:PLDδ-
eGFP were introduced into pldδ and Col-0 via Agrobacterium-mediated transforma-
tion (Clough and Bent, 1998).
The expression and localization of the PLDα1-eGFP and PLDδ-eGFP fusion
proteins were examined by confocal microscopy using a Zeiss LSM710 microscope
(Wang et al., 2013). Confocal images were processed using the ZEN software (2009
edition) from Carl Zeiss and Adobe Photoshop CC.
76
2.5.3 Pathogen Infection, Disease Phenotyping, and Quantification
Three powdery mildew isolates were used. Isolate Gc UCSC1 was maintained
on live Col-0 and Col-nahG plants, Gc UMSG1 on live sow thistle plants (Wen
et al., 2011), and Gc UMSG3 on live tobacco plants for fresh inocula. Inoculation,
visual scoring of disease reaction phenotypes and conidiophore quantification were
done as previously described (Xiao et al., 2005). Briefly, for conidiophore quantifi-
cation, ∼6 leaves per genotype were collected from sparsely and evenly inoculated
6-week-old plants at 4 days post-inoculation (dpi), cleared in a clearing solution
(ethanol:phenol:acetic acid:glycerol=8:1:1:1, v/v/v/v), and stained by trypan blue
solution (250µg ml−1 in lactic acid:glycerol:H2O=1:1:1, v/v/v) for visualizing the
fungal structure under the microscope. For each experiment, the total number of
conidiophores per fungal colony was counted for at least 20 colonies per genotype.
Data combined from three independent experiments were presented in a boxplot.
For spore quantification, 4–6 leaf samples (∼150 mg leaves per sample) per genotype
from 6- to 7-week-old plants at 10–13 dpi were collected. A spore suspension of each
sample was made by vortexing the leaves for 1 min in 40 ml H2O (0.02% Silwet L-77)
and used (diluted if necessary for susceptible genotypes) for spore counting using a
hemocytometer under a dissecting microscope. Spore counts were normalized to the
fresh weight of the corresponding leaf samples. All data analyses were done in R (R
Core Team, 2017), and graphics were generated using ggplot2 (Wickham, 2009).
Assays with oomycete strains Hyaloperonospora arabidopsidis Noco2 and Emwa1,
and bacterial strains Pseudomonas syringae pv.maculicola (Pma) ES4326, Pma avr-
77
Rpm1, Pma avrRps4, and Pma ∆hrcC were done according to previous reports
(Bonardi et al., 2011, Tornero and Dangl, 2001).
2.5.4 In situ Detection of H2O2 Accumulation and Callose Deposition
In situ H2O2 production and accumulation in the haustorium-invaded epider-
mal cells were stained by 1 mg ml−1 DAB (3,3’-diaminobenzidine) solution (Thordal-
Christensen et al., 1997) and assessed by microscopy. Briefly, to make 20 ml 1 mg
ml−1 DAB solution, 20 mg DAB was first dissolved in ∼300µl H2O with 3–5 drops
of 10N HCl by vortexing, then H2O was added to 19 ml, pH was ajusted to 3.8,
and the final volume was brought up to 20 ml with H2O. The petiole of a detached
leaf was inserted in a well of a 96-well plate containing 300µl DAB solution. Leaf
samples were incubated in a growth chamber with adequate light for 6–8 h, then
collected and immersed in a clearing solution (ethanol:H2O:acetic acid:glycerol =
8:1:1:1, v/v/v/v) for 24–48 h with one change of the solution. The cleared leaves
were then mounted on microscope slides, stained by a few drops of 0.5% Trypan blue
staining solution (250µg/ml trypan blue in lactic acid:glycerol:H2O=1:1:1, v/v/v)
for fungus and examined under microscope.
Callose deposition at the fungal penetration sites and around the haustorium
was detected and evaluated by aniline blue staining. Briefly, leaf samples were
collected 3 days after inoculation, immersed in freshly prepared aniline blue solu-
tion (0.01% aniline blue in 150 mM KH2PO4 solution with a pH of 9.5), incubated
overnight in dark and observed under an epifluorescence microscope. All light mi-
78
croscopy images were viewed and processed using Zeiss Imager A1.
2.5.5 Determination of Endogenous SA, JA, and ABA Concentra-
tions
Sample collection:For each experiment, three leaf samples of 6- to 7-week-
old plants (4–5 leaves per sample, ∼150 mg) per genotype were harvested both
before pathogen inoculation and at 5 dpi for determining endogenous SA, JA, and
abscisic acid (ABA) concentrations simultaneously. Samples were then subjected to
phytohormone analyses as described previously for auxins (Blakeslee and Murphy,
2016, Novák et al., 2012), with the following modifications for the analysis of SA,
JA, and ABA.
SA, JA and ABA Extraction: Samples were ground in liquid nitrogen
in 1.7 ml centrifuge tubes using plastic pestle. Approximately 40 mg of the ground
tissue was then subjected to a secondary grinding in liquid nitrogen, and samples
were extracted with 1 ml 40 mM sodium phosphate buffer (pH 7.0). A 10 ng aliquot
of d4-SA (C/D/N Isotopes Inc., Quebec, Canada, part #D-1156), 50 ng of d5-JA
(C/D/N Isotopes Inc., Quebec, Canada, part #D-6936), and 50 ng of d6-ABA (Ol-
ChemIm, Ltd., Olomouc, Czech Rebuplic, part #0342722) were added into each
sample as internal standards. Samples were buffer-extracted at 4 ◦C on a lab rota-
tor for 20 min, centrifuged at 12,000 g for 15 min, and supernatants were collected
and transferred to fresh 1.7 ml centrifuge tubes. The pH of supernatants was then
adjusted using HCl and samples were further purified via solid-phase extraction.
79
Eluted samples were dried under nitrogen gas, re-dissolved in 100µl of methanol,
and filtered through 0.2 µm PTFE filters (Fisher Scientific, Pittsburgh, PA, USA
part #03-391-4E).
LC-MS/MS analysis: 1µl of each re-dissolved sample was injected onto an
Agilent 1260 infinity high-pressure liquid chromatography system. Compounds were
separated using an Agilent Poroshell 120 EC-C18 (3.5 × 50 mm, 2.7µm) column and
an acidified water: methanol buffer system (Buffer A: 0.1% acetate, 5% methanol
in water; Buffer B: 0.1% acetate in methanol). Gradient conditions were as follows:
hold at 2% B for 1.5 min, 2 min at 2-60% B, 4.5 min at 60-98% B, hold at 98% B for
3.5 min, and then back to 2% B for 1 min. Eluted samples were further separated
and quantified through the coupled Agilent 6460 triple quadrupole dual mass spec-
trometer equipped with an electrospray ionization (ESI) source. Compounds were
quantified in negative ion mode. ESI source parameters were set as follows: gas
temperature at 250 ◦C, gas flow rate at 10 L min−1, nebulizer at 60 psi, sheath gas
temperature at 400 ◦C, sheath gas flow at 12 L min−1, capillary at 4500 V, nozzle
voltage at 500 V. Retention and mass transitions for SA, JA, and ABA were verified
using authentic standards. Specific mass transitions (precursor ion→product ion
pairs, m/z) monitored for each phytohormone were: ABA, 263→153, 263→203; JA,
209→59; SA, 137→93, 137→65.
80
2.5.6 qRT-PCR Analysis
Three leaf samples of 6- to 7-week-old plants (∼100 mg) per genotype were
harvested before and at 5 dpi after Gc UCSC1 infection. Total RNA was isolated for
each sample using TRIzol©R Reagent and reverse transcribed using SuperScriptTM
III Reverse Transcriptase (Invitrogen, Thermo Fisher Scientific Inc.). For each ex-
periment, qRT-PCR was performed with three biological replicates per treatment
and three technical replicates per sample using the Applied Biosystems 7300 Real-
Time PCR System with SYBRTM Green PCR Master Mix (Thermo Fisher Scientific
Inc.). The transcript levels of the target genes were normalized to that of UBC9
(Ubiquitin conjugating enzyme 9, AT4G27960). Data were analyzed using the Ap-
plied Biosystems 7300 Real-Time PCR System Software and comparative ∆∆Ct
method (Livak and Schmittgen, 2001). Primers used are listed in Table 2.2.
2.5.7 JA sensitivity assay
Arabidopsis root response to MeJA assay was adapted from previous descrip-
tion (Xiao et al., 2004a). Briefly, seeds were surface sterilized, plated on MS medium
containing various concentrations of MeJA (0µm, 5µm, 10µm, 25µm, 50µm) , cold-
treated (4 ◦C) for 2 days and transfered to short-day (8 h light at ∼125µmol m−2
s−1, 16 h dark) condition. Images of the seedlings were taken at day 10 and root
length was measured using ImageJ (Schneider et al., 2012).
81
Table 2.2: Primers used in Chapter 2
Primer ID Sequence (5’ → 3’) Purpose
Adapter Cloning
PLDα1-pF caccGGATCCGGCTTCGCTTTTGGGTTTTCT cacc, BamHI PLDα1 genomic + promoter
PLDα1-F caccATGGCGCAGCATCTGTTGCA cacc PLDα1 genomic
PLDα1-R1 TTAGGTTGTAAGGATTGGAGGCA no PLDα1 genomic w/o stop codon
for C-terminal fusion with YFP
PLDα1-R2 GGTTGTAAGGATTGGAGGCAGGTA no PLDα1 genomic w/ stop codon
PLDδ-F caccGGATCCATGGCGGAGAAAGTATCGGA cacc, BamHI PLDδ genomic
PLDδ-R GCGAATTCTTACGTGGTTAAAGTGTCAGGAAGA EcoRI PLDδ genomic w/ stop codon
PLDδ-R2 CGTGGTTAAAGTGTCAGGAAGAGCCA no PLDδ genomic w/o stop codon
for C-terminal fusion with YFP
H-PLDδ-pF caccAAGCTTGTCTCAGCCCATACAGCTCA cacc, HindIII PLDδ promoter
S-PLDδ-pR GTACTAGTGGTTACAACAATTCAGGTGGAA SpeI PLDδ promoter
Gene locus Genotyping
PLDα1-RP CAAGGCTGCAAAGTTTCTCTG PLDα1 WT allele: PLDα1-RP/LP
PLDα1-LP ATTAAGTGCAGGGCATTGATG At3g15730 T-DNA: PLDα1-RP/LBa1
PLDδ-RP TCCGTTTGACCAGATCCATAG PLDδ WT allele: PLDδ-RP/LP
PLDδ-LP TTGCGATTATTACCAACAGCC At4g35790 T-DNA: PLDδ-RP/LBa1
LBa1 GCCATCGCCCTGATAGACGGTT - Salk T-DNA primer
EDS6 GTGGAAACCAAATTTGACATTAG EDS1 Genotyping of eds1-2.
EDS4 GGCTTGTATTCATCTTCTATCC At3g48090 WT allele: 1500 + 750 bp
105/E2 ACACAAGGGTGATGCGAGACA Mut allele:1500 + 600 bp
pad4-1F GCGATGCATCAGAAGAGCA PAD4 WT allele: 260 +108 bp
pad4-1R GCGTTGTGCTCGCGTATCT At3g52430 after Bsm FI digestion.
sid2-2 F5 TTCTTCATGCAGGGGAGGAG SID2 WT allele: F6/R5 (879 bp)
sid2-2 F6 CAACCACCTGGTGCACCAGC At1g74710 Mut allele: F5/R5 (581 bp)
sid2-2 R5 AAGCAAAATGTTTGAGTCAGCA
RPW8.1-F ATGCCGATTGGTGAGCTTGCGATA RPW8.1 RPW8.1 transgene
RPW8.1-R TCAAGCTCTTATTTTACTACAAGC
RPW8.2-F ATGATTGCTGAGGTTGCCGCA RPW8.2 RPW8.2 transgene
RPW8.2-R TCAAGAATCATCACTGCAGAACGT
Gene locus qRT-PCR
AtPR1-F AGAGGCAACTGCAGACTCATACAC At2g14610 PR1: AtPR1-F/R
AtPR1-R AGCCTTCTCGCTAACCCACAT
AtPDF1.2-F TGTTCTCTTTGCTGCTTTCGACGC At5g44420 PDF1.2: AtPDF1.2-F/R
AtPDF1.2-R TGTGTGCTGGGAAGACATAGTTGC
AtUBC9-F CAGTGGAGTCCTGCTCTCACAA At4g27960 UBC9: AtUBC9-F/R
AtUBC9-R CATCTGGGTTTGGATCCGTTA
82
Chapter 3: Mutant Screens to Identify Novel Immune Components
against Powdery Mildew in Arabidopsis
3.1 Abstract
Progress made in the past two decades in deciphering molecular mechanisms
of the plant immune system has greatly advanced our knowledge in plant-pathogen
interactions. However, how plants mount appropriate defense responses against in-
vading (potential) pathogens with different levels of adaptation, and how potential
pathogens overcome different layers of plant immunity to establish infection via ma-
nipulating host immunity and nutrients are still not well characterized. Inspired
by the results on the two PLD genes characterized in Chapter 2, a large-scale ge-
netic screen was designed and performed in the background of a super-susceptible
Arabidopsis eds1-2pad4-1sid2-2 (eps) triple mutant for susceptible to non-adapted
powdery mildew (snap) and compromised immunity yet poor infection (cipi) mutants
upon powdery mildew infection. Of ∼50 mutants isolated, 23 (5 snaps and 18 cipis)
with a non-segregating infection phenotype were prioritized for further characteriza-
tion and candidate gene mapping using two whole-genome sequencing (WGS)-based
approaches: (i) Bulked segregant analysis combined with WGS (BSA-WGS) of F2
83
backcrossed populations; and (ii) Comparative whole-genome sequencing (CWGS) of
mutants with similar phenotypes. By using BSA-WGS, a nonsynonymous mutation
in MAP KINASE PHOSPHATASE1 (MKP1) was found to be the causal mutation
in cipi1, which was validated by knocking out MKP1 in eps using CRISPR/Cas9.
By using CWGS, five independent mutations in MILDEW RESISTANCE LOCUS
O2 (MLO2) were identified to be the causal mutations in five cipis with further
evidence from lack of genetic complementation between the five cipi mutants. In-
terestingly, snap1 has been found to be susceptible to strawberry powdery mildew
Podosphaera aphanis, which is a distant, non-adapted pathogen incapable of infect-
ing eps and other snap mutants. While the causal mutation mapping for snap1
and other mutants are still in progress, the findings and valuable genetic materials
from this genetic screen will undoubtedly not only lead to a better understanding
towards the multi-layered plant immunity and host-adaptation mechanisms of pow-
dery mildew, but may also help design novel strategies for targeted engineering of
antifungal resistance in crops using new gene-editing technologies.
84
3.2 Introduction
Plant diseases caused by various pathogenic microbes result in 15-30% crop
loss each year at a global scale, posing a serious threat to food security (Strange and
Scott, 2005). Pathogens suppress immune responses of host plants and steal nutri-
ents from them. Despite tremendous progress towards understanding the molecular
mechanisms of host immunity and pathogenesis in recent years, how plants mount
appropriate defense responses against (potential) pathogens with different levels of
adaptation, and how pathogens manipulate host metabolism and derive nutrients
from host cells are still not well characterized. The identification of PLDα1 and
PLDδ as novel components modulating plant immunity via a novel yet uncharacter-
ized pathway(s) (Chapter 2) supports this statement and indicates that there are
hidden layers of plant immunity remaining to be characterized. To identify novel
components of a genetic network or pathway in general, an unbiased, whole-genome
scale forward genetic screen is undoubtedly one of the most powerful approaches.
However, it has become increasingly difficult to identify new plant genes involved in
immunity by mutagenizing wild-type plants since conventional genetic screens for
mutants with altered immunity responses using a wild-type plant has been largely
saturated. To identify novel/hidden immune components or nutrient factors manip-
ulated by plant pathogens, more sensitive genetic screening systems are desirable.
During genetic characterization of the pldα1 and pldδ mutants, I made the
triple mutant eds1-2pad4-1sid2-2 (eps) (Chapter 2). The eps mutant contains mu-
tations that disrupt the functions of three key immune components, ENHANCED
85
DISEASE SUSCEPTIBILITY 1 (EDS1) and its homologous & interacting part-
ner PHYTOALEXIN DEFICIENT 4 (PAD4) (Falk et al., 1999, Jirage et al., 1999,
Feys et al., 2001, Wagner et al., 2013), and SALICYLIC ACID INDUCTION DE-
FICIENT 2 (SID2) (Wildermuth et al., 2001). Given that EDS1/PAD4 positively
regulate plant basal defense as well as ETI by both promoting and working in paral-
lel with SID2-generated SA signaling, the eps mutant is defective in major immune
signaling pathways (Feys et al., 2005, Venugopal et al., 2009, Zhu et al., 2011, Wag-
ner et al., 2013, Cui et al., 2017). As expected, eps is super-susceptible to the
adapted powdery mildew (PM) Gc UCSC1. Interestingly, it exhibits moderate sus-
ceptibility to the non-adapted PM Gc UMSG1 but remains completely resistant to
the non-adapted barley PM Blumeria graminis f. sp. hordei (Bgh) (Fig. 3.1) and
strawberry PM Podosphaera aphanis (Pa) UMSG4 (a PM species recently identified
by the Xiao lab) (Fig. 3.10). Though Gc UMSG1 is a nonhost pathogen by def-
inition due to lack of sporulation (i.e. no reproduction) on Arabidopsis wild-type
accessions, it has already overcome the penetration resistance of Arabidopsis and
can be viewed as a poorly-adapted PM (Wen et al., 2011). By contrast, Bgh rarely
if ever penetrates Arabidopsis epidermal cells (Collins et al., 2003). Therefore, it
appears that the nonhost resistance of Arabidopsis against Bgh may contain ad-
ditional layers and is more robust compared to that against Gc UMSG1. These
suggest that while EDS1, PAD4, and/or SID2 serve an important role in basal,
post-penetration resistance against Gc UMSG1, additional immunity mechanisms
independent of EDS1, PAD4, and SID2 could provide adequate protection against
Bgh in this severely immunocompromised eps mutant. Alternatively, Arabidopsis
86
Figure 3.1: Infection phenotypes of eps to different powdery
mildews. Representative plant images showing the infection phenotypes
of the triple mutant eps to powdery mildews with different levels of
adaptation on wild type Col-0: the well-adapted Gc UCSC1, the non-
adapted Gc UMSG1 and the more distant non-adapted barley powdery
mildew Bgh.
may simply lack a host nutrient signal that is required for early differentiation of
Bgh.
Based on the above observations and reasoning, I initiated a forward genet-
ics project using the artificial pathosystem eps and Gc UMSG1 as the screening
system, aiming to uncover hidden layers of plant immunity or other susceptibil-
ity factors independent of EDS1, PAD4, and SID2, and/or nutrient components
manipulated by PM fungi. The moderately susceptible phenotype of eps to Gc
UMSG1 enables a screening for mutants displaying further enhanced disease sus-
ceptibility (“eds”) or enhanced disease resistance (“edr”) to Gc UMSG1 in an EMS-
mutagenized eps population. The isolated “eds” and “edr” mutants were then tested
with the well-adapted PM Gc UCSC1, the poorly-adapted Gc UMSG1, and the
completely non-adapted Bgh or strawberry PM Pa UMSG4 for phenotypic verifica-
87
tion/characterization. For clear nomenclature of the mutants, the “eps-eds” mutants
are designated as susceptible to non-adapted powdery mildew (snap) mutants and the
“eps-edr” mutants as compromised immunity yet poor infection (cipi) mutants. It is
conceivable that the “eds” phenotype of a snap mutant may be attributable to the
loss of a positive regulator or ectopic (over)activation of a negative regulator of plant
immunity. Likewise, the “edr” phenotype of a cipi mutant may be due to ectopic
(over)activation of a positive regulator or loss of a negative regulator of plant immu-
nity, or loss of a host susceptibility factor such as a nutrient regulator/transporter
that is targeted by the PM pathogens for pathogenesis.
3.3 Results
3.3.1 The design and implementation of a sensitive mutant screen
The design of the mutant screen is illustrated in Fig. 3.2. About 200 mg (about
10,000 seeds) of eps seeds were mutagenized in 0.2 % EMS (ethyl methanesulfonate).
An M1 population of over 6,000 plants were obtained and divided into 102 pools
with 30 to 90 plants per pool. The M2 plants from the 102 pools were then screened
with Gc UMSG1 for mutants showing either “eds” (snap mutants) or “edr” (cipi
mutants) pool by pool. After this first round of screening, about 144 tentative snap
and 332 tentative cipi M2 mutants were isolated. Then, a second round of screening
was conducted with both Gc UMSG1 and Gc UCSC1 on the M3 progenies of each
individual putative mutant. At least 12 M3 plants of each line were tested for
each PM isolate and the infection phenotypes were scored. The results can inform
88
0.2% EMS
eps seeds
~6000 M1 plants
~102 pools of M2 seeds
(30–90 plants/pool)
M2 plants
(~30,000 indivuduals)
eps-eds eps-edr 
(susceptible to non-adapted (compromised immunity 
 powdery mildew, snap) yet poor infection, cipi)
1st round of screen
Gc UMSG1
M3 seeds from 
snap or cipi individuals
Individual M3 lines
2nd round of screen
Gc UCSC1
Gc UMSG1
snap cipi cipi cipi
Mapping causal mutations using 
whoe-genome sequencing-assisted approaches 
Figure 3.2: Scheme of the sensitive mutant screen. M1 plants in
102 trays were grown to maturity and seeds were collected separately to
make 102 M2 pools. M2 plants (∼5/M1 plant) were screened for mutants
with altered disease infection phenotypes.
89
us of (i) the mutation type (recessive or dominant, homozygous or heterozygous)
and (ii) the spectrum of the disease susceptibility or resistance of the mutants,
and thus guide phenotype-based comparative whole-genome sequencing to identify
causal mutations. Based on the infection phenotypes of all tested M3 progenies
to both PM isolates, the snap mutants were further defined as those that were
susceptible to both PM isolates and the cipi mutants as those resistant to either
isolate (Fig. 3.2).
3.3.2 Infection phenotypes of the mutants to two powdery mildews
and an oomycete
A total of 23 mutant lines including 5 snap and 18 cipi mutants were selected
for further characterization after the second round of screening, because these M3
progenies exhibited homogeneous “edr” or “eds” phenotypes upon infection with Gc
UCSC1 and Gc UMSG1, indicating that the causal mutations in the 23 mutants have
become homozygous. The disease reaction (DR) scores of the infection phenotypes
of these mutants to both PMs are listed in Table. 3.1. The DR scores are rated
on a scale of 0 to 5 with 0 being immune without any visible fungal mass and 5
being super susceptible with a thick layer of white powder covering an entire leaf.
Some mutants (cipi1, cipi7, cipi13, snap1, snap2, snap3, snap4, and snap5) showed
similar DR phenotypes in response to infection from either PM isolate, indicating
that the implicated genes play a general role in plant-PM interactions. Some other
cipi mutants (cipi4, cipi5, cipi6, cipi8, cipi9, cipi10, cipi14, cipi16, cipi17, and
90
cipi18) showed greater resistance to Gc UMSG1 than to Gc UCSC1, suggesting
that the defense mechanisms activated by the mutations may be partially suppressed
by the well-adapted PM pathogen. Intriguingly, a few cipi mutants (cipi2, cipi3,
cipi11, cipi12, and cipi15) showed better resistance to Gc UCSC1 than that to Gc
UMSG1, implying that such mutations may activate an immune mechanism(s) that
is more effective in restricting the well-adapted pathogen. Alternatively, loss of
a susceptibility factor may impose a more significant impact on the well-adapted
pathogen since it is more specialized (thus more dependent) on such factors for
invasive growth.
To further define the scope of resistance, the cipi mutants were subjected
to infection of a virulent oomycete, Hyaloperonospora arabidopsidis (Hpa) isolate
Noco2. Results from two independent infection tests showed that the majority of
the cipi mutants were as susceptible as eps to Hpa Noco2. Fungus-like oomycete
pathogens are more evolutionarily related to algae. However, they have evolved a
similar invasive strategy as PM and rust fungi, which is to form haustoria in host
cells for nutrient extraction. Unlike PM fungi which only invade host epidermal
cells, oomycete pathogens primarily infect mesophyll cells. In addition, they require
much higher relative (> 90%) humidity for infection. Such distinctive features of
oomycetes relative to PM may explain why most of the cipi mutants do not confer
resistance to Hpa Noco2. Nevertheless, 6 cipi mutants (cipi6, cipi7, cipi8, cipi9,
cipi10, and cipi16) showed obvious reduced susceptibility or enhanced resistance to
Hpa Noco2 (Table 3.1).
The information on the degrees and spectra of disease resistance or suscep-
91
tibility of the cipi or snap mutants was then used to prioritize my efforts towards
identification of the causal mutations through genome sequencing as detailed in the
next section.
3.3.3 Mapping and identifying causal mutations by whole-genome
sequencing
Mapping causal mutations of the mutants using traditional techniques can be
very tedious and time-consuming. With the ever-decreasing cost for whole-genome
sequencing, mapping by sequencing has become a popular and affordable strategy
for simultaneous mapping and identification of causal mutations efficiently. To map
the candidate causal mutations in the 23 mutants, two whole-genome sequencing
(WGS)-based strategies were employed (Fig. 3.3).
The first strategy combines bulked segregant analysis with WGS (BSA-WGS)
(James et al., 2013). It is achieved by deep sequencing of the pooled genomes from
bulked recombinants derived from an F2 backcross population followed by analyz-
ing genetic marker recombination frequencies to map causal mutations (Fig. 3.3).
In such F2 populations, all EMS-induced mutations serve as genetic markers (i.e.
SNPs between “eds” or “edr” mutants and the eps parental line) for assessing their
possible association with the “eds” or “edr” phenotypes. This approach led to the
identification of a nonsynonymous mutation in MAP KINASE PHOSPHATASE1
MKP1 (At3g55207) to be a candidate causal mutation of cipi1.
The second strategy uses comparative WGS of all the cipi and snap mutants
92
Two Whole-Genome Sequencing (WGS)-based Mapping Approaches
Mapping by bulked segregant analysis Mapping by comparative WGS of mutants 
with WGS of backcrossed populations showing similar infection phenotypes
Backcrossing 
mutants Identifying homozygous M3 mutants showing
to the eps X similar powdery mildew infection phenotypes
parental line eps Mutant (M3)
F1 ……
Selfing
F2 segregating population for mapping
DNA #1 DNA #2 …… DNA #n
Sequencing
Gc UMSG1 or Gc UCSC1 Identifying candidate causal mutations 
Selecting mutant plants for DNA extraction that occurred in the same genes 
via comparative WGS analysis
Pooled genomes for sequencing
Figure 3.3: Schematics of whole-genome sequencing-assisted
mapping of causal mutations.
93
Table 3.1: Summary of cipi and snap mutants subjected to WGS analysis
DR scores
Mutant ID Gc Gc Hpa Candidate Encoded protein Mutation type
UMSG1 UCSC1 Noco2 gene ID
Col-0 0 3 3
eps 3 4.5 4.5
cipi1 1 1 – At4g20160 MAP kinase phosphatase 1 nonsynonymous
cipi2 1.5 0.5 5 At1g11310 MLO2 creating an ex-
onic donor splice
site
cipi3 1.5 0.5 5 At1g11310 MLO2 abolishing an ac-
ceptor splice site
cipi4 0.5 1 4 At4g38600 UBIQUITIN-PROTEIN LIGASE 3 nonsynonymous
(UPL3)
cipi5 1 1.5 4 At4g38600 UPL3 nonsynonymous
cipi6 0.5 1 1 unknown
cipi7 1.5 1.5 1 At2g31990 or GT15 (Golgi-localized exotosin) or nonsynonymous
At4g20160 Golgin
cipi8 0.5 1.5 1 At1g46696 Hypothetical protein nonsynonymous
cipi9 0.5 1.5 1 At1g46696 Hypothetical protein nonsynonymous
cipi10 1.5 2.5 1.5 At2g31990 GT15 (Golgi-localized exotosin) nonsynonymous
cipi11 1.5 0.5 5 At1g11310 MLO2 nonsynonymous
cipi12 2 1 5 At1g11310 MLO2 nonsynonymous
cipi13 0.5 0.5 4 At4g20160 Golgin nonsynonymous
cipi14 0.5 1.5 4 unknown
cipi15 1.5 1 5 At1g11310 MLO2 nonsynonymous
cipi16 1.5 2.5 2 At4g38600 or UPL3 nonsynonymous
At1g46696 hypothetical protein nonsynonymous
cipi17 0.5 4 5 At1g27770, AUTOINHIBITED CA2+-ATPASE nonsynonymous
or At4g35620 1 or CYCLIN B2;2 or DEMETER-
or At3g10010 LIKE 2
cipi18 0.5 4 5 At1g27770, AUTOINHIBITED CA2+-ATPASE nonsynonymous
or At4g35620 1 or CYCLIN B2;2 or DEMETER-
or At3g10010 LIKE 2
snap1 4.5 4.5 – unknown (susceptible to non-adapted Pa
UMSG4)
snap2 4.5 5 – At3g52140 FRIENDLY MITOCHONDRIA
(FMT, mitochondrial quality
control)
snap3 4.5 5 – At3g52140 FRIENDLY MITOCHONDRIA
(FMT, mitochondrial quality
control)
snap4 4 4.5 – unknown
snap5 4.5 5 – unknown
cipi, compromised immunity yet poor infection
snap, susceptible to non-adapted powdery mildew
WAS, whole-genome sequencing.
DR scores, disease reaction scores, a higher score indicates more susceptibility.
Mutants potentially in the same complementation group are highlighted in the same color.
94
to reveal independent mutations that occur in the same genes. Under the assump-
tion that a saturated mutagenesis and genetic screen may lead to generation of
two or more independent causal mutations in the same gene, it is possible for me
to identify candidate implicated genes directly by conducting whole-genome scale
comparative analysis of all EMS-induced mutations between mutants with similar
phenotypes (Fig. 3.3). Since the candidate causal mutations can be directly identi-
fied without the need to make bulked segregant pools, which involves backcrossing
and selfing, and phenotyping of the segregating F2 populations, this strategy can
greatly facilitate the initial mapping step.
DNA samples were prepared from each of the mutants and sent to BGI (Shen-
zhen, China) for sequencing. The genome sequences were analyzed by the Xiao lab
members (Drs. Ying Wu and Wanpeng Wang). The detailed results are summarized
in Table 3.1.
However, this strategy cannot obviate gene identification and verification by
other methods such as genetic complementation. To perform genetic complemen-
tation, the DR data in Table 3.1 were used to infer likely genetic complementation
groups first. Mutants in the same tentative groups were then crossed with each other
and F1 hybrid plants were tested with PMs. While disease phenotyping for most of
the F1 hybrids is still ongoing, this approach has already led to the identification
of 5 independent mutations in MLO2 (At1g11310) being responsible for the “edr”
phenotypes in cipi2, cipi3, cipi11, cipi12, and cipi15.
95
3.3.4 CIPI1 encodes MAPK Phosphatase 1, a negative regulator of
plant immunity
3.3.4.1 Identification of the causal mutation in cipi1
The cipi1 mutant is the very first “edr” mutant isolated. The infection phe-
notype of cipi1 to both Gc UMSG1 and Gc UCSC1 was further tested with the M3
generation and all 12 M3 progenies showed “edr”, indicating that the mutation is
homozygous and the resistance activated is effective in restricting both well-adapted
and poorly adapted PM fungi (Fig. 3.4A). Microscopic examination showed that PM
has very limited growth and sporulation on the cipi1 mutants, and the resistance is
not associated with programmed cell death (PCD) or hypersensitive response (HR)
(Fig. 3.4B). Since no other validated cipi mutants were available at the time, the
BSA-WGS strategy was taken to map the candidate causal mutation. The cipi1
mutant was first backcrossed to the eps parental line followed by one generation of
selfing to establish an F2 mapping population. The F1 plants were as susceptible as
eps to both Gc UMSG1 and Gc UCSC1, indicating that the causal mutation in cipi1
is recessive. Indeed, roughly a quarter (∼50) of the F2 population (∼200 individu-
als) showed similar “edr” against Gc UMSG1 to cipi1. Genomic DNAs from these
∼50 F2 recombinants were pooled and sequenced with a genome-wide coverage of
∼30×. Sequence analysis found a C→T mutation in the second exon of At3g55207
in cipi1 to be the most possible candidate mutation (Fig. 3.4C). At3g55207 encodes
a mitogen-activated protein kinase (MAPK) phosphatase previously characterized
96
A eps cipi1 B Gc UCSC1
eps
cipi1
C MKP1 (At3g55207)
C725T
E1 E2 E3 E4
D MKP1 (784 aa)
Catalytic Gelsolin 
domain domain CaMBD 1 CaMBD 2
157–288 323–392 445–469 669–692
NH2- -COOH
S242F
Figure 3.4: Identification of MKP1 as a negative regulator of
plant immunity. (A) Representative disease phenotypes of eps and
cipi1 mutants upon infection by Gc UMSG1 (upper panel) and Gc
UCSC1 (lower panel). (B) Microscopic images showing fungal struc-
tures of Gc UCSC1 stained by trypan blue on the leaf surface of eps
and cipi1 at 9 dpi. Note that no PM-induced cell death is found. Bars,
100µm. (C) Schematic of the Arabidopsis MKP1 gene showing the na-
ture and position of the EMS-induced missense mutation found in cipi1,
which is the 725th base from the ATG start codon. (D) Protein domain
structure of the MKP1, showing the the amino acid change in cipi1.
97
Gc UCSC1 Gc UMSG1
as MKP1 (Ulm et al., 2001). The candidate causal mutation is predicted to result
in an amino acid change (S242F) in the catalytic domain of MKP1(Fig. 3.4D)
MAPK signaling cascades are among the earliest signaling events in PTI and
ETI (Meng and Zhang, 2013, He et al., 2018). Proper control of the magnitude
and duration of the MAPK signaling is essential for appropriate defense outcome
without detrimental effects on the plants. The Arabidopsis MKP1 has been shown to
negatively regulate MAPK signaling in PTI against bacteria Pseudomonas syringae
pv tomato DC3000 by dephosphorylating and inactivating MPK6 (Bartels et al.,
2009, Anderson et al., 2011). However, its role in PTI against other pathogens
has not been reported. Thus, my results on identifying MKP1 as CIPI1 in the
eps background in relation to PM infection has broadened the pathocontext under
which MKP1 may function.
3.3.4.2 Verification of the role of MKP1 in immunity by CRISPR/Cas9-
induced targeted mutagenesis
To determine whether loss-of-function of MKP1 is indeed responsible for the
“edr” phenotype in cipi1, a CRISPR/Cas9 construct expressing two guide RNA
(gRNA) sequences specifically targeting MKP1 was made and introduced into eps
through Agrobacterium-mediated transformation. The two protospacers to which
the two gRNAs match are 111 bp apart in the second exon of MKP1 (Fig. 3.5A).
A PM infection test showed that five of the eight T1 plants transgenic for this
CRISPR/Cas9 construct displayed similar “edr” against Gc UMSG1 seen in cipi1
98
A C725T
MKP1
111bp protospacer 2 PAM
protospacer 5’-…CACCTCGTTCACATCATAACAGCAAGGC…AACACCTAGCGGGAACAAAACCGGGGAG…-3’
locations 3’-…GTGGAGCAAGTGTAGTATTGTCGTTCCG…TTGTGGATCGCCCTTGTTTTGGCCCCTC…-5’
PAM protospacer 1
B T1 lines haboring CRISPR-induced mutations in MKP1 in eps
eps T1-1 T1-3 T1-4 T1-2 T1-5
C 46bp
WT 5’-…CCTCG-TTCACATCATAACAGCAAGGC…AACACCTAGCGGGAAC--AAAACCGGGGAG…GATAAG…-3’
T1-1, 3, & 4 Premature Stop5’-…CCTCG-TTCACATCATAACAGCAAGGC…AACACCTAGCGGGAAC-AAAAACCGGGGAG…GATAAG…-3’ +1bp
homozygous
T1-2 -144bp 5’-…CCTCG-TT-------------------…----------------------CCGGGGAG…GATAAG…-3’
biallelic in frame
5’-…CCTCGTTTCACATCATAACAGCAAGGC…AACACCTAG--------------CGGGGAG…GATAAG…-3’ -11bp
T1-5 5’-…CCTCG-TT-------------------…---------------------ACCGGGGAG…GATAAG…-3’ -143bp
biallelic
5’-…CCTCGTTTCACATCATAACAGCAAGGC…AACACCTAGCGGGAACAAAAAACCGGGGAG…GATAAG…-3’ +3bp
D Col-0 mkp1 (Col-0, CRISPR-knockout) E Col-0 mkp1-1 (Ws)
10 dpi 0 dpi 10 dpi 10 dpi 0 dpi 10 dpi
Figure 3.5: Knocking out MKP1 by CRISPR/Cas9 results in
leaf necrosis and PM resistance. (A) Locations of the two proto-
spacers in MKP1 targeted by the two gRNAs used in the CRISPR/Cas9
construct, highlighted in blue with corresponding PAM in magenta. (B)
Infection phenotypes of plants of indicated genotypes to Gc UMSG1
at 12 dpi. Note mkp1 mutants showed “edr” compared to eps control.
(C) Sequences of mutated alleles from the five T1 lines in (B) showing
CRISPR-induced indel mutations in MPK1. Premature stop codons are
marked by red hexagons. (D) Representative plants showing the phe-
notypes of T1 Col-0 plants transgenic for the CRISPR/Cas9 targeting
MKP1 before (0 dpi) and after infection with Gc UCSC1 (10 dpi). (E)
Representative plants of the mkp1-1(Ws) mutant before (0 dpi) and after
infection with Gc UCSC1 (10 dpi).
99
Gc UCSC1 Gc UMSG1
Gc UCSC1
(Fig. 3.5B). As expected, sequencing of the gRNA target regions in MKP1 detected
CRISPR-induced indel mutations in all of the five T1 “edr” plants. The primers
for PCR and sequencing are listed in Table 3.2. Specifically, the T1-1, T1-3, and
T1-4 transgenic plants are homozygous for a single “A” nucleotide insertion in the
second gRNA target site, creating a premature stop codon (Fig. 3.5C). Meanwhile,
both T1-2 and T1-5 contain biallelic mutations in MKP1. One allele in T1-2 has
an in-frame deletion of 144 bp between the two gRNA target sites, resulting in an
internal deletion of 48 amino acids in MKP1; the other allele contains one base
(“T”) insertion in the first gRNA target site (resulting in a premature stop codon)
plus an additional 12 bp deletion in the second gRNA target site. As for T1-5,
each allele encodes a premature stop codon with one allele having a 143 bp deletion
between the two gRNA target sites and the other allele harboring a “T” and two
“AA” insertions in the first and second gRNA target site, respectively (Fig. 3.5C).
These data provide compelling genetic evidence that CIPI1 encodes MKP1 and loss
of MKP1 results in activation of defense in cipi1.
To further consolidate the above genetic data and assess the defense pheno-
types caused by loss of MKP1 in the Col-0 wild-type background, the CRSPR/Cas9
construct expressing the two gRNAs targeting MKP1 was also introduced into Col-
0 by Agrobacterium-mediated transformation. Nine of the 48 T1 plants displayed
leaf necrosis at 4 weeks-old before pathogen infection, which is consistent with a
previous finding that the mkp1-1 T-DNA mutant in Col-0 background also shows
autoimmune-like responses (Bartels et al., 2009). Targeted sequencing of MKP1 am-
plified from these resistant Col-0 T1 plants confirmed that they all contain biallelic
100
frameshift mutations in MKP1 (data not shown). However, the mkp1-1 in Ws back-
ground, where it was originally identified from, does not have these autoimmune-like
phenotypes (Fig. 3.5D, Ulm et al., 2001). This phenotypic difference in the näıve
plants has been found to be due to a TIR-NLR, SNC1 (SUPPRESSOR OF npr1-1,
CONSTITUTIVE) that is specifically present in Col-0 but absent in Ws (Bartels
et al., 2009).
Interestingly, despite this phenotypic difference prior to pathogen infection,
plants of both the mkp1(Col-0) and mkp1-1(Ws) (Fig. 3.5D, E) showed complete
resistance to Gc UCSC1 but also massive fungus-induced cell death, which was not
detectable in cipi1 plants (Fig. 3.4B).
Taken together, the above genetic data indicate that: (i) MKP1 is a negative
regulator of plant immunity and pathogen-induced programmed cell death; and (ii)
the immune responses modulated by MKP1 appears to be partially or completely
independent of EDS1/PAD4 and SA, and programmed cell death (PCD).
3.3.5 Five cipi mutants contain mutations in MLO2
3.3.5.1 Isolation of five cipi mutants with better resistance to Gc
UCSC1 than to Gc UMSG1
Among all the cipi mutants, cipi2, cipi3, cipi11, cipi12, and cipi15 exhibit sim-
ilar and interesting resistance phenotypes to the two PMs. While all five mutants
show remarkable resistance to the well-adapted PM Gc UCSC1 (Fig. 3.6B, Ta-
ble 3.1), they are only moderately resistant to the poorly-adapted PM Gc UMSG1
101
A Col-0 eps cipi2 cipi3 cipi11 cipi12 cipi15
B
C eps cipi3 D F1 cipi2 x cipi3 cipi11 x cipi3 cipi11 x cipi12 cipi15 x cipi12
E cipi3Col-0 eps
Epidermal cell infection Trichome infection
Figure 3.6: Infection Phenotypes of the five cipi mutants con-
taining mutations in MLO2. (A, B) Representative plants of the
five cipi mutants along with Col-0 and the eps parental line showing in-
fection phenotypes to Gc UMSG1 (A) and Gc UCSC1 (B) at 11–13dpi.
(C) Representative leaves from eps and cipi3 plants inoculated with Gc
UCSC1 at 20 dpi. (D) Infection phenotypes of F1 hybrids between the
indicated mutants at 12 dpi to Gc UCSC1. (E) Microscopic images show-
ing fungal structures of Gc UCSC1 stained by trypan blue on the leaf
surface (upper panel) or haustoria in the epidermal cells (lower panel) of
the indicated genotypes at 11 dpi. Note that the trichome cells support
better PM sporulation in the cipi3 mutant. Arrowheads in red indicate
rod-shaped conidiophores, which are spore-producing structures stained
in dark blue by trypan blue. Arrows in magenta indicate haustoria in-
side the epidermal cells. Dashed line in red highlights a trichome cell.
Bars, 100µm.
102
Gc UCSC1 Gc UCSC1 Gc UMSG1
Gc UCSC1
Gc UCSC1
(Fig. 3.6A, Table 3.1). This is intriguing because Gc UMSG1 cannot even complete
its life cycle on Col-0, yet it can grow better than Gc UCSC1 on these cipi mutants.
This observation implies that the product(s) of the mutated gene(s) in these five
cipi mutants is more critical to the success of Gc UCSC1, which may reflect one
step of Gc UCSC1’s adaptation process. Interestingly, sporadic whitish colonies are
easily visible to the naked eye on the Gc UCSC1-inoculated leaves of these mu-
tants at 10–13 dpi (Fig. 3.6B). These whitish colonies become more prominent and
seem to be associated with trichomes at a later stage around 20 dpi (Fig. 3.6C). A
closer examination under a microscope revealed that though some Gc UCSC1 could
penetrate the epidermal cells to form seemingly normal haustoria and establish mi-
crocolonies on the leaf surface, they barely sporulate except a few, of which most
are associated with trichomes on these five cipi mutants as represented by cipi3 at
11dpi (Fig. 3.6E). Meanwhile, Col-0 and eps plants could support profuse sporula-
tion of Gc UCSC1 and large amount of haustoria in the pavement cells revealed by
removing fungal mycelia on the surface (Fig. 3.6E, lower panel). This observation
suggests that trichomes might be relatively more susceptible to Gc UCSC1. Given
that trichomes are leaf hairs that are believed to serve as physical barriers to insect
and pathogen invasion (Clauss et al., 2006, Frerigmann et al., 2012), this observation
is also surprising and worth further investigation.
103
3.3.5.2 Identification of five independent mlo2 alleles
To identify causal mutations in these cipi mutants, we decided to directly se-
quence and compare the genomes of the five mutants with the anticipation of finding
independent mutations in the same gene(s). Sequence analysis indeed revealed that
each of the five cipi mutants contains a different mutation in At1G11310, a gene
encoding MLO2, one member of the Mildew Locus O (MLO) protein family, indi-
cating that these mutations in MLO2 are responsible for the resistance in these five
cipi mutants. To further confirm this result, I performed genetic complementation
test between the mutants and found that all F1 hybrids showed strong resistance to
Gc UCSC1 (Fig. 3.6D). Based on these compelling genetic data, it can be concluded
that functional impairment of MLO2 results in PM resistance that is independent
of EDS1, PAD4, and SID2.
The Mlo gene was first identified in barley as a susceptibility factor for barley
PM (Büschges et al., 1997) and loss of its functional homologs has been found to
render PM resistance in many plants species (Kusch and Panstruga, 2017) including
Arabidopsis in which loss of MLO2 as a major player along with MLO6 and MLO12
results in penetration resistance to adapted PM fungi (Consonni et al., 2006). How-
ever, the molecular basis of mlo-based PM resistance remains a mystery till today
(Kuhn et al., 2017).
Detailed sequence analysis showed that each of the three mlo2 alleles in mu-
tants cipi11, cipi12, and cipi15 contains a mis-sense mutation resulting in an amino
acid substitution (Fig. 3.7A). The mlo2 alleles found in cipi11 and cipi15 are known
104
A MLO2 (At1g11310)
cipi12 E7K cipi11 G66R (mlo2-8) cipi15 D287N (mlo2-11)
cipi3 cipi2
E1 E2 E3 E4 E5 E6 E7 E8 E9 E10 E11 E12 E13 E14
B cipi3 Premature Stop C cipi2 Premature Stop
E1 E2 E3 E4 E9 E10 E11 E12 E13 E14
normal splicing cipi3 cipi2 normal splicing
WT: 5’-…AAGCAGgtaact…attcagAGCTTA…-3’ WT: 5’-…TTTGGCAAG…ACGgtacaa...ttgcagAATGCA…-3’
- 2nt - 47nt
Mut: 5’-…AAGCAGgtaact…attcaaAGCTTA…-3’ Mut : 5’-…TTTGGTAAG…ACGgtacaa...ttgcagAATGCA…-3’
cryptic acceptor cryptic donor 
splice site splice site
splicing at cryptic site in cipi3 splicing at cryptic site in cipi2
D E cipi3 F cipi2
H2N- H H N-2N- 2
cipi12
E7K
1 2 3 4 5 6 7 extracellular 1 1 2 3 4 5
cytoplamic
-COOH
cipi11
G66R cipi15 -COOH -COOH
(mlo2-8) D287N 
(mlo2-11)
Figure 3.7: Identification of the causal mol2 mutations in the five
cipi mutants. (A) Schematic of the Arabidopsis MLO2 gene showing
the positions of the five single nucleotide changes induced by EMS found
in the five cipi mutants, of which cipi11, cipi12, and cipi15 each contains
a missense mutation. E indicates exon. (B, C) Schematic illustration of
the positions of the cryptic splice sites and the resulting premature stop
codons caused by the mutations in cipi3 (B) and cipi2 (C). Premature
stop codons are marked by red hexagons. (D–F) Graphs showing the
topology of the full-length MLO2 protein with three amino acid muta-
tions found in cipi12, cipi11, and cipi15 highlighted in red (D), and the
putative truncated MLO2 produced in cipi3 (E) or cipi2 (F). The blue
horizontal bar represents the plasma membrane lipid bilayer. The gray
vertical bars represent the transmembrane domains of MLO2.
105
alleles. The cipi11 mlo2 contains a G→A mutation that changes glycine to arginine
(G66R) in the second transmembrane domain (Fig. 3.7D), which is identical to the
previously reported mlo2-8 allele (Consonni et al., 2006). The cipi15 mlo2 harbors
another G→A mutation resulting in an aspartic acid to asparagine (D287N) substi-
tution in the second cytoplasmic loop (Fig. 3.7D), which is identical to the mlo2-11
allele reported before (Consonni et al., 2006). Both mlo2-8 and mlo2-11 were first
identified from an EMS-mutagenesis screen for powdery mildew resistance mutant
(pmr in Arabidopsis and are also known as pmr2-2 and pmr2-1, respectively (Vogel
and Somerville, 2000, Consonni et al., 2006). The cipi12 mlo2 allele contains a G→A
mutation resulting in a glutamic acid to lysine (E7K) exchange in the N-terminal
extracellular domain (Fig. 3.7D), which is a novel mlo2 allele, close to that (R10W)
in the barley mlo-9 allele (Reinstädler et al., 2010)
Interestingly, the mlo2 mutations in cipi2 and cipi3 lead to mis-spliced MLO2
transcripts. The G→A mutation in the cipi3 mlo2 allele occurred in the acceptor
splice site in the 2nd intron, activating an immediate downstream exonic cryptic
acceptor splice site in the beginning of the 3rd exon, which in turn causes a frameshift
resulting in a premature stop codon (Fig. 3.7B). The seemingly synonymous C→T
substitution in cipi2 mlo2 creates an exonic cryptic donor splice site in the 10th exon
of MLO2, resulting in a 47 nt deletion of the 10th exon starting from the mutation
site, generating a premature stop codon in the 12th exon (Fig. 3.7C). The activation
of these cryptic splice sites were verified by detection of the predicted mis-spliced
transcripts in cipi2 and cipi3 in the transcriptome of their F1 hybrid plants. These
aberrant MLO2 transcripts in cipi2 and cipi3 are predicted to produce truncated
106
proteins of 380 or 97 amino acids, respectively (Fig. 3.7E, F). Alternatively, they
could be eliminated through nonsense-mediated mRNA decay (Baker and Parker,
2004).
3.3.5.3 Knocking out MLO6 and MLO12 in cipi3
The Arabidopsis genome contains 15 MLO genes, among which MLO2, MLO6,
and MLO12 form a phylogenetic clade. These three are co-orthologs of the barley
Mlo gene with unequal contributions to PM susceptibility. Knocking out of MLO6
and MLO12 renders mlo2 mutant with complete resistance against adapted PMs,
though the mlo6, mlo12 single, and mlo6mlo12 double mutants show wild-type-
like susceptibility (Consonni et al., 2006). Since our cipi mlo2 mutants in the eps
background showed less resistance to Gc UMSG1 than to Gc UCSC1 (Fig. 3.6A,
B), I was curious to know if knocking out MLO6 and MLO12 in the cipi mlo2
mutants would also result in complete resistance against Gc UMSG1. To test this,
a CRIPSR/Cas9 construct with two gRNAs each targeting either MLO6 or MLO12
was made and introduced into cipi3. Both gRNAs target the third exon of the
corresponding gene (Fig. 3.8). A preliminary PM infection test identified 3 out
of 24 T1 transgenic plants to be completely resistant against Gc UMSG1 (data
not shown). Targeted sequencing revealed that both MLO6 and MLO12 in all the
three T1 plants received indel mutations that resulted in premature stop codons
(Fig. 3.8). The primers for PCR and sequencing are listed in Table 3.2. Thus, it
appears that loss of MLO2, MLO6, and MLO12 in eps more or less recapitulated
107
A
MLO6 (At1g61560)
protospacer PAM 35bp
5’-…GTATGGTAAGAAAGATGTC--CCAAAGGAAGATGA…AGTTAGT…-3’
3’-…CATACCATTCTTTCTACAG--GGTTTCCTTCTACT…TCAATCA…-5’
B T1-1 Premature Stop
homozygous 5’-…GTATGGTAAGA---------------------TGA…AGTTAGT…-3’-19bp
T1-7
homozygous 5’-…GTATGGTAAGAAAGATGTC----------------…AGTTAGT…-3’
-31bp
T1-5 5’-…GTATGGTAAGAAAGATGTCCTCCAAAGGAAGATGA…AGTTAGT…-3’+2bp
homozygous
C
MLO12 (At2g39200)
5’-…AGTCCTCG-C-CGAAATTTAGCCACCAAGGGTTATGAC…-3’
3’-…TCAGGAGC-G-GCTTTAAATCGGTGGTTCCGTTATGAC…-5’
PAM protospacer
D
T1-1 5’-…AGTCCTCGCC-CGAAATTTAGCCACCAAGGGTTATGAC…-3’ +1bp
homozygous
-34bp 3rd intron
T1-7
biallelic 5’-…AGTCCTCG-C-----------------------------AG……GGGAAAGTAGC-3’
5’-…AGTCCTCG-CACGAAATTTAGCCACCAAGGGTTATGAC…-3’ +1bp
T1-5
biallelic 5’-…AGTCCTCG-CACGAAATTTAGCCACCAAGGGTTATGAC…-3’ +1bp
5’-…AGTCCTCG-CTCGAAATTTAGCCACCAAGGGTTATGAC…-3’ +1bp
Figure 3.8: Indel mutations induced by CRISPR/Cas9 in mlo6
and mlo12 in cipi3. (A, C) Schematics of the Arabidopsis MLO6 (A)
and MLO12 (C) gene showing the positions of the protospacers to which
the gRNAs expressed from the corresponding CRISPR/Cas9 constructs
match. (B, D) Sequences of mutated alleles from 3 T1 eps lines show-
ing CRISPR-induced simultaneous indel mutations in MLO6 (B) and
MLO12 (D). Premature stop codons are marked by red hexagons.
108
the resistance phenotypes found with the mlo2mlo6mlo12 triple mutant (Consonni
et al., 2006). This preliminary result further support the notion that host MLO(s)
is a critical susceptibility factor for PM fungi, although one cannot formally exclude
the possibility that MLOs serve as important negative regulator of plant immunity
and loss of this regulatory step ectopically activates EDS1-PAD4-SID1-independent
cell wall-based resistance against PM pathogens.
3.3.5.4 MLO2-GFP exhibits focal accumulation at the PM penetra-
tion site
To provide additional genetic evidence for the conclusion that mutations in
MLO2 are responsible for the PM resistance in the five cipi mutants, and determine
the precise subcellular localization of MLO2 in PM-infected cells, I have introduced a
C-terminal GFP-tagged MLO2 driven by either its native promoter (pMLO2:MLO2-
GFP) or the constitutive 35S promoter (p35S:MLO2-GFP) into the cipi3 mutant.
Interestingly, the GFP signal can be detected in PM-infected leaves of T1 plants
transgenic for pMLO2:MLO2-GFP but not p35S:MLO2-GFP, suggesting that the
native MLO2 promoter is required for efficient expression of MLO2. Consistent
with this, all 24 T1 cipi3 plants transgenic for pMLO2:MLO2-GFP restored the
susceptibility to Gc UCSC1, as observed both macroscopically and microscopically
(Fig. 3.9A, B). This result further demonstrates that the PM resistance in cipi3 was
indeed caused by functional impairment of MLO2. In addition, it also indicates that
the MLO2-GFP fusion protein is functional and proper MLO2 expression (from its
109
A p35S:MLO2-GFP pMLO2:MLO2-GFP C pMLO2:MLO2-GFP
/cipi3 /cipi3 /cipi3
P
B D
P
H
Sporelings
Figure 3.9: MLO2-GFP is functional and exhibits focal accumu-
lation around PM penetration sites. (A) Representative photos
showing Gc UCSC1-infected cipi3 T1 plants transgenic for p35S:MLO2-
GFP or pMLO2:MLO2-GFP at 11dpi. (B) Representative microscopic
images showing the sporelings and a mycelial network of Gc UCSC1
stained by propidium iodide on the leaf surface of the indicated trans-
genic lines. Bars, 50µm. (C, D) Representative confocal images showing
subcellular localization of MLO2-GFP expressed from its native pro-
moter in a Gc UCSC1-infected cell. P, penetration site; arrows, vesicle-
like structures surrounding the penetration site (C) or the haustorium
(D); H, haustorium. Bars, 10µm
110
native promoter) is required for successful invasion of PM fungi.
Confocal imaging of the infected leaves of T1 plants transgenic for pMLO2:MLO2-
GFP revealed that GFP signal was primarily detected at the penetration site sur-
rounding the neck region of the Gc UCSC1 haustoria while small vesicle-like struc-
tures were occasionally visible around the penetration site (Fig. 3.9C, D). The focal
accumulation of MLO2 at the PM penetration site is similar to what was reported for
barley Mlo (Bhat et al., 2005) and consistent with the hypothetical role of MLO2 as
a host membrane protein being recruited by PM pathogens for cell wall penetration.
3.3.6 snap1 is susceptible to a more distant, non-adapted PM Po-
dosphaera aphanis
To further characterize the putative snap mutants that showed “eds” to Gc
UMSG1, they were tested with the barley PM Bgh and none could support any
fungal growth visible to the naked eye (data not shown). This suggests that de-
spite the loss of EDS1, PAD4, and SID2, nonhost resistance against Bgh remains
effective the five snap mutants. Pa UMSG4 is a PM species infectious on straw-
berry. The Xiao lab recently identified and purified one Pa isolate designated Pa
UMSG4. Interestingly, eps plants still exhibits adequate nonhost resistance to Pa
UMSG4, which prevents Pa UMSG4 from sporulating (Fig. 3.10). Since Pa UMSG4
is evolutionarily closer to Gc UCSC1 compared to Bgh, it is possible that some of
the snap mutants may be able to support infection from Pa UMSG4. To test this
possibility, 11 snap mutants were infected with Pa UMSG4 and snap1 was found
111
Pa UMSG4
eps snap1
eps snap1
Figure 3.10: Mutant snap1 is susceptible to strawberry mildew
Pa UMSG1. Representative eps and snap1 plants infected with straw-
berry mildew Podosphaera aphanis (Pa) UMSG4 at 12dpi.
112
to be moderately susceptible (Fig. 3.10). The Xiao lab recently found that knock-
ing out PENETRATION 2 (PEN2) in the eps background rendered the quadruple
mutant moderately susceptible to Pa UMSG4 (data not shown). PEN2 encodes
a myrosinase that produces an antifungal glucosinonate products contributing to
penetration resistance against non-host pathogens (Lipka et al., 2005). To check
whether the susceptible phenotype of snap1 is due to a mutation in the PEN2 gene,
we sequenced the PEN2 gene from snap1 and found no mutation. This suggests that
a mutation in another gene playing a similar role in penetration resistance occurred
in snap1. We are currently building a F2 backcrossed population for identifying this
gene using the BSA-WGS approach.
3.4 Discussion
With help from other lab members, I have performed a large-scale genetic
screen in the immunocompromised eds1-2pad4-1sid2-2 (eps) triple mutant back-
ground and isolated a total of ∼50 cipi and snap mutants displaying significantly
altered PM infection phenotypes, of which 23 (18 cipis and 5 snaps) were priori-
tized for further causal mutation characterization (Table 3.1). Of the snap mutants,
at least one mutant is susceptible to the non-adapted strawberry PM Pa UMSG4.
Of the cipi mutants isolated, I have so far identified six causal mutations with one
in MAP Kinase Phosphatase 1 (MKP1, At3g55207) and the other five in MLO2
(At1g11310). Meanwhile, eight candidate causal mutations have been identified for
12 other cipi mutants and 2 additional snap mutants, and the corresponding genes
113
will be functionally characterized in the near future.
The identification of novel immunity components and/or susceptibility fac-
tors of PM fungi through this novel and sensitive genetic screen will likely not only
lead to a better understanding of host-immune mechanisms in general, but also en-
able engineering of crop resistance against fungal pathogens using new gene-editing
technologies such as CRISPR/Cas9.
3.4.1 The eps-Gc UMSG1 artificial pathosystem enables a sensitive
genetic screen
To unravel hidden immune components of the multi-layered immune system
in plants against haustorium-forming pathogens and identify host susceptibility fac-
tors of these pathogens, I conducted a forward genetic screen using an artificial
pathosystem. This artificial pathosystem is made possible by the creation of the im-
munocompromised eps triple Arabidopsis mutant and the utilization of Gc UMSG1,
a non-adapted PM of wild type Arabidopsis (Fig. 3.1). With this pathosystem, I
aimed to screen for compromised immunity yet poor infection (cipi) mutants and
susceptible to non-adapted powdery mildew (snap) mutants (Fig. 3.2). This artifi-
cial pathosystem-based genetic screening system has several advantages over classic
genetic screens. First, using eps as parental line with three major known immu-
nity components EDS1, PAD4, and SID2 removed greatly increases the chances of
identifying genes involved in hidden immune mechanisms. Second, the eps mutant
is moderately susceptible to the non-adapted PM Gc UMSG1 (Fig. 3.1), possessing
114
the potential to become more susceptible or resistant to allow screening for mutants
with both “eds” and “edr” phenotypes. Third, the “eds” and “edr” phenotypes
are macroscopically visible and examinable, allowing efficient screens for mutants
by the naked eye. Last, Gc UMSG1 is maintained on its adapted host sow thistle,
which has much bigger leaves compared to Arabidopsis, thus providing abundant
PM spores for large-scale inoculation experiments. Indeed, all the above four factors
contributed to what I believe to be a very efficient and fruitful genetic screen, which
led to the isolation of a total of ∼50 cipi and snap mutants.
As reasoned earlier in 3.2 Introduction, the “edr” phenotype of a cipi mutant
may be due to ectopic (over)activation of a positive regulator or, or loss of a negative
regulator of plant immunity, or loss of a host susceptibility factor required for PM
pathogenesis and the implicated mechanisms must be independent of the EDS1,
PAD4, and SID2-controlled defense signaling pathways. Supporting this reasoning,
the “edr” phenotype of cipi1 is due to loss of a negative regulator of plant immunity,
MKP1 (Fig. 3.4), and that of cipi2, cipi3, cipi11, cipi12, and cipi15 is due to loss of a
host susceptibility factor MLO2 known to be recruited by PM pathogens (Fig. 3.6 &
3.7). I have also identified seven additional candidate genes whose mutations might
be responsible for the “edr” phenotypes of 12 cipi mutants (Table 3.1). Conversely,
though the nature of the causal mutation has not been definitively characterized
for any of the snap mutants so far, the causal mutations responsible for the “eds”
phenotype of a snap mutant may lead to ectopic (over)activation of a negative
regulator, or loss of a positive regulator of plant immunity. Both CRISPR and
BSA-WGS approaches will be used to identify the causal mutations in the snap and
115
the remaining cipi mutants in the future (Table 3.1). This may lead to isolation of
host nutrient regulators/transporters that are targeted by the PM pathogens.
3.4.2 Plant defense mechanisms independent of EDS1/PAD4 and SA
Host immunity in plants against biotrophic and semi-biotrophic pathogens is
associated with elevated SA levels, induced expression of PR genes, and production
of reactive oxygen species, which often culminates in programmed cell death (PCD)
at the site of infection known as the hypersensitive response (HR) to restrict the
spreading of the invading pathogen (Thomma et al., 2011, Cui et al., 2015). How-
ever, due to lack of EDS1, PAD4, and SID2 in the eps background, resistance in the
cipi mutants does not show strict association with these typical defense responses.
Although detailed molecular characterization has yet to be done with any of the cipi
mutants, data from this study suggest the existence of independent defense mecha-
nisms that function together or in parallel with EDS1/PAD4 or SA as exemplified
by the cipi1 mutation in MKP1 and five other cipi mutations in MLO2. Though
MKP1 and MLO2 have been studied quite extensively and represent two distinct
mechanisms, my data provide new insights into each of the two mechanisms.
3.4.2.1 Negative regulation of PTI signaling by MKP1
MKP1 has been shown to be a negative regulator of PTI against bacterial
pathogen Pst DC3000 by controlling inappropriate activation of MPK6 (Bartels
et al., 2009, Anderson et al., 2011, Jiang et al., 2017). The mkp1-1 T-DNA inser-
116
tion mutant was originally identified in the Arabidopsis accession Ws background
and developmentally the mkp1-1(Ws) plants are indistinguishable from the Ws wild
type (Ulm et al., 2001). Interestingly, when the T-DNA insertion in mkp1-1 (Ws)
was introgressed into Col-0, the mkp1-1(Col-0) plants showed dwarfism, early senes-
cence, constitutive defense responses and enhanced resistance to virulent bacterial
pathogens (Bartels et al., 2009). Elevated SA levels were found to be responsi-
ble for the aberrant phenotypes as knocking out EDS1 or PAD4 or expressing the
SA-depleting NahG gene could largely suppress the autoimmune-like responses in
mkp1(Col-0). It was further shown that this autoimmune-like responses seen in
mkp1-1(Col-0) was largely due to the presence of the Col-0-specific TIR-NLR SUP-
PRESSOR OF npr1-1, CONSTITUTIVE (SNC1), which is absent in the Ws back-
ground (Bartels et al., 2009). However, the precise role of SA-signaling in mkp1-1-
triggered defense responses is unclear.
I found here that the mkp1(Col-0) mutants generated by CRISPR/Cas9 were
also resistant to biotrophic fungal PM pathogen Gc UCSC1 (Fig. 3.5D). Intriguingly,
mutations in MKP1 still render PM resistance in the absence of EDS1, PAD4, and
SID2, and PCD (cipi1 and mkp1eps mutants generated by CRISPR/Cas9, Fig. 3.4A,
3.4B, & 3.5B). In addition, though the mkp1(Ws) mutant plants are normal in devel-
opment, they displayed PM-induced cell death 10 days after being challenged with
Gc UCSC1 (Fig. 3.5E). It is clear that: (i) PM resistance caused by loss of MKP1 is
due to derepression of (most likely PTI) defense mechanisms that are independent
of SA signaling; (ii) autoimmune-like responses such as HR-like cell death or PCD in
mkp1(Col-0) requires EDS1, PAD4, and SID2 perhaps through SNC1, which can be
117
uncoupled from PM resistance; (iii) SNC1 is not responsible for mkp1-mediated re-
sistance, as mkp1-1(Ws) is still resistant to PM in the absence of SNC1; (iv) though
SNC1 is responsible for the aberrant developmental phenotypes including PCD in
mkp1(Col-0), it is not responsible for PM-induced PCD in mkp1-1(Ws) that most
likely engages EDS1 and/or PAD4 and/or SID2 through an unknown mechanism.
Thus, it appears that MKP1 negatively regulates PTI and PCD in parallel
or conjunction to ensure appropriate activation of costly defense responses. While
MKP1 most likely represses PTI via dephosphorylation of MPK6 (and other sub-
strates) in the MAPK pathway, how it curbs PCD is not clear. It is possible that
perturbation of MKP1 derepresses two separate components – one connects to SNC1
(in Col-0), resulting in constitutive immune responses including PCD; and the other
one leads to pathogen-induced PCD (in both Col-0 and Ws), both of which require
EDS1 and/or PAD4 and/or SID2. Future experiments are needed to define the
regulatory network of MKP1 and assess the effectiveness of mkp1-triggered defense
against other pathogens in the eps background.
3.4.2.2 Possible mechanisms of mlo-based powdery mildew resistance
The PM resistance conferred by mlo2 shown both previously and here (cipi mlo2
and mlo2mlo6mlo12eps) is truly remarkable (Fig. 3.6). However, the molecular basis
of mlo-based resistance is currently unknown. There are two possible mechanistic
explanations.
First, MLO2 and its close family members are negative regulators of PTI
118
and/or some cell wall-based defense mechanisms. This is supported by spontaneous
cell death and senescence-like chlorosis and necrosis exhibited by the mlo2 and
mlo2mlo6mlo12 mutants (Consonni et al., 2006). Consistent with this, these defense-
associated developmental phenotypes are abolished in the cipi mlo2 quadruple and
mlo2mlo6mlo12eps hextuple mutants (Fig. 3.6). In addition, mutations in several
genes involved in different defense mechansims have been found to partially suppress
mlo2-mediated resistance (Consonni et al., 2010, Lorek et al., 2013). However,
despite extensive efforts, none of the “suppressors” of mlo2 identified by previous
forward and reverse genetic approaches is able to suppress mlo2mlo6mlo12-based
resistance (Kuhn et al., 2017).
Second, more likely, MLO2 and its close family members serve as critical sus-
ceptibility factors of PM fungi for host cell entry. This hypothesis is well-supported
by the following two observations. The first is that mlo-based resistance against PM
fungi appears to be highly specific to PM fungi. Although it was reported that the
mlo2mlo6mlo12 triple mutant could support reduced host cell entry rates by Col-
letotrichum higginsianum, a fungal pathogen that directly penetrates leaf epidermal
cells (Acevedo-Garcia et al., 2017), the resistance is not comparable to that against
PM fungi. Consistent with this, the five cipi mlo2 mutants were as susceptible as
eps to a virulent oomycete pathogen (Table 3.1). The second is that mlo2-based
resistance in the five cipi mlo2 mutants is more effective against the well-adapted
PM Gc UCSC1 than to the non-adapted Gc UMSG1. This makes sense because the
adapted PM may rely more on and make better use of host MLO2 for penetration
and establishing colonization due to long-term co-evolution with its host. On the
119
contrary, if loss of MLO2 activates a defense mechanism, non-adapted PM fungi
should be more sensitive to plant defenses, , which is not the case and opposite to
what I observed.
Though there is no clear evidence to demonstrate why and how PM fungi use
MLO for host cell entry, the biological functions of MLO7 so far characterized might
provide some clues. MLO7, also known as NORTIA, has been found to work with
FERONIA, a receptor-like kinase, for controlling pollen tube perception in synergids
(Jones et al., 2017) and expressing MLO2-GFP from the MLO7 promoter can largely
rescue the reduced fertility phenotype in mlo7 (Kessler et al., 2010). This means that
MLO2 can perform the same molecular function as MLO7 if expressed in the right
place. If ectopic expression of MLO7 in leaves of the mlo2mlo6mlo12eps hextuple
mutant can restore susceptibility to PM fungi, the molecular and cell biological
mechanisms underlying MLO-assisted host cell-entry may mirror those for pollen
tube perception by ovules. After all, the fungal hyphal tip growth and penetration
of host cell wall are physically and topologically similar to ovule-directed pollen tube
growth and entry of the ovule. Compatible with this notion, MLO2-GFP was found
to accumulate focally around the fungal penetration site (Fig. 3.9), which is also
similar to pollen tube-induced polarized localization of MLO7-GFP at the pollen
tube entry site of the ovule (Kessler et al., 2010).
To elucidate the molecular basis of MLO as a critical host susceptibility factor
of PM, it is necessary to further define the molecular functions of MLO proteins
(MLO2 in particular) for plant growth and development and responses to exter-
nal stimuli. Also, identification of a true suppressor of mlo-based complete resis-
120
tance should be revealing. For this, the mlo2mlo6mlo12esp hextuple mutant and Gc
UMSG1 may be used as a pathosystem for screening plant mutants that can support
infection of wild type or mutants of Gc UMSG1 that can overcome mlo-mediated
penetration resistance, because the immune-compromised hextuple would more eas-
ily show up infection to a wild type or mutant Gc UMSG1 if the MLO requirement
can be (partially) met or compensated. Relevant to this, I already did a suppressor-
screening earlier using EMS-mutagenized seeds of cipi2 and cipi3 and identified
five mutants that at least partially restored susceptibility to Gc UCSC1 (data not
shown). It is worth testing if the these five suppressors can also suppress resistance
in the mlo2mlo6mlo12eps background. In addition, the seemingly preferential tri-
chome cell infection in cipi mlo2 mutants (Fig. 3.6E) and mlo2mlo6mlo12eps (data
not shown) by PM fungi is another intriguing observation that may help resolve the
MLO mystery. It is possible that some other MLO members may be more highly
expressed in trichome cells (hence partially compensate the loss of MLO2 alone or
along with MLO6 and MLO12). Alternatively, the cell wall of trichomes may be
compositionally different from that of epidermal pavement cells, hence it is easier for
PM to penetrate. Future experiments will be directed to testing these speculations.
3.4.3 Towards better understanding multi-layered non-host resistance
The goals of the new genetic screen are to unravel additional/hidden layers
of plant immunity and/or host susceptibility factors essential for PM pathogenesis.
Although it is not entirely novel, the identification of MKP1 as a negative regulator
121
of PTI and MLO2 as a PM susceptibility factor has demonstrated the power and
efficacy of the new pathogenetic system. The remaining cipi and snap mutants will
serve as invaluable starting genetic materials for understanding multi-layered plant
immunity.
Among these mutants, snap1 is the most exciting because it can support infec-
tion from Pa UMSG4 which cannot establish infection on eps (Fig. 3.10), indicating
a breakdown of nonhost resistance against non-adapted pathogens. Nonhost resis-
tance protects most plants from most potential pathogens, however the underlying
mechanisms are poorly characterized in general. In Arabidopsis, PENETRATION1
(PEN1), PEN2, and PEN3 have been demonstrated in contributing to penetration
resistance against non-adapted PM through two independent mechanisms. While
the PEN1 syntaxin, together with its SNARE partners, mediates focal exocytosis
of antimicrobial materials to the penetration site (Collins et al., 2003, Kwon et al.,
2008), PEN2, a myrosinase, produces an antifungal glucosinonate product that is
then transported by the PEN3 ATP-binding cassette transporter to the site of in-
fection (Lipka et al., 2005, Stein et al., 2006, Bednarek et al., 2009). Consistent
with this, knocking out both PEN1 and PEN2 in eps renders the quintuple mu-
tant susceptible to Pa UMSG4; however, knocking out PEN1 alone in eps does
not (data not shown). This suggests that the role of the mutated gene in snap1
is probably more prominent than that of PEN1. PLDδ is another gene that has
been shown to contribute to penetration resistance (Pinosa et al., 2013) as well as
post-penetration resistance (Chapter 2, Zhang et al., 2018). However, neither does
the eps/pldδ quadruple mutant support visible fungal growth of Pa UMSG1 (not
122
shown). Since PEN2 in snap1 bears no mutation as shown by targeted sequencing,
it is highly possible that the susceptibility of snap1 to Pa UMSG4 is caused by
a novel mutation that either disrupts a positive regulator or activates a negative
regulator of nonhost resistance. Identification of the causal mutation in snap1 is in
progress; further functional characterization of SNAP1 will hopefully help unravel
another layer of the complex nonhost resistance in plants.
3.5 Material and Methods
3.5.1 Plant lines and growth conditions
See Chapter 2.
3.5.2 Pathogen Infection, Disease Phenotyping, and Quantification
See Chapter 2.
3.5.3 Plant Transformation and Microscopy
See Chapter 2.
3.5.4 EMS mutagenesis
About 10,000 eps seeds (approx. 200 mg) were weighted and placed in a 250 ml
glass flask. Add 15 ml dH2O followed by 30µl EMS (0.2 %). The flask was sealed
with parafilm and shook overnight on a rocker in the fume hood. EMS solution was
removed by pipetting 5 M NaOH, left in the hood overnight and disposed. Seeds
123
were washed 15× with dH2O and blotted dry on filter paper. They were then
mixed with 250 g fine sand of similar size and aliquoted into about 50 parts. Each
aliquot of ∼5 g sand with seeds was then sown evenly in Sunshine Mix #1 (Maryland
Plant and Suppliers) in a typical 1020 tray (28 cm W×54 cm L×6 cm D). Seedling
were grown in a greenhouse under long day (16 h light at 125µmol m−2 s−1, 8 h
dark) condition for about two months till maturity. Seeds from 30–90 M1 plants
were collected together to make one M1 seed pool. A total of 102 M1 seed pools
were obtained. For mutant screening, about 150–450 M2 plants per pool (∼5×
coverage of the M1 plants) were prepared in one or two flats for inoculation with Gc
UMSG1 at 5–6 weeks old. Disease reaction phenotypes (DR scores) were evaluated
macroscopically. Potential cipi and snap mutants were transplanted into individual
pots for seeds collection. M3 progenies were tested with both Gc UMSG1 and Gc
UCSC1.
3.5.5 Preparation of CRISPR DNA constructs
Plasmids with an Arabidopsis egg cell-specific promoter-driven Cas9, pHEE2A-
TRI and pHEE401E, developed by Wang et al. were used for making CRISPR con-
structs with 2 guide RNAs (gRNAs) (Wang et al., 2015). The two CRISPR/Cas9
constructs with one targeting MKP1 (MKP1-2gRNAs/pHEE401E) and the other
one targeting both MLO6 and MLO12 (MLO6-MLO12-gRNAs/pHEE401E) were
prepared according to the method described by Wang et al. (Wang et al., 2015).
Briefly, first, the gRNA sequences with high predicted on-target activities and low
124
off-target effects were picked using Benchling (ben, 2017). To target MKP1 in eps
background, two gRNAs about 150 bp apart targeting the second exon of MKP1
were used. To mutate MLO6 and MLO12 with the same CRISPR/Cas9 construct
in cipi3, the two gRNAs targeting each gene were cloned into the same construct.
Then a cassette containing both gRNAs with a scaffold RNA, a U6-26 terminator
and a U6-29 promoter in between (gRNA1–gRNA-Sc–U6-26t–U6-29p-gRNA2) were
cloned from pHEE2A-TRI with MKP1-gRNA1-BsF/MKP1-gRNA2-BsR or MLO6-
gRNA1-BsF/MLO12-gRNA1-BsR by Q5 DNA polymerase (New England Biolabs,
M0491L). The primers are listed in Table 3.2. Each primer contains a gRNA (in
blue) and a BsaI recognition site (in red) for cloning. Subsequent cloning of this
cassette into pHEE401E was through Golden Gate reactions (mix 50 ng purified
cassette from PCR, 200 ng pHEE401E, 1.5µl 10xT4 DNA ligase buffer (Promega,
M1804), 1.5µl 10x BSA, 1µl BsaI (New England Biolabs, R0535S), 1µl T4 DNA
ligase (Promega, M1804), and add water up to 15µl); incubate for 5 hours at 37 ◦C,
then 5 min at 50 ◦C, and 10 min at 80 ◦C to terminate reaction). Successful recom-
binant clones were then used for Agrobacterium-mediated plant transformation.
3.5.6 Genotyping
F1 hybrids (cipi2×cipi3, cipi11×cipi3, cipi11×cipi12, cipi15×cipi12) of the
cipi-mlo2 mutants were confirmed by PCR-base genotyping. All primers used for
genotyping are listed in Table 3.2. To detect the mlo2 allele from cipi2, the fragment
containing the mutation was amplified by MLO2-e9F/MLO2-e11R followed by MwoI
125
digestion (New England Biolabs, R0573S). The wild type allele can be digested
(161+242 bp), while the cipi2 mutant allele is intact (403 bp). Similarly, to detect
the mlo2 allele from cipi3, the fragment containing the mutation was amplified
by MLO2-e2F/MLO2-e3R followed by HindIII digestion (New England Biolabs,
R3104S). While the wild type allele is intact (341 bp), the cipi3 mutant allele is
digested (179+162 bp). The rest mlo2 alleles from 11, 12, and 15 were amplified
with MLO2-e2F/MLO2-e3R, MLO2-e1F/MLO2-e3R, and MLO2-e6F/MLO2-e8R,
respectively, followed by sequencing to detect the respective mutations.
3.5.7 Genome Sequencing and Mapping
For whole-genome sequencing, 23 genomic DNA samples were isolated from
leaf tissues harvested from one pool of ∼50 F2 backcrossed individuals derived from
cipi1×eps and 22 the rest of the mutants listed in Table 3.1 using NucleoSpin©R
Plant II kit (MACHEREY-NAGEL, #740770.50). About 2µg of each DNA sample
was sent to BGI (Beijing Genomics Institute, Shenzhen, China) for deep sequencing
using Illumina Hiseq 2000 plantform at an average coverage of ∼50× generating
100 bp paired end reads. The subsequent sequence analysis was done according to
the method presented in (Wang et al., 2018). Briefly, sequencing reads were mapped
against the TAIR10 Arabidopsis reference genome using Bowtie (Langmead et al.,
2009) and variants were called using SAMtools (Li et al., 2009). Only G→A and
C→T conversions predominantly caused by EMS mutagenesis were picked for further
analysis. The effect of each EMS-induced SNP (single-nucleotide polymorphism) on
126
the corresponding gene was annotated using snpEffect (Cingolani et al., 2012). The
effects of the SNPs were classified into three categories: very high (STOP GAINED,
SPLICE SITE DONOR, SPLICE SITE ACCEPTOR), high (NON SYNONYMOUS
CODING, START GAINED, STOP LOST), and other (INTERGENIC, INTRON,
3PRIME UTR, 5PRIME UTR, SYNONYMOUS CODING). Coding and splicing
variants with high or very high effects were compared across different libraries.
Genes that have variants from multiple independent libraries were listed as candi-
date genes (Table 3.1).
3.5.8 Accession Numbers
Sequences of the genes used in this chapter can be found in the Arabidopsis In-
formation Resoruce (https://www.arabidopsis.org) or GenBank/EMBL (https:
//www.ncbi.nlm.nih.gov/genbank/) databases. The accession numbers of the
genes only used this Chapter are as follows: At3g55270 (MKP1), At1g11310 (MLO2),
At1g61560 (MLO6), and At2g39200 (MLO12).
127
Table 3.2: Primers used in Chapter 3
Purpose Primer ID Sequence (5’ → 3’)
MKP1-gRNA1-BsF ATATATGGTCTCGATTGACCTAGCGGGAACAAAACCGGTTTTAGAGCTAGAAATAGC
Making MKP1-gRNA2-BsR ATTATTGGTCTCGAAACCCGGTAAGGGGAGAGGTAAGCAATCTCTTAGTCGACTCTAC
CRISPR
constructs MLO6-gRNA1-BsF ATATATGGTCTCGATTGTGGTAAGAAAGATGTCCCAAGTTTTAGAGCTAGAAATAGC
MLO12-gRNA1-BsR ATTATTGGTCTCGAAACCGCCGAAATTTAGCCACCAACAATCTCTTAGTCGACTCTAC
MKP1-tpF caccATGGTGGGAAGAGAGGATGCGAT
Sanger
Sequencing- MKP1-242mR TATTACCAACGAGGTAAATCGTGAAACCCCTT
based
CRISPR- MLO6-tpF caccATGGCGGATCAAGTTAAAGAA
induced
Indel MLO6-e4R TGACAAACCGCAAGAACAAA
Detection
MLO12-tpF caccATGGCAATAAAAGAGCGATCA
MLO12-e4R TGCAGCTGGTGGATACCATA
MLO2-e1F TCAAAAGAAGAACACGAAACTCTG
MLO2-e2F GAAGCACAAGCAGGCTCTTTTT
MLO2-e3R GGGTTTATCTCCATCTCCATCATCTT
Genotyping MLO2-e6F GGGACACATCTTTTGGGAGA
MLO2-e8R GCACAGCGACAAACCAGATA
MLO2-e9F ACCTCTGGTTACCATTCATTCC
MLO2-e11R TCCAGGCAAAGAATGCAAGT
*Sequences in red are BsaI recognition site.
**Sequences in blue are gene-specific gRNA sequences.
128
Chapter 4: Conclusions and Future Directions
Food security, by definition, exits when all people, at all times, have physical,
social and economic access to sufficient, safe and nutritious food which meets their
dietary needs and food preferences for an active and healthy life (1996 World Food
Summit, Food and Nations, 2005). The global demand for food is projected to
increase by ∼70% to ensure food security for a population of 9.1 billion by 2050
(Alexandratos et al., 2012). Furthermore, factors such as climate change and biotic
and abiotic stresses continue to escalate food insecurity. For example, plant diseases
alone can cause an average of 15% loss of the global crop production at preharvest
stage every year (Dangl et al., 2013). The challenge for humanity is: can we produce
enough food to feed 9 billion people in 2050 with limited arable land and water
resources? The key to meeting this challenge is to improve agricultural performance.
Agriculture especially modern industrial crop production favors large-scale
monocropping, which can achieve higher yields while enabling easy management to
reduce labor, but also often increases the risk of epidemics of plant diseases/pests.
To protect crops from pathogens and insects in such agricultural settings, large
amounts of pesticides have to be used, which may cause human health problems
and impose cost to the environment. How to resolve such a complex issue? One
129
important approach is to breed crop cultivars with broad-spectrum and durable re-
sistance against constantly evolving pests to enable sustainable agriculture. Tradi-
tional breeding by introgression of R (resistance) genes to crops has been successful,
but at the same time tedious, time-consuming, and limited to sexually compatible
relatives. The advances of modern gene-editing technologies such as CRISPR has
largely removed these constraints, leaving the only major limitation to be identify-
ing the genes (and the mechanisms by which they function) controlling disease/pest
resistance in plants.
Tremendous progress has been made in the past two decades towards decipher-
ing the molecular mechanisms of the plant immune system. However, how plants
mount spatiotemporally appropriate defenses against pathogens, and how pathogens
overcome host immunity and extract nutrients from host cells to establish infection
are still not well characterized. During my PhD studies, I have been working towards
the identification and characterization of novel plant immune components using the
Arabidopsis-Powdery mildew pathosystem through both reverse and forward genet-
ics. Indeed, my work has revealed hidden plant immune components/layers against
powdery mildew fungi, which should contribute to a better understanding of the
multi-layered plant immune system and powdery mildew host adaptation.
In Chapter 2, through a reverse genetic screen to assess the involvement of
lipid metabolizing enzymes in plant immunity, PLDα1 and PLDδ have been found
to play opposing roles in modulating basal, post-penetration resistance against pow-
dery mildew through an uncharacterized mechanism that is independent of the key
immune components EDS1/PAD4, salicylic acid (SA), and jasmonic acid (JA).
130
Inspired by this finding, as detailed in Chapter 3, I designed and conducted
a forward genetic screen using a sensitive artificial pathosystem aiming to identify
novel/hidden immune components or host susceptible factors manipulated by pow-
dery mildew. This pathosystem, which consists of a super-susceptible Arabidopsis
eds1-2pad4-1sid2-2 (eps) triple mutant as host and powdery mildew species from
different hosts with different levels of adaptation on Arabidopsis as pathogen, has
enabled the identification of mutations that render either enhanced disease resis-
tance (edr) or enhanced disease susceptibility (eds) independent of the mechanisms
controlled by EDS1, PAD4 and SID2. A total of over 50 “edr” mutants named com-
promised immunity yet poor infection (cipi) and “eds” mutants named susceptible to
non-adapted PM (snap) were isolated and phenotypically characterized. Mapping
of the causal mutations for 23 non-segregating mutant lines (18 cipi and 5 snap
mutants) has so far led to the identification of MAP KINASE PHOSPHATASE1
(MKP1), MILDEW RESISTANCE LOCUS 2 (MLO2). While MKP1 is a nega-
tive regulator of a plant immune signaling pathway, MLO2 appears to be a critical
susceptibility factor of powdery mildew fungi, both of which act independently of
EDS1, PAD4, and SID2.
While the roles of MKP1 and MLO2 in plant immunity have been reported,
genetic data from my work added interesting and novel insights onto the underly-
ing mechanisms. More excitingly, eight additional candidate CIPI genes have been
identified and are currently under verification by CRISPR-based targeted mutage-
nesis (Table 3.1). In addition, it is worth pointing out that of the snap mutants,
snap1 could support visible infection from strawberry powdery mildew, a non-host
131
powdery mildew pathogen that cannot infect the eps parental line, suggesting that
the causal mutation results in breakdown of an additional layer of non-host resis-
tance in snap1. Hence, identification of SNAP1 may unravel a novel cell wall-based
defense mechanism plants use to fight against non- or poorly-adapted fungi.
The identification of novel immune components by my work hopefully will
enrich and broaden our current knowledge of the multi-layered plant innate immune
system. However, like many other scientific discoveries, these findings I made raise
more intriguing questions than they solve, which in turn fuels the progress of science,
getting us one step closer to the truth each time.
Specifically, the original findings I made in my thesis project invites the fol-
lowing questions for future studies. (1) Why do the two PLD isoforms with similar
enzymatic activity play opposing roles in plant immunity? (2) What is the molecular
and biochemical basis of their distinct biological functions? (3) Could any of the im-
mune components/mechanisms (to be) identified from the forward genetic screen be
connected to the PLD-mediated defense mechanisms? (4) How may the novel immu-
nity layers be integrated or connected with the well-characterized PAMP-triggered
immunity and effector-triggered immunity? (5) What roles do these new immune
components play in fighting against pathogens with different levels of adaptation
(i.e. host resistance versus non-host resistance)? (6) Whether and why are MLO
proteins recruited by powdery mildew fungi for host entry? In other words, what
are the molecular/cellular functions of MLO proteins in plants that are hijacked by
powdery mildew for breaching the host cell wall? (7) Are there other susceptibility
factors such as nutrient regulators/transporter critical for pathogenesis of powdery
132
mildew and other fungi?
Apparently, characterization of the additional CIPI and SNAP genes based on
their respective cipi and snap mutant phenotypes may help address one or more of
the above questions in the near future. Hopefully, new knowledge and genetic ma-
terials from my thesis work will not only fuel continuing research towards dissecting
the multi-layered plant resistance against powdery mildew in the Xiao lab but also
contribute to our understanding of plant innate immunity in general. Finally, with
rapid advances in the development of genome-editing tools, it is possible that some
of the new genes identified in my thesis project may be used for developing crop
cultivars with broad-spectrum and durable disease resistance via precision breeding,
thereby contributing to a sustainable agriculture much needed for ensuring food
security in the coming decades.
133
Appendix A: Co-authored Non-thesis Manuscripts Published or In-
preparation
A.1 Published Manuscripts
A.1.1
Homologues of the RPW8 Resistance Protein Are Localized to the Ex-
trahaustorial Membrane that Is Likely Synthesized De Novo
Robert Berkey, Yi Zhang, Xianfeng Ma, Harlan King, Qiong Zhang, Wenming
Wang, and Shunyuan Xiao
Plant physiology, pages pp01539, 2016.
Abstract
Upon penetration of the host cell wall, the powdery mildew fungus develops a feed-
ing structure named the haustorium in the invaded host cell. Concomitant with
haustorial biogenesis, the extrahaustorial membrane (EHM) is formed to separate
the haustorium from the host cell cytoplasm. The Arabidopsis resistance protein
RPW8.2 is specifically targeted to the EHM where it activates haustorium-targeted
resistance against powdery mildew. RPW8.2 belongs to a small family with six
members in Arabidopsis (Arabidopsis thaliana). Whether Homologs of RPW8 (HR)
134
1 to HR4 are also localized to the EHM and contribute to resistance has not been de-
termined. Here, we report that overexpression of HR1, HR2, or HR3 led to enhanced
resistance to powdery mildew, while genetic depletion of HR2 or HR3 resulted in
enhanced susceptibility, indicating that these RPW8 homologs contribute to basal
resistance. Interestingly, we found that N-terminally YFP-tagged HR1 to HR3 are
also EHM-localized. This suggests that EHM-targeting is an ancestral feature of
the RPW8 family. Indeed, two RPW8 homologs from Brassica oleracea tested also
exhibit EHM-localization. Domain swapping analysis between HR3 and RPW8.2
suggests that sequence diversification in the N-terminal 146 amino acids of RPW8.2
probably functionally distinguishes it from other family members. Moreover, we
found that N-terminally YFP-tagged HR3 is also localized to the plasma membrane
and the fungal penetration site (the papilla) in addition to the EHM. Using this
unique feature of YFP-HR3, we obtained preliminary evidence to suggest that the
EHM is unlikely derived from invagination of the plasma membrane, rather it may
be mainly synthesized de novo.
135
A.1.2
Dual and Opposing Roles of Xanthine Dehydrogenase in Defense-Associated
Reactive Oxygen Species Metabolism in Arabidopsis
Xianfeng Ma, Wenming Wang, Florian Bittner, Nadine Schmidt, Robert Berkey,
Lingli Zhang, Harlan King, Yi Zhang, Jiayue Feng, Yinqiang Wen, Liqiang Tan,
Yue Li, Qiong Zhang, Ziniu Deng, Xingyao Xiong, Shunyuan Xiao
The Plant Cell, 28(5): 1108-1126, 2016
Abstract
While plants produce reactive oxygen species (ROS) for stress signaling and pathogen
defense, they need to remove excessive ROS induced during stress responses in or-
der to minimize oxidative damage. How can plants fine-tune this balance and meet
such conflicting needs? Here, we show that XANTHINE DEHYDROGENASE1
(XDH1) in Arabidopsis thaliana appears to play spatially opposite roles to serve
this purpose. Through a large-scale genetic screen, we identified three missense
mutations in XDH1 that impair XDH1s enzymatic functions and consequently af-
fect the powdery mildew resistance mediated by RESISTANCE TO POWDERY
MILDEW8 (RPW8) in epidermal cells and formation of xanthine-enriched autoflu-
orescent objects in mesophyll cells. Further analyses revealed that in leaf epidermal
cells, XDH1 likely functions as an oxidase, along with the NADPH oxidases RbohD
and RbohF, to generate superoxide, which is dismutated into H2O2. The resulting
enrichment of H2O2 in the fungal haustorial complex within infected epidermal cells
helps to constrain the haustorium, thereby contributing to RPW8-dependent and
136
RPW8-independent powdery mildew resistance. By contrast, in leaf mesophyll cells,
XDH1 carries out xanthine dehydrogenase activity to produce uric acid in local and
systemic tissues to scavenge H2O2 from stressed chloroplasts, thereby protecting
plants from stress-induced oxidative damage. Thus, XDH1 plays spatially specified
dual and opposing roles in modulation of ROS metabolism during defense responses
in Arabidopsis.
137
A.1.3
Lipids in Salicylic Acid-mediated Defense in Plants: Focusing on the
Roles of Phosphatidic Acid and Phosphatidylinositol 4-Phosphate
Qiong Zhang and Shunyuan Xiao
Frontiers in Plant Science, 6(387), 2015
Abstract
Plants have evolved effective defense strategies to protect themselves from vari-
ous pathogens. Salicylic acid (SA) is an essential signaling molecule that mediates
pathogen-triggered signals perceived by different immune receptors to induce down-
stream defense responses. While many proteins play essential roles in regulating
SA signaling, increasing evidence also supports important roles for signaling phos-
pholipids in this process. In this review, we collate the experimental evidence in
support of the regulatory roles of two phospholipids, phosphatidic acid (PA) and
phosphatidylinositol 4-phosphate (PI4P), and their metabolizing enzymes in plant
defense, and examine the possible mechanistic interaction between phospholipid
signaling and SA-dependent immunity with a particular focus on the immunity-
stimulated biphasic PA production that is reminiscent of and perhaps mechanisti-
cally connected to the biphasic reactive oxygen species (ROS) generation and SA
accumulation during defense activation.
138
A.1.4
Dominant Negative RPW8.2 Fusion Proteins Reveal the Importance of
Haustorium-oriented Protein Trafficking for Resistance against Powdery
Mildew in Arabidopsis
Qiong Zhang, Robert Berkey, Zhiyong Pan, Wenming Wang, Yi Zhang, Xianfeng
Ma, Harlan King and Shunyuan Xiao
Plant Signaling & Behavior, 10(3):e989766, 2015
Abstract
Powdery mildew fungi form feeding structures called haustoria inside epidermal cells
of host plants to extract photosynthates for their epiphytic growth and reproduc-
tion. The haustorium is encased by an interfacial membrane termed the extrahaus-
torial membrane (EHM). The atypical resistance protein RPW8.2 from Arabidopsis
is specifically targeted to the EHM where RPW8.2 activates haustorium-targeted
(thus broad-spectrum) resistance against powdery mildew fungi. EHM-specific lo-
calization of RPW8.2 suggests the existence of an EHM-oriented protein/membrane
trafficking pathway during EHM biogenesis. However, the importance of this specific
trafficking pathway for host defense has not been evaluated via a genetic approach
without affecting other trafficking pathways. Here, we report that expression of
EHM-oriented, nonfunctional RPW8.2 chimeric proteins exerts dominant negative
effect over functional RPW8.2 and potentially over other EHM-localized defense pro-
teins, thereby compromising both RPW8.2- mediated and basal resistance to pow-
dery mildew. Thus, our results highlight the importance of the EHM-oriented pro-
139
tein/membrane trafficking pathway for host resistance against haustorium-forming
pathogens such as powdery mildew fungi.
140
A.1.5
A Comprehensive Mutational Analysis of the Arabidopsis Resistance
Protein RPW8.2 Reveals Key Amino Acids for Defense Activation and
Protein Targeting
Wenming Wang, Yi Zhang, Yingqiang Wen, Robert Berkey, Xianfeng Ma, Zhiyong
Pan, Dipti Bendigeri, Harlan King, Qiong Zhang, and Shunyuan Xiao
The Plant Cell, 25(10):4242-4261, 2013
Abstract
The Arabidopsis thaliana RESISTANCE TO POWDERY MILDEW8.2 (RPW8.2)
protein is specifically targeted to the extrahaustorial membrane (EHM) encasing the
haustorium, or fungal feeding structure, where RPW8.2 activates broad-spectrum
resistance against powdery mildew pathogens. How RPW8.2 activates defenses at a
precise subcellular locale is not known. Here, we report a comprehensive mutational
analysis in which more than 100 RPW8.2 mutants were functionally evaluated for
their defense and trafficking properties. We show that three amino acid residues (i.e.,
threonine-64, valine-68, and aspartic acid-116) are critical for RPW8.2-mediated cell
death and resistance to powdery mildew (Golovinomyces cichoracearum UCSC1).
Also, we reveal that two arginine (R) or lysine (K)enriched short motifs (i.e., R/K-
R/K-x-R/K) make up the likely core EHM-targeting signals, which, together with
the N-terminal transmembrane domain, define a minimal sequence of 60 amino acids
that is necessary and sufficient for EHM localization. In addition, some RPW8.2
mutants localize to the nucleus and/or to a potentially novel membrane that wraps
141
around plastids or plastid-derived stromules. Results from this study not only reveal
critical amino acid elements in RPW8.2 that enable haustorium-targeted trafficking
and defense, but also provide evidence for the existence of a specific, EHM-oriented
membrane trafficking pathway in leaf epidermal cells invaded by powdery mildew.
142
A.2 Manuscripts In Preparation
A.2.1
Comparative Genome Analyses Reveal Sequence Features Reflecting Dis-
tinct Modes of Host-adaptation between Dicot and Monocot Powdery
Mildew
Ying Wu, Xianfeng Ma, Zhiyong Pan, Shiv D. Kale, Yi Song, Harley King, Qiong Zhang,
Christian Presley, Xiuxin Deng, Cheng-I Wei, and Shunyuan Xiao
Abstract
Background: Powdery mildew (PM) is one of the most important and widespread
plant diseases caused by biotrophic fungi. Notably, while monocot PM fungi exhibit
high-level of host-specialization, many dicot PM fungi display a broad host range.
To understand such distinct modes of host-adaptation, we sequenced the genomes
of four dicot PM biotypes belonging to Golovinomyces cichoracearum or Oidium
neolycopersici. Results: We compared genomes of the four dicot PM together with
those of Blumeria graminis f.sp. hordei, B. graminis f.sp. tritici, and Erysiphe neca-
tor infectious on barley, wheat and grapevine, respectively. We found that despite
having a similar gene number (∼6500), the PM genomes vary from 120 to 222 Mb
in size. This high-level of genome size variation is indicative of highly differential
transposon activities in the PM genomes. While the total number of genes in any
given PM genome is only about half of that in the genomes of closely related as-
comycete fungi, most (∼93%) of the ascomycete core genes (ACGs) can be found in
143
the PM genomes. Yet, 186 ACGs were found absent in at least two of the seven PM
genomes, of which 35 are missing in some dicot PM biotypes, but present in the two
monocot PM fungi, indicating remarkable, independent and perhaps ongoing gene
loss in different PM lineages. Consistent with this, we found that only 3129 (2889
singular) genes are shared by all the seven PM genomes, the remaining genes are
lineage- or biotype-specific. Strikingly, whereas the two monocot PM fungi possess
up to 534 genes encoding candidate secreted effector proteins (CSEPs) with fami-
lies containing up to 21 members, all the five dicot PM fungi have only 116 - 178
genes encoding CSEPs with limited gene amplification. Conclusions: Compared
to monocot PM fungi, dicot PM fungi have a much smaller effectorome. This is
consistent with their contrasting modes of host-adaption: while the monocot PM
fungi show a high-level of host specialization, which may reflect an advanced host-
pathogen arms-race, the dicot PM fungi tend to practice polyphagy, which might
have lessened selective pressure for escalating an arms-race with a particular host.
144
A.2.2
Preferential Binding of Phosphatidylinositol 3-Phosphate Establishes Lo-
calization Specificity of RPW8.2 to the Extrahaustorial Membrane
Qiong Zhang, Shiv Kale, Yi Zhang, Harley King, Xianfeng Ma, Robert Berkey, and
Shunyuan Xiao
Abstract
The broad-spectrum disease resistance protein RPW8.2 from Arabidopsis is specif-
ically targeted to the extrahaustorial membrane (EHM) to constrain the fungal
feeding structure. A previous study identified two basic residue-enriched EHM-
targeting motifs in RPW8.2 to be critical for EHM-targeting. However, the un-
derlying molecular mechanism is not known. It has been suggested that phospho-
inositides are subcellular landmarks; their asymmetric distribution may determine
the polarity of membrane/protein trafficking. Using biosensors (i.e. fluorescent
proteins that bind a specific phosphoinositide), we found that while phosphatidyli-
nositol 4,5-bisphosphate (PI(4,5)P2) is enriched in the plasma membrane in plant
cells, phosphatidylinositol 3-phosphate (PI3P) is enriched in the tonoplast tightly
wrapping around the EHM or possibly enriched the EHM itself. These observations
suggest a potential role of PI3P in EHM biogenesis and/or EHM-oriented traffick-
ing. Interestingly, lipid filter binding assays showed that RPW8.2 preferentially
binds PI3P among all phosphoinositides. Subsequent liposome binding assays also
demonstrated that RPW8.2 binds PI3P-containing liposomes with a much higher
145
affinity (with a dissociation constant of ∼300 nM) compared to PI4P (2µM) or PI5P
(20muM) determined by surface plasmon resonance using RPW8.2 expressed from
Pichia pastoris. Importantly, RPW8.2 mutants in which the two EHM-targeting
motifs are replaced by six amino acids NAAIRS showed significantly lower binding
affinity to PI3P-containing liposomes, supporting an important role of PI3P bind-
ing in RPW8.2s EHM-targeting. Consistent with this, expression of a dominant
negative form of phosphoinositide-3-OH kinase (VPS34; At1g60490; PI3K) appears
to interfere with EHM-targeting of RPW8.2. Taken together, our results suggest
that preferential binding of PI3P contributes to RPW8.2s specific localization to
the EHM and that PI3P may play a role in EHM biogenesis and/or EHM-oriented
vesicle trafficking.
146
A.2.3
A Qb SNARE at the TGN is Essential for RPW8.2-mediated Powdery
Milew Resistance
Xianfeng Ma, Qiong Zhang, Ying Wu, Harley King, Marcela Rojas-Pierce, Xingyao
Xiong, and Shunyuan Xiao
Abstract
Many fungal and oomycete pathogens use similar feeding structures called haustoria
to secrete effectors to suppress host immunity and steal nutrients from plant cells.
The extrahaustorial membrane (EHM) represents the host-pathogen interface and
a critical battleground. The broad-spectrum resistance protein RPW8.2 from Ara-
bidopsis is specifically targeted to the EHM where it activates haustorium-focused
defenses. The origin and biogenesis of the EHM is not known. By using RPW8.2
and other RPW8 family members as tools, we demonstrated that the EHM is most
likely synthesized de novo, which is consistent with our observation that RPW8.2
is targeted via the Trans Golgi Network (TGN) to the EHM during EHM bio-
genesis. To understand how RPW8.2 is precisely targeted to the EHM, we have
recently conducted a large-scale genetic screen and found that loss of VTI11, a Qb
SNARE at the TGN required for vesicle trafficking from the TGN to the vacuole,
results in mis-targeting of RPW8.2 to the plasma membrane and loss of resistance
to powdery mildew. Interestingly, vti11 mutant plants in the absence of RPW8.2
exhibits enhanced disease resistance to adapted powdery mildew, suggesting that a
147
TGN-localized SNARE complex containing VTI111 is essential for correct sorting
of RPW8.2 and perhaps other membrane proteins involved in plant immunity at the
TGN.
148
Bibliography
Benchling [biology software]. retrieved from https://benchling.com., 2017. URL
https://benchling.com.
Aarts, N, Metz, M, Holub, E, Staskawicz, B. J, Daniels, M. J, and Parker, J. E.
Different requirements for eds1 and ndr1 by disease resistance genes define at
least two r gene-mediated signaling pathways in arabidopsis. Proceedings of the
National Academy of Sciences, 95(17):10306–10311, 1998.
Acevedo-Garcia, J, Gruner, K, Reinstädler, A, Kemen, A, Kemen, E, Cao, L,
Takken, F. L, Reitz, M. U, Schäfer, P, O’Connell, R. J, et al. The powdery
mildew-resistant arabidopsis mlo2 mlo6 mlo12 triple mutant displays altered in-
fection phenotypes with diverse types of phytopathogens. Scientific reports, 7(1):
9319, 2017.
Adam, L and Somerville, S. C. Genetic characterization of five powdery mildew
disease resistance loci in arabidopsis thaliana. The Plant Journal, 9(3):341–356,
1996.
Alexandratos, N, Bruinsma, J, et al. World agriculture towards 2030/2050: the 2012
revision. Technical report, ESA Working paper FAO, Rome, 2012.
Anderson, J. C, Bartels, S, Besteiro, M. A. G, Shahollari, B, Ulm, R, and Peck,
S. C. Arabidopsis map kinase phosphatase 1 (atmkp1) negatively regulates mpk6-
mediated pamp responses and resistance against bacteria. The Plant Journal, 67
(2):258–268, 2011.
Asai, T, Tena, G, Plotnikova, J, Willmann, M. R, Chiu, W.-L, Gomez-Gomez,
L, Boller, T, Ausubel, F. M, and Sheen, J. Map kinase signalling cascade in
arabidopsis innate immunity. Nature, 415(6875):977, 2002.
Assaad, F. F, Qiu, J.-L, Youngs, H, Ehrhardt, D, Zimmerli, L, Kalde, M, Wanner,
G, Peck, S. C, Edwards, H, Ramonell, K, et al. The pen1 syntaxin defines a novel
cellular compartment upon fungal attack and is required for the timely assembly
of papillae. Molecular biology of the cell, 15(11):5118–5129, 2004.
149
Axtell, M. J and Staskawicz, B. J. Initiation of rps2-specified disease resistance in
arabidopsis is coupled to the avrrpt2-directed elimination of rin4. Cell, 112(3):
369–377, 2003.
Baker, K. E and Parker, R. Nonsense-mediated mrna decay: terminating erroneous
gene expression. Current opinion in cell biology, 16(3):293–299, 2004.
Bargmann, B. O and Munnik, T. The role of phospholipase D in plant stress
responses. Current opinion in plant biology, 9(5):515–522, 2006.
Bargmann, B. O, Laxalt, A. M, Riet, B. t, Schouten, E, Van Leeuwen, W, Dekker,
H. L, De Koster, C. G, Haring, M. A, and Munnik, T. Lepldβ1 activation and
relocalization in suspension-cultured tomato cells treated with xylanase. The
Plant Journal, 45(3):358–368, 2006.
Bari, R and Jones, J. D. Role of plant hormones in plant defence responses. Plant
molecular biology, 69(4):473–488, 2009.
Bartels, S, Anderson, J. C, Besteiro, M. A. G, Carreri, A, Hirt, H, Buchala, A,
Métraux, J.-P, Peck, S. C, and Ulm, R. Map kinase phosphatase1 and protein
tyrosine phosphatase1 are repressors of salicylic acid synthesis and snc1-mediated
responses in arabidopsis. The Plant Cell, 21(9):2884–2897, 2009.
Bartels, S, Besteiro, M. A. G, Lang, D, and Ulm, R. Emerging functions for plant
map kinase phosphatases. Trends in plant science, 15(6):322–329, 2010.
Bartsch, M, Gobbato, E, Bednarek, P, Debey, S, Schultze, J. L, Bautor, J, and
Parker, J. E. Salicylic acid–independent enhanced disease susceptibility1 signaling
in arabidopsis immunity and cell death is regulated by the monooxygenase fmo1
and the nudix hydrolase nudt7. The Plant Cell Online, 18(4):1038–1051, 2006.
Bednarek, P, Píslewska-Bednarek, M, Svatoš, A, Schneider, B, Doubskỳ, J,
Mansurova, M, Humphry, M, Consonni, C, Panstruga, R, Sanchez-Vallet, A,
Molina, A, and Schulze-Lefert, P. A glucosinolate metabolism pathway in liv-
ing plant cells mediates broad-spectrum antifungal defense. Science, 323(5910):
101–106, 2009.
Behnia, R and Munro, S. Organelle identity and the signposts for membrane traffic.
Nature, 438(7068):597, 2005.
Berkey, R, Zhang, Y, Ma, X, King, H, Zhang, Q, Wang, W, and Xiao, S. Homologues
of the rpw8 resistance protein are localized to the extra-haustorial membrane that
is likely synthesized de novo. Plant physiology, pages pp–01539, 2016.
Bernoux, M, Moncuquet, P, Kroj, T, Dodds, P. N, et al. A novel conserved mech-
anism for plant nlr protein pairs: the ‘integrated decoy’hypothesis. Frontiers in
plant science, 5:606, 2014.
150
Bhat, R. A, Miklis, M, Schmelzer, E, Schulze-Lefert, P, and Panstruga, R. Recruit-
ment and interaction dynamics of plant penetration resistance components in a
plasma membrane microdomain. Proceedings of the National Academy of Sciences
of the United States of America, 102(8):3135–3140, 2005.
Bhullar, N. K, Zhang, Z, Wicker, T, and Keller, B. Wheat gene bank accessions as
a source of new alleles of the powdery mildew resistance gene pm3: a large scale
allele mining project. BMC plant biology, 10(1):88, 2010.
Bi, G, Zhou, Z, Wang, W, Li, L, Rao, S, Wu, Y, Zhang, X, Menke, F. L, Chen,
S, and Zhou, J.-M. Receptor-like cytoplasmic kinases directly link diverse pat-
tern recognition receptors to the activation of mitogen-activated protein kinase
cascades in arabidopsis. The Plant Cell, pages tpc–00981, 2018.
Blakeslee, J. J and Murphy, A. S. Microscopic and biochemical visualization of aux-
ins in plant tissues. Environmental Responses in Plants: Methods and Protocols,
pages 37–53, 2016.
Bonardi, V, Tang, S, Stallmann, A, Roberts, M, Cherkis, K, and Dangl, J. L.
Expanded functions for a family of plant intracellular immune receptors beyond
specific recognition of pathogen effectors. Proceedings of the National Academy of
Sciences, 108(39):16463–16468, 2011.
Boutrot, F and Zipfel, C. Function, discovery, and exploitation of plant pattern
recognition receptors for broad-spectrum disease resistance. Annual review of
Phytopathology, 55:257–286, 2017.
Büschges, R, Hollricher, K, Panstruga, R, Simons, G, Wolter, M, Frijters, A, van
Daelen, R, van der Lee, T, Diergaarde, P, Groenendijk, J, et al. The barley mlo
gene: a novel control element of plant pathogen resistance. Cell, 88(5):695–705,
1997.
Chanda, B, Xia, Y, Mandal, M. K, Yu, K, Sekine, K.-T, Gao, Q.-m, Selote, D,
Hu, Y, Stromberg, A, Navarre, D, et al. Glycerol-3-phosphate is a critical mobile
inducer of systemic immunity in plants. Nature genetics, 43(5):421, 2011.
Chaturvedi, R, Venables, B, Petros, R. A, Nalam, V, Li, M, Wang, X, Takemoto,
L. J, and Shah, J. An abietane diterpenoid is a potent activator of systemic
acquired resistance. The Plant Journal, 71(1):161–172, 2012.
Chen, S, Songkumarn, P, Liu, J, and Wang, G.-L. A versatile zero background
t-vector system for gene cloning and functional genomics. Plant physiology, 150
(3):1111–1121, 2009.
Choi, J, Tanaka, K, Cao, Y, Qi, Y, Qiu, J, Liang, Y, Lee, S. Y, and Stacey, G.
Identification of a plant receptor for extracellular atp. Science, 343(6168):290–
294, 2014.
151
Chung, E.-H, Da Cunha, L, Wu, A.-J, Gao, Z, Cherkis, K, Afzal, A. J, Mackey,
D, and Dangl, J. L. Specific threonine phosphorylation of a host target by two
unrelated type iii effectors activates a host innate immune receptor in plants. Cell
host & microbe, 9(2):125–136, 2011.
Cingolani, P, Platts, A, Wang, L. L, Coon, M, Nguyen, T, Wang, L, Land, S. J, Lu,
X, and Ruden, D. M. A program for annotating and predicting the effects of single
nucleotide polymorphisms, snpeff: Snps in the genome of drosophila melanogaster
strain w1118; iso-2; iso-3. Fly, 6(2):80–92, 2012.
Clauss, M. J, Dietel, S, Schubert, G, and Mitchell-Olds, T. Glucosinolate and
trichome defenses in a natural arabidopsis lyrata population. Journal of chemical
ecology, 32(11):2351–2373, 2006.
Clough, S. J and Bent, A. F. Floral dip: a simplified method foragrobacterium-
mediated transformation ofarabidopsis thaliana. The plant journal, 16(6):735–
743, 1998.
Collier, S. M, Hamel, L.-P, and Moffett, P. Cell death mediated by the n-terminal
domains of a unique and highly conserved class of nb-lrr protein. Molecular plant-
microbe interactions, 24(8):918–931, 2011.
Collins, N. C, Thordal-Christensen, H, Lipka, V, Bau, S, Kombrink, E, Qiu, J.-
L, Hückelhoven, R, Stein, M, Freialdenhoven, A, Somerville, S. C, and Schulze-
Lefert, P. Snare-protein-mediated disease resistance at the plant cell wall. Nature,
425(6961):973–977, 2003.
Consonni, C, Humphry, M. E, Hartmann, H. A, Livaja, M, Durner, J, Westphal, L,
Vogel, J, Lipka, V, Kemmerling, B, Schulze-Lefert, P, et al. Conserved requirement
for a plant host cell protein in powdery mildew pathogenesis. Nature genetics, 38
(6):716, 2006.
Consonni, C, Bednarek, P, Humphry, M, Francocci, F, Ferrari, S, Harzen, A, van
Themaat, E. V. L, and Panstruga, R. Tryptophan-derived metabolites are re-
quired for antifungal defense in the arabidopsis mlo2 mutant. Plant physiology,
152(3):1544–1561, 2010.
Cui, H, Tsuda, K, and Parker, J. E. Effector-triggered immunity: from pathogen
perception to robust defense. Annual Review of Plant Biology, 66:487–511, 2015.
Cui, H, Gobbato, E, Kracher, B, Qiu, J, Bautor, J, and Parker, J. E. A core function
of eds1 with pad4 is to protect the salicylic acid defense sector in arabidopsis
immunity. New Phytologist, 213(4):1802–1817, 2017.
Dangl, J. L and Jones, J. D. Plant pathogens and integrated defence responses to
infection. Nature, 411(6839):826–833, 2001.
Dangl, J. L, Horvath, D. M, and Staskawicz, B. J. Pivoting the Plant Immune
System from Dissection to Deployment. Science, 341(6147):746–751, 2013.
152
de Torres Zabela, M, Fernandez-Delmond, I, Niittyla, T, Sanchez, P, and Grant, M.
Differential expression of genes encoding Arabidopsis phospholipases after chal-
lenge with virulent or avirulent Pseudomonas isolates. Molecular plant-microbe
interactions, 15(8):808–816, 2002.
Devoto, A, Piffanelli, P, Nilsson, I, Wallin, E, Panstruga, R, von Heijne, G, and
Schulze-Lefert, P. Topology, subcellular localization, and sequence diversity of
the mlo family in plants. Journal of Biological Chemistry, 274(49):34993–35004,
1999.
Dewdney, J, Reuber, T. L, Wildermuth, M. C, Devoto, A, Cui, J, Stutius, L. M,
Drummond, E. P, and Ausubel, F. M. Three unique mutants of arabidopsis
identify eds loci required for limiting growth of a biotrophic fungal pathogen. The
Plant Journal, 24(2):205–218, 2000.
Di Paolo, G and De Camilli, P. Phosphoinositides in cell regulation and membrane
dynamics. Nature, 443(7112):651, 2006.
Ding, Y, Sun, T, Ao, K, Peng, Y, Zhang, Y, Li, X, and Zhang, Y. Opposite roles of
salicylic acid receptors npr1 and npr3/npr4 in transcriptional regulation of plant
immunity. Cell, 2018.
Eliáš, M, Potockỳ, M, Cvrčková, F, and Žárskỳ, V. Molecular diversity of phospho-
lipase d in angiosperms. BMC genomics, 3(1):2, 2002.
Ellis, J. G, Dodds, P. N, and Lawrence, G. J. Flax rust resistance gene speci-
ficity is based on direct resistance-avirulence protein interactions. Annu. Rev.
Phytopathol., 45:289–306, 2007.
Falk, A, Feys, B. J, Frost, L. N, Jones, J. D, Daniels, M. J, and Parker, J. E. Eds1,
an essential component of r gene-mediated disease resistance in arabidopsis has
homology to eukaryotic lipases. Proceedings of the National Academy of Sciences,
96(6):3292–3297, 1999.
Feechan, A, Anderson, C, Torregrosa, L, Jermakow, A, Mestre, P, Wiedemann-
Merdinoglu, S, Merdinoglu, D, Walker, A. R, Cadle-Davidson, L, Reisch, B, et al.
Genetic dissection of a tir-nb-lrr locus from the wild north american grapevine
species muscadinia rotundifolia identifies paralogous genes conferring resistance to
major fungal and oomycete pathogens in cultivated grapevine. The Plant Journal,
76(4):661–674, 2013.
Feys, B. J, Moisan, L. J, Newman, M.-A, and Parker, J. E. Direct interaction
between the arabidopsis disease resistance signaling proteins, eds1 and pad4. The
EMBO journal, 20(19):5400–5411, 2001.
Feys, B. J, Wiermer, M, Bhat, R. A, Moisan, L. J, Medina-Escobar, N, Neu, C,
Cabral, A, and Parker, J. E. Arabidopsis senescence-associated gene101 stabilizes
and signals within an enhanced disease susceptibility1 complex in plant innate
immunity. The Plant Cell, 17(9):2601–2613, 2005.
153
Figueroa, M, Hammond-Kosack, K. E, and Solomon, P. S. A review of wheat
diseases—a field perspective. Molecular plant pathology, 19(6):1523–1536, 2018.
Flor, H. H. Current status of the gene-for-gene concept. Annual review of phy-
topathology, 9(1):275–296, 1971.
Food and Nations, A. O. O. T. U. Trade reforms and food security: Conceptualizing
the linkages. Food & Agriculture Organi, 2005.
Frerigmann, H, Böttcher, C, Baatout, D, and Gigolashvili, T. Glucosinolates are
produced in trichomes of arabidopsis thaliana. Frontiers in plant science, 3:242,
2012.
Frye, C. A, Tang, D, and Innes, R. W. Negative regulation of defense responses in
plants by a conserved MAPKK kinase. Proceedings of the National Academy of
Sciences, 98(1):373–378, 2001.
Fu, Z. Q and Dong, X. Systemic acquired resistance: turning local infection into
global defense. Annual review of plant biology, 64:839–863, 2013.
Fu, Z. Q, Yan, S, Saleh, A, Wang, W, Ruble, J, Oka, N, Mohan, R, Spoel, S. H,
Tada, Y, Zheng, N, et al. Npr3 and npr4 are receptors for the immune signal
salicylic acid in plants. Nature, 486(7402):228, 2012.
Gao, M, Liu, J, Bi, D, Zhang, Z, Cheng, F, Chen, S, and Zhang, Y. Mekk1,
mkk1/mkk2 and mpk4 function together in a mitogen-activated protein kinase
cascade to regulate innate immunity in plants. Cell research, 18(12):1190, 2008.
Garćıa, A. V, Blanvillain-Baufumé, S, Huibers, R. P, Wiermer, M, Li, G, Gobbato,
E, Rietz, S, and Parker, J. E. Balanced nuclear and cytoplasmic activities of eds1
are required for a complete plant innate immune response. PLoS pathogens, 6(7):
e1000970, 2010.
Gil, F and Gay, J. Ultrastructural and physiological properties of the host interfacial
components of haustoria of erysiphe pisi in vivo and in vitro. Physiological Plant
Pathology, 10(1):1–12, 1977.
Glazebrook, J. Contrasting mechanisms of defense against biotrophic and
necrotrophic pathogens. Annu. Rev. Phytopathol., 43:205–227, 2005.
Gómez-Gómez, L and Boller, T. Fls2: an lrr receptor–like kinase involved in the
perception of the bacterial elicitor flagellin in arabidopsis. Molecular cell, 5(6):
1003–1011, 2000.
Gonsalves, D and Ferreira, S. Transgenic papaya: a case for managing risks of
papaya ringspot virus in hawaii. Plant health progress, 10, 2003.
Grant, M. R, Godiard, L, Straube, E, Ashfield, T, Lewald, J, Sattler, A, Innes,
R. W, and Dangl, J. L. Structure of the arabidopsis rpm1 gene enabling dual
specificity disease resistance. Science, 269(5225):843–846, 1995.
154
Guo, L, Devaiah, S. P, Narasimhan, R, Pan, X, Zhang, Y, Zhang, W, and Wang, X.
Cytosolic glyceraldehyde-3-phosphate dehydrogenases interact with phospholipase
Dδ to transduce hydrogen peroxide signals in the Arabidopsis response to stress.
The Plant Cell Online, 24(5):2200–2212, 2012.
Hann, D. R and Rathjen, J. P. Early events in the pathogenicity of pseudomonas
syringae on nicotiana benthamiana. The Plant Journal, 49(4):607–618, 2007.
Hartmann, M, Zeier, T, Bernsdorff, F, Reichel-Deland, V, Kim, D, Hohmann, M,
Scholten, N, Schuck, S, Bräutigam, A, Hölzel, T, et al. Flavin monooxygenase-
generated n-hydroxypipecolic acid is a critical element of plant systemic immunity.
Cell, 173(2):456–469, 2018.
He, Y, Zhou, J, Shan, L, and Meng, X. Plant cell surface receptor-mediated
signaling–a common theme amid diversity. J Cell Sci, 131(2):jcs209353, 2018.
Hillmer, R. A, Tsuda, K, Rallapalli, G, Asai, S, Truman, W, Papke, M. D, Sakak-
ibara, H, Jones, J. D, Myers, C. L, and Katagiri, F. The highly buffered arabidop-
sis immune signaling network conceals the functions of its components. PLoS
genetics, 13(5):e1006639, 2017.
Hong, Y, Zhao, J, Guo, L, Kim, S.-C, Deng, X, Wang, G, Zhang, G, Li, M, and
Wang, X. Plant phospholipases d and c and their diverse functions in stress
responses. Progress in lipid research, 2016.
Hu, Y, Li, Y, Hou, F, Wan, D, Cheng, Y, Han, Y, Gao, Y, Liu, J, Guo, Y, Xiao,
S, et al. Ectopic expression of arabidopsis broad-spectrum resistance gene rpw8.
2 improves the resistance to powdery mildew in grapevine (vitis vinifera). Plant
Science, 267:20–31, 2018.
Hückelhoven, R and Panstruga, R. Cell biology of the plant–powdery mildew inter-
action. Current opinion in plant biology, 14(6):738–746, 2011.
Hückelhoven, R, Trujillo, M, and Kogel, K.-H. Mutations in ror1 and ror2 genes
cause modification of hydrogen peroxide accumulation in mlo-barley under attack
from the powdery mildew fungus. Molecular plant pathology, 1(5):287–292, 2000.
Huh, S. U, Cevik, V, Ding, P, Duxbury, Z, Ma, Y, Tomlinson, L, Sarris, P. F,
and Jones, J. D. Protein-protein interactions in the rps4/rrs1 immune receptor
complex. PLoS pathogens, 13(5):e1006376, 2017.
Humphry, M, Reinstaedler, A, Ivanov, S, Bisseling, T, and Panstruga, R. Durable
broad-spectrum powdery mildew resistance in pea er1 plants is conferred by natu-
ral loss-of-function mutations in psmlo1. Molecular plant pathology, 12(9):866–878,
2011.
Ichimura, K, Casais, C, Peck, S. C, Shinozaki, K, and Shirasu, K. Mekk1 is required
for mpk4 activation and regulates tissue-specific and temperature-dependent cell
death in arabidopsis. Journal of Biological Chemistry, 281(48):36969–36976, 2006.
155
James, G. V, Patel, V, Nordström, K. J, Klasen, J. R, Salomé, P. A, Weigel, D, and
Schneeberger, K. User guide for mapping-by-sequencing in arabidopsis. Genome
Biol, 14(6):R61, 2013.
Jiang, L, Anderson, J. C, Besteiro, M. A. G, and Peck, S. C. Phosphorylation of
arabidopsis map kinase phosphatase 1 (mkp1) is required for pamp responses and
resistance against bacteria. Plant physiology, 175(4):1839–1852, 2017.
Jirage, D, Tootle, T. L, Reuber, T. L, Frost, L. N, Feys, B. J, Parker, J. E, Ausubel,
F. M, and Glazebrook, J. Arabidopsis thaliana pad4 encodes a lipase-like gene
that is important for salicylic acid signaling. Proceedings of the National Academy
of Sciences, 96(23):13583–13588, 1999.
Johansson, O. N, Fahlberg, P, Karimi, E, Nilsson, A. K, Ellerström, M, and Anders-
son, M. X. Redundancy among phospholipase d isoforms in resistance triggered by
recognition of the pseudomonas syringae effector avrrpm1 in arabidopsis thaliana.
Frontiers in plant science, 5, 2014.
Jones, D. S, Yuan, J, Smith, B, Willoughby, A, Kumimoto, E. L, and Kessler,
S. A. Mildew resistance locus function in pollen-ovary reception is linked to its
oligomerization and subcellular. Plant physiology, pages pp–00523, 2017.
Jones, J. D and Dangl, J. L. The plant immune system. Nature, 444(7117):323–329,
2006.
Jones, J. D, Vance, R. E, and Dangl, J. L. Intracellular innate immune surveillance
devices in plants and animals. Science, 354(6316):aaf6395, 2016.
Jørgensen, I. H. Discovery, characterization and exploitation of mlo powdery mildew
resistance in barley. Euphytica, 63(1-2):141–152, 1992.
Jung, H. W, Tschaplinski, T. J, Wang, L, Glazebrook, J, and Greenberg, J. T.
Priming in systemic plant immunity. Science, 324(5923):89–91, 2009.
Kaku, H, Nishizawa, Y, Ishii-Minami, N, Akimoto-Tomiyama, C, Dohmae, N, Takio,
K, Minami, E, and Shibuya, N. Plant cells recognize chitin fragments for de-
fense signaling through a plasma membrane receptor. Proceedings of the National
Academy of Sciences, 103(29):11086–11091, 2006.
Karimi, M, Inzé, D, and Depicker, A. GatewayTM vectors for agrobacterium-
mediated plant transformation. Trends in plant science, 7(5):193–195, 2002.
Kessler, S. A, Shimosato-Asano, H, Keinath, N. F, Wuest, S. E, Ingram, G,
Panstruga, R, and Grossniklaus, U. Conserved molecular components for pollen
tube reception and fungal invasion. Science, 330(6006):968–971, 2010.
Kim, Y, Tsuda, K, Igarashi, D, Hillmer, R. A, Sakakibara, H, Myers, C. L, and
Katagiri, F. Mechanisms underlying robustness and tunability in a plant immune
signaling network. Cell host & microbe, 15(1):84–94, 2014.
156
Koh, S, André, A, Edwards, H, Ehrhardt, D, and Somerville, S. Arabidopsis thaliana
subcellular responses to compatible erysiphe cichoracearum infections. The Plant
Journal, 44(3):516–529, 2005.
Kreuze, J. F and Valkonen, J. P. Utilization of engineered resistance to viruses
in crops of the developing world, with emphasis on sub-saharan africa. Current
opinion in virology, 26:90–97, 2017.
Krinke, O, Flemr, M, Vergnolle, C, Collin, S, Renou, J.-P, Taconnat, L, Yu, A,
Burketová, L, Valentová, O, Zachowski, A, et al. Phospholipase D activation
is an early component of the salicylic acid signaling pathway in Arabidopsis cell
suspensions. Plant Physiology, 150(1):424–436, 2009.
Kuhn, H, Lorek, J, Kwaaitaal, M, Consonni, C, Becker, K, Micali, C, Ver Loren van
Themaat, E, Bednarek, P, Raaymakers, T. M, Appiano, M, Bai, Y, Meldau,
D, Baum, S, Conrath, U, Feussner, I, and Panstruga, R. Key components of
different plant defense pathways are dispensable for powdery mildew resistance
of the arabidopsis mlo2 mlo6 mlo12 triple mutant. Frontiers in Plant Science, 8:
1006, 2017. ISSN 1664-462X. doi: 10.3389/fpls.2017.01006. URL https://www.
frontiersin.org/article/10.3389/fpls.2017.01006.
Kusch, S and Panstruga, R. mlo-based resistance: an apparently universal “weapon”
to defeat powdery mildew disease. Molecular Plant-Microbe Interactions, 30(3):
179–189, 2017.
Kusner, D. J, Barton, J. A, Wen, K.-K, Wang, X, Rubenstein, P. A, and Iyer,
S. S. Regulation of phospholipase d activity by actin actin exerts bidirectional
modulation of mammalian phospolipase d activity in a polymerization-dependent,
isoform-specific manner. Journal of Biological Chemistry, 277(52):50683–50692,
2002.
Kwon, C, Neu, C, Pajonk, S, Yun, H. S, Lipka, U, Humphry, M, Bau, S, Straus, M,
Kwaaitaal, M, Rampelt, H, et al. Co-option of a default secretory pathway for
plant immune responses. Nature, 451(7180):835–840, 2008.
Langmead, B, Trapnell, C, Pop, M, and Salzberg, S. L. Ultrafast and memory-
efficient alignment of short dna sequences to the human genome. Genome biology,
10(3):R25, 2009.
Lemmon, M. A. Phosphoinositide recognition domains. Traffic, 4(4):201–213, 2003.
Li, H, Handsaker, B, Wysoker, A, Fennell, T, Ruan, J, Homer, N, Marth, G, Abeca-
sis, G, and Durbin, R. The sequence alignment/map format and samtools. Bioin-
formatics, 25(16):2078–2079, 2009.
Li, Y, Zhang, Y, Wang, Q.-X, Wang, T.-T, Cao, X.-L, Zhao, Z.-X, Zhao, S.-L,
Xu, Y.-J, Xiao, Z.-Y, Li, J.-L, et al. Resistance to powdery mildew8. 1 boosts
pattern-triggered immunity against multiple pathogens in arabidopsis and rice.
Plant biotechnology journal, 2017.
157
Lipka, V, Dittgen, J, Bednarek, P, Bhat, R, Wiermer, M, Stein, M, Landtag, J,
Brandt, W, Rosahl, S, Scheel, D, et al. Pre-and postinvasion defenses both con-
tribute to nonhost resistance in Arabidopsis. science, 310(5751):1180–1183, 2005.
Livak, K. J and Schmittgen, T. D. Analysis of relative gene expression data using
real-time quantitative pcr and the 2- δδct method. methods, 25(4):402–408, 2001.
Lorek, J, Griebel, T, Jones, A. M, Kuhn, H, and Panstruga, R. The role of ara-
bidopsis heterotrimeric g-protein subunits in mlo2 function and mamp-triggered
immunity. Molecular Plant-Microbe Interactions, 26(9):991–1003, 2013.
Ma, X.-F, Li, Y, Sun, J.-L, Wang, T.-T, Fan, J, Lei, Y, Huang, Y.-Y, Xu, Y.-J,
Zhao, J.-Q, Xiao, S, et al. Ectopic expression of resistance to powdery mildew8.
1 confers resistance to fungal and oomycete pathogens in arabidopsis. Plant and
Cell Physiology, 55(8):1484–1496, 2014.
Manohar, M, Tian, M, Moreau, M, Park, S.-W, Choi, H. W, Fei, Z, Friso, G, Asif,
M, Manosalva, P, von Dahl, C. C, et al. Identification of multiple salicylic acid-
binding proteins using two high throughput screens. Frontiers in plant science,
5:777, 2015.
Meng, X and Zhang, S. Mapk cascades in plant disease resistance signaling. Annual
review of phytopathology, 51:245–266, 2013.
Nakagami, H, Soukupová, H, Schikora, A, Zárskỳ, V, and Hirt, H. A mitogen-
activated protein kinase kinase kinase mediates reactive oxygen species homeosta-
sis in arabidopsis. Journal of Biological Chemistry, 281(50):38697–38704, 2006.
Narusaka, M, Shirasu, K, Noutoshi, Y, Kubo, Y, Shiraishi, T, Iwabuchi, M, and
Narusaka, Y. Rrs1 and rps4 provide a dual resistance-gene system against fungal
and bacterial pathogens. The Plant Journal, 60(2):218–226, 2009.
Novák, O, Hényková, E, Sairanen, I, Kowalczyk, M, Posṕı̌sil, T, and Ljung, K.
Tissue-specific profiling of the arabidopsis thaliana auxin metabolome. The Plant
Journal, 72(3):523–536, 2012.
Park, S.-W, Kaimoyo, E, Kumar, D, Mosher, S, and Klessig, D. F. Methyl salicylate
is a critical mobile signal for plant systemic acquired resistance. Science, 318
(5847):113–116, 2007.
Pendaries, C, Tronchère, H, Racaud-Sultan, C, Gaits-Iacovoni, F, Coronas, S, Ma-
nenti, S, Gratacap, M.-P, Plantavid, M, and Payrastre, B. Emerging roles of
phosphatidylinositol monophosphates in cellular signaling and trafficking. Ad-
vances in enzyme regulation, 45(1):201–214, 2005.
Pieterse, C. M, Van der Does, D, Zamioudis, C, Leon-Reyes, A, and Van Wees, S. C.
Hormonal modulation of plant immunity. Annual review of cell and developmental
biology, 28:489–521, 2012.
158
Piffanelli, P, Zhou, F, Casais, C, Orme, J, Jarosch, B, Schaffrath, U, Collins, N. C,
Panstruga, R, and Schulze-Lefert, P. The barley mlo modulator of defense and
cell death is responsive to biotic and abiotic stress stimuli. Plant physiology, 129
(3):1076–1085, 2002.
Pinosa, F, Buhot, N, Kwaaitaal, M, Fahlberg, P, Thordal-Christensen, H, Eller-
ström, M, and Andersson, M. X. Arabidopsis Phospholipase Dδ Is Involved in
Basal Defense and Nonhost Resistance to Powdery Mildew Fungi. Plant physiol-
ogy, 163(2):896–906, 2013.
Qian, L.-H, Zhou, G.-C, Sun, X.-Q, Lei, Z, Zhang, Y.-M, Xue, J.-Y, and Hang,
Y.-Y. Distinct patterns of gene gain and loss: diverse evolutionary modes of nbs-
encoding genes in three solanaceae crop species. G3: Genes, Genomes, Genetics,
7(5):1577–1585, 2017.
Qin, C and Wang, X. The arabidopsis phospholipase d family. characterization
of a calcium-independent and phosphatidylcholine-selective pldζ1 with distinct
regulatory domains. Plant physiology, 128(3):1057–1068, 2002.
Qiu, J.-L, Zhou, L, Yun, B.-W, Nielsen, H. B, Fiil, B. K, Petersen, K, MacKinlay, J,
Loake, G. J, Mundy, J, and Morris, P. C. Arabidopsis mitogen-activated protein
kinase kinases mkk1 and mkk2 have overlapping functions in defense signaling
mediated by mekk1, mpk4, and mks1. Plant physiology, 148(1):212–222, 2008.
R Core Team. R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria, 2017. URL https:
//www.R-project.org/.
Reinstädler, A, Müller, J, Czembor, J. H, Piffanelli, P, and Panstruga, R. Novel
induced mlo mutant alleles in combination with site-directed mutagenesis reveal
functionally important domains in the heptahelical barley mlo protein. BMC plant
biology, 10(1):31, 2010.
Rietz, S, Stamm, A, Malonek, S, Wagner, S, Becker, D, Medina-Escobar, N, Co-
rina Vlot, A, Feys, B. J, Niefind, K, and Parker, J. E. Different roles of enhanced
disease susceptibility1 (eds1) bound to and dissociated from phytoalexin deficient4
(pad4) in arabidopsis immunity. New Phytologist, 191(1):107–119, 2011.
Robatzek, S, Bittel, P, Chinchilla, D, Köchner, P, Felix, G, Shiu, S.-H, and Boller, T.
Molecular identification and characterization of the tomato flagellin receptor lefls2,
an orthologue of arabidopsis fls2 exhibiting characteristically different perception
specificities. Plant molecular biology, 64(5):539–547, 2007.
Robert-Seilaniantz, A, Grant, M, and Jones, J. D. Hormone crosstalk in plant
disease and defense: more than just jasmonate-salicylate antagonism. Annual
review of phytopathology, 49:317–343, 2011.
159
Roberts, A, Mackie, A, Hathaway, V, Callow, J, and Green, J. Molecular differenti-
ation in the extrahaustorial membrane of pea powdery mildew haustoria at early
and late stages of development. Physiological and Molecular Plant Pathology, 43
(2):147–160, 1993.
Rodas-Junco, B, Muñoz-Sánchez, J, Vázquez-Flota, F, and Hernández-Sotomayor,
S. Salicylic-acid elicited phospholipase d responses in capsicum chinense cell cul-
tures. Plant Physiology and Biochemistry, 90:32–37, 2015.
Saha, U, Mazumder, M, Mukherjee, A, Parveen, S, Mondal, B, Maji, S. R, and
Basu, D. A critical analysis of phosphatidic acid mediated resistance response
in sinapis alba against alternaria brassicicola. Physiological and Molecular Plant
Pathology, 94:90–99, 2016.
Saville, A. C, Martin, M. D, and Ristaino, J. B. Historic late blight outbreaks
caused by a widespread dominant lineage of phytophthora infestans (mont.) de
bary. PloS one, 11(12):e0168381, 2016.
Schneider, C. A, Rasband, W. S, and Eliceiri, K. W. Nih image to imagej: 25 years
of image analysis. Nat methods, 9(7):671–675, 2012.
Serrano, M, Wang, B, Aryal, B, Garcion, C, Abou-Mansour, E, Heck, S, Geisler,
M, Mauch, F, Nawrath, C, and Métraux, J.-P. Export of salicylic acid from the
chloroplast requires the multidrug and toxin extrusion-like transporter eds5. Plant
physiology, 162(4):1815–1821, 2013.
Shao, Z.-Q, Zhang, Y.-M, Hang, Y.-Y, Xue, J.-Y, Zhou, G.-C, Wu, P, Wu, X.-Y,
Wu, X.-Z, Wang, Q, Wang, B, et al. Long-term evolution of nucleotide-binding
site-leucine-rich repeat genes: understanding gained from and beyond the legume
family. Plant physiology, 166(1):217–234, 2014.
Shao, Z.-Q, Wang, B, and Chen, J.-Q. Tracking ancestral lineages and recent expan-
sions of nbs-lrr genes in angiosperms. Plant signaling & behavior, 11(7):2095–109,
2016.
Shubchynskyy, V, Boniecka, J, Schweighofer, A, Simulis, J, Kvederaviciute, K,
Stumpe, M, Mauch, F, Balazadeh, S, Mueller-Roeber, B, Boutrot, F, et al. Protein
phosphatase ap2c1 negatively regulates basal resistance and defense responses to
pseudomonas syringae. Journal of experimental botany, 68(5):1169–1183, 2017.
Skou, J.-P. Callose formation responsible for the powdery mildew resistance in
barley with genes in the ml-o locus. Journal of Phytopathology, 104(1):90–95,
1982.
Skou, J.-P, Jørgensen, J. H, and Lilholt, U. Comparative studies on callose formation
in powdery mildew compatible and incompatible barley. Journal of Phytopathol-
ogy, 109(2):147–168, 1984.
160
Spoel, S. H and Dong, X. How do plants achieve immunity? defence without
specialized immune cells. Nature reviews immunology, 12(2):89, 2012.
Srichumpa, P, Brunner, S, Keller, B, and Yahiaoui, N. Allelic series of four pow-
dery mildew resistance genes at the pm3 locus in hexaploid bread wheat. Plant
Physiology, 139(2):885–895, 2005.
Staswick, P. E, Su, W, and Howell, S. H. Methyl jasmonate inhibition of root growth
and induction of a leaf protein are decreased in an arabidopsis thaliana mutant.
Proceedings of the National Academy of Sciences, 89(15):6837–6840, 1992.
Stein, M, Dittgen, J, Sánchez-Rodŕıguez, C, Hou, B.-H, Molina, A, Schulze-Lefert,
P, Lipka, V, and Somerville, S. Arabidopsis pen3/pdr8, an atp binding cassette
transporter, contributes to nonhost resistance to inappropriate pathogens that
enter by direct penetration. The Plant Cell, 18(3):731–746, 2006.
Strange, R. N and Scott, P. R. Plant disease: a threat to global food security.
Annual review of phytopathology, 43, 2005.
Suarez-Rodriguez, M. C, Adams-Phillips, L, Liu, Y, Wang, H, Su, S.-H, Jester, P. J,
Zhang, S, Bent, A. F, and Krysan, P. J. Mekk1 is required for flg22-induced mpk4
activation in arabidopsis plants. Plant physiology, 143(2):661–669, 2007.
Takai, R, Isogai, A, Takayama, S, and Che, F.-S. Analysis of flagellin perception
mediated by flg22 receptor osfls2 in rice. Molecular Plant-Microbe Interactions,
21(12):1635–1642, 2008.
Takamatsu, S. Phylogeny and evolution of the powdery mildew fungi (erysiphales,
ascomycota) inferred from nuclear ribosomal dna sequences. Mycoscience, 45(2):
147–157, 2004.
Takemoto, D, Jones, D. A, and Hardham, A. R. Gfp-tagging of cell components
reveals the dynamics of subcellular re-organization in response to infection of
arabidopsis by oomycete pathogens. The Plant Journal, 33(4):775–792, 2003.
Testerink, C and Munnik, T. Phosphatidic acid: a multifunctional stress signaling
lipid in plants. Trends in plant science, 10(8):368–375, 2005.
Testerink, C and Munnik, T. Molecular, cellular, and physiological responses to
phosphatidic acid formation in plants. Journal of Experimental Botany, 62(7):
2349–2361, 2011.
Tetiana, K, Oksana, I, and Sergii, K. Involvement of phospholipase D and NADPH-
oxidase in salicylic acid signaling cascade. Plant Physiology and Biochemistry,
2013.
Thomma, B. P, Nürnberger, T, and Joosten, M. H. Of pamps and effectors: the
blurred pti-eti dichotomy. The plant cell, 23(1):4–15, 2011.
161
Thordal-Christensen, H, Zhang, Z, Wei, Y, and Collinge, D. B. Subcellular localiza-
tion of h2o2 in plants. h2o2 accumulation in papillae and hypersensitive response
during the barley—powdery mildew interaction. The Plant Journal, 11(6):1187–
1194, 1997.
Tornero, P and Dangl, J. L. A high-throughput method for quantifying growth
of phytopathogenic bacteria in arabidopsis thaliana. The Plant Journal, 28(4):
475–481, 2001.
Trdá, L, Fernandez, O, Boutrot, F, Héloir, M.-C, Kelloniemi, J, Daire, X, Adrian,
M, Clément, C, Zipfel, C, Dorey, S, et al. The grapevine flagellin receptor vvfls2
differentially recognizes flagellin-derived epitopes from the endophytic growth-
promoting bacterium burkholderia phytofirmans and plant pathogenic bacteria.
New Phytologist, 201(4):1371–1384, 2014.
Tricoll, D. M, Carney, K. J, Russell, P. F, McMaster, J. R, Groff, D. W, Hadden,
K. C, Himmel, P. T, Hubbard, J. P, Boeshore, M. L, and Quemada, H. D. Field
evaluation of transgenic squash containing single or multiple virus coat protein
gene constructs for resistance to cucumber mosaic virus, watermelon mosaic virus
2, and zucchini yellow mosaic virus. Nature Biotechnology, 13(12):1458, 1995.
Tsuda, K, Sato, M, Stoddard, T, Glazebrook, J, and Katagiri, F. Network properties
of robust immunity in plants. PLoS Genet, 5(12):e1000772, 2009.
Tsuda, K, Mine, A, Bethke, G, Igarashi, D, Botanga, C. J, Tsuda, Y, Glazebrook,
J, Sato, M, and Katagiri, F. Dual regulation of gene expression mediated by
extended mapk activation and salicylic acid contributes to robust innate immunity
in arabidopsis thaliana. PLoS Genetics, 9(12):e1004015, 2013.
Ulm, R, Revenkova, E, di Sansebastiano, G.-P, Bechtold, N, and Paszkowski, J.
Mitogen-activated protein kinase phosphatase is required for genotoxic stress relief
in arabidopsis. Genes & Development, 15(6):699–709, 2001.
Van der Does, D, Leon-Reyes, A, Koornneef, A, Van Verk, M. C, Rodenburg, N,
Pauwels, L, Goossens, A, Körbes, A. P, Memelink, J, Ritsema, T, et al. Salicylic
acid suppresses jasmonic acid signaling downstream of scfcoi1-jaz by targeting
gcc promoter motifs via transcription factor ora59. The Plant Cell, 25(2):744–
761, 2013.
van der Hoorn, R. A and Kamoun, S. From guard to decoy: a new model for
perception of plant pathogen effectors. The Plant Cell, 20(8):2009–2017, 2008.
Venugopal, S. C, Jeong, R.-D, Mandal, M. K, Zhu, S, Chandra-Shekara, A, Xia,
Y, Hersh, M, Stromberg, A. J, Navarre, D, Kachroo, A, et al. Enhanced disease
susceptibility 1 and salicylic acid act redundantly to regulate resistance gene-
mediated signaling. PLoS Genetics, 5(7):e1000545, 2009.
162
Vlot, A. C, Dempsey, D. A, and Klessig, D. F. Salicylic acid, a multifaceted hormone
to combat disease. Annual review of phytopathology, 47:177–206, 2009.
Vogel, J and Somerville, S. Isolation and characterization of powdery mildew-
resistant arabidopsis mutants. Proceedings of the National Academy of Sciences,
97(4):1897–1902, 2000.
von Malek, B, van der Graaff, E, Schneitz, K, and Keller, B. The arabidopsis male-
sterile mutant dde2-2 is defective in the allene oxide synthase gene encoding one
of the key enzymes of the jasmonic acid biosynthesis pathway. Planta, 216(1):
187–192, 2002.
von Röpenack, E, Parr, A, and Schulze-Lefert, P. Structural analyses and dynamics
of soluble and cell wall-bound phenolics in a broad spectrum resistance to the
powdery mildew fungus in barley. Journal of Biological Chemistry, 273(15):9013–
9022, 1998.
Wagner, S, Stuttmann, J, Rietz, S, Guerois, R, Brunstein, E, Bautor, J, Niefind,
K, and Parker, J. E. Structural basis for signaling by exclusive eds1 heteromeric
complexes with sag101 or pad4 in plant innate immunity. Cell host & microbe,
14(6):619–630, 2013.
Wang, C and Wang, X. A novel phospholipase D of Arabidopsis that is activated by
oleic acid and associated with the plasma membrane. Plant physiology, 127(3):
1102–1112, 2001.
Wang, N and Trivedi, P. Citrus huanglongbing: a newly relevant disease presents
unprecedented challenges. Phytopathology, 103(7):652–665, 2013.
Wang, W, Sijacic, P, Xu, P, Lian, H, and Liu, Z. Arabidopsis tso1 and myb3r1 form
a regulatory module to coordinate cell proliferation with differentiation in shoot
and root. Proceedings of the National Academy of Sciences, page 201715903, 2018.
Wang, W, Devoto, A, Turner, J. G, and Xiao, S. Expression of the membrane-
associated resistance protein RPW8 enhances basal defense against biotrophic
pathogens. Molecular plant-microbe interactions, 20(8):966–976, 2007.
Wang, W, Wen, Y, Berkey, R, and Xiao, S. Specific targeting of the Arabidop-
sis resistance protein RPW8. 2 to the interfacial membrane encasing the fungal
haustorium renders broad-spectrum resistance to powdery mildew. The Plant
Cell Online, 21(9):2898–2913, 2009.
Wang, W, Zhang, Y, Wen, Y, Berkey, R, Ma, X, Pan, Z, Bendigeri, D, King, H,
Zhang, Q, and Xiao, S. A comprehensive mutational analysis of the Arabidopsis
resistance protein RPW8. 2 reveals key amino acids for defense activation and
protein targeting. The Plant Cell Online, 25(10):4242–4261, 2013.
163
Wang, X. Multiple forms of phospholipase D in plants: the gene family, catalytic
and regulatory properties, and cellular functions. Progress in lipid research, 39
(2):109–149, 2000.
Wang, X. Lipid signaling. Current Opinion in Plant Biology, 7(3):329–336, 2004.
Wang, X, Devaiah, S. P, Zhang, W, and Welti, R. Signaling functions of phosphatidic
acid. Progress in lipid research, 45(3):250–278, 2006.
Wang, Z.-P, Xing, H.-L, Dong, L, Zhang, H.-Y, Han, C.-Y, Wang, X.-C, and Chen,
Q.-J. Egg cell-specific promoter-controlled crispr/cas9 efficiently generates ho-
mozygous mutants for multiple target genes in arabidopsis in a single generation.
Genome biology, 16(1):1–12, 2015.
Wen, Y, Wang, W, Feng, J, Luo, M.-C, Tsuda, K, Katagiri, F, Bauchan, G, and
Xiao, S. Identification and utilization of a sow thistle powdery mildew as a poorly
adapted pathogen to dissect post-invasion non-host resistance mechanisms in Ara-
bidopsis. Journal of experimental botany, 62(6):2117–2129, 2011.
Wickham, H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New
York, 2009. ISBN 978-0-387-98140-6. URL http://ggplot2.org.
Wiermer, M, Feys, B. J, and Parker, J. E. Plant immunity: the eds1 regulatory
node. Current opinion in plant biology, 8(4):383–389, 2005.
Wildermuth, M. C, Dewdney, J, Wu, G, and Ausubel, F. M. Isochorismate synthase
is required to synthesize salicylic acid for plant defence. Nature, 414(6863):562–
565, 2001.
Wu, Y, Zhang, D, Chu, J. Y, Boyle, P, Wang, Y, Brindle, I. D, De Luca, V, and
Després, C. The arabidopsis npr1 protein is a receptor for the plant defense
hormone salicylic acid. Cell reports, 1(6):639–647, 2012.
Xiao, S, Dai, L, Liu, F, Wang, Z, Peng, W, and Xie, D. Cos1: an arabidopsis
coronatine insensitive1 suppressor essential for regulation of jasmonate-mediated
plant defense and senescence. The Plant Cell, 16(5):1132–1142, 2004a.
Xiao, S, Ellwood, S, Calis, O, Patrick, E, Li, T, Coleman, M, and Turner, J. G.
Broad-spectrum mildew resistance in Arabidopsis thaliana mediated by RPW8.
Science, 291(5501):118–120, 2001.
Xiao, S, Brown, S, Patrick, E, Brearley, C, and Turner, J. G. Enhanced transcription
of the Arabidopsis disease resistance genes RPW8. 1 and RPW8. 2 via a salicylic
acid–dependent amplification circuit is required for hypersensitive cell death. The
Plant Cell Online, 15(1):33–45, 2003.
Xiao, S, Emerson, B, Ratanasut, K, Patrick, E, O’Neill, C, Bancroft, I, and Turner,
J. G. Origin and maintenance of a broad-spectrum disease resistance locus in
Arabidopsis. Molecular biology and evolution, 21(9):1661–1672, 2004b.
164
Xiao, S, Calis, O, Patrick, E, Zhang, G, Charoenwattana, P, Muskett, P, Parker,
J. E, and Turner, J. G. The atypical resistance gene, RPW8, recruits components
of basal defence for powdery mildew resistance in Arabidopsis. The Plant Journal,
42(1):95–110, 2005.
Xie, D.-X, Feys, B. F, James, S, Nieto-Rostro, M, and Turner, J. G. Coi1: an
arabidopsis gene required for jasmonate-regulated defense and fertility. Science,
280(5366):1091–1094, 1998.
Xing, L, Hu, P, Liu, J, Witek, K, Zhou, S, Xu, J, Zhou, W, Gao, L, Huang, Z,
Zhang, R, et al. Pm21 from haynaldia villosa encodes a cc-nbs-lrr that confers
powdery mildew resistance in wheat. Molecular plant, 2018.
Yamaguchi, T, Kuroda, M, Yamakawa, H, Ashizawa, T, Hirayae, K, Kurimoto, L,
Shinya, T, and Shibuya, N. Suppression of a phospholipase d gene, ospldβ1, acti-
vates defense responses and increases disease resistance in rice. Plant physiology,
150(1):308–319, 2009.
Yan, S and Dong, X. Perception of the plant immune signal salicylic acid. Current
Opinion in Plant Biology, 20:64–68, 2014.
Young, S. A, Wang, X, and Leach, J. E. Changes in the plasma membrane distri-
bution of rice phospholipase D during resistant interactions with Xanthomonas
oryzae pv oryzae. The Plant Cell Online, 8(6):1079–1090, 1996.
Yu, D, Chen, Q, Huang, W, Wan, S, and Zhan, J. Cloning, bioinformatic analysis
and expression pattern of phospholipase d gene family in vitis vinifera. Current
Bioinformatics, 13(1):42–49, 2018.
Zhang, J, Shao, F, Li, Y, Cui, H, Chen, L, Li, H, Zou, Y, Long, C, Lan, L, Chai,
J, et al. A pseudomonas syringae effector inactivates mapks to suppress pamp-
induced immunity in plants. Cell host & microbe, 1(3):175–185, 2007.
Zhang, Q and Xiao, S. Lipids in salicylic acid-mediated defense in plants: focusing
on the roles of phosphatidic acid and phosphatidylinositol 4-phosphate. Fron-
tiers in Plant Science, 6(387), 2015. ISSN 1664-462X. doi: 10.3389/fpls.2015.
00387. URL http://www.frontiersin.org/plant-microbe_interaction/10.
3389/fpls.2015.00387/abstract.
Zhang, Q, Berkey, R, Pan, Z, Wang, W, Zhang, Y, Ma, X, King, H, and Xiao, S.
Dominant negative RPW8.2 fusion proteins reveal the importance of haustorium-
oriented protein trafficking for resistance against powdery mildew in Arabidopsis.
Plant Signaling & Behavior, 10(3):e989766, 2015. doi: 10.4161/15592324.2014.
989766. URL http://dx.doi.org/10.4161/15592324.2014.989766. PMID:
25830634.
165
Zhang, Q, Berkey, R, Blakeslee, J. J, Lin, J, Ma, X, King, H, Liddle, A, Guo,
L, Munnik, T, Wang, X, et al. Arabidopsis phospholipase dα1 and dδ oppo-
sitely modulate eds1-and sa-independent basal resistance against adapted pow-
dery mildew. Journal of Experimental Botany, page ery146, 2018.
Zhang, W, Qin, C, Zhao, J, and Wang, X. Phospholipase dα1-derived phosphatidic
acid interacts with abi1 phosphatase 2c and regulates abscisic acid signaling.
Proceedings of the National Academy of Sciences of the United States of America,
101(25):9508–9513, 2004.
Zhang, Y.-M, Shao, Z.-Q, Wang, Q, Hang, Y.-Y, Xue, J.-Y, Wang, B, and Chen,
J.-Q. Uncovering the dynamic evolution of nucleotide-binding site-leucine-rich
repeat (nbs-lrr) genes in brassicaceae. Journal of integrative plant biology, 58(2):
165–177, 2016a.
Zhang, Y, Zhu, H, Zhang, Q, Li, M, Yan, M, Wang, R, Wang, L, Welti, R, Zhang, W,
and Wang, X. Phospholipase Dα1 and phosphatidic acid regulate NADPH oxidase
activity and production of reactive oxygen species in ABA-mediated stomatal
closure in Arabidopsis. The Plant Cell Online, 21(8):2357–2377, 2009.
Zhang, Z, Wu, Y, Gao, M, Zhang, J, Kong, Q, Liu, Y, Ba, H, Zhou, J, and Zhang,
Y. Disruption of pamp-induced map kinase cascade by a pseudomonas syringae
effector activates plant immunity mediated by the nb-lrr protein summ2. Cell
host & microbe, 11(3):253–263, 2012.
Zhang, Z, Liu, Y, Huang, H, Gao, M, Wu, D, Kong, Q, and Zhang, Y. The nlr
protein summ2 senses the disruption of an immune signaling map kinase cascade
via crck3. EMBO reports, page e201642704, 2016b.
Zhao, J. Phospholipase d and phosphatidic acid in plant defence response: from
protein–protein and lipid–protein interactions to hormone signalling. Journal of
experimental botany, page eru540, 2015.
Zhao, J, Devaiah, S. P, Wang, C, Li, M, Welti, R, and Wang, X. Arabidopsis phos-
pholipase Dβ1 modulates defense responses to bacterial and fungal pathogens.
New Phytologist, 2013.
Zheng, L, Krishnamoorthi, R, Zolkiewski, M, and Wang, X. Distinct ca2+ binding
properties of novel c2 domains of plant phospholipase dα and β. Journal of
Biological Chemistry, 275(26):19700–19706, 2000.
Zheng, L, Shan, J, Krishnamoorthi, R, and Wang, X. Activation of plant phospho-
lipase dβ by phosphatidylinositol 4, 5-bisphosphate: characterization of binding
site and mode of action. Biochemistry, 41(14):4546–4553, 2002.
Zheng, X.-y, Spivey, N. W, Zeng, W, Liu, P.-P, Fu, Z. Q, Klessig, D. F, He, S. Y,
and Dong, X. Coronatine promotes pseudomonas syringae virulence in plants by
activating a signaling cascade that inhibits salicylic acid accumulation. Cell host
& microbe, 11(6):587–596, 2012.
166
Zhong, Y and Cheng, Z.-M. M. A unique rpw8-encoding class of genes that origi-
nated in early land plants and evolved through domain fission, fusion, and dupli-
cation. Scientific reports, 6:32923, 2016.
Zhou, F, Kurth, J, Wei, F, Elliott, C, Valè, G, Yahiaoui, N, Keller, B, Somerville, S,
Wise, R, and Schulze-Lefert, P. Cell-autonomous expression of barley mla1 con-
fers race-specific resistance to the powdery mildew fungus via a rar1-independent
signaling pathway. The Plant Cell, 13(2):337–350, 2001.
Zhou, F, Menke, F. L, Yoshioka, K, Moder, W, Shirano, Y, and Klessig, D. F. High
humidity suppresses ssi4-mediated cell death and disease resistance upstream of
map kinase activation, h2o2 production and defense gene expression. The Plant
Journal, 39(6):920–932, 2004.
Zhou, N, Tootle, T. L, Tsui, F, Klessig, D. F, and Glazebrook, J. Pad4 functions
upstream from salicylic acid to control defense responses in arabidopsis. The Plant
Cell, 10(6):1021–1030, 1998.
Zhu, S, Jeong, R.-D, Venugopal, S. C, Lapchyk, L, Navarre, D, Kachroo, A, and
Kachroo, P. Sag101 forms a ternary complex with eds1 and pad4 and is re-
quired for resistance signaling against turnip crinkle virus. PLoS pathogens, 7
(11):e1002318, 2011.
Zipfel, C, Kunze, G, Chinchilla, D, Caniard, A, Jones, J. D, Boller, T, and Felix, G.
Perception of the bacterial pamp ef-tu by the receptor efr restricts agrobacterium-
mediated transformation. Cell, 125(4):749–760, 2006.
Zou, S, Wang, H, Li, Y, Kong, Z, and Tang, D. The nb-lrr gene pm60 confers
powdery mildew resistance in wheat. New Phytologist, 218(1):298–309, 2018.
167