ABSTRACT 
 
 
 
 
Title of Dissertation: NEISSERIA GONORRHOEAE MODULATES 
INFECTIVITY BASED ON PROPERTIES OF 
HUMAN CERVICAL EPITHELIA AND PHASE 
VARIABLE BACTERIAL SURFACE 
STRUCTURES 
  
 Qian Yu 
Doctor of Philosophy, 2019 
  
Dissertation Directed by: Dr. Wenxia Song, Professor 
Department of Cell Biology and Molecular 
Genetics 
 
 
 
Neisseria gonorrhoeae (GC) infection in the human female reproductive tract causes 
various clinical outcomes, from no symptom to severe complications. The major 
barrier to a better understanding of GC infection in women is the lack of experimental 
system closely mimicking in vivo infection. Here, I developed a human cervical tissue 
explant model, which maintains the heterogeneity of the cervical epithelium. Using 
this model, my thesis research examined the impact of the heterogeneity of the 
cervical epithelium and the phase variation of GC surface structures on GC infectivity. 
My research revealed that GC preferentially colonize the ectocervix and the 
transformation zone (TZ), but exclusively penetrate into the subepithelial tissues of the 
TZ and endocervix. Pili are essential for GC colonization in all regions of the cervix. 
 
 
Expression of Opa isoforms that bind to the host receptors CEACAM (OpaCEA) 
enhances GC colonization in the ecto/endocervix but inhibits GC penetration into the 
endocervix. However, GC infectivity in the TZ does not respond to Opa phase 
variation, due to the low expression level and intracellular location of CEACAMs in 
the TZ epithelial cells. OpaCEA enhances GC colonization in the ecto/endocervix by 
inhibiting epithelial exfoliation and suppresses GC penetration into the endocervical 
subepithelium by inhibiting GC-induced disassembly of the apical junction. Opa-
mediated modulation of GC infectivity depends on the immune receptor tyrosine-
based inhibitory motif (ITIM) of CEACAM1 and its downstream phosphatase SHP. 
The effect of epithelial cell polarity on GC invasion was studied using a cell line 
model. My results show that GC invade more efficiently into non-polarized than 
polarized epithelial cells without changing the adhesion efficiency. Opa (phase 
variable) expression enhances both adhesion and invasion in both non-polarized and 
polarized cells. In non-polarized cells, Opa expression induces F-actin accumulation 
and microvilli elongation underneath GC microcolonies, suggesting an actin-mediated 
uptake of GC. In contrast, GC expressing no Opa reduce F-actin and demolish 
microvilli underneath microcolonies in both polarized epithelial cell line and 
endocervical epithelial cells potentially by increasing calcium flux, NMII activation 
and the redistribution of actin nucleation factor Arp2/3 from the apical surface. Taking 
together, my research demonstrates that both the heterogeneity of the cervical 
epithelium and the phase variation of bacterial surface structures regulate GC 
infectivity in the human cervix, either dominated by colonization or penetration, 
consequently influencing the clinical outcomes of the infection. 
 
 
 
 
 
 
 
 
 
 
 
 
 
NEISSERIA GONORRHOEAE MODULATES INFECTIVITY BASED ON 
PROPERTIES OF HUMAN CERVICAL EPITHELIA AND PHASE VARIABLE 
BACTERIAL SURFACE STRUCTURES 
 
 
 
 By 
 
 
Qian Yu 
 
 
 
 
 
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,  
2019 
 
 
 
 
 
 
 
 
 
Advisory Committee: 
Professor Wenxia Song, Ph.D. Chair 
Professor Norma Andrews, Ph.D. 
Professor Kevin McIver, Ph.D. 
Professor Daniel. C. Stein, Ph.D. 
Professor Xiaoping Zhu, Ph.D. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
© Copyright by 
Qian Yu 
2019 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Dedication 
I would like to dedicate this dissertation to my parents, Xiaoqiang Yu, and Qianqian 
Zhang, who always support me. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ii 
 
Acknowledgment 
I would like to thank Dr. Wenxia Song, my mentor, for her invaluable guidance, 
insights, and expertise on my thesis project and life. I would also like to thank Dr. 
Daniel Stein for his expertise and insights on my thesis project, and my committee 
members, Dr. Norma Andrews, Dr. Kevin McIver, and Dr. Silvia Muro, and Dr. 
Xiaoping Zhu for all their input and help on my thesis project. 
 
I would like to take this opportunity to thank all of my lab mates in both the Song lab 
and Stein lab, who made this journal so much easier. I would like to thank my lab 
mate, Liang-Chun Wang, for his friendship and all the support and help. I have to 
thank my former and current lab mates Katie, Michelle, Heather, Senthil, Salsawi, 
Sofia, Yanan, Anshuman, Jurri, Betty, Ebby, Melanie, and Grace, for being 
supportive, and being good friends. Last but not least, I would like to thank Amy 
Beaven from the imaging core to help me with all the fluorescence microscopy, and 
Timothy K Maugel for all the help and work on the TEM and SEM.  
 
 
 
 
 
iii 
 
Table of Contents 
 
Dedication ......................................................................................................................ii 
Acknowledgment .......................................................................................................... iii 
Table of Contents .......................................................................................................... iv 
List of Abbreviations .................................................................................................. viii 
List of Figures ............................................................................................................... xi 
Chapter 1: Introduction ................................................................................................. 1 
1.1 Gonorrhea and Neisseria gonorrhoeae ......................................................... 1 
1.2 Neisseria gonorrhoeae infection in females ........................................................ 2 
1.3 The cytoarchitecture of cervical epithelium ........................................................ 5 
1.4 Neisseria gonorrhoeae and its virulence factors ................................................ 9 
1.5 Neisseria gonorrhoeae engaged host receptors and downstream signaling .... 11 
1.6 Neisseria gonorrhoeae interaction with human epithelial cells ....................... 21 
1.7 Epithelial cell-cell junction and its regulation .................................................. 26 
1.8 The role of the actin cytoskeleton in Neisseria gonorrhoeae infection ............. 34 
1.9 The advantages and disadvantages of various current GC infection models ... 38 
Rational ....................................................................................................................... 41 
Central hypothesis ...................................................................................................... 43 
iv 
 
Aim 1 ........................................................................................................................... 43 
Aim 2 ........................................................................................................................... 43 
Aim 3 ........................................................................................................................... 43 
Chapter 2: Infectivity of Neisseria gonorrhoeae in the human cervix varies with the 
properties of cervical epithelial cells and bacterial phase variable surface structures 44 
2.1 Abstract ............................................................................................................. 44 
2.2 Introduction ...................................................................................................... 45 
2.3 Materials and Methods ..................................................................................... 48 
2.4 Results ............................................................................................................... 54 
2.4.1 GC exhibit distinct infectivity patterns on different mucosal surfaces of 
the human cervix ............................................................................................... 54 
2.4.2 Pili are required for GC colonization, and Opa phase variation regulates 
GC infectivity in the human cervix .................................................................. 59 
2.4.3 The OpaH-dependent regulation of GC infectivity involves CEACAM1L 
expression in cervical epithelial cells ............................................................... 65 
2.4.4 The OpaH-dependent regulation of GC infectivity involves activation of 
SHP ..................................................................................................................... 69 
2.4.5 OpaH expression inhibits GC-induced cervical epithelial exfoliation in a 
CEACAM1L and SHP-dependent manner ..................................................... 74 
2.4.6 GC infection manipulates cell-cell junction in cervical epithelial cells in 
a CEACAM1L and SHP-dependent manner .................................................. 79 
v 
 
2.5 Discussion ......................................................................................................... 85 
Chapter 3: The polarity of epithelial cells regulates Neisseria gonorrhoeae infectivity
..................................................................................................................................... 92 
3.1 Abstract ............................................................................................................. 92 
3.2 Introduction ...................................................................................................... 93 
3.2 Materials and Methods ..................................................................................... 95 
3.4 Results ............................................................................................................. 101 
3.4.1 GC invade into non-polarized epithelial cells more efficiently than into 
polarized epithelial cells without altering the level of GC adhesion ........... 101 
3.4.2 Expression of Opa enhances GC adhesion and invasion in both non-
polarized and polarized epithelial cells .......................................................... 104 
3.4.3 GC inoculation induces differential membrane remodeling in non-
polarized and polarized epithelial cells .......................................................... 106 
3.4.4 GC inoculation induces differential redistribution of F-actin at the 
luminal surface of non-polarized and polarized epithelial cells .................. 108 
3.4.5 GC induce F-actin redistribution in endocervical epithelial cells, which 
involves NMII ................................................................................................... 111 
3.4.6 GC induce the redistribution of the actin nucleation factor Arp2/3 in 
endocervical epithelial cells, which involves NMII ....................................... 115 
3.3.7 GC inoculation induces the elevation of the intracellular calcium level in 
polarized but not non-polarized epithelial cells ............................................ 117 
vi 
 
3.5 Discussion ....................................................................................................... 120 
Chapter 4: Conclusions, discussion, and future direction ......................................... 126 
4.1 Summary ......................................................................................................... 126 
4.2 Working Model ............................................................................................... 129 
4.3 Future Direction ............................................................................................. 136 
4.3.1 Highlights ................................................................................................ 136 
4.3.2 Question 1. Does the mucosal microenvironment of the human cervix 
affect GC infectivity and how? ....................................................................... 137 
4.3.3 Question 2. Does GC infection induce differential immune responses in 
different cervical regions?............................................................................... 138 
4.3.4 Question 3. How to reconstruct the heterogeneous cervical in vitro? 139 
Bibliography ............................................................................................................. 140 
 
 
 
 
 
 
vii 
 
List of Abbreviations 
ARPC actin-related protein complex 
AJ adherens junction 
ASGP-R asialoglycoprotein receptor 
ASM acidic sphingomyelinase 
CEACAM carcinoembryonic antigen-related cell adhesion molecule 
CR3 complement receptor 3 
DAG diacylglycerol 
DGI disseminated gonococcal infection 
EGFR epidermal growth factor receptor 
ERK extracellular signal-related kinase 
FBS fetal bovine serum 
FRT female reproductive tract 
GCK gonococcal media base 
GC Neisseria gonorrhoeae 
HEC-1-B human endometrial adenocarcinoma cell line 
HSPG heparan sulfate proteoglycans 
IP3 inositol triphosphate 
viii 
 
JAM junctional adhesion molecule 
LOS lipooligosaccharide 
MAPK mitogen-activated protein kinase 
MLC myosin light chain 
MLCK myosin light chain kinase 
MOI multiplicity of infection 
NMII non-muscle myosin II 
NPRs nucleation promoting factors  
N-WASP neuronal Wiskott–Aldrich syndrome protein  
Opa opacity-associated proteins 
PC-PLC phosphatidylcholine-specific phospholipase C 
PI3K phosphoinositide 3-kinase 
PID pelvic inflammatory disease 
Pil pili 
PIP2 phosphoinositol 4, 5 bisphosphate 
PIP3 phosphoinositol 3, 4, 5 bisphosphate 
PKC protein Kinase C 
PLC phospholipase C 
ix 
 
pMLC phosphorylated myosin light chain 
pSHP1 phosphorylated protein tyrosine phosphatase 1B, PTP1B 
pSHP2 phosphorylated protein tyrosine phosphatase 2B, PTP2B 
ROCK Rho-associated coiled-coil kinase 
RTK receptor tyrosine kinase 
SH2 Src homology 2 
SCJ squamocolumnar junction 
SHP1 protein tyrosine phosphatase 1B, PTP1B 
SHP2 protein tyrosine phosphatase 2B, PTP2B 
T3SS type III secretion system  
T84 human colorectal adenocarcinoma cell line 
TEER transepithelial resistance 
TJ tight junction 
TZ transformation zone 
ZO zonula occludens 
 
 
 
x 
 
List of Figures 
Figure 1 - Unique features of female gonorrhea 
Figure 2 - Distinct cytoarchitecture of the cervical epithelium at different anatomical 
regions  
Figure 3 - Structure and downstream signaling of carcinoembryonic antigen-related 
cell adhesion molecule (CEACAM) triggered by Opa 
Figure 4 - Structure and downstream signaling of heparan sulfate proteoglycans 
(HSPG) triggered by Opa 
Figure 5 - Organization and regulation of the tight junction (TJ) and E-cadherin-based 
the adherens junction (AJ) 
Figure 6 - Organization of cell-cell junction complexes in the human cervical 
epithelium 
Figure 7 - Human cervical tissue explants maintain the in vivo characteristics in 
culture 
Figure 8 - GC exhibit different levels of colonization and subepithelial penetration in 
the ectocervical, TZ, and endocervical epithelium 
Figure 9 - Pili are required for GC colonization, and Opa expression differentially 
regulates GC infectivity in the three cervical regions.  
Figure 10 - Opa variants differentially regulate GC colonization and penetration into 
the three cervical regions 
xi 
 
Figure 11 - Regulatory roles of OpaH depend on the expression of CEACAML in the 
human cervical epithelial cells 
Figure 12 - Treatment of the SHP inhibitor NSC-87877 has no significant effect on 
GC growth 
Figure 13 - Human cervical tissue explants show no significant staining of SHP1 
Figure 14 - SHP activation is involved in the regulatory role of OpaH in GC infectivity 
in the human cervix 
Figure 15 - OpaH inhibits GC-induced shedding of the cervical epithelial cells in a 
CEACAM1L and SHP-dependent manner 
Figure 16 - OpaH inhibits GC-induced junction disruption in the cervical epithelial 
cells in a CEACAM1L and SHP-dependent manner 
Figure 17 - GC inoculation disrupts E-cadherin-based cell-cell junction  
Figure 18 - GC invade into non-polarized epithelial cells more efficiently than into 
polarized epithelial cells without altering the level of GC adhesion. 
Figure 19 - Expression of Opa enhances the level of adhesion and invasion in non-
polarized and polarized epithelial cells 
Figure 20 - GC inoculation induces differential remodeling of the plasma membrane 
of non-polarized and polarized epithelial cells 
Figure 21- GC inoculation induces differential redistribution of F-actin at GC adherent 
sites in non-polarized and polarized epithelial cells 
xii 
 
Figure 22 - GC inoculation had no significant effect on F-actin redistribution in 
ectocervical epithelial cells. 
Figure 23 - GC induce F-actin redistribution in endocervical epithelial cells, which 
involves NMII 
Figure 24 - GC induce Arp2/3 redistribution in endocervical epithelial cells, which 
involves NMII 
Figure 25 - GC inoculation induces the elevation of intracellular calcium levels in 
polarized but not in non-polarized T84 cells 
Figure 26 - Working model for Neisseria gonorrhoeae interaction with the three 
cervical regions, ectocervix, transformation zone (TZ), and endocervix 
 
 
 
 
 
 
 
 
 
xiii 
 
Chapter 1: Introduction 
1.1 Gonorrhea and Neisseria gonorrhoeae 
Gonorrhea is the second most commonly reported sexually transmitted infection in the 
world, as well as in the U.S. Each year, there are an estimated 78 million new 
infections with gonorrhoeae worldwide [1]. In 2017, a total of 555,608 new cases 
reported in the US, which is 18.6% increase since 2016, and 75.2% increase since the 
historic low in 2009 [2]. The rates of reported gonorrhea cases are the highest among 
teenagers and young adults. Furthermore, the reported rates of gonorrhea are masked 
by the asymptomatic nature of gonococcal infection [3]. Particularly, 50 to 80% of 
gonococcal infection in the female reproductive tract (FRT) is asymptomatic [4]. 
 
Gonorrhea is caused by an exclusive human pathogen, gram-negative bacteria 
Neisseria gonorrhoeae (GC). The cases of GC infection that are resistant to currently 
available antibiotics are on the rise, which complicates gonorrhea treatment [2]. The 
decrease of susceptibility to ceftriaxone, cefixime, and azithromycin has been reported 
in recent years [5]. The only treatment of uncomplicated gonorrhea recommended by 
CDC is the combination of ceftriaxone and azithromycin or doxycycline [6]. The 
increased resistance of GC to available antimicrobials underscores the urgent need to 
develop a new treatment for gonorrhea.  
 
1 
 
GC primarily infect urethra in male, inducing inflammatory responses and causing 
acute urethritis with purulent discharge [4]. A smaller percentage of male infection 
develop asymptomatic gonococcal urethritis, compared to female infection. However, 
there has been an increase in asymptomatic male infection in recent years [7]. In 
females, GC primarily infect lower genital tract when transmitted from men, causing 
asymptomatic cervicitis [4]. 45% of such asymptomatic infection in lower 
reproductive tract lead to ascending infection in the upper genital tract, causing pelvic 
inflammatory disease (PID), which can lead to permanent fallopian tube scarring and 
blockage, and subsequently infertility and ectopic pregnancy [4]. It has been shown 
that gonorrhea can increase the risk of coinfection with other sexually transmitted 
infections such as Chlamydia, and HIV [8]. The risk may be even higher in the view 
of the high percentage of asymptomatic, subclinical female infection.   
 
Asymptomatic female infection which does not get treatment and the increase in 
antibiotics resistance of GC makes gonorrhea an emerging public health crisis.  
 
1.2 Neisseria gonorrhoeae infection in females 
In females, GC primarily initiate infection in the female reproductive tract (FRT) from 
the mucosal surface of cervical epithelia [3, 9, 10]. GC are first deposited on vaginal 
epithelia. However, vaginitis caused by GC infection is rarely reported [3]. Vagina 
connects the outside of the body with the cervix, the gate of the female reproductive 
tract (FRT) that is often blocked by mucus plug. Ectocervix (Ecto) is the outmost part 
2 
 
of the cervix, connecting the cervix with vagina. The mucosal surface of both the 
vagina and the ectocervix is colonized by microbiota [11]. Ecto leads to the 
endocervix (endo) through transformation zone (TZ), which leads to the upper genital 
tract, composed of uterine, fallopian tube and ovary. Though previously believed to be 
sterile, several studies showed that the endocervix and upper genital tract also harbor 
microbial communities and is accessible by pathogens [12-14]. GC colonization in 
endo has been reported at a relatively low level compared to ectocervix. GC 
colonization in endo can lead to bacterial dissemination into the upper genital tract, 
also potentially causes inflammation, leading to serious complications, such as pelvic 
inflammatory disease (PID) and acute salpingitis [3, 4, 9]. Subepithelial GC was 
observed in the endocervix of patients’ biopsies [15], suggesting the penetration of GC 
from the mucosal surface of the endo into subepithelial tissue. The penetrated GC 
potentially trigger immune responses causing symptomatic infection and have 
opportunities to further disseminate through bloodstream causing disseminated 
gonococcal infection (DGI) [16]. Moreover, it has been shown that the serum level of 
progesterone, one of the sex hormones controlling the menstrual cycle, is significantly 
higher in the asymptomatic patients than the symptomatic patients [17]. It indicates the 
impact of menstrual cycle on GC infection progression, which adds more 
complications to female gonorrhea. However, how the localized GC infection in the 
cervix leads to such variety of clinical outcome is unknown.  
 
 
 
3 
 
 
 
 
Figure 1. Unique features of female gonorrhea 
A schematic diagram of Neisseria gonorrhoeae (GC) infection in the female 
reproductive tract (FRT). After first deposited in the vagina, GC initiate infection from 
the cervix, the gate of the FRT. GC have to overcome the barrier of the cervix and to 
ascend to the upper genital tract, which can cause pelvic inflammatory disease (PID). 
To disseminate into the bloodstream, GC need to penetrate across the epithelium of 
the FRT, causing disseminated gonococcal infection (DGI). Cervical gonorrhea can 
increase the risk of coinfection with HIV and other sexually transmitted infections 
(STI). 
 
 
4 
 
1.3 The cytoarchitecture of cervical epithelium 
The cervix is the gate of the female reproductive tract, protecting the sterile upper 
genital tract from the unsterile vagina, which is exposed to the outer environment. The 
cervix is also the initiation site of GC infection. To establish infection, GC need to 
overcome the cervical epithelium, the first line of defense. The cervical epithelium 
exhibits distinct morphology at different anatomical regions. The ectocervix, 
connecting with the vagina, is covered by multiple layers of non-polarized stratified 
squamous epithelial cells. At the bottom of the ectocervical epithelium, a single layer 
of round basal cells attached to the basal membrane. The basal cells can divide and 
differentiate to form the next few layers of parabasal cells. Further differentiation of 
parabasal cells forms the intermediate layers of polygonal cells, which further 
differentiate into the large and flattened cells forming the superficial layers. The 
formation of intermediate and superficial layers is subjected to the change of female 
hormone.  
 
Ascending to the upper genital tract, endocervix is covered by a single layer of 
polarized columnar epithelial cells. The endocervical epithelium does not form a 
flattened surface. In contrast, it is thrown into several longitude folds forming 
invaginations into the cervical stroma, also referred to as endocervical glands. The 
endocervical epithelial cells secrete mucus, forming a barrier against pathogen 
entrance. The components and amount of mucus secreted by the endocervical 
epithelial cells also change dramatically during the menstrual cycle [18], which further 
complicates the microenvironment of the endocervix.  
5 
 
 
The site where the endocervical epithelium meets with ectocervical epithelium is 
called the squamocolumnar junction (SCJ). Transformation zone (TZ) is the region 
located between newly formed SCJ and the original SCJ. The TZ is formed at the 
onset of puberty, when the endocervical epithelium proliferates under the influence of 
estrogen and extends into ectocervix, pushing the original SCJ into the ectocervix. The 
eversion of endocervical epithelium into the ectocervical region triggers squamous 
metaplasia. In the portion of endocervical epithelium invading into ectocervix, the 
reserve cells underneath the endocervical epithelium start to proliferate and gradually 
differentiate into ectocervical epithelial cells. The new SCJ is formed between the 
endocervical epithelium and the metaplastic TZ epithelium.  At the ecto end of TZ, TZ 
epithelium eventually becomes indistinguishable from the ectocervical epithelium. 
Thus, the cytoarchitecture of TZ is a transformation from multiple layers (ecto) to a 
single layer (endo), and from non-polarized (ecto) to polarized (endo). The length of 
TZ varies with age, hormonal status and from person to person.  
 
 
 
 
 
 
 
6 
 
 
 
 
 
 
 
 
7 
 
 
 
Figure 2. Distinct cytoarchitecture of the cervical epithelium at different 
anatomical regions (modified based on the Colposcopy and treatment of cervical 
intraepithelial neoplasia: a beginners’ manual, and [19]). 
A) A schematic diagram of the cytoarchitecture of cervical epithelium. The ectocervix 
is covered by the multiple-layered stratified squamous epithelial cells. The endocervix 
is lined with the single-layer columnar epithelial cells. The transformation zone (TZ) 
is the region located between ecto and endocervix. B) A schematic representation of 
the formation of TZ. TZ is located between the original squamocolumnar junction 
(SCJ) and the newly formed SCJ after puberty: (a) before puberty, (b) after puberty 
and at early reproductive age, (c) around 30 years old, and (d) perimenopausal. C) The 
TZ epithelium is formed through the squamous metaplasia. (a) polarized endocervical 
epithelium with reserve cells (arrows) underneath, (b) proliferation of reserve cells 
forming layers of reserve cells underneath the original endocervical epithelium, (c) 
further proliferation and differentiation of reserve cells forming immature metaplastic 
epithelium, and (d) mature squamous metaplastic epithelium indistinguishable from 
the original squamous epithelium. 
 
 
 
 
8 
 
Although data from clinical researches using patient biopsies and papsmear has shown 
the colonization in all three cervical regions, and penetration in TZ and endocervix 
[15, 20], little is known about the mechanism used by GC to infect different types of 
cervical epithelial cells in vivo. The questions that does the complicated 
cytoarchitecture of cervical epithelium have any impact on the various clinical 
outcomes of the localized cervical GC infection remains not answered.  
 
1.4 Neisseria gonorrhoeae and its virulence factors 
To survive in the various microenvironment of the cervix and facilitate infection, the 
surface molecules of GC undergo phase and antigenic variation. Pili are filamentous 
structure expressed on the surface of gram-negative bacteria. GC express the Type IV 
pili, which undergo both phase and antigenic variation [21]. Type IV pili are 
composed of PilE, a major pilus subunit protein, PilC, a minor pilus-associated 
protein, and PilT. The pilE gene can go through antigenic variation by recombination 
of pilE locus with one or more silent pil loci (pilS) [22]. Most strains harbor two pilC 
genes that are highly homogeneous, expressing two different proteins, PilC1 and PilC2 
[23]. PilC protein has been identified as the adhesive component of pili. It is located at 
the tip of pili and mediates initial attachment of GC to host cells [24, 25]. PilT protein 
is involved in the GC twitching motility and pili retraction, thus mediating 
microcolony formation and the bringing GC close to host cell membrane [26-28].  Pili 
interact with complement receptor 3 (CR3) expressed on the mucosal surface of the 
FRT [29].  
 
9 
 
Opacity associated proteins (Opa) are a group of integral outer membrane proteins. A 
single bacterial cell can express up to 11 different Opa genes. Each of the 11 different 
Opa genes can be turned on or off independently due to the phase variation [30, 31]. 
The predicted secondary structure of Opa proteins has a β-barrel structure formed by 8 
anti-parallel membrane spanning β-strands with four extracellular loops. Antigenic 
variations of Opa proteins are predominantly observed in the central two hypervariable 
(HV) loops [31]. Opa proteins can be grouped into two classes according to the 
receptors on human cells they bind to. The majority of the 11 Opa proteins interact 
with carcinoembryonic antigen-related cell adhesion molecules (CEACAMs), such as 
OpaH [21, 32]. OpaC, representing other class, interacts with heparan sulphate 
proteoglycan (HSPG) [33]. Studies showed that GC isolates from female patients at 
different stages of the menstrual cycle express different isoforms of Opa [34]. 
However, how the heterogeneous expression of Opa protein affects GC infection in 
the human cervix is not fully understood.  
 
As Gram-negative bacteria, GC possess lipooligosaccharide (LOS) within its outer 
membrane, which undergoes antigenic variation. The structure of LOS is composed of 
three oligosaccharide chains tethered to a lipid A core [35]. The terminal 
oligosaccharide of LOS varies, depending on the expression of glycosyl transferases 
involved in LOS biosynthesis [36, 37]. It has been shown that LOS function as a 
ligand to the human asialoglycoprotein receptor (ASGP-R), contributing to the uptake 
of GC into primary human urethral epithelial cells [38]. Our lab has previously shown 
10 
 
that the terminal oligosaccharide structure of LOS alters the level of invasion into 
human endometrium cell line ME-180 [39]. 
 
Porin (PorB) is the most abundant protein in the outer membrane of GC [4]. PorB 
forms water-filled channel within gonococcal outer membrane, through which ion and 
small nutrients can transverse. However, PorB does not undergo phase and antigenic 
variation. It is present as one of the two isotypes, PI.A or PI.B [4]. It has been shown 
that porin can translocate into the plasma membrane of the eukaryotic cell [40] and 
further the mitochondrial membrane [41]. PorB has been shown to can manipulate 
phagosome maturation [42, 43] and triggers apoptosis [43, 44]. However, the anti-
apoptotic effect of PorB was also observed in the human urethral epithelium, which 
was thought to be a strategy for GC to survive and proliferate within epithelial cells 
[45].  
 
1.5 Neisseria gonorrhoeae engaged host receptors and downstream signaling 
Each of the GC surface virulence factors interacts with specific receptors on host cells 
triggering downstream signaling to facilitate infection. It has been shown that CR3 
plays important roles in GC infection as a receptor for pili, involving porin and LOS 
as well. CR3 is a member of the integrin family, known as integrin αMβ2, and is 
expressed by professional phagocytes. CR3 is the receptor for iC3b, a component of 
the complement system, and mediates phagocytosis of complement-opsonized 
microorganisms [46]. Dr. Apicella’s groups first showed the expression of CR3 in 
primary human ectocervical and endocervical epithelial in culture and in the human 
11 
 
cervical tissue. GC adhered to primary cervical epithelial cells colocalized with CR3 
antibody staining [29]. The cleavage products of C3, C3b-α, C3b-β, iC3bα1i, iC3bα2i, 
and C3dg, were found to be deposited onto GC in the infected primary cervical 
epithelial cells as well as in the cervical secretion specimen from patients. The 
deposition of C3b on GC surface occurs through the lipid A core of LOS [47]. Pili 
bind to the I-domain of CR3 in cooperating with porin and iC3b-opsonization, which 
is required for adhesion and subsequent invasion into primary cervical epithelial cells. 
[48, 49]. 
 
The majority of the 11 Opa proteins has been shown interacting with 
Carcinoembryonic antigen-related cell adhesion molecules (CEACAM) [50]. 
CEACAMs are first discovered as cell adhesion molecules in cancer cells, hence the 
name. CEACAM proteins mediate intercellular adhesion through homophilic and/or 
heterophilic interaction [51]. In humans, multiple isoforms have been identified. All 
CEACAM proteins belong to the immunoglobulin (Ig) superfamily and share similar 
overall structure at the N-terminus, one variable (V)-like domain followed by none or 
several constant C2-like Ig domains [51, 52]. At the C-terminus, different isoforms 
possess different structures, an intracellular domain or a glycosylphosphatidylinositol 
(GPI) anchor [52]. In human, CEACAM1, CEACAM3, CEACAM5, and CEACAM6 
have been shown to bind to Neisserial Opa proteins [53], among which CEACAM1 
and CEACAM3 possess a cytoplasmic domain. The long splicing form of CEACAM1 
(CEACAM1L) has an immunoreceptor tyrosine-based inhibitory motif (ITIM) at the 
C-terminus of its cytoplasmic domain. In contrast, CEACAM3 has an immunoreceptor 
12 
 
tyrosine-based activation motif (ITAM) [52]. These features suggest the ability of 
CEACAM1 and CEACAM3 to transducing signals upon GC infection. 
 
CEACAM1, CEACAM3, and CEACAM6 are expressed in granulocytes. Lee et al. 
showed that upon Opa-CEACAM1 binding, GC recruited CEACAM1 to the sites of 
attached GC on the surface of primary CD4+T cells. Upon T cell receptor (TCR) 
ligation, CEACAM1 was phosphorylated on the tyrosine residues harbored in the 
ITIM. The phosphorylated ITIM recruited and activated the tyrosine phosphatase 
SHP-1 and SHP-2, which inhibited the phosphorylation of ZAP-70, a key adaptor 
protein in T cell signaling pathway, consequently inhibiting T cell activation [54]. 
Using human myelomonocytic cell line JOSK-M cells, Hauck et al. showed that 
inoculation with GC that express CEACAM-binding Opa triggered activation of acid 
sphingomyelinase (ASM) and downstream signaling involving Src-like tyrosine 
kinases Hck and Fgr and Jun-N-terminal kinases (JNK), eventually lead to activation 
of the Rho GTPase Rac1 and bacteria uptake [55].  CEACAM3 is expressed 
exclusively in neutrophils. Dr. Gray-Owen’s group has done a series of studies 
showing that upon Opa-CEACAM3 interaction, tyrosine residues in the ITAM are 
phosphorylated by Src-family kinases, which is necessary for GC uptake by 
neutrophils [56]. The phosphorylated ITAM recruited and triggered the 
phosphorylation of the cytoplasmic tyrosine kinase Syk and activated downstream 
signaling pathways including phospholipase C (PLC), small GTPase and 
phosphatidylinositol 3-kinase (PI3K). This signaling activation led to the 
rearrangement of the actin cytoskeleton and GC uptake [57-59]. 
13 
 
 
Epithelial cells express CEACAM1, CEACAM5, and CEACAM6. Using CEACAM1 
transfected T293 cells, Voges et al. showed that the Opa-CEACAM1 dependent 
uptake of GC by epithelial cells requires PI3K activity and the cytoplasmic domain 
and the extracellular IgC2-like domain of CEACAM1 [60]. Using a CEACAM5 
transgenic mouse model, ME-180 cells, and CEACAM1 or CEACAM6 transfected 
T293 cells, Muenzer et al. showed that Opa-CEACAM1 interaction inhibited 
epithelial exfoliation through upregulation of CD105, one of the transforming growth 
factor- β1 receptor family (TGFβ1R), and subsequent activation of integrin β1. The 
inhibition of epithelial exfoliation through this pathway required the cytoplasmic 
domain of CEACAM1 [61, 62].  
 
The expression and distribution of various epithelial CEACAMs in human cervix has 
not been studied in detail. Islam et al. detected CEACAM1 on the apical surface of 
endocervical epithelial cells, but a low level of CEACAM1 and a moderate level of 
CEACAM5 in the intermediate and superficial layers of the ectocervical epithelium. 
They did not detect any expression of CEACAM6 along the cervical epithelium [63]. 
However, whether the differential expression of CEACAMs in different cervical 
regions affects host signaling response and GC infection process in corresponding 
regions, as well as clinical outcomes of GC infection,  has not been well studied. 
 
 
14 
 
 
 
 
 
 
 
 
 
 
 
15 
 
 
Figure 3. Structure and downstream signaling of carcinoembryonic antigen-
related cell adhesion molecule (CEACAM) triggered by Opa (Modified based on 
[52, 64]). 
A) A schematic diagram of the CEACAM structure. At the N-terminus, one variable 
(V)-like domain is followed by none (CEACAM3 and 4) or several constant C2-like 
Ig domains (CEACAM1, 5 and 6). At the C-terminus, CEACAM 1, 3 and 4 possess a 
cytoplasmic domain.  CEACAM5 and 6 are GPI anchored proteins. B) In the 
phagocytic cells, such as JOSK-M and neutrophils (a), Opa interaction with 
CEACAM1, 6 or 3 leads to activation of Src family kinases Hck and Fgr through 
acidic sphingomyelinase (ASM) resulting in tyrosine phosphorylation (P-Tyr), and 
activation of the Rho GTPase Rac1. This pathway eventually leads to phagocytic 
bacterial uptake through cytoskeleton rearrangement. In T cells (b), Opa-CEACAM1 
interaction activates SHP1/2 leading to the inhibition of T cell activation through 
inhibiting ZAP-70. In epithelial cells (c), Opa triggered downstream signaling 
involves CEACAM1, 5 and 6. This pathway leads to CD105 and integrin activation 
through PI3K, resulting in the enhanced focal adhesion of epithelial cells. Full arrows, 
direct activation of the downstream pathway. The dashed arrows, unknown 
downstream pathway. 
 
 
 
 
16 
 
Heparan sulphate proteoglycans (HSPG) is the other host receptors for Opa proteins. 
Proteoglycans (PG) are membrane-anchored core proteins, modified with 
glycosaminoglycans (GAGs) side chains. The side chains of HSPG are composed of 
heparan sulphates (HS). A specific isoform of Opa proteins (OpaHSPG) interacts with 
HSPG on the mucosal surface of human epithelial cells [65]. Opa-HSPG interactions 
have been shown to mediate intimate adherence and invasion of GC into various 
epithelial cell lines. This process is facilitated by extracellular matrix, such as 
fibronectin (FN) in HEp-2 cell [66] and vitronectin (VN) in Chinese hamster ovary 
(CHO) cells [67] and Hela cells [66, 68]. Using Chang conjunctiva cells, Grassme et 
al. showed that the Opa-HSPG mediated GC uptake is dependent on the activity of 
phosphatidylcholine phospholipase C (PC-PLC), which generates diacylglycerol 
(DAG). DAG appears to activate acidic sphingomyelinase (ASM), which produces 
ceramide and promote GC uptake [69]. Syndecan is a major group of transmembrane 
HSPG. Freissler et al. showed expression of syndecan-1 and syndecan-4 in various 
epithelial cell lines, including the human endometrial cell line ME-180, human 
conjunctival epithelial cells (Chang), human endometrial cell line Hec-1-B and human 
cervical cell line Hela. They found that Opa-HSPG mediated GC uptake requires the 
cytoplasmic tail of both syndecans and the possible involvement of protein kinase C 
(PKC) and phosphatidylinositol-4,5-bisphosphate [PI(4, 5)P2] in this process [70]. 
 
 
 
 
 
17 
 
 
 
 
 
 
 
 
 
 
 
 
 
18 
 
 
 
 
Figure 4. Structure and downstream signaling of heparan sulfate proteoglycans 
(HSPG) triggered by Opa (Modified base on [71] and doi:10.4172/2161-
0703.1000157). 
A) A schematic diagram of the HSPG structure using syndecan as an example. HSPG 
is composed of a membrane-anchored core protein and several heparan sulphates (HS) 
side chains. B) GC Opa-induced HSPG signaling has been studied in Chang 
conjunctival epithelial cells (a), where Opa-HSPG interaction triggers activation of 
phosphatidylcholine phospholipase C (PC-PLC), leading to acidic sphingomyelinase 
(ASM) activation through diacylglycerol (DAG) and eventually uptake of GC. In Hela 
and CHO cells (b), Opa-HSPG interaction, facilitated by the vitronectin (VN) bridged 
integrin αvβ5 signaling leads to protein kinase C (PKC) activation. In Hep-2 larynx 
carcinoma cells (c), OpaC-HSPG interaction, facilitated by the fibronectin (FN) 
bridged integrin α5β1 signaling leads to PKC activation. Both pathways (b, c) are 
involved in the efficient uptake of GC in corresponding cell lines. Full arrows, direct 
activation of the downstream pathway. The dashed arrows, unknown downstream 
pathway. 
 
 
19 
 
The human asialoglycoprotein receptor (ASGP-R) is a membrane glycoprotein 
belonging to C-type lectin. ASGP-R recognizes terminal asialylated galactose or N-
acetylgalactosamine residues and mediates clathrin-dependent endocytosis [72]. Using 
hepatoma cell line HepG2, Porat at al. showed that GC producing LOS with lacto-N-
neotetraose-terminal bind specifically to ASGP-R, and the binding enhanced the 
expression of ASGP-R [73]. GC-induced increase in the surface level of ASGP-R was 
also observed in primary human urethral epithelial cells (PHUECs) [74]. Furthermore, 
the interaction of ASGP-R with the lacto-N-neotetraose-terminal of gonococcal LOS 
has been shown to play a role in GC invasion into PHUECs in culture, by recruiting 
clathrin to the adherence site of GC microcolonies. Inhibition of clathrin-mediated 
endocytosis inhibited GC invasion into PHUECs [38].  
 
The members of the ErbB family, epithelial growth factor receptor (EGFR, or ErbB1) 
and ErbB2, are commonly expressed in epithelial cells and have been shown to play a 
role in bacterial infection including GC. Our lab has previously shown that GC 
inoculation induced cellular redistribution of EGFR and ErbB2 from the basolateral 
surface to the apical surface under GC adherent sites of polarized epithelial cells and 
the activation of EGFR and ErbB2. However, GC do not directly engage EGFR and 
ErbB2, instead transactivate them by inducing the transcription of EGFR ligands. GC-
induced transactivation of EGFR and ErbB2 contributes to GC invasion into human 
endometrial epithelial cell line HEC-1-B and ME180 [75] and GC transmigration 
across the polarized human colonic epithelial cell line T84 and HEC-1-B cell by 
disruption of the apical junction [76]. GC has been shown to activate several 
20 
 
downstream signaling cascades of EGFR, including mitogen-activated protein 
kinase/extracellular signal-related kinase (MAPK/ERK) [77, 78], phospholipase C γ 
(PLCγ) and calcium flux, and PI3K [79]. Dr. So’s group found that GC recruited 
EGFR together with f-actin, ezrin, I-CAM and CD44 in A431 cells, in a PilE/PilC 
dependent manner [80]. Inoculation of GC activated PI3 kinase and Akt, which is 
involved in the pilT-dependent invasion of GC into A431 cells [81]. However, 
whether GC trigger above pathways through EGFR activation remains unknown. 
 
While GC have been shown to bind to various host surface receptors through their 
surface molecules and trigger signaling in various experimental systems, whether 
these occur in vivo, such as during GC infection of the three cervical epithelium with 
heterogeneous cytoarchitecture, is unknown. The expression level and distribution of 
host receptors may vary in different cervical regions. It is reasonable to hypothesize 
that GC interaction with cervical epithelial cells and host cell signaling responses vary 
with the expression of host receptors and the intrinsic signaling machinery in the 
different types of cervical epithelial cells they encounter. 
 
1.6 Neisseria gonorrhoeae interaction with human epithelial cells 
Epithelial cells, which lines the outer surface of organs protecting hosts from the outer 
environment, are the front line defense against bacterial infection. When GC is first 
deposited onto the lumen of the genital, Pili, the long filamentous structure extending 
from the surface of GC, establish the initial interaction with the mucosal surface of 
epithelia [24]. The retraction of pili overcomes the electrostatic force at the luminal 
21 
 
surface of epithelia and brings the bacteria closer to the luminal surface of epithelia 
[82]. It has been shown that, only at this close distance, Opa proteins can interact with 
their receptors on the epithelial cells, CEACAM and HSPG, which establish an 
intimate adherence or colonization [21]. LOS also contributes to the intimate 
adherence by interacting with ASGP-R [73]. 
 
Following colonization, GC can invade into epithelial cells intracellularly. 
Intracellular GC has been observed in the gram stained cervical papsmear samples 
[83]. Fichorova et al. showed that immortalized endocervical, ectocervical, and 
vaginal epithelial cells derived from the primary cell can uptake GC [84]. GC invasion 
into epithelial cells in the fallopian tube organ culture has also been observed [85]. 
The Opa-CEACAM interaction has also been shown to contribute to GC invasion in 
the CEACAM-transfected Hela cell line [86]. Opa-HSPG mediated invasion has been 
shown in various epithelial cell lines such as ME-180 and CHO-K1 [67, 87]. Our lab 
has previously shown that the lacto-N-neotetrose in LOS is required for GC invasion 
[39]. Furthermore, GC can manipulate the distribution and activity of the epidermal 
growth factor receptor (EGFR) in the polarized HEC-1-B cells to facilitate invasion 
[75]. 
 
The fate of GC inside the epithelial cells remains unclear. It has been shown that a 
subpopulation of GC can survive intracellularly by resisting the proteolytic activity of 
host lysosomes with the help of Porin [42]. The type 2 IgA1 protease secreted by GC 
22 
 
can also promote intracellular survival of GC [88]. Using primary human cervical 
epithelial (pex) cells, Edwards et al showed GC intracellular survival is affected by the 
level of the female sex hormones estrogen and progesterone in culture [89]. 
 
The intracellular survival of GC may eventually lead to GC transcytosis or penetration 
into subepithelial tissues. McGee et al. first observed penetration of GC into 
subepithelial tissue in fallopian tube organ culture by transmitted electron microscopy 
(TEM) [85, 90]. GC transmigration across epithelial cells has also been observed in 
various cell line culture. Ilver et al. described the transcellular passage of GC in three 
cell line cultures, HEC-1-B (endometrial adenocarcinoma), ME-180 (cervical 
carcinoma) and Chang conjunctiva, which involves pili phase variation [91].  Data 
from our lab showed that GC can disrupt the apical junction in polarized T84 and 
HEC-1-B cell by activating EGFR, which facilitates transmigration [76]. Wang et al. 
showed the transcytosis of GC expressing CEACAM-binding Opa through the 
polarized epithelial T84 cell requires the cytoskeleton and motor proteins [92]. Opa-
CEACAM interaction has been shown to mediates GC transcytosis across T84 cell 
culture [93]. However, data from our lab and Dr. Stein’s lab show that Opa expression 
reduces the level of GC transmigration in polarized human epithelial cell line models 
as well as human endocervical tissue culture [94, 95]. Stein et al. suggest that Opa 
expression reduces the level of transmigration across T84 cell culture by inducing 
CEACAM redistribution [94]. Our lab showed that GC can transmigrate across 
polarized T84 epithelial cells and polarized endocervical epithelial cells in tissue 
explants by activating the actin motor non-muscle myosin II (NMII), and the 
23 
 
expression of CEACAM-binding Opa reduces the level of transmigration [95]. The 
results of studies on the role of GC surface molecules on infection vary with GC 
strains and infection models that were used. Despite the inconsistency, taken together, 
these data showed that GC can use phase variation of its surface molecules to 
manipulate epithelial cell responses and facilitate transmigration. 
 
GC in the paracellular space between epithelial cells have also been observed in three 
different cell line culture by TEM [91], suggesting another route of transmigration 
across the epithelium. Ward et al. observed a population of GC in between epithelial 
cells in fallopian tube organ culture [96, 97]. However, whether GC penetrate into 
subepithelial tissues preferentially through an intracellular route or a paracellular route 
is unknown. The mechanism underlying GC penetration remains to be elucidated.   
 
GC has been shown to induce exfoliation of epithelial cells to facilitate penetration 
and infection in fallopian tube organ culture [98], cell line culture[62], mouse model 
[61], and endocervical tissue explant model [95]. However, understanding of the 
underlying mechanism remains limited. Dr. Hauck’s group first showed that Opa-
CEACAM interaction inhibited epithelial exfoliation by increase the expression of 
CD105 and downstream activation of integrin β1 in both human cell line and 
transgenic mouse model that expresses CEACAM [61, 62]. The results from our lab 
showed that expression of CEACAM-binding Opa reduces epithelial exfoliation by 
inhibition of GC-induced calcium flux, NMII activation, and epithelial cell-cell 
24 
 
junction disruption in cell line and endocervical tissue explant models [95]. 
Exfoliation naturally occurs during epithelial turnover, which is controlled by a 
balance between eliminating damaged cells and replenishing new cells while 
maintaining the integrity of the epithelium [99]. Epithelial exfoliation is also a part of 
the defense mechanism against infection: eliminating epithelial cells colonized or 
damaged by pathogens. However, epithelial exfoliation induced by pathogens may 
leave unsealed gaps in the epithelium, especially the single-layered columnar 
epithelium, which may provide a portal for pathogens to reach subepithelial tissues.  A 
similar process has been observed in the case of uropathogenic Escherichia coli 
(UPEC) infection in the bladder [100, 101]. However, epithelial exfoliation does not 
seem to be sufficient to eliminate UPEC infection. In contrast, it may expose the 
underlying epithelium to UPEC. The mechanism of epithelium turnover and pathogen-
mediated manipulation of this process for infections have been studied extensively in 
intestine epithelium. However, the turnover of the cervical epithelium and its 
regulation is unknown.  
 
Despite the numerous studies on GC-epithelial interaction discussed above, the 
mechanism by which GC infect the heterogeneous cervical epithelium remains 
elusive, partly because most previous research was done in experimental systems that 
did not mimic human cervical epithelium. Consequently, whether GC modulate the 
above infection processes, colonization, invasion, penetration, and epithelial 
exfoliation, in different types of cervical epithelial cells in the same way as shown 
using currently available infection models is unknown. 
25 
 
1.7 Epithelial cell-cell junction and its regulation 
The integrity of epithelium is essential for the protection of the host from the 
pathogen. Epithelial cells are organized into the epithelium by the formation of 
junctional complexes with neighboring cells [102]. Epithelial junction formation keeps 
the neighboring cells in strict contact to form single- or multiple-layered three 
dimensional architecture. There are two major types of epithelial junction complexes, 
adherens junction (AJ) and tight junction (TJ). 
 
In epithelial cells, one of the major components of adherens junction is a single pass 
transmembrane protein, E-cadherin, and various adaptor proteins regulate and connect 
E-cadherin to the actin cytoskeleton [103]. The extracellular domain of E-cadherin 
consists of five repetitive subdomains called cadherin repeats (EC). Each EC contains 
the calcium-binding sequence. E-cadherin mediates cell-cell contact by intercellular 
homophilic interaction of EC domains in a calcium-dependent manner. Thus, adherens 
junction is characterized by a 10-20 nm distance between the plasma membrane of two 
neighboring cells. The cytoplasmic domain of E-cadherin binds a number of proteins, 
collectively called catenins. β-catenin is associated with E-cadherin and also α-catenin 
through p120 catenin (p120ctn). α-catenin has been shown to link the E-cadherin-
catenin complex to the actin cytoskeleton through unknown components. The binding 
of E-cadherin with catenins regulates the stability of the adherens junction [104]. The 
binding of E-cadherin with β-catenin is enhanced by the phosphorylation of three 
serine residues (S684, S686, S692) in the cytoplasmic domain of E-cadherin, which is 
mediated by casein kinase II (CKII) and glycogen synthase kinase 3β (GSK-3β) [105-
26 
 
107]. In contrast, the phosphorylation of tyrosine residues (Y489, Y654) in β-catenin 
reduces the affinity to E-cadherin, and the phosphorylation at Y142 reduces the 
affinity to α-catenin. These phosphorylation has been shown involving abelson (Abl) 
kinase at Y489 [108], pp60c-src at Y654 [109], and Fer kinase at Y142 [110]. The 
tyrosine phosphorylation of β-catenin is balanced by tyrosine phosphatase, such as 
SHP1 (PTP1B, protein tyrosine phosphatase 1B) [107, 111]. Without binding to β-
catenin, the cytoplasmic domain of E-cadherin becomes unstructured, which leads to 
its internalization and degradation [112-114].  
 
Similar to the adherens junction, the tight junction is composed of transmembrane 
proteins, occludins, claudins and junctional adhesion molecules (JAMs), and adaptor 
proteins connecting transmembrane proteins to the cytoskeleton. Occludin and claudin 
are tetraspan proteins with two extracellular loops, mediating intercellular homophilic 
interactions with neighboring cells. JAMs are single-pass transmembrane proteins, 
which have two extracellular immunoglobulin-like domains [115]. The tight junction 
is characterized by an intercellular space ranging from 0.2 to 0.5 μm. Zonula 
occludens (ZOs) are the best-studied adaptor proteins in tight junction. ZO proteins are 
peripheral membrane-associated components of the tight junction, which belongs to 
the membrane-associated guanylate kinase (MAGuK) homolog family. The C-
terminus of occludin and claudin bind to the PDZ-1 domain of ZO proteins, which is 
required for directing occludin and claudin to the tight junction [116]. JAMs also 
interact with ZO proteins. The permeability of the tight junction is regulated by a wide 
range of growth factors, cytokines, and hormones. Protein kinase C (PKC) mediated 
27 
 
phosphorylation of occludin and tyrosine phosphorylation of ZO-1 have been 
suggested to decrease the epithelial permeability [116].  Actomyosin rings associated 
with the tight junction and it’s regulation by myosin light chain (MLC) also play 
important roles in tight junction regulation [117]. Two functions have been proposed 
for the tight junction, fence function, and gate function. The fence function is the 
ability of the tight junction to prevent the lateral movement of membrane molecules 
between the apical and basolateral domains. Gate function is the ability of the tight 
junction to control the paracellular permeability of the epithelium [104]. 
 
In the epithelium consisting of single-layered polarized columnar epithelial cells, such 
as the intestinal epithelium, the apical junction (AJ) is formed to maintain the integrity 
and polarity of the epithelium, forming apical and basolateral domain [102]. The 
apical and basolateral membrane have distinct membrane and protein components 
which are generated and maintained by the polarized membrane trafficking machinery. 
The polarized trafficking machinery includes secretory and endocytic pathways that 
sort and delivers different proteins and lipids to the apical or basolateral membranes 
[118]. The apical junction is composed of both the tight junction and the adherens 
junction and is located close to the lumen surface of polarized epithelial cells, with the 
tight junction at the apical end of the lateral membrane domain. Assembly of the 
adherens junction leads to the formation of the tight junction to stabilize epithelium. 
However, the depletion of E-cadherin by RNA interference in Madin-Darby Canine 
Kidney (MDCK) Epithelial Cells did not lead to the disruption of the tight junction, 
28 
 
indicating that the adherens junction formation plays a role in apical junction 
assembly, but not required for maintenance [119].  
 
Rho family small GTPases, such as Rho, Rac, and Cdc42 and Rap1, play important 
roles in cell-cell junction formation, maintenance, and dynamics. During the adhesion 
junction establishment and maintenance, the initial homophilic trans-interaction of E-
cadherin between neighboring cells eventually activates Rac through recruitment of 
the tyrosine kinase c-Scr. The recruitment of c-Src leads to phosphorylation of Crk 
adaptor protein and C3G, one of the Rap1-GEF (Guanine nucleotide exchange 
factors), and subsequent activation of Rap1. Rap1 activation mediates PI3K activation, 
and eventually Rac activation through Vav2 [120, 121]. Such Rac activation together 
with Cdc42 stabilizes adherens junction by inhibiting E-cadherin endocytosis through 
reorganization of the actin cytoskeleton [120, 122]. Once stable adhesion junction is 
established, Cdc42 comes into play in the regulation of adhesion junction stability by 
controlling adhesion junction components turnover. Cdc42 also activates N-WASP, 
and together they interact with Arp2/3 complex to promote polymerization of the 
branched actin filament. This process has been shown to be required to maintain 
adherens junction integrity.  
 
 
 
29 
 
 
Figure 5. Organization and regulation of the tight junction (TJ) and the adherens 
junction (AJ) (Modified based on [121]) 
The tight junction is located at the apical end of the lateral membrane. Tight junctions 
are composed of transmembrane proteins, JAM, claudin and occludin, and an adaptor 
protein, ZO. ZO proteins connect tight junction to the actin cytoskeleton. E-cadherin is 
the major transmembrane proteins of the adherens junction in epithelial cells. E-
cadherin is connected to p120 catenin, which interacts with β-catenin and α-catenin. β-
catenin and α-catenin connect the adherens junction to the actin cytoskeleton and 
regulates the assembly and disassembly of the adherens junction through small 
GTPases and formin and Arp2/3 mediated actin polymerization.  
 
30 
 
The distinct cytoarchitecture of endo and ectocervical epithelia is maintained by 
differential epithelial cell-cell junction complexes and organization. The multiple 
layers of stratified squamous epithelial cells covering the ectocervix are held together 
mostly by adherens junction [123]. Visualized by SEM, the luminal layer of 
ectocervix is flattened and appears to be detached from the layers underneath [123, 
124]. Shown by TEM and immunofluorescence, the most luminal layers (three to four 
layers) are cornified and lose cell-cell contact, thus permeable to mucosal IgG. The 
adherens junction staining is absent from the luminal layers and only detected three to 
four layers from the luminal surface, present through the intermediate, suprabasal and 
basal layers. Staining of tight junction proteins, claudin-1, JAM 3, and ZO-1 also was 
also detected most abundantly in the intermediate, suprabasal and basal layers of the 
ectocervical epithelium [123, 125].  
 
The single-layered polarized epithelium covering the endocervix possesses the 
classical apical junction. SEM visualized abundant microvilli at the apical surface of 
endocervical epithelial cells [126]. TEM visualized the tight and adherens junctions 
located close to the apical surface between neighboring cells with the tight junction on 
the top, showing a classical columnar epithelial morphology. Because of the formation 
of the tight junction, the endocervical epithelium is not permeable to luminal IgG. 
Immunofluorescence microscopy showed staining of claudin-1, occludin, ZO-1, and 
E-cadherin between epithelial cells, but not at the basal membrane attached to the 
extracellular matrix (ECM) [123]. 
 
31 
 
 
  
Figure 6. Organization of cell-cell junction complexes in the human cervical 
epithelium [123] 
A schematic diagram of the cell-cell junction organization in the endocervix and 
ectocervix. A) The paracellular space of the endocervical epithelium is sealed by 
classic apical junctions composed of the tight junction (claudin, occludin, JAM, and 
ZO-1) and the adherens junction (E-cadherin). Neighboring cells are also connected 
by desmosomes. B) The parabasal and basal layers of epithelial cells in the ectocervix 
form the adherens junction, but the most luminal layers of the ectocervical epithelium 
gradually lose cell-cell contact and the expression of adherens junction proteins. 
 
 
32 
 
It has been shown that cell-cell junction is a frequent target of gut pathogens to breach 
the epithelial barrier and to establish infection. Enteropathogenic E. coli (EPEC) has 
been shown to deliver effector proteins, EspM, EspF, and Map, into intestinal 
epithelial cells causing disruption of the tight junction and consequently the epithelial 
integrity [127]. EspF has been shown interacting with N-WASP and the endocytic 
regulators sorting nexin 9 (SNX9), modifying the tight junction through the actin 
cytoskeleton and the endocytic pathway [128]. Map has been shown to activate 
Cdc42, modifying cell-cell junction through actin remodeling [127]. EspM has been 
shown to disrupt the tight junction by activating RhoA [129] and subsequent NMII. 
Similarly, Salmonella Typhimurium (S. Typhimurium) and Helicobacter pylori (H. 
pylori) also deliver effector proteins into intestinal epithelial cells, such as SopB and 
SopE for S. Typhimurium and CagA for H. pylori which target myosin light chain 
(MLC) to disrupt the tight junction and the adherens junction [99].  
 
However, the effect of GC infection on cell-cell junction has not been studied in 
details. Our lab has shown that GC can disrupt the apical junction to facilitate GC 
transmigration in both polarized HEC-1-B and T84 cell lines. GC induce the junction 
disruption by activating EGFR [76] and NMII [95].  However, it remains unknown 
whether GC disrupt epithelial cell-cell junctions of the human cervical epithelial cells 
in vivo using similar mechanisms. The distinct properties of cell-cell junctions in the 
ecto and endocervical epithelia potentially alter the strategy that GC use to infect 
corresponding cervical regions. Moreover, the organization of cell-cell junction in the 
33 
 
TZ epithelium has not been examined, despite the fact that penetration [15] and high 
level of colonization of GC at the TZ has been shown [20] [124].  
 
1.8 The role of the actin cytoskeleton in Neisseria gonorrhoeae infection 
The actin cytoskeleton regulates the remodeling of the plasma membrane in response 
to pathogen infection at pathogen-host cell interface. It also regulates the dynamics of 
cell-cell junction assembly and disassembly, which can also be exploited by pathogens 
to facilitate infection. 
 
The actin cytoskeleton has been shown to play an important role in GC infection in 
epithelial cells, but the detailed mechanism is still lacking. Edwards et al. showed that 
GC inoculation induced extensive membrane ruffling and cytoskeleton rearrangement 
in the primary endocervical and ectocervical epithelial cells [130]. Griffiss et al. 
showed that GC use both Opa and Pili to induce microvilli elongation at the surface of 
HEC-1-B cells [131]. Our lab has previously shown that the lacto-N-neotetraose in 
LOS induced elongation of microvilli warping around GC and form the intimate 
interaction between GC outer membrane and the plasma membrane of ME180 cells, 
which leads to the uptake of GC [39]. Grassme et al. showed that Opa promoted 
bacterial entry in an actin polymerization-dependent manner, as treatment of 
cytochalasin D that disrupts the actin cytoskeleton inhibited GC invasion [132]. Dr. 
So’s group showed that pili also played a role in inducing dramatic membrane 
microvilli and cortical plaque [133, 134]. There is no effector protein secreted by GC 
discovered so far. GC entry into epithelial cells is shown to be mediated by the 
34 
 
bacterial surface molecule and its interaction with host receptors, which fits the zipper 
mechanism. But the membrane ruffling and microvilli formation observed by multiple 
groups argue against it [59]. As the membrane ruffling and microvilli formation are 
mediated by the actin cytoskeleton, GC targeted actin regulators have not been 
identified. The role and underlying mechanism of actin reorganization for GC 
infection remain unclear.  
 
The plasma membrane remodeling induced by pathogens has been studied the most in 
the process of bacteria entry into non-phagocytic cells, by stimulating endogenous 
uptake processes, such as phagocytosis and macropinocytosis. Such processes all 
involve plasma membrane protrusion driven by Arp2/3 complex-mediated actin 
polymerization [135]. Arp2/3 complex is one of the actin nucleators and is composed 
of seven subunits, including Arp2, Arp3, and five additional subunits, ARPC1-ARPC5 
(actin related protein complex 1-5).  Arp2/3 complex is recruited and positioned at the 
sides of existing actin filament by actin nucleation-promoting factors (NPF), such as 
N-WASP and cortactin. This allows the Arp2/3 complex to initiate actin 
polymerization to generate branched actin networks, which drives membrane 
protrusion [136]. The Arp2/3 complex-mediated bacteria uptake is divided into two 
types of mechanisms, zipper and trigger mechanism [135]. The zipper mechanism is 
best studied in Listeria monocytogenes uptake, which is induced by the interaction of 
bacteria surface molecules with host receptors. In this case, induced plasma membrane 
protrusion formed by actin cytoskeleton reorganization is limited and closely 
surrounding the bacteria. The trigger mechanism is best studied in Salmonella enterica 
35 
 
serovar Typhimurium (Salmonella) and Shigella flexneri. It is stimulated by the 
injection of bacterial effector proteins into host cells through type III secretion system 
(T3SS). Injected effector proteins trigger actin reorganization, forming actin-rich 
membrane ruffles, which leads to the uptake of bacteria by macropinocytosis. 
 
The actin cytoskeleton is linked to the apical junction and the adherens junction in 
polarized and non-polarized epithelial cells. The perijunctional actin filaments form 
the cortical belt at the apical junction, which is required for formation of the apical 
junction and the maintenance of the polarized epithelial barrier. Disruption of the actin 
cytoskeleton by cytochalasin D or latrunculin A increases the permeability of 
polarized epithelial monolayers [137].  E-cadherin, the major component of the 
adherens junction, has been shown to be coimmunoprecipitated with Arp2/3 complex 
and cortactin [121]. The adaptor protein of the adherens junction, α-catenin, links E-
cadherin to the actin cytoskeleton as monomers, while homodimers of α-catenin have 
been shown to interact with Arp2/3 complex and regulate actin polymerization [104, 
107, 138]. Arp2/3, N-WASP, and cortactin have been found to be localized at the tight 
junction as well [121]. Taken together, the dynamics of the actin cytoskeleton at the 
apical junction can regulate the dynamics of apical junction assembly and 
disassembly, consequently modulating epithelial cell polarity and barrier function.  
 
The actin motor protein myosin is a major component in the perijunctional actin ring. 
Myosin converts the energy from ATP hydrolysis to mechanical force, mediating 
36 
 
static tension and contractility of actin filament [139]. Non muscle myosin II (NMII) 
is the major type of myosin expressed in the epithelial cells and is enriched at the 
apical junction. The structure of NMII is similar to the myosin expressed in muscle 
cells, consisting of two heavy chains, which are responsible for the ATP hydrolysis, 
actin filament binding and oligomerization, and two essential light chains and two 
regulatory light chains (MLC), which are responsible for NMII activation by 
phosphorylation. Multiple kinases have been reported to phosphorylate MLC, 
including myosin light chain kinase (MLCK) and Rho-dependent coiled-coil kinase 
(ROCK). MLCK is activated by calcium-bound calmodulin. ROCK is activated by the 
small GTP-binding protein RhoA. The phosphorylation of MLC mediated by both 
ROCK and MLCK has been linked to apical junction disassembly and increased 
epithelial permeability. H. pylori and EPEC have been shown to disrupt the apical 
junction by inducing MLCK mediated NMII activation [99, 139, 140]. Our lab has 
shown that GC induce activation of NMII by MLCK, leading to apical junction 
disruption and GC transmigration across polarized epithelial cells [95].  
  
Whether myosin-mediated contraction plays any role in the cell-cell junction of non-
polarized or stratified epithelial cells, like the ectocervical epithelial cells, is unknown. 
Furthermore, how non-polarized epithelial cells change morphology by reorganization 
of the actin cytoskeleton in response to pathogens are also not well studied. Polarized 
epithelial cells are known to form densely packed microvilli at the apical surface, 
which are support by dense bundled F-actin, and these microvillar actin bundles are 
inserted into the actin terminal web underneath [141]. In non-polarized epithelial cells, 
37 
 
data from our lab and Dr. Stein’s lab show that the non-polarized ME180 cells form 
irregular microvilli-like plasma membrane protrusion at the luminal surface, which 
indicates the different organization of actin cytoskeleton at luminal surface in non-
polarized epithelial cells. Infection of polarized epithelium with EPEC induces 
microvilli disruption [142]. In contrast, GC inoculation, as described above, induces 
microvilli elongation in non-polarized epithelial cells. The difference in the actin 
cytoskeleton organization suggests different responses to GC infection in non-
polarized ectocervical and polarized endocervical epithelial cells. 
 
1.9 The advantages and disadvantages of various current GC infection models 
One of the major research barriers of GC pathogenesis is the limitation of 
experimental models to closely mimic human infection. The majority of research on 
GC infection relies on immortalized mammalian cell lines. Cell lines are readily 
available, easy to maintain and manipulate in laboratories. However, being cancerous 
cell itself and completely out of the context of tissue architecture may alter the 
morphology, protein components and receptor distribution and expression compared 
to the in vivo scenario in human tissues. For example, the endocervix is covered with 
polarized epithelial cells. However, the cell line derived from cervical epithelial cancer 
ME-180 completely lose the ability to polarize and form the apical junction in culture. 
The origin of some cell lines used is not the natural infection sites of GC. For 
example, the colon epithelial cell line T84 has been used to study GC infection in 
polarized epithelial cells, because of the ability of T84 cells to polarize in culture. 
Consequently, different cell lines used for studies generate different and sometimes 
38 
 
conflicting results. It is unclear whether results obtained from cell line models can be 
applicable for explaining infection in vivo.  
 
Primary cervical epithelial cell culture has been developed and has generated 
interesting and potentially physiological relevant data [29, 130]. However, primary 
cells isolated from the human cervix, again are removed from their tissue architecture 
and are not capable of polarizing to the level as epithelial cells in the endocervix. 
Fallopian tube organ culture has been enabling us to better understand GC infection in 
this particular location, which associates with PID [85, 143]. However, the fallopian 
tube epithelium is distinct from the cervical epithelium. The fallopian tube organ 
culture maintains the polarity and columnar nature of epithelial cells and dispersion 
between non-ciliated and ciliated epithelial cells, unlike the cervix, where epithelial 
cells progressively change from non-polarized stratified to polarized columnar 
epithelial cells. Furthermore, the current experimental systems are highly controlled 
and consistent, which does not reflect the complicated microenvironment in vivo, 
potentially driving the phase and antigenic variation of the virulence determinants of 
GC. Thus results obtained from these systems may be overly simplified.  
 
Since GC is a human exclusive pathogen, the humanized mouse model has been 
established by hormone treatment [5, 144] and by expressing the human CEACAM 
transgene [61, 145]. Vaginal inoculation of GC in these mouse models results in GC 
colonization for up to 10-40 days [5], thus does not cause the infection observed in 
39 
 
women. Besides CEACAM, mice do not express several other human specific 
receptors, such as CR3, which have been shown to play important roles in GC 
infection. While the mouse model has significantly advanced our understanding of GC 
pathogenesis in females, we still do not know whether data gathered from mouse 
models are applicable to human female infection.  
 
Therefore, an experimental system that closely mimics the heterogeneity of the 
cervical epithelium in its native tissue organization and environment is essential to 
understand the initiation of GC infection from the human cervix and the mechanisms 
leading to the wide range of clinical outcomes. 
 
 
 
 
 
 
 
 
 
 
40 
 
Rational 
This research aims to understand the mechanism by which GC establish infection at 
the heterogeneous mucosal surface of the human cervix, the gate of the female 
reproductive tract (FRT).  
 
The unique features of GC infection in females make gonorrhea a severe threat to 
women’s health. GC initiate infection at the mucosal surface of the cervix, the gate of 
FRT. The properties of the epithelium at different cervical regions vary significantly 
from non-polarized multiple-layered to polarized single-layered epithelial cells. 
Majority of GC infection in the lower part of the FRT (50-80%) is asymptomatic, thus 
untreated, which allows GC to further ascend to the upper genital tract, causing severe 
complications. The penetration of GC through the cervical epithelium into sub-
epithelium tissues, which has been shown previously, can lead to symptomatic and 
disseminated gonococcal infection (DGI). However, whether the heterogeneous 
cervical epithelia affect GC infection and how asymptomatic, localized GC infection 
in the cervix lead to a wide range of clinical outcomes is unknown. Thus, I focus my 
Ph.D. research on GC infection in human cervix. Because there is no experiment 
system that can mimic the heterogeneous epithelia of the human cervix, previous 
studies using cell line and mouse models give rise to conflicting results. To overcome 
this barrier, I developed a human cervical tissue explant model that preserves the in 
vivo properties of the human cervical epithelium and mimics in vivo GC infection 
processes when inoculated with the bacteria. This tissue explant model enabled me to 
41 
 
examine the impact of distinct properties of cervical epithelial cells at different 
anatomical regions on GC infectivity and infection mechanisms and the role of pili 
and Opa phase variation in GC infection of the cervix. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
42 
 
Central hypothesis 
GC modulate infectivity, colonization or penetration, based on the GC surface 
structures and properties of epithelial cells, with which they interact, by inducing 
differential signaling and cytoskeleton responses through different host receptors. GC 
colonization may only lead to localized asymptomatic infection, while GC penetration 
into the subepithelial tissue can lead to symptomatic and disseminated infection. 
 
Aim 1 
To determine the impact of the heterogeneous epithelia of the cervix on GC 
infectivity. 
Aim 2 
To determine the roles of the Opa phase variation and Opa interaction with the host 
receptor CEACAM in GC infection in different cervical regions.  
Aim 3 
To determine the effect of epithelial cell polarity on GC infectivity. 
 
 
43 
 
Chapter 2: Infectivity of Neisseria gonorrhoeae in the human 
cervix varies with the properties of cervical epithelial cells and 
bacterial phase variable surface structures 
 
2.1 Abstract  
The epithelium is the front line defense against pathogens in the outer environment. 
The heterogeneity of cervical epithelium further complicates the environment GC 
need to overcome to establish infection. To mimic the heterogeneous cervical 
epithelium, the human cervical tissue explant model was first established. GC colonize 
and penetrate the most in the transformation zone (TZ) of the cervix. In the ectocervix, 
GC colonize at a high level without any penetration. In the endocervix, the level of 
colonization is low, but penetration is observed. Pili is required for colonization in all 
three cervical regions. Expression of CEACAM-binding Opa, OpaH, enhances 
colonization in the ecto- and endocervix, where CEACAM is abundantly expressed at 
the luminal surface but inhibits penetration solely in the endocervix. In contrast, 
expression of different Opa proteins does not affect GC infectivity in the TZ, which 
expresses little CEACAM. This OpaH-CEACAM interaction-mediated differential 
infectivity involves ITIM domain of CEACAM1 and the subsequent activation of 
SHP. Such CEACAM1-SHP signaling pathway leads to the inhibition of cell-cell 
junction disruption, epithelial exfoliation, and eventually the penetration across the 
cervical epithelium into subepithelium tissue. Therefore, GC modulate its infectivity 
between colonization and penetration based on both the properties of cervical 
epithelial cells and phase variation of bacterial surface structure. 
44 
 
2.2 Introduction  
Gonorrhea, caused by Gram-negative bacterium Neisseria gonorrhoeae (GC), is the 
second most common sexually transmitted infection (STI) in the United States. The 
unique features of female infection make gonorrhea a severe threat to women’s health. 
50-80% of female infection is asymptomatic [2], leading to susceptibility to severe 
complications such as pelvic inflammatory disease (PID), fallopian tubes scarring and 
blockage, with subsequent ectopic pregnancy and infertility. An upsurge of multidrug-
resistant strains of GC further complicates gonorrhea treatment [5]. However, the 
mechanism by which GC and other STI pathogens interact with the human female 
reproductive tract (FRT) and cause a wide range of clinical outcomes remains elusive. 
A primary obstacle for a better understanding of STIs is a lack of infection models that 
mimic human infection.  
 
In women, GC initiate infection primarily from the cervix, the gate of the female 
reproductive tract (FRT) [10]. The cervical epithelium is the first line of defense GC 
need to overcome to establish infection, which exhibits distinct properties at different 
anatomical regions. The ectocervix is the outmost part of the human cervix, line with 
multiple-layered stratified epithelial cells, and exposed to the outer environment. The 
endocervix, connecting to the upper genital tract, is the upper portions of the human 
cervix, covered by a single-layered polarized columnar epithelial cells. The 
squamocolumnar junction (SCJ), or transformation zone (TZ), is a specialized region 
between the ecto and endocervix. The morphology of the epithelium covering the TZ 
transforms from multiple-layered to single-layered, and from noon-polarized stratified 
45 
 
to polarized columnar epithelial cells. The ecto and endocervical epithelium display 
distinct organization of cell-cell junction complexes, which maintains the integrity and 
regulates the dynamics of cervical epithelia [123]. Clinical studies suggest that GC 
colonization at the ectocervical epithelium leads to asymptomatic infection [3]. In 
contrast, clinical studies using patient biopsies only found subepithelial GC in the TZ 
and endocervix [15]. However, how the heterogeneity of the mucosal surface of the 
cervix impacts GC infection and subsequent clinical outcomes has not been examined. 
 
GC have evolved multiple mechanisms to cope with the heterogeneous mucosal 
surface in humans, including phase and antigenic variation of their surface molecules. 
GC can vary the structure of the lipooligosaccharide (LOS) [36, 146], turn the 
expression of pili on and off [23], and express any of 11 opacity-associated (Opa) 
protein isoforms at any time [30].  It has been shown that pili mediate the initial 
interaction of GC with the mucosal surface of epithelial cells [147] and that the 
interaction of Opa proteins with their host receptors mediate an intimate interaction. 
Most Opa isoforms bind to carcinoembryonic antigen-related cell adhesion molecules 
(CEACAMs) (OpaCEA) [148] while others bind to heparan sulfate proteoglycans 
(HSPG) (OpaHSPG) [33]. It has been shown that OpaCEA-CEACAM interaction 
inhibited epithelial exfoliation, thus increased GC colonization in CEACAM 
transgenic mouse model [61]. Among the CEACAM isoforms expressed in epithelial 
cells, CEACAM1L is the only one harbors an immune-receptor tyrosine-based 
inhibitory motif (ITIM) motif in its cytoplasmic domain; while its splicing isoform 
CEACAM1S does not. The differential expression of CEACAM isoforms in different 
46 
 
anatomic regions of the FRT has been shown in patient biopsies [63]. However, the 
role of each of GC surface structures and their interactions with human receptors in 
GC pathogenesis has not been systematically examined in the human cervical 
epithelia. 
 
Here, I established the human cervical tissue explant model to examine the impact of 
the heterogeneous human cervical epithelia and the phase variation of pili and Opa on 
GC infection. The human cervical tissue explants preserve the distinct cytoarchitecture 
of the three cervical regions in vivo and mimic GC infectivity observed in patient 
biopsies. Using this model, I found that the outcome of GC infection depends on the 
type of cervical epithelial cells GC interact with and the Opa isoforms GC express. 
Expression of OpaCEA drives GC colonization on the mucosal surface of the ecto and 
endocervix, which express CEACAM. Furthermore, OpaCEA expression suppresses 
GC penetration across the endocervical epithelium through inhibition of GC-induced 
cell-cell junction disassembly and epithelial shedding. The low expression level of 
CEACAM in the TZ epithelial cells abolishes the regulatory role of OpaCEA 
expression. The above results provide a potential mechanism for the various clinical 
outcomes derived from localized cervical GC infection in women. 
 
 
 
47 
 
2.3 Materials and Methods 
Neisseria Strains 
N. gonorrhoeae strain MS11 that expressed both pili and Opa (MS11Pil+Opa+) was 
obtained from Dr. Herman Schneider, Walter Reed Army Institute for Research. 
Derivatives of this strain, MS11ΔOpa, MS11 OpaH (OpaCEA, CEACAM-binding), and 
MS11OpaC (OpaHSPG, HSPG-binding) have previously been described [149]. 
MS11Opa+ and MS11ΔOpa strains express similar LOS [149]. MS11 Pil+Opa+ and 
Pil-Opa+ colonies were identified based on their morphology using a dissecting light 
microscope. GC were grown on plates with GC media (Difco, BD Bioscience, 
Franklin Lakes, NJ) and 1% Kellogg’s supplement [150] for 16-18 h before 
inoculation. The concentration of GC in suspension was determined using a 
spectrophotometer and inoculated with epithelial cells at MOI ~10. 
 
Human cervical tissue explants 
The tissue explants were cultured as previously described [151]. Briefly, cervical 
tissues were obtained from patients undergoing voluntary hysterectomies and received 
within 24 h post-surgery. Age of patients ranges from 28 to 40. Supporting muscle 
tissue was removed by using a carbon steel surgical blade. Remaining cervical tissue 
samples were cut into ~2.5 cm (L) X 0.6 cm (W) X 0.3 cm (H) pieces, incubated in 6-
well tissue culture plate in cervical tissue culture media conataining CMRL-1066 
(GIBCO), with 5% heat-inactivated fetal bovine serum, 2 mM L-glutamine, bovine 
insulin (1 μg/ml, Akron Biotech) and penicillin/streptomycin for 24 h and then in 
48 
 
antibiotic-free media for another 24 h, before inoculation with GC as previously 
described [152].  
 
Immunofluorescence analysis of human cervical tissue explants 
Individual pieces of cervical tissue explants were inoculated with GC at MOI ~10. The 
number of epithelial cells at the mucosal surface is determined by the area of the 
mucosal surface of the individual piece and the average luminal area of individual 
cervical epithelial cells. When indicated, tissue pieces were inoculated with GC with 
or without the SHP inhibitor NSC-87877 (20 μM, EMD Millipore). GC were collected 
by using a sterile applicator to swab GC from the plate and re-suspend GC in pre-
warmed antibiotic-free cervical tissue culture media. An aliquot of the GC suspension 
was added directly onto each tissue piece to make an MOI = 10 bacteria to 1 epithelial 
cell. The cervical tissue explants were incubate at 37 °C with 5% CO2 with gentle 
shaking for 24 h. The infected cervical tissue explants were rinsed with antibiotic-free 
cervical tissue culture medium at 6 and 12 h to remove un-adhered bacteria and 
replace with fresh media without bacteria. Tissue explants were then fixed 24 h post 
inoculation, embedded in 20% gelatin, cryopreserved, sectioned by cryostat across the 
luminal and basal surfaces of the epithelium, stained for F-actin (Cytoskeleton), E-
cadherin (BD Bioscience), CEACAMs (cross-react with CEACAM1, 3, and 6, Santa 
Cruz Biotechnology, Dallas, TX), SHP1 (Santa Cruz Biotechnology), SHP2 (Santa 
Cruz Biotechnology), phospho-SHP2 Tyr542 (Thermo Fisher Scientific) and GC by 
specific antibodies, and nuclei by Hoechst (Life Technologies), and analyzed using 
confocal fluorescence microscope (Zeiss LSM 710, Carl Zeiss Microscopy LLC) as 
49 
 
previously described [152]. Images of the luminal surface were randomly acquired 
from the ectocervix to the endocervix as single images or Z-series of 0.57 μm/image, 
and 3D composites obtained using Zeiss Zen software.  
 
The level of GC colonization was quantified by two methods using confocal images: 
(1) percentage of the number of GC-associated luminal epithelial cells versus the total 
number of luminal epithelial cells through visual inspection, and (2) fluorescence 
intensity (FI) of GC staining per μm2 at the luminal surface using the NIH ImageJ 
software. The level of GC penetration was determined by two methods: (1) percentage 
of the number of epithelial cells associated with GC staining at and underneath the 
basal membrane versus of the total number of epithelial cells through visual 
inspection, and (2) percentage of the FI of GC staining at sub-epithelial tissue versus 
the total FI of GC staining at both the subepithelium and the luminal surface in each 
randomly acquired image. 
 
The level of epithelial exfoliation in the endocervix was determined by the percentage 
of the number of epithelial cells localized on top of the epithelial monolayer versus the 
total number of epithelial cells through visual inspection. Level of epithelial 
exfoliation in the ectocervix and the transformation zone (TZ) was determined by two 
methods: (1) the percentage of the thickness (μm) of the epithelium in ectocervical and 
TZ tissue explants with or without inoculation of GC, and (2) the percentage of 
50 
 
number of epithelial layers in ectocervical and TZ tissue explants with or without 
inoculation of GC. 
 
The recruitment of CEACAM staining to GC microcolony was evaluated by the 
percentage of the number of GC microcolonies colocalizing with CEACAM staining 
underneath versus the total number of GC microcolonies through visual inspection. 
 
The redistribution of E-cadherin staining from the cell-cell junction to the cytoplasm 
was evaluated by the fluorescence intensity ratios (FIR) of E-cadherin staining at cell-
cell junction versus that in the cytoplasm in individual epithelial cells using confocal 
images and the NIH ImageJ software.  
 
Each of the above results was generated from cervical tissue explants of 3-5 human 
subjects and >7 randomly acquired images from each cervical region and each tissue 
explant. 
 
Epithelial cells and CEACAM1 transfection 
HEC-1-B, a human endometrial adenocarcinoma cell line (ATCC), was maintained in 
Eagles MEM alpha medium supplemented with 10% heat-inactivated FBS. Cells were 
seeded at 3.5 x105 per 60 mm diameter cell culture dish, (Thermo Fisher Scientific) 
and cultured for 2 days before transfection. Cells were transiently transfected with 5 
51 
 
μg of a plasmid containing CEACAM1L or CEACAM1S cDNA using lipofectamine 
3000 reagents (Thermo Fisher Scientific). Transfected cells were seeded at 6x104 per 
transwell (6.5 mm diameter, 3 μm pore size, Corning) and cultured for 2 days before 
inoculation with GC at MOI ~10. 
 
T84, a human colorectal carcinoma cell line (ATCC), was maintained in DMEM:Ham 
F12 (1:1) supplemented with 7% heat-inactivated FBS. Cells were seeded at 6x104 per 
transwell (6.5 mm diameter, 3 μm pore size, Corning) and cultured for ~10 days until 
the transepithelial electrical resistance (TEER) reached ~2000Ω before GC 
inoculation. TEER was measured using a Millicell ERS volt-ohm meter (EMD 
Millipore). 
 
Immunofluorescence analysis of epithelial cells 
Transfected HEC-1-B cells were inoculated with GC at MOI of ~10 for 6 h. Cells 
were rinsed, fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, 
and stained with anti-β-catenin (EMD Millipore), anti-CEACAMs (Santa Cruz 
Biotechnology), anti-GC antibodies, and Hoechst for nuclei. Images were randomly 
acquired using CFM. The level of GC adherence to cells with or without CEACAM 
transfection was compared by the fluorescence intensity ratio (FIR) of GC staining in 
CEACAM-positive cells versus CEACAM-negative cells in the same image using the 
same ROI. Epithelial exfoliation was quantified using xz images by the percentage of 
HEC-1-B cells moving above the monolayer versus the total number of epithelial 
52 
 
cells. The redistribution of β-catenin staining from the cell-cell junction to the 
cytoplasm was quantified as described for the E-cadherin in tissue explants. The 
results were obtained from 3-5 independent experiments and >5 randomly acquired 
images of each condition and each experiment. 
 
GC transmigration assay 
The assay was performed as previously described [76]. Briefly, polarized T84 
epithelial cells that were pretreated with or without the SHP inhibitor NSC-87877 (20 
μM, EMD Millipore) for 1 h were inoculated apically with GC at MOI ~10 for 6 h. 
The basolateral media were collected and cultured by plating, and the resulting 
colonies were counted as transmigrated bacteria. 
 
Statistical analysis 
Statistical significance was assessed by using the Student’s t-test by Prism software 
(GraphPad Software, San Diego, CA). P-values were determined using unpaired t test 
with Welch’s correction in comparison with no infection controls. 
 
Ethics statement 
Human cervical issue explants were obtained from the National Disease Research 
Interchange (NDRI, Philadelphia, PA). Human cervical tissues used were anonymized. 
53 
 
The use of human tissues for this research has been approved by the Institution 
Review Board of the University of Maryland. 
 
2.4 Results 
2.4.1 GC exhibit distinct infectivity patterns on different mucosal surfaces of the 
human cervix 
We sought to preserve the heterogeneity of cervical epithelium by using human 
cervical tissue explants. Cervical tissue explants were cultured mucosal side up. Thin 
sections of cryo-preserved cervical tissue were sectioned, stained and analyzed using 
three-dimensional confocal fluorescence microscopy (3D-CFM). Images showing both 
the mucosal and subepithelial sides of the ectocervix, TZ and endocervix were 
analyzed. I identified epithelial cells on the tissue section by staining cytokeratin (Fig. 
7 A), the marker of epithelial cells. The cell-cell junction was visualized by staining 
adherence junction protein E-cadherin (Fig. 7 B). I first used a 10X object lens to scan 
the mucosal surface of the cervical section from ecto to endocervix continuously. The 
three cervical regions were identified by the distinct features of the epithelial cells: 1) 
ectocervix, multiple-layered non-polarized epithelial cell, 2) TZ, multiple-layered 
cubical epithelial cells, 3) endocervix, single-layered polarized epithelial cells. I found 
that the human cervical tissue explant model can preserve the overall morphology of 
cervical epithelium (Fig. 7 B).  
 
 
 
54 
 
 
 
Figure 7. Human cervical tissue explants maintain the in vivo characteristics in 
culture. 
Human cervical tissue explants were cultured for three days and cryopreserved. Tissue 
sections across the luminal and basal surface of epithelia were collected, stained for 
DNA, E-cadherin, cytokeratin and/or F-actin, and analyzed using CFM. (A) 
Representative images of the mucosal epithelial regions of cervical tissue explants 
obtained by combining >30 images acquired using the 10X objective. Dashed lines 
indicate the boundary between the endocervix and the TZ and between the TZ and the 
ectocervix. Scale bar, 100 m. (B) Representative images of the epithelium of three 
cervical regions. Scale bar, 20 m. 
 
55 
 
To determine whether the heterogeneity of cervical epithelium affects GC infectivity 
in the same way or differently, I inoculated cervical tissue explants from the same 
patient sample with or without MS11 Pil+Opa+, a GC strain expressing phase variable 
opa genes, for 24 h. Each tissue explant contains all three cervical regions and 
inoculated with GC at MOI~10. Unassociated bacteria were rinsed at 6 and 12 h post 
inoculation. Cervical tissue explants were cryopreserved and sectioned across the 
luminal surface to subepithelial tissue. Tissue sections were stained for GC, F-actin, 
and nuclei and analyzed by CFM. I used two methods to quantify the level of GC 
colonization using CFM images (Fig. 8 B, C): 1) the percentage of number of luminal 
epithelial cells associated with GC staining (Fig. 8 B), and 2) the fluorescence 
intensity (FI) of GC staining per μm2 of the luminal surface of the cervix (Fig. 8 C). 
The two methods showed consistent results. MS11Pil+Opa+ colonize the mucosal 
surface of all three cervical regions. However, the level of colonization in the 
ectocervix and TZ is significantly higher than the endocervix (Fig. 8 B, C). Similarly, 
I used two methods to quantify the level of penetration: 1) percentage of the number of 
epithelial cells associated with GC staining below basal membrane (Fig. 8 D), and 2) 
percentage of FI of GC staining at and below basal membrane versus total FI of GC 
staining (Fig. 8 E). Penetrating GC were only detected in the TZ and endocervix, and 
the level of penetration into the subepithelial tissue in the TZ is significantly higher 
than into the endocervix. These results are consistent with clinical observation, 
indicating GC infection in human cervical tissue explant model mimics GC infection 
in vivo. It further suggests that GC modifies its infectivity based on the different 
properties of the cervical epithelial cells they encounter. 
56 
 
 
 
 
 
 
57 
 
 
 
 
Figure 8. GC exhibit different levels of colonization and subepithelial penetration 
in the ectocervical, TZ, and endocervical epithelium.  
Human cervical tissue explants were incubated with MS11 Pil+Opa+ for 24 h 
(MOI~10), washed at 6 and 12 h post-inoculation, and cryopreserved. Tissue sections 
were collected, processed, and analyzed using CFM. (A) Representative images of the 
three cervical regions with or without GC infection. Scale bar, 20 m. Arrowheads, 
colonizing GC. Arrows, penetrating GC. (B and C) The level of GC colonization was 
quantified of by the percentage (±SEM) of luminal epithelial cells with GC attaching 
to the luminal surface (B) and by fluorescence intensity (FI) (±SEM) of GC staining 
per m2 at the luminal surface (C). (D and E) The level of GC penetration was 
quantified by the percentage (±SEM) of epithelial cells associating with GC at 
subepithelia (D) and by the percentage of GC FI (±SEM) at the subepithelia relative to 
the total GC FI at the epithelium (E). n=4. >20 randomly selected fields and >200 cells 
per cervix and condition. *p<0.05; **p<0.01; ***p<0.001. 
 
 
 
 
58 
 
2.4.2 Pili are required for GC colonization, and Opa phase variation regulates 
GC infectivity in the human cervix 
Pili and Opa are major phase variable virulence determinants of GC. To examine the 
roles of Pili in GC infection in the cervical epithelium, I first compared the level of 
colonization and penetration of piliated (Pil+) and non-piliated (Pil-) MS11Opa+ in 
the cervical tissue explant. Compared to the MS11Pil+Opa+, MS11Pil-Opa+ failed to 
colonize efficiently in all three cervical regions (Fig. 9 A, B-D). And no penetration of 
MS11Pil-Opa+ in all three cervical regions was observed (Fig. 9. A, E-G). To 
examine the roles of Opa proteins in GC infectivity, I inoculated cervical tissue 
explants with Pil+ GC expressing phase variable Opa (Opa+) and all 11 Opa isoforms 
deleted (ΔOpa). The deletion of Opa significantly reduced the level of colonization in 
the ecto and endocervix, but not in the TZ (Fig. 9 A, B-D). In contrast, deletion of Opa 
significantly increased the level of penetration in the endocervix (Fig. 9 A, G). The 
above results indicate the Pili is essential to establish colonization in all three cervical 
regions; while the expression of phase variable Opa proteins regulates GC infectivity 
differently in the three cervical regions. 
 
 
 
 
 
 
59 
 
 
 
 
 
 
 
60 
 
 
 
 
 
Figure 9. Pili are required for GC colonization, and Opa expression differentially 
regulates GC infectivity in the three cervical regions.  
Human cervical tissue pieces were inoculated with GC MS11 Pil+Opa+, Pil+ΔOpa, or 
Pil-Opa+, processed and analyzed as described in Fig. 8. (A) Representative images of 
three cervical regions with or without GC infection. Arrowheads, colonizing GC. 
Arrows, penetrating GC. Scale bar, 20 µm. (B-D) Levels of GC colonization at the 
ectocervix (B), TZ (C), and endocervix (D) by GC FI per m2 (±SEM). (E-G) Levels 
of GC penetration into the subepithelia of the ectocervix (E), TZ (F), and endocervix 
(G) by the percentage of GC FI (±SEM) at the subepithelia relative to the total GC FI 
at the epithelium. n=2~4. >20 randomly selected fields and >200 cells per cervix and 
condition. *p<0.05; **p< 0.01; ***p<0.001. 
 
 
 
 
 
 
 
61 
 
One GC bacterium can express up to 11 different Opa isoforms, which can be 
categorized into two groups: 1) CEACAM-binding OpaCEA, represented by OpaH, 2) 
HSPG-biding OpaHSPG, represented by OpaC. Majority of Opa proteins expressed are 
CEACAM-binding Opa. I hypothesize that Opa proteins regulate GC infectivity in 
different cervical regions through phase variation, switch between dominant OpaCEA 
and dominant OpaHSPG expression. To test this hypothesis, I inoculated human cervical 
tissue explants with Pil+ MS11 expressing OpaH and OpaC that do not phase vary, as 
well as MS11Pil+ ΔOpa. Expression of OpaH but not OpaC significantly enhanced the 
level of colonization in the ecto and endocervix, but not in the TZ (Fig. 10 A, B-D). In 
contrast, expression of OpaH inhibited penetration only in the endocervix (Fig. 10 A, 
G). All three Opa isogenic strains did not penetrate in the ectocervix (Fig. 10 A, E); 
while penetrating at a high level in the TZ (Fig. 10 A, F). GC expressing OpaC show a 
similar level of colonization and penetration as GC ΔOpa. The above data suggest that 
OpaCEA selectively enhanced the level of colonization in the ecto and endocervix, and 
inhibited the level of penetration solely in the endocervix. Expression of OpaHSPG does 
not significantly affect GC infectivity in the human cervix. Intriguingly, expression of 
different Opa isoforms does not affect GC infectivity in TZ. Taken together, GC 
infectivity is regulated by both Opa phase variation and properties of different types of 
cervical epithelial cell. 
 
 
 
62 
 
 
 
63 
 
 
 
 
 
Figure 10. Opa variants differentially regulate GC colonization and penetration 
into the three cervical regions. 
Human cervical tissue pieces were incubated with piliated (Pil+) GC MS11 OpaH, 
OpaC, and ΔOpa, processed and analyzed as described in Fig. 8. (A) Representative 
images of three cervical regions with or without GC infection. Arrowheads, colonizing 
GC. Arrows, penetrating GC. Scale bar, 20 µm. (B-D) Levels of GC colonization at 
the ectocervix (B), TZ (C), and endocervix (D) by GC FI per m2 (±SEM). (E-G) 
Levels of GC penetration into the subepithelia of the ectocervix (E), TZ (F), and 
endocervix (G) by the percentage of GC FI (±SEM) at the subepithelia relative to the 
total GC FI at the epithelium. n=3~4. >20 randomly selected fields and >200 cells per 
cervix and condition. **p< 0.01; ***p<0.001. 
 
 
 
 
 
 
64 
 
2.4.3 The OpaH-dependent regulation of GC infectivity involves CEACAM1L 
expression in cervical epithelial cells 
My finding that OpaH expression affects GC infectivity in different cervical regions 
suggests the differential role of its epithelial receptor CEACAM in different cervical 
regions. To test this hypothesis, I first examined the expression and distribution of 
CEACAM by CEACAM fluorescence intensity (FI) per cell. I used an antibody cross 
react with CEACAM1, 3 and 6 (Fig. 11 A, B). Without inoculation with GC, the TZ 
epithelial cells express low-level CEACAM, with most puncta staining in the 
cytoplasma. The endocervical epithelial cells show strong staining at the apical 
surface. The ectocervical epithelial cells show strong staining at luminal surface and 
strong puncta staining in cytoplasma. The lack of CEACAM expression provides a 
potential explanation of the similar infectivity displayed by GC OpaH, OpaC and 
ΔOpa in the TZ.  
 
To answer the question that whether the expression of OpaCEA regulates GC infectivity 
through differential interaction with CEACAM in three cervical regions, I examined 
the recruitment of CEACAM by percentage of GC microcolonies colocalized with 
CEACAM staining underneath. In the endocervix, around 58% microcolonies of GC 
Pil+ OpaH colocalized with CEACAM staining (Fig. 11 C, D, arrowheads), 
significantly higher than the 20% in the TZ, which expresses low-level CEACAM. It 
is also significantly higher than the level induced by GC Pil+ΔOpa and GC Pil+OpaC, 
which displays less than 10% colocalization in both endocervix and TZ. These 
65 
 
observations indicate a direct interaction of OpaCEA with CEACAM and its role in GC 
infectivity modulation. 
 
To test the hypothesis further, I utilized a human endometrium epithelial cell line, 
HEC-1-B, which does not express endogenous CEACAM [153]. CEACAM1 is the 
only CEACAM expressed in epithelial cells that has a cytoplasmic domain. 
CEACAM1L harbors an ITIM motif in the cytoplasmic domain; while the isoform 
CEACAM1S lacks it. I transiently transfected HEC-1-B with CEACAM1L and 
CEACAM1S and inoculated transfected cells with MS11Pil+OpaH and ΔOpa for 6 h 
on transwell. I examined the level of adhesion of GC by the GC fluorescent intensity 
ratio (FIR) at the area with CEACAM1 expression to that without CEACAM1 
expression using the same ROI in the same image with all z series slides projected 
onto it (Fig. 11 E, F). GC Pil+OpaH adhered to the CEACAM1L-expressing area 
twice as much as the area without CEACAM1L expression, while GC Pil+ΔOpa 
adhered to the area with or without CEACAM1L expression at a similar level. In the 
CEACAM1S transfected cells, both MS11Pil+OpaH and ΔOpa adhere to the area with 
or without CEACAM1S expression similarly (FIR ~ 1). Taken together, the above 
data suggest the OpaCEA-dependent regulation of GC infectivity in different cervical 
regions involves CEACAM1L expression. 
 
 
 
 
66 
 
 
 
 
 
67 
 
 
 
 
 
Figure 11. Regulatory roles of OpaH depend on the expression of CEACAML in 
the human cervical epithelial cells.  
(A) Representative CFM images of uninfected human cervical tissue explants stained 
with the anti-CEACAM antibody (reacting with CEACAM 1, 3, and 6). (B) Levels of 
CEACAM expression measured by the FI (±SEM) of CEACAM per cell. (C) 
Representative CFM images of human cervical tissue explants inoculated with GC 
MS11Pil+ OpaH, OpaC, and ΔOpa and stained for GC and CEACAMs. Arrowhead, 
GC microcolonies colocalizing with CEACAM staining. (D) The spatial relationship 
of GC with CEACAMs was quantified by the percentage of GC microcolonies 
colocalizing with CEACAM staining. (E) Representative CFM images of HEC-1-B 
cells transiently transfected with or without CEACAM1L or 1S and infected with GC 
MS11 Pil+ OpaH or Pil+Opa GC. White dashed lines circle CEACAM-expressing 
cells. Yellow dashed lines circle cells that do not express CEACAMs. (F) Levels of 
GC adherence to CEACAM1L or 1S-expressing cells were compared to those that did 
not express CEACAMs by the FIR (±SEM) of GC on CEACAM-positive cells versus 
CEACAM-negative cells in the same images using the same ROI.  Scale bar, 20 µm. 
n=3~4. >4 randomly selected fields and >200 cells per condition and cervixes. **p< 
0.01; ***p<0.001. 
 
 
68 
 
2.4.4 The OpaH-dependent regulation of GC infectivity involves activation of 
SHP 
The enhanced adherence of MS11Pil+OpaH in CEACAM1L-expressing cells 
compared to CEACAM1S-expressing cells suggests the role of the ITIM motif in the 
cytoplasmic domain of CEACAM1L. CEACAM1L has been shown to recruit and 
activate the tyrosine phosphatase SHP1/2 [154]. Thus, I hypothesize that SHP1/2 play 
a role downstream of OpaCEA-CEACAM1L dependent regulation of GC infectivity. I 
used an SHP1/2 enzymatic inhibitor NSC-87877 to test this hypothesis [155]. The 
treatment of GC alone with the inhibitor did not affect the growth rate of GC (Fig. 12). 
The treatment of the cervical tissue explant with SHP1/2 inhibitor significantly 
reduced the level of colonization of OpaH in the ecto and endocervix but not the TZ 
compared to the inoculation with GC Pil+OpaH alone (Fig. 14 A, B). Treatment with 
the inhibitor also enhanced the level of penetration of GC Pil+OpaH in the endocervix 
but had no effect on penetration in the ecto and TZ (Fig. 14 A, C). Similarly, treatment 
of polarized T84 cell monolayer with the inhibitor increase the level of GC Pil+OpaH 
transmigration (Fig. 14 D). These observations suggest that the OpaCEA-CEACAM1L 
dependent regulation of GC infectivity involves the subsequent activation of SHP1/2. 
 
To test whether the inoculation with MS11Pil+OpaH induces SHP1/2 activation in 
cervical epithelial cells, I examined the phosphorylation and redistribution of SHP1/2 
with or without inoculation with GC Pil+OpaH with or without inhibitor. The staining 
of SHP1 in the cervical tissue showed little signal (Fig. 13), thus I focused on the 
phosphorylation of SHP2 (pSHP2) and its redistribution. I only detected a significant 
69 
 
signal of pSHP2 staining in the ectocervix inoculated with GC Pil+OpaH, which was 
abolished by the treatment with the inhibitor (Fig. 14 E). I quantified it by the 
fluorescence intensity (FI) of pSHP2 per cell in the three cervical regions (Fig. 14 F). 
Staining of SHP2 was detected in all three cervical regions (Fig. 14 G). The ectocervix 
displayed an overall cytoplasmic staining pattern with no significant difference 
detected with or without inoculation with GC Pil+ OpaH with or without inhibitor. In 
contrast, in the TZ and endocervix, SHP2 staining was observed both at the luminal 
surface and in cytoplasma, with the FIR of SHP2 at the luminal surface to cytoplasma 
around 2. Inoculation of GC Pil+OpaH increased the level of SHP2 staining at the 
luminal surface in the endocervix to FIR ~ 3, but not in the TZ. This was abolished by 
the treatment with NSC-87877 (Fig. 14 G, H). The above data indicate that GC 
Pil+OpaH can induce the activation of SHP2 in ecto and endocervical but not the TZ 
epithelial cells, and expression of OpaCEA enhances GC colonization and inhibits 
penetration by inducing CEACAM1L-mediated SHP activation. 
 
 
 
 
 
 
 
 
 
 
70 
 
 
Figure 12. Treatment of the SHP inhibitor NSC-87877 has no significant effect on 
GC growth (Conducted by Sofia Di Benigno).  
GC MS11 Pil+OpaH was cultured in GC media (with 1% Kellogg’s supplement and 
1% NaHCO3) in the absence or presence of NSC-87877 (20 M). The bacterial CFU 
was numerated at 6, 12 and 24 h. Shown are average CFU (±SEM) of three 
independent experiments.
 
Figure 13. Human cervical tissue explants show no significant staining of SHP1. 
Representative images of the three regions of uninfected cervical tissue explants 
stained for SHP1, F-actin, and DNA (A) and Jurkat human T cells stained for SHP1 as 
a positive control. Scale bar, 20 µm. 
71 
 
 
 
 
 
72 
 
 
 
 
Figure 14. SHP activation is involved in the regulatory role of OpaH in GC 
infectivity in the human cervix. 
(A) Representative CFM images of human cervical tissue explants inoculated with GC 
MS11 Pil+OpaH with or without the SHP inhibitor NSC87877 (20 µM) and stained 
for GC, F-actin, and DNA. Arrowheads, colonizing GC. Arrows, penetrating GC. (B) 
Levels of GC colonization at the ectocervix, TZ and endocervix by GC FI per µm2 
(±SEM) of the luminal surface. (C) Levels of GC penetration into the subepithelia of 
the ectocervix, TZ and endocervix measured as the percentage of GC FI (±SEM) at the 
subepithelium. (D) The level of transmigration of GC Pil+OpaH and Pil+ΔOpa across 
polarized T84 epithelial cells treated with or without the SHP inhibitor NSC-87877 
(20 µM) is showed as the fold increase of GC CFU in the basal medium compared to 
the CFU of transmigrated GC Pil+OpaH without SHP inhibitor treatment. (E and G) 
Representative CFM images of tissue explants inoculated with GC Pil+OpaH with or 
without the SHP inhibitor and stained for GC, DNA, SHP2, or pSHP2. (F) Levels of 
pSHP2 staining were quantified by pSHP2 FI per cells (±SEM). (H) The redistribution 
of SHP2 from the cytoplasm to the luminal surface quantified by the FIR (±SEM) at 
the luminal surface versus the cytoplasm using FI line profiles cross luminal surface 
(yellow dotted lines). Scale bar, 20 µm. n=3~4. >20 randomly selected fields and >200 
cells per cervix and condition. *p<0.05; **p< 0.01; ***p<0.001. 
 
 
73 
 
2.4.5 OpaH expression inhibits GC-induced cervical epithelial exfoliation in a 
CEACAM1L and SHP-dependent manner 
Clinical research shows that GC infection can induce cervical epithelial exfoliation. It 
has been shown that OpaCEA inhibited epithelial exfoliation in CEACAM transfected 
cell line and mouse model [61, 62]. However, whether OpaCEA-CEACAM interaction 
regulates epithelial exfoliation in the cervical epithelium in the same way or 
differently is unknown. To answer the questions, I first examined the level of 
exfoliation induced by GC Pil+OpaH, OpaC or ΔOpa in the three cervical regions. 
Inoculation with GC induced epithelial exfoliation in all three cervical regions (Fig. 15 
A, E, H). The superficial stratified layers of epithelial cells in the ectocervix were 
exfoliated in layers (Fig. 15 A). I plot the line profile of F-actin staining across the 
ectocervical epithelium (Fig. 15 A, white dash lines, B), to measure the thickness (μm) 
and the number of layers of the ectocervical epithelial cells. I quantified the level of 
epithelial exfoliation in the TZ and ectocervix by the percentage of the thickness (Fig. 
15 C), and the number of epithelial layers remained (Fig. 15 D), compared to the 
tissue without GC inoculation. In the ectocervix, expression of OpaH reduced the level 
of exfoliation by 50% comparing to OpaC, and ΔOpa. In contrast, all three Opa 
isogenic strains induced around 50% epithelial exfoliation in the TZ (Fig. 15 F, G). In 
the endocervix, epithelial cells were exfoliated as individual cells (Fig. 15 H, 
arrowhead), thus I quantified the level of exfoliation by the percentage of cells located 
above the epithelium versus total number of epithelial cells (Fig. 15 I). Inoculation of 
GC Pil+ OpaH induced 13% exfoliation, significantly lower than the 27% and 33% for 
GC Pil+ OpaC and ΔOpa. Inoculation with MS11 Pil-Opa+ did not induce significant 
74 
 
epithelial exfoliation in all three cervical regions (data not shown). Therefore, the 
expression of OpaCEA inhibits GC-induced exfoliation in the endo and ectocervix, 
which may contribute to the enhanced, and the inhibited penetration in the endocervix. 
In the TZ, potentially due to the low level of CEACAM expression, the expression of 
OpaCEA did not affect the level of epithelial exfoliation as with the infectivity. 
 
To determine whether the inhibition of epithelial exfoliation by OpaCEA involves 
CEACAM1L, I inoculated HEC-1-B cells transiently transfected with CEACAM1L 
and CEACAM1S with GC Pil+ OpaH and ΔOpa. Using the x-z view CFM images 
cutting across the luminal surface to the basal level, I quantified the level of 
exfoliation by the percentage of HEC-1-B cells located on top of the monolayer versus 
the total number of epithelial cells (Fig. 15 J, K). Without transfection, both OpaH and 
ΔOpa induced around 12% exfoliation. Without inoculation with GC, transfection of 
both constructs did not induce a significant basal level of exfoliation (around 4%). The 
transfection of CEACAM1S did not affect the level of exfoliation induced by both 
OpaH and ΔOpa, compared to the cells without transfection. In contrast, transfection 
of CEACAM1L reduced OpaH-induced exfoliation, but not ΔOpa-induced, to the 
basal level.  Taken together, the inhibition of epithelial exfoliation by OpaCEA 
expression depends on the CEACAM1L expression, thus potentially involves SHP as 
well. 
 
75 
 
To examine the role of SHP activation in the epithelial exfoliation, I quantified the 
level of exfoliation in the three cervical regions with or without NSC-87877 treatment, 
the SHP1/2 activation inhibitor. The treatment with inhibitor significantly increased 
the level of exfoliation induced by GC Pil+ OpaH in the ecto and endocervix, but not 
TZ (Fig. 15 L-O). Thus, the above data suggest that the OpaCEA-dependent inhibition 
of epithelial exfoliation involves CEACAM1L and subsequent activation of SHP1/2. 
 
 
76 
 
 
 
 
 
77 
 
 
 
Figure 15. OpaH inhibits GC-induced shedding of the cervical epithelial cells in a 
CEACAM1L and SHP-dependent manner.  
(A, D, and F) Representative images of the ectocervical (A), TZ (D), and endocervical 
(F) regions of tissue explants inoculated with or without GC Pil+ OpaH, OpaC or 
Opa. Arrowheads, shedding cells. (B) Representative line profiles of F-actin FI along 
dashed lines in A, indicating the number of epithelial layers and the thickness of the 
epithelium remained. (C and E) Levels of epithelial shedding in the ectocervix (C) and 
the TZ (E) quantified by the percentage (±SEM) of the thickness of the epithelium and 
the number of epithelial layers remained. (G) Levels of epithelial shedding in the 
endocervix quantified by the percentage (±SEM) of epithelial cells moving above the 
epithelial monolayer. (H) Representative images of HEC-1-B cell monolayers that 
transfected with or without CEACAM1L or 1S and inoculated with or without GC 
Pil+OpaH and Pil+Opa. Dashed lines, the surface of monolayers. Arrowheads, cells 
moving above the epithelial monolayer. (I) Shedding of HEC-1-B cells quantified by 
the percentage (±SEM) of cells moving above the monolayers. (J) Representative 
CFM images of human cervical tissue explants inoculated with GC Pil+OpaH with or 
without the SHP inhibitor NSC-87877 (20 μM). Arrowheads, shedding cells. (K-M) 
Levels of epithelial shedding in the ectocervical (K), TZ (L), and endocervical (M) 
regions quantified as above. Scale bar, 20 µm. n=3~4. >20 randomly selected fields 
and >200 cells per condition and cervix. *p<0.05; **p< 0.01; ***p<0.001. 
 
78 
 
2.4.6 GC infection manipulates cell-cell junction in cervical epithelial cells in a 
CEACAM1L and SHP-dependent manner 
Our lab has previously shown that inoculation of GC Pil+ΔOpa induced junction 
disassembly leading to epithelial exfoliation and GC penetration across polarized 
epithelial cells, and Opa expression inhibited these processes [95]. To determine 
whether OpaCEA expression induces a similar effect on the cell-cell junction in the 
three cervical regions, I examined the distribution of adherens junction protein E-
cadherin in the three cervical regions (Fig. 16 A, Fig. 17). I evaluated the integrity of 
E-cadherin based cell-cell junction by the FIR of E-cadherin at the cell-cell junction 
(average of the value of red dots) to that at the cytoplasm (average of the value of 
yellow dots) (Fig. 16 B). Without inoculation with GC, E-cadherin staining was 
primarily localized at the cell-cell junction in the intermedium and basal layers in the 
ectocervix (Fig. 16 A). Inoculation with all three Opa variants did not change the 
junction: cytoplasma FIR of E-cadherin significantly in the ectocervix (Fig. 16 C), 
indicating that ectocervical epithelium is organized and regulated by another 
mechanism. In the TZ, E-cadherin staining was localized at the cell-cell junction; in 
the endocervix, at the lateral surface of polarized endocervical epithelial cells. In both 
TZ and endocervix, inoculation with GC Pil+ΔOpa and OpaC induced a significant 
decrease of junction: cytoplasma FIR (~2) of E-cadherin compared to no inoculation 
control (~4). However, expression of OpaH restored the junction: cytoplasma FIR of 
E-cadherin only in the endocervix (~3.5) but not in TZ (~2) (Fig. 16 C). The above 
data suggest that the expression of OpaCEA inhibited cell-cell junction disruption in the 
endocervix, but not in TZ, and ectocervix. 
79 
 
To determine whether the OpaCEA-mediated inhibition of cell-cell junction disruption 
also involves CEACAM1L, I examined the redistribution of β-catenin, an adaptor 
protein associated with E-cadherin in adherens junction, in the HEC-1-B cells 
transfected with CEACAM1L and CEACAM1S (Fig. 16 D). Using the same method 
as to quantify redistribution of E-cadherin, inoculation of both CEACAM1S and 
CEACAM1L transfected cells with GC Pil+ ΔOpa significant decreased the FIR, 
indicating cell-cell junction disruption. In contrast, inoculation of CEACAM1L 
transfected cells with GC Pil+ OpaH restored the FIR in the CEACAM1L-expressing 
cells alone (Fig. 16 E), indicating the inhibition of cell-cell junction disruption 
requires both OpaCEA and CEACAM1L. 
 
To determine whether the OpaCEA- CEACAM1L dependent inhibition of cell-cell 
junction disruption involves SHP1/2 activation, I examined E-cadherin distribution in 
the cervical tissue treated with or without NSC-87877 using the same method. With or 
without the treatment with inhibitor, GC Pil+ OpaH inoculation did not affect the E-
cadherin distribution in the ectocervix and TZ. In contrast, in the endocervix, 
treatment with the inhibitor-induced a significant decrease of E-cadherin FIR, 
indicating the OpaH-CEACAM1L dependent inhibition of junction disruption 
involves SHP (Fig. 16 F, G). 
 
Taking together, the above data suggest that the inhibition of cell-cell junction 
disruption by OpaCEA involves CEACAM1 and subsequent SHP1/2 activation. In the 
80 
 
endocervix, the inhibited junction disruption by OpaCEA can lead to the decrease 
epithelial exfoliation, thus decreased penetration, and increased colonization. In the 
TZ, the low level of CEACAM expression abolishes the OpaCEA-CEACAM pathway 
and subsequent SHP1/2 activation, thus the expression of which Opa or no Opa does 
not affect GC infectivity. In the ectocervix, the luminal expression of CEACAM1L 
and SHP activation may contribute to the enhanced colonization and inhibited 
exfoliation mediated by OpaCEA, but not through inhibition of junction disruption. 
 
 
 
 
 
81 
 
 
 
 
 
82 
 
 
 
 
 
Figure 16. OpaH inhibits GC-induced junction disruption in the cervical 
epithelial cells in a CEACAM1L and SHP-dependent manner.  
(A) Representative CFM images of human cervical tissue explants inoculated with GC 
Pil+ OpaH, OpaC or Opa and stained for GC and E-cadherin (E-cad). (B and C) 
Disruption of cell-cell junctions was determined by the FIR (±SEM) of E-cad at the 
cell-cell border (average of red dots in B) relative to the cytoplasm (average of yellow 
dots in B) using FI line profiles crossing cell-cell junction (B). (D) Representative 
CFM images of HEC-1-B cells transfected with or without CEACAM1L or 1S, 
inoculated with GC Pil+OpaH or Pil+Opa, and stained for GC, CEACAM, and β-
catenin. Dashed lines, cells expressing CEACAM1L or 1S. (E) Disruption of cell-cell 
junctions was determined by the FIR (±SEM) of β-catenin at the cell-cell border to the 
cytoplasm. (F) Representative CFM images of human cervical tissue explants 
inoculated with GC Pil+OpaH with or without the SHP inhibitor (20 M). (G) 
Disruption of cell-cell junctions was determined by the FIR (±SEM) of E-cad at the 
cell-cell border to the cytoplasm. Scale bar, 20 µm. n=3~4. >20 randomly selected 
fields and >50 cells per condition and per cervix. **p< 0.01; ***p<0.001. 
 
 
 
 
 
 
83 
 
 
 
 
 
 
 
 
 
Figure 17. GC inoculation disrupts E-cadherin-based cell-cell junction.  
Representative 3D reconstruction images of the TZ and endocervical epithelium in 
human cervical tissue explants that were inoculated with or without Pil+OpaH or 
Pil+Opa GC and stained for GC and E-cadherin. Scale bar, 20 m. 
 
 
 
 
 
84 
 
2.5 Discussion 
To study the GC infection in the female reproductive tract, I first developed a human 
cervical tissue explant model, which not only closely mimics in vivo morphology of 
cervical epithelium, but mimics the in vivo GC infectivity as well. Using this model, I 
showed that GC modifies infectivity based on the different properties of cervical 
epithelial cells they encounter by phase varying bacterial surface molecules. When 
exposed to all three cervical regions, GC colonize more in the ectocervix and TZ than 
endocervix; while only penetrate in the TZ and endocervix. Pili are required for 
colonization in all three cervical regions. Expression of OpaCEA enhances colonization 
in ecto and endocervix, where CEACAM is abundantly expressed at the luminal 
surface of the epithelial cells. Expression of OpaCEA inhibits penetration in the 
endocervix, by engaging CEACAMs, leading to subsequent activation of SHP1/2 and 
inhibition of junction disruption. The unresponsiveness of TZ to Opa phase variation 
can be accounted for by the lack of CEACAM expression. The enhanced colonization 
by OpaCEA expression in the ecto and endocervix can enhance the localized 
asymptomatic infection; while the high penetration in the TZ and enhanced 
penetration by the lack of OpaCEA in endocervix provides an explanation for the switch 
to the penetrating symptomatic GC infection which may further lead to DGI. 
 
Transformation zone (TZ) of the cervix is a specialized region in between the ecto and 
endocervix. The formation of TZ starts at the puberty when the endocervix invades 
into the ectocervix. The invasion triggers metaplasia of the part of endocervical tissue 
now located in the ectocervical region to gradually transform into ectocervix. Two 
85 
 
theories of the morphology of TZ are available. The first theory calls TZ the 
squamocolumnar junction (SCJ) instead. It shows as an abrupt junction where the 
polarized endocervical epithelium meets with multiple layered squamous ectocervical 
epithelium. The small region flanks the squamocolumnar junction is defined as TZ. In 
this theory the boundary between the ectocervix and TZ, and TZ and the endocervix 
are blurry. In other words, the range of the region flanks squamocolumnar junction is 
not defined. The second theory shows a region going through metaplasia where the 
epithelium is composed of multiple layers of epithelial cells. And the TZ epithelial 
cells are neither as polarized and columnar as in the endocervix nor as stratified and 
squamous as in the ectocervix. Based on this theory, the TZ is easy to define based on 
the morphology of epithelial cells. I also observe a junction between the TZ and 
ectocervix in this model. Out of the total 22 samples I used for this research, the 
morphology of the TZ in two samples fits the first theory, where the boundary of the 
TZ is hard to define. Thus, data of these two samples were not included. The rest 20 
samples have TZ with the morphology described in the second theory. Although the 
length the TZ varies based on the age of the patients, a clear region of TZ can be 
defined. Thus, data presented here can be applied to the GC infection in the cervix 
with a gradual change of epithelial morphology from the ecto to endocervix. 
 
I showed that TZ is the most vulnerable cervical region to GC infection with a high 
level of colonization and penetration. The Opa phase variation fails to regulate 
infectivity in this region, which can be explained by the lack of CEACAM expression. 
The similar behavior of OpaHSPG and ΔOpa eliminates the role of HSPG as well. These 
86 
 
observations led me to ask the question that what the mechanism of the strong 
infectivity of GC in TZ is. One possibility is the signaling triggered by pili through its 
receptor CR3. The non-piliated GC fails to colonize all three cervical regions, making 
it difficult to study the detailed role of pili in GC infection. The other possibility is the 
dynamics of epithelial turnover in TZ. TZ is known to be a dynamic region, where the 
reserve cells proliferate and differentiate. This dynamic nature of TZ can be utilized 
by GC to infect. Although the data presented here obtained from the cervix with the 
morphology of TZ fitting the second theory, the results may be also applicable to the 
first theory, with the squamocolumnar junction a soft spot for GC infection.  
 
I observed both a strong luminal expression of CEACAM and puncta staining in the 
intermedium layers of the ectocervix. The luminal expression provides an explanation 
of the enhanced colonization by OpaCEA expression. The lack of OpaCEA expression 
reduced the level of colonization but enhanced exfoliation without inducing any 
penetration. This can be explained by the multiple-layered structure of the ectocervical 
epithelium. Of all the samples, despite the dramatic exfoliation induced by OpaHSPG 
and ΔOpa, and the thinness of the ectocervical epithelium left, I have never observed a 
breach of the ectocervical epithelium. It reinforces my central hypothesis that both the 
properties of cervical epithelium and GC phase variable surface molecules affect 
infectivity. It also agrees with the asymptomatic GC infection in ectocervix. In vivo, 
ectocervix is covered by dense normal microbiota. When first deposited, GC need to 
first compete with the normal microbiota to establish a niche, in other words, 
87 
 
colonization, and to proliferate. The successful colonization in ectocervix provides the 
opportunity for GC to ascend to the upper genital tract. 
 
The upper genital tract, the upper portion of TZ and endocervix, is sterile. Thus, when 
GC make it to the TZ and endocervix along the continuity of cervical epithelium, the 
possibility of a small amount of pathogens present causing infection increases. The 
relatively low level of colonization and penetration presented in endocervix agrees 
with the clinical observation. The inhibition of exfoliation and further penetration by 
OpaCEA expression indicates the role of Opa phase variation in regulation of switch 
between localized colonization to disseminated penetration, thus switch between 
asymptomatic to symptomatic infection. 
 
The OpaCEA-CEACAM interaction has long been established, but its role in GC 
infection in the female reproductive tract is not completely understood. Islem et al. 
showed differential expression of CEACAM1, 5, 6, three epithelial CEACAM, in the 
female reproductive tract, indicating the role of CEACAM in the regulation of GC 
infectivity in the three cervical regions. Here, I used an antibody reacting with 
CEACAM1, 3, 6. CEACAM3 is expressed in neutrophil exclusively. CEACAM1 is 
the only one has a cytoplasmic domain and harbors an ITIM motif in the cytoplasmic 
domain, which makes it a perfect candidate for signaling transduction. I used the 
CEACAM1L and CEACAM1S transfected Hec-1-B cells. The enhanced adherence 
and inhibited exfoliation requires both OpaCEA and CEACAM1L, which confirms the 
88 
 
CEACAM1L-dependent regulatory role of OpaCEA of GC infectivity in three cervical 
regions. It has been shown that the ITIM motif of CEACAM1L recruits and activates 
SHP1/2. The treatment of SHP inhibitor, NSC-87877, abolished the OpaCEA-
dependent enhancement of colonization and the inhibition of exfoliation in the ecto 
and endocervix. It also abolishes the inhibition of penetration in endocervix. Overall 
data suggest the engagement of OpaCEA-induced CEACAM1L-SHP pathway in the 
regulation of GC infectivity. 
 
The invasion of GC into epithelial cells intracellularly has been shown in various cell 
line models by invasion assay and by TEM as well. However, using the human 
cervical tissue explants together with the confocal microscopy, it remains technically 
difficult to distinguish intracellular and extracellular bacteria. On the other hand, I 
rarely observed intracellular GC staining, which does not agree with the previous 
observation. However, the ability of GC to disruption cell-cell junction in various cell 
line models as well as in the human cervical tissue explant model indicates that GC is 
also capable of transmigration or penetration across epithelial cells paracellularly, 
without entering into cytoplasma. 
 
My data showed the different organization of E-cadherin based cell-cell junction in the 
three cervical regions, and the differential effect of GC inoculation on E-cadherin 
redistribution. The disruption of the cell-cell junction has been shown in a close 
relationship with epithelial exfoliation [99]. In the ectocervix, E-cadherin is expressed 
89 
 
at the cell-cell junction in the intermedium and basal layers, but not in the superficial 
layers. Despite the dramatic exfoliation induced by OpaHSPG and ΔOpa and the 
thinness of the ectocervical epithelium left, the very superficial layers were always 
devoid of E-cadherin staining and the no redistribution of E-cadherin is detected. 
These data indicated that the exfoliation and response of cell-cell junction to GC 
inoculation in the ectocervical epithelium is mediated by other mechanisms. In TZ, E-
cadherin is expressed at cell-cell junction throughout all the layers. GC inoculation 
induced redistribution of E-cadherin protein into cytoplasma, indicating the junction 
disassembly and junction disruption. Because of the lack of CEACAM expression in 
TZ, Opa phase variation did not affect the level of E-cadherin redistribution, and a 
potentially similar level of exfoliation. In endocervix, E-cadherin based adhesion 
junction is part of the apical junction. Expression of OpaCEA inhibited E-cadherin 
redistribution into cytoplasma, restoring the integrity of cell-cell junction, thus 
inhibited exfoliation and penetration. Since endocervical epithelium is only composed 
of one layer of epithelial cells, the exfoliation induced by GC inoculation can 
potentially create breach on the epithelial layer, thus a portal for GC to penetrate into 
subepithelial tissue and triggers immune response causing symptomatic infection and 
DGI. 
 
Based on the data presented in this chapter, I propose the following working model. 
GC colonize at a high level in the ectocervix without penetration thus result in the 
localized asymptomatic infection. The Opa phase variation favors OpaCEA expression 
in response to the high expression of CEACAM at the luminal surface to enhance the 
90 
 
colonization and the chance to get into the upper genital tract. Once in the endocervix, 
the colonization level is relatively low, but penetration across the single layer of 
epithelial cells occurs. Opa phase variation regulates the infectivity between 
colonization and penetration, thus asymptomatic and symptomatic, localized and 
disseminated gonococcal infection. OpaCEA expression enhances colonization and 
inhibits exfoliation and penetration by engaging CEACAM1L-SHP pathway with 
subsequent inhibited junction disruption. TZ is localized between the ecto and 
endocervix and does not express very much CEACAM, thus did not respond to Opa 
phase variation, and is the most vulnerable to colonization and penetration. 
 
 
 
 
 
 
 
 
 
 
91 
 
Chapter 3: The polarity of epithelial cells regulates Neisseria 
gonorrhoeae infectivity 
 
3.1 Abstract  
To establish infection in the female reproductive tract from the cervix, GC encounter 
both non-polarized stratified ectocervical epithelial cells and polarized columnar 
endocervical epithelial cells. The studies presented in Chapter two have shown that 
distinct properties of cervical epithelial cells affect GC infectivity. However, whether 
the different polarity of epithelial cells, which associates with cell morphology and 
actin cytoskeleton organization, contribute to the differential infectivity of GC in the 
ecto and endocervix is unknown. To address this question, I utilized a non-polarized 
and polarized epithelial cell model generated from the same cell line. While GC 
adhered to both non-polarized and polarized epithelial cells to a similar level. GC 
invasion into non-polarized epithelial cells was more efficient than into polarized 
epithelial cells. The expression of phase variable Opa (Opa+) enhanced GC adhesion 
and invasion, compared to an Opa deletion mutant (ΔOpa). Electron microscopy 
observed microvilli elongation in non-polarized epithelial cells inoculated with Opa+ 
GC, which was concurrent with F-actin recruitment at GC adherent sites. In contrast, 
massive disruption of microvilli was observed in polarized epithelial cells inoculated 
with GC ΔOpa, which was concurrent with NMII-dependent reduction of F-actin at 
GC adherent sites. GC inoculation only induced intracellular calcium in polarized, but 
not in non-polarized epithelial cells. Overall, my data demonstrated that the level of 
92 
 
epithelial cell polarity regulates GC infectivity and the function of Opa in GC 
infection. 
 
3.2 Introduction  
Successful colonization at the mucosal surface is the first step for microbial pathogens 
to establish infection. The epithelium is the first line of defense to protect the host 
against the microbial pathogens. Neisseria gonorrhoeae (GC), the pathogen causing 
the sexually transmitted infection (STI) gonorrhea, initiates infection from the mucosal 
surface of the human cervix, the gate of the female reproductive tract (FRT). The 
human cervical epithelia are composed of multiple-layered non-polarized stratified 
epithelial cells at the ectocervix and single-layered polarized columnar epithelial cells 
at the endocervix. The polarity of epithelial cells is an essential property distinguishing 
the ecto and endocervical epithelial cells. However, whether the epithelial cell polarity 
affects GC infection in the non-polarized ecto and polarized endocervical epithelial 
cells is unknown. 
 
The polarity of polarized epithelial cells is maintained and regulated by the apical 
junction (AJ) and the perijunctional actomyosin ring [117]. One essential component 
of the perijunctional actomyosin ring is the actin motor protein non-muscle myosin II 
(NMII), which regulates the dynamics of the apical junction by contracting the 
perijunctional actin cytoskeleton [156]. The role of the actin-nucleating factor Arp2/3 
in regulating the dynamics of the apical junction through perijunctional actin 
93 
 
cytoskeleton has been suggested [121]. In the non-polarized epithelial cells, the 
Arp2/3 complex mediates the polymerization of branched actin filament and plays an 
essential role in remodeling of the plasma membrane in response to pathogens [135]. 
Furthermore, polarized epithelial cells are known to form densely packed microvilli at 
the apical surface, which is supported by the bundled F-actin [141]. Similarly, in the 
human cervix, the polarized columnar endocervical epithelial cells form the apical 
junction between neighboring cells [123] and densely packed microvilli at the apical 
surface [124]. In contrast, the mucosal surface of non-polarized ectocervical 
epithelium appears smooth [124].  
 
It has been shown that GC can induce membrane ruffling, microvilli elongation and 
the rearrangement of the actin cytoskeleton at the plasma membrane in multiple non-
polarized epithelial cell lines [130, 131, 133]. Our lab has previously shown that GC 
can induce cell-cell junction disassembly in both polarized epithelial cell line [76] and 
the human endocervical tissue explant [95]. However, whether GC induce the similar 
response of actin cytoskeleton and the underlying mechanism in the non-polarized 
ecto and polarized endocervical epithelial cells are unknown.  
 
Here, I used a non-polarized and polarized epithelial cell model generated from the 
same cell line to study the impact of epithelial cell polarity on GC infection processes 
and its underlying mechanism. I found that GC invade into non-polarized epithelial 
cells more efficiently than into polarized epithelial cells without changing the level of 
94 
 
GC adhesion. And the expression of Opa enhances GC adhesion and invasion 
compared to an Opa deletion mutant (ΔOpa). Such differential GC infectivity is 
concurrent with differential morphology change of plasma membrane and F-actin 
reorganization induced by GC in the non-polarized and polarized epithelial cells.  
 
3.2 Materials and Methods  
Neisseria Strains 
N. gonorrhoeae strain MS11 that expressed both phase variable Pili and Opa (Pil+ 
Opa+) and the MS11 opa deletion mutant (Pil+ ΔOpa) were used [149]. GC were 
grown on GC media plates supplemented with 1% Kellogg’s supplement (GCK) for 
16–18 h before inoculation. Pil+ colonies were identified based on colony morphology 
using a dissecting light microscope. GC was suspended and the concentration of 
bacteria was determined by using a spectrophotometer. GC were inoculated with 
epithelial cells at MOI ~10. 
 
Epithelial Cells 
Human colorectal carcinoma cell line T84 cells (ATCC) was maintained in Dulbecco's 
modified Eagle's medium: Ham F12 (1:1) supplemented with 7% heat-inactivated 
FBS. Cells were maintained at 37°C with 5% CO2. Cells were seeded at 6x104 (6.5 
mm diameter transwell inserts, Corning, Lowell, MA, USA) and cultured for ~10 days 
until transepithelial electrical resistance (TEER) reached ~2000 Ω for polarized 
epithelial cells. Cells were seeded at 1.2 x105 (6.5 mm diameter transwell inserts) and 
95 
 
cultured for ~2 days for non-polarized epithelial cells. TEER was measured using a 
Millicell ERS volt-ohm meter (Millipore, Bedford, MA, USA). 
 
Immunofluorescence analysis of epithelial cells 
Non-polarized and polarized T84 cells were inoculated with GC for 6 h. Cells were 
rinsed, fixed with 4% paraformaldehyde, permeabilized, and stained with phalloidin 
(Life Technology, Grand Island, NY, USA), anti-GC antibodies, and hoechst for 
nuclei (Life Technologies). Cells were analyzed by confocal fluorescence microscopy 
(Zeiss LSM 710, Carl Zeiss Microscopy LLC, Thornwood, NY, USA). Z-series of 
images were obtained in 0.57 μm slices from the apical to the basal of the monolayer, 
and three-dimensional (3D) composites obtained.  
 
The redistribution of F-actin induced by GC inoculation at the luminal surface was 
examined by using the x-z view images cutting across the luminal surface to the basal 
level. Fluorescence intensity (FI) profiles of F-actin were generated by drawing a line 
underneath GC microcolonies using the ImageJ software. The fluorescence intensity 
ratio (FIR) of F-actin underneath GC microcolonies to the neighboring area without 
GC was quantified. 
 
 
 
96 
 
Human cervical tissue explants 
The tissue explants were cultured using a previously published protocol [152]. Briefly, 
cervical tissues were obtained from patients undergoing voluntary hysterectomies and 
received within 24 h post-surgery. Age of patients ranges from 28 to 40. Supporting 
muscle tissue was removed by using a carbon steel surgical blade. Remaining cervical 
tissue samples were cut into ~2.5 cm (L) X 0.6 cm (W) X 0.3 cm (H) pieces, incubated 
in 6-well tissue culture plate in cervical tissue culture media conataining CMRL-1066 
(GIBCO), with 5% heat-inactivated fetal bovine serum, 2 mM L-glutamine, bovine 
insulin (1 μg/ml, Akron Biotech) and penicillin/streptomycin for 24 h and then in 
antibiotic-free media for another 24 h, before inoculation with GC as previously 
described.   
 
Immunofluorescence analysis of human cervical tissue explants 
Individual pieces of cervical tissue explants were inoculated with GC at MOI ~10. The 
number of epithelial cells at the mucosal surface was determined by the area of the 
mucosal surface of the individual piece and the average luminal area of individual 
cervical epithelial cells. When indicated, the tissue explants were pretreated with MLC 
kinase inhibitor ML-7 (10 μM, Calbiochem, San Diego, CA, USA) for 1 h, and 
inoculated with GC with or without the inhibitors for 24 h as previously described. 
Briefly, GC were collected by using a sterile applicator to swab GC from the plate and 
re-suspend GC in pre-warmed antibiotic-free cervical tissue culture media. An aliquot 
of the GC suspension was added directly onto each tissue piece to make an MOI = 10 
97 
 
bacteria to 1 epithelial cell. The cervical tissue explants were incubate at 37 °C with 
5% CO2 with gentle shaking for 24 h. The infected cervical tissue explants were 
rinsed with antibiotic-free cervical tissue culture medium at 6 and 12 h to remove un-
adhered bacteria and replace with fresh media without bacteria. Tissue explants were 
then fixed 24 h post inoculation, embedded in 20% gelatin, cryopreserved, sectioned 
by cryostat across the luminal and basal surfaces of the epithelium, stained for F-actin 
(Cytoskeleton), Arp2 (Santa Cruz Biotechnology), GC by specific antibodies, and 
nuclei by Hoechst (Life Technologies), and analyzed using confocal fluorescence 
microscope (Zeiss LSM 710, Carl Zeiss Microscopy LLC) as previously described 
[95]. Images of the luminal surface were randomly acquired in the endocervix as 
single images or Z-series of 0.57 μm/image, and 3D composites obtained using Zeiss 
Zen software.  
 
The redistribution of F-actin induced by MS11 Pil+Opa+ and ΔOpa was evaluated by 
two methods: 1) the percentage of GC microcolonies with decrease F-actin staining 
underneath versus total number of GC microcolonies, 2) the fluorescence intensity 
ratio (FIR) of F- actin at the apical surface to the lateral surface in individual 
endocervical epithelial cells. The redistribution of Arp2 induced by MS11 Pil+ ΔOpa 
was evaluated by the fluorescence intensity ratio (FIR) of Arp2 at the apical surface to 
the lateral surface in individual endocervical epithelial cells.  
 
 
98 
 
GC adherence and invasion assays 
Non-polarized and polarized T84 cells were inoculated apically with GC for 3 h for 
adherence assay and 6 h for invasion assays. For adherence, cells were intensively 
rinsed, lysed and plated on GCK plates to determine the CFU. For invasion, cells were 
treated with gentamicin (100 μg/ml) for 2 h, rinsed, lysed and plated. Cell-associated 
bacteria that were resistant to gentamicin treatment were counted as invaded GC. 
 
Calcium imaging 
Non-polarized and polarized T84 cells were seeded at 1×105 per transwell on the 
underside of transwells [95] and cultured for ~2 days for non-polarized, and ~10 days 
for polarized epithelial cells. Cells were pretreated with or without the intracellular 
Ca2+ inducer thapsigargin (10 μM, Sigma) for 1 h and incubated with GC (MOI ~10) 
apically with or without the inhibitors for 4 h. Cells were incubated with the 
fluorescent calcium indicator Fluo-4 (100 μM, Life Technologies) for 1 h. Confocal x-
z view images were acquired with the membrane dye CellMask (5 mg/ml, Life 
Technology) using Leica TCS SP5X confocal microscope (Leica Microsystems, 
Buffalo Grove, IL). 
 
To quantify the level of intracellular calcium, the cytoplasmic area of the individual 
cell was manually selected based on the Cell Mask staining in randomly acquired 
confocal images, and the mean fluorescent intensity (MFI) of Fluo-4 in the 
cytoplasmic region was measured using the NIH ImageJ software. 
 
99 
 
Ultrastructure analysis 
For transmitted electron microscopy (TEM), non-polarized and polarized T84 cells 
were apically incubated with GC for 6 h with MOI ~50. The monolayer of epithelial 
cells on the transwell inserts was first fixed with 2% glutaraldehyde, and secondary 
fixed with 1% osmium tetroxide. Samples were rinsed with distilled water and post-
fixed with 2% uranyl acetate. The samples were dehydrated in a 35% to 100% graded 
ethanol series, and infiltrated with graded EPON 812 resin (Araldite/Medcast; Ted 
Pella, Redding, Ca.) series. The transwell membranes were embedded in flat silicone 
molds (Polyscience, Inc. Warrington, PA) overnight at 60°C. Cross sections of the 
cells were obtained by sectioning through the embedded transwell membranes. 
Sections were placed on bare 150-mesh copper grids, stained with uranyl acetate and 
lead citrate, and viewed with a ZEISS 10CA electron microscope (ZEISS, Thornwood, 
NY). 
 
For scanning electron microscopy (SEM), non-polarized and polarized T84 cells were 
apically incubated with GC for 6 h with MOI ~50. The monolayer of epithelial cells 
on the transwell inserts was first fixed with 2% glutaraldehyde, and secondary fixed 
with 1% osmium tetroxide. The samples were dehydrated in a graded ethanol series, 
dried by critical point drying method, and coated with gold-palladium alloy. Samples 
were visualized with Amray 1820D microscope (20 kV) and Hitachi S4700 
microscope (5 kV). 
 
 
100 
 
Statistical analysis 
Statistical significance was assessed by using Student’s t-test by Prism software 
(GraphPad Software, San Diego, CA). P-values were determined using unpaired t test 
with Welch’s correction in comparison with no infection controls. 
 
Ethics statement 
Human cervical tissue explants were obtained from the National Disease Research 
Interchange (NDRI, Philadelphia, PA). Human cervical tissues used were anonymized. 
The use of human tissues has been approved by the Institution Review Board of the 
University of Maryland. 
 
3.4 Results  
3.4.1 GC invade into non-polarized epithelial cells more efficiently than into 
polarized epithelial cells without altering the level of GC adhesion 
To mimic the non-polarized and polarized nature of the ecto and endocervical 
epithelial cells, I established the non-polarized and polarized epithelial cell model 
using human colonic epithelial cell line T84. T84 is the only cell line that can be 
polarized to the level similar to that of endocervical epithelial cells in vivo. And none 
of available female genital epithelial cell line can be polarized to that level. T84 cells 
also express the host receptors for GC Opa proteins as well [93, 157]. T84 cells were 
cultured on transwells for ~10 days, when the transepithelial resistance reached 2000, 
to mimic polarized epithelial cells and for two days (transepithelial resistance <300) to 
mimic non-polarized epithelial cells. While the 2-day culture of T84 cells showed flat 
101 
 
morphology and non-polarized staining of junction proteins E-cadherin and ZO-1, the 
10-day culture of T84 cells displayed columnar morphology and polarized staining of 
E-cadherin and ZO-1 staining at the apical junction.  
 
To answer the question whether GC infect non-polarized and polarized epithelial cells 
in the same way, I compared the level of adhesion and invasion of piliated Opa phase 
variable MS11 (Pil+Opa+) and its Opa deletion mutant (Pil+ΔOpa) in non-polarized 
and polarized epithelial cells. To measure GC adherence, epithelial cells were 
incubated with GC for 3 h. Epithelial cells were intensively washed to remove non-
adhered GC, lysed, and the lysates were cultured to numerated the colony forming 
units (CFU). Both Pil+Opa+ and Pil+ ΔOpa Opa adhered to the non-polarized and 
polarized epithelial cells at the same level (Fig. 18 A). GC invasion was determined 
using gentamicin resistant assay post 6 h inoculation. In contrast to GC adhesion, there 
were significantly more gentamicin resistant Pil+Opa+ and Pil+ ΔOpa GC in the non-
polarized than the polarized epithelial cells (Fig. 18 B). This suggests that GC invade 
into the non-polarized epithelial cells more efficiently than into polarized epithelial 
cells independently of GC adhesion. 
 
 
 
 
 
102 
 
 
 
 
 
Figure 18. GC invade into non-polarized epithelial cells more efficiently than into 
polarized epithelial cells without altering the level of GC adhesion.  
Two-day non-polarized and 10-day polarized T84 cells were apically inoculated with 
Pil+Opa+ or ΔOpa GC (MOI~10). (A) Adherent GC (±SEM) were quantified by 
culturing the lysates of infected epithelial cells 3 h post incubation.  (B) Invaded GC 
(±SEM) were quantified by gentamicin resistance assay 6 h post inoculation. The 
results were generated from 4–6 independent experiments. *p<0.05; **p< 0.01; 
***p<0.001. 
 
 
 
103 
 
3.4.2 Expression of Opa enhances GC adhesion and invasion in both non-
polarized and polarized epithelial cells  
Chapter 2 has shown that both properties of epithelial cells and the phase variable GC 
surface molecules affect GC infectivity. To determine if Opa expression differentially 
affected GC infectivity in non-polarized and polarized epithelial cells, I compared the 
level of adhesion and invasion of Pil+ Opa+ and Pil+ΔOpa GC in non-polarized and 
polarized epithelial cells. Pil+ Opa+ GC exhibited a significantly higher level of 
adhesion and invasion in both non-polarized and polarized T84 cells than Pil+ΔOpa 
GC (Fig. 19 A, B). In addition, both GC strains adhered to and invade into non-
polarized epithelial cells much more efficiently than into polarized epithelial cells 
(Fig. 19 A). The above results indicate that Opa proteins play a similar role in GC 
infection of non-polarized and polarized epithelial cells.   
 
 
 
 
 
 
 
 
 
104 
 
 
 
 
 
Figure 19. Expression of Opa enhances the level of adhesion and invasion in non-
polarized and polarized epithelial cells. 
Two-day non-polarized and 10-day polarized T84 cells were apically inoculated with 
Pil+ Opa+ or Pil+ ΔOpa for 3 or 6 h (MOI~10), and the numbers of adherent GC 
(±SEM) (A) and invading GC (±SEM) (B) were determined as described above. n=4–
6. *p<0.05; **p< 0.01; ***p<0.001. 
 
 
 
 
105 
 
3.4.3 GC inoculation induces differential membrane remodeling in non-polarized 
and polarized epithelial cells 
Previous studies showed that GC inoculation can induce membrane ruffling and 
microvilli elongation in primary cervical epithelial cell culture [130], patient biopsies 
[130], and cell lines [39, 131, 132, 134]. This membrane remodeling has been 
implicated for GC invasion. The different levels of GC adhesion and invasion in non-
polarized and polarized epithelial cells lead to a hypothesis that Opa can regulate GC 
infectivity by inducing differential membrane remodeling in non-polarized and 
polarized epithelial cells. To test this hypothesis, I conducted transmitted electron 
microscopy (TEM) (Fig. 20 A) and scanning electron microscopy (SEM) (Fig. 20 B) 
of non-polarized T84 cells to compare with polarized T84 cell images obtained by 
previous graduate students Liang-Chun Wang (TEM) and Senthil V. Bhoopalan 
(SEM). Without GC inoculation, non-polarized T84 cells possess short irregular 
membrane protrusions on their luminal membrane (Fig. 20 A, arrows), while polarized 
T84 cells have densely packed microvilli at the apical surface (Fig. 20). The 
inoculation of GC Pil+Opa+ induced elongation of membrane protrusions in non-
polarized cells (Fig. 20 open arrows) but did not change the morphology of microvilli 
of the polarized cell. Pil+Opa+ appeared to establish intimate interaction with the 
surface of non-polarized but only sat on the top of microvilli at the apical surface of 
polarized epithelial cells (Fig. 20 arrowheads). In contrast to GC Pil+Opa+, GC 
Pil+ΔOpa did not induce membrane protrusion in non-polarized cells or sit on the top 
of microvilli of polarized cells. Instead, GC Pil+ ΔOpa inoculation abolished 
microvilli of both non-polarized and polarized epithelial cells, enabling Pil+ΔOpa to 
106 
 
form intimate interactions with the luminal membrane of epithelial cells as well (Fig. 
20 arrowheads). Thus, the expression of Opa induces membrane remodeling in non-
polarized epithelial cells but inhibits GC-induced disruption of microvilli in polarized 
epithelial cells.  
 
 
Figure 20. GC inoculation induces differential remodeling of the plasma 
membrane of non-polarized and polarized epithelial cells. 
Two-day non-polarized and 10-day polarized T84 cells were apically incubated with 
Pil+Opa+ or ΔOpa for 6 h (MOI~50). Cells were fixed and processed for transmitted 
electron microscopy (TEM) (A) and scanning electron microscopy (SEM) (B). Shown 
are representative images. (TEM of polarized T84 cells was obtained by Liang-Chun 
Wang, and SEM of polarized T84 cells was obtained by Senthil Velan Bhoopalan). 
 
 
107 
 
3.4.4 GC inoculation induces differential redistribution of F-actin at the luminal 
surface of non-polarized and polarized epithelial cells 
Disruption of the actin cytoskeleton by cytochalasin D inhibited GC invasion into but 
not adhesion to Chang epithelial cells [132]. I postulated that GC induce different 
morphological changes in the plasma membrane of non-polarized and polarized 
epithelial cells by differentially remodeling the actin cytoskeleton underneath, which 
leads to different levels of GC adhesion and invasion. To test this, I inoculated non-
polarized and polarized T84 cells with GC Pil+Opa+ and Pil+ΔOpa, for 6 h. Infected 
cells were fixed, stained for F-actin, GC, and nuclei, and analyzed by confocal 
fluorescence microscopy (CFM). The remodeling of the actin cytoskeleton was 
evaluated by the mean fluorescent intensity ratio (FIR) of F-actin staining underneath 
GC microcolonies to the immediately neighboring region at the cell surface. In non-
polarized T84 cells, the FIRs of F-actin staining underneath both Pil+Opa+ and 
Pil+ΔOpa microcolonies were significantly bigger than 1 (Fig. 21 A, open arrows), 
indicating the recruitment of the actin cytoskeleton to GC adherent sites. The FIR of 
F-actin underneath Pil+Opa+ GC (FIR~ 1.5) was significantly higher than that 
underneath Pil+ΔOpa (FIR~ 1.2) (Fig. 21 B), indicating that Opa expression enhances 
F-actin accumulation in non-polarized epithelial cells. In contrast, in polarized T84 
cells, the FIRs of F-actin staining underneath both Pil+Opa+ and Pil+ΔOpa GC 
microcolonies at the apical surface were smaller than one and significantly lower than 
those in non-polarized epithelial cells (Fig. 21 A, stars in x-y view, arrows in x-z 
view). Furthermore, the F-actin FIR in Pil+ΔOpa-infected cells (~ 0.5) was 
significantly lower than that of Pil+Opa+ infected polarized epithelial cells (~ 0.7) 
108 
 
(Fig. 21 B). Taken together, these data suggest that GC induce F-actin recruitment to 
adherent sites in non-polarized epithelial cells, which is associated with microvilli 
elongation, but F-actin reduction at GC adherent sites in polarized epithelial cells, 
which is associated with microvilli damaging. Opa expression enhances F-actin 
recruitment and reduction at GC adherent sites of non-polarized and polarized 
epithelial cells respectively.  
 
 
 
 
 
 
 
 
 
 
 
 
 
109 
 
 
 
Figure 21. GC inoculation induces differential redistribution of F-actin at GC 
adherent sites in non-polarized and polarized epithelial cells. 
Two-day non-polarized and 10-day polarized T84 cells were apically incubated with 
Pil+Opa+ or ΔOpa for 6 h (MOI~10). Cells were fixed and stained for F-actin and GC, 
and analyzed using 3D-CFM. (A) Representative images of x-y views of the luminal 
surface and x-z views of sections cross the luminal to the basal surface are shown. (B) 
The level of F-actin redistribution was quantified by mean fluorescence intensity ratio 
(FIR) (±SEM) of F-actin underneath GC microcolony to the adjacent luminal area. 
n=3, *p<0.05; **p< 0.01; ***p<0.001.  
 
 
110 
 
3.4.5 GC induce F-actin redistribution in endocervical epithelial cells, which 
involves NMII 
To determine if GC-induced differential actin reorganization in non-polarized and 
polarized epithelial cells occurs in the human cervix, I used the human cervical tissue 
explant model. Without GC inoculation, F-actin staining is evenly distributed under 
the plasma membrane of individual epithelial cells across all the layers in the 
ectocervix (Fig. 22). In contrast, F-actin staining in the endocervical epithelium is 
highly polarized to the apical surface with much less staining at the lateral and basal 
surface (Fig. 23 A). Unlike in the non-polarized T84 cells, GC inoculation (both 
Pil+Opa+ and Pil+ΔOpa) had no significant effect on F-actin redistribution in 
ectocervical epithelial cells (Fig. 22). In contrast, decreased F-actin staining was 
observed underneath GC microcolonies in endocervical epithelial cells (Fig. 23 A, 
arrows). By visual inspection, I found ~70% Pil+ΔOpa but ~40% Pil+Opa+ 
microcolonies with decreased F-actin staining underneath (Fig. 23 B).  
 
In addition to the reduction of F-actin staining underneath GC microcolonies, the level 
of F-actin staining under the lateral membrane increased, which was concurrent with a 
decrease in the height of endocervical columnar epithelial cells inoculated with 
Pil+ΔOpa. These observations suggest a decrease in the polarity of endocervical 
epithelial cells. To quantify this decrease, I measured the FIR of F-actin at the apical 
surface (Fig. 23 C, yellow dotted line) relatively to the lateral surface (Fig. 23 C, 
yellow dotted line). Inoculation of both Pil+Opa+ and Pil+ΔOpa induced significant 
decreases in the F-actin apical to the lateral FIR, compared to no inoculation control. 
111 
 
Furthermore, the apical to lateral F-actin FIR in Pil+ΔOpa GC-infected endocervical 
epithelial cells (~2) was significantly lower than Pil+Opa+ GC-infected cells (~3) 
(Fig. 23 C). Taking together, my data suggest that GC can induce actin depolarization 
beneath their contact sites and disrupt the polarized actin organization in the 
endocervical epithelial cells, but do not significantly change actin organization in 
ectocervical epithelial cells. Opa expression weakens the ability of GC to remodel the 
actin cytoskeleton. 
 
Previous studies from our lab showed that GC induced activation of the actin motor 
protein NMII through its light chain kinase (MLCK) and NMII activation involved in 
GC-induced exfoliation and penetration in the human endocervix [95]. To address the 
question if NMII activation plays any role in the GC-induced actin redistribution in 
polarized epithelial cells, I examined F-actin distribution in epithelial cells of cervical 
tissue explants treated with the MLCK inhibitor ML-7. The treatment with ML-7 
significantly decreased the percentage of Pil+ΔOpa microcolonies with decreased F-
actin underneath (from ~70% to ~ 40%) (Fig. 23 A, B, arrowheads). The apical 
polarity of F-actin, measured by F-actin apical to the lateral FIR, and columnar 
morphology of endocervical epithelial cells were also partially restored by ML-7 (Fig. 
23 C). These results indicate that GC-induced actin remodeling in the columnar 
endocervical epithelial cells involves NMII activation. The inhibition of NMII 
activation can be a potential mechanism by which Opa inhibits actin remodeling in 
polarized endocervical epithelial cells. 
 
112 
 
 
 
  
Figure 22. GC inoculation had no significant effect on F-actin redistribution in 
ectocervical epithelial cells. 
Human ectocervical tissue pieces were incubated with Pil+Opa+ and Pil+ΔOpa 
(MOI~10) for 24 h, with unassociated GC washed off at 6 and 12 h. Tissues were 
fixed, stained for F-actin, DNA, and GC, and analyzed using 3D-CFM. Representative 
images that intercept both the apical and basolateral surfaces are shown (Scale bar, 20 
μm). n=3, *p<0.05; **p< 0.01; ***p<0.001. 
 
 
113 
 
 
Figure 23. GC induce F-actin redistribution in endocervical epithelial cells, which 
involves NMII. 
Human endocervical tissue pieces were untreated or pretreated with ML-7 for 1 h and 
incubated with Pil+Opa+ and ΔOpa (MOI~10) with or without inhibitor for 24 h at 
37˚C, with unassociated GC washed off at 6 and 12 h. Tissues were fixed, stained for 
F-actin, DNA, and GC, and analyzed using 3D-CFM. (A) Representative images that 
intercept both the apical and basolateral surfaces are shown (Scale bar, 20 μm). The 
level of F-actin redistribution was quantified by the percentage of GC microcolonies 
with decreased F-actin staining (±SEM) (B), and by the FIR of F-actin at the apical 
surface to the lateral surface of endocervical epithelial cells (±SEM) (C). n=2, 
*p<0.05; **p< 0.01; ***p<0.001. 
114 
 
3.4.6 GC induce the redistribution of the actin nucleation factor Arp2/3 in 
endocervical epithelial cells, which involves NMII 
Arp2/3 complex, which mediates polymerization of branched actin filaments, has been 
implicated for the uptake of various pathogens [135]. I hypothesize that GC regulate 
actin remodeling in polarized endocervical epithelial cells through the Arp2/3 
complex. To test this hypothesis, I inoculated human endocervical tissue explants with 
Pil+Opa+ and Pil+ΔOpa GC for 24 h and stained tissue sections for Arp2. The Arp2 
distribution in columnar endocervical epithelial cells was quantified by the Arp2 FIR 
at the apical to the lateral membrane as described in Figure 23 C. Without infection, 
Arp2 staining concentrated beneath the apical membrane, with an FIR >2 (Fig. 24 A, 
B). Inoculation with GC Pil+ΔOpa significantly decreased the apical polarization of 
Arp2 (FIR< 1.5). Inoculation with GC Pil+Opa+ also decreased the Arp2 polarity 
(FIR~ 1.7), but the decrease was significantly lower than that induced by GC 
Pil+ΔOpa (Fig. 24 A arrowheads, B). The treatment of the MLCK inhibitor ML-7 
inhibited Arp2 redistribution induced by Pil+ΔOpa back to the level of Pil+ Opa+ 
infected cells (FIR~ 2)(Fig. 24 A, arrowheads). My data indicate that GC disrupt the 
apical polarized distribution of Arp2/3 in endocervical epithelial cells, which depends 
on GC-induced NMII activation. Opa expression inhibits GC-induced redistribution of 
Arp2/3 
 
 
 
115 
 
 
 
Figure 24. GC induce Arp2/3 redistribution in endocervical epithelial cells, which 
involves NMII. 
Human endocervical tissue explants were untreated or pretreated with ML-7 for 1 h 
and incubated with Pil+Opa+ and Pil+ΔOpa (MOI~10) with or without inhibitor for 
24 h at 37˚C, with unassociated GC washed off at 6 and 12 h. Tissues were fixed, 
sectioned, stained for Arp2, DNA and GC, and analyzed using CFM. (A) 
Representative images that intercept both the apical and basolateral surfaces are shown 
(Scale bar, 20 μm). (B) The level of Arp2 redistribution was quantified by the FIR of 
Arp2 staining at the apical relative to the lateral surface (±SEM). n=2 *p<0.05; **p< 
0.01; ***p<0.001. 
 
 
116 
 
3.3.7 GC inoculation induces the elevation of the intracellular calcium level in 
polarized but not non-polarized epithelial cells 
Data from my experiments and previous lab members have shown that GC inoculation 
induces the increase of intracellular calcium level in polarized T84, which leads to the 
NMII activation, apical junction disruption, epithelial exfoliation, and eventually 
penetration into the subepithelium. Opa expression inhibits these processes [95]. 
However, whether GC also induce the calcium flux similarly in non-polarized 
epithelial cells is unknown. To address this question, I compared the intracellular 
calcium levels in non-polarized and polarized T84 cells with or without inoculation 
with GC Pil+Opa+ and Pil+ΔOpa by the mean fluorescence intensity (MFI) of Fluo-4, 
a calcium dye, in the cytoplasma region of individual cells (Fig. 25 A). Epithelial cells 
treated with thapsigargin, which induces intracellular calcium by inhibition of calcium 
pumping into endoplasmic reticula, serves as a positive control. Consistent with my 
previous results, inoculation with GC Pil+Opa+ and Pil+ΔOpa both increased the 
intracellular calcium level in polarized epithelial cells, with the level in Pil+ΔOpa-
infected cells significantly higher than Pil+Opa+-infected cells (Fig. 25 B). In contrast, 
in non-polarized epithelial cells, cells inoculated with both Pil+Opa+ and Pil+ΔOpa 
did not show significant changes in the level of the intracellular calcium (Fig. 25 B). 
This result suggests that GC induce different upstream signaling which triggers 
distinct actin reorganization and leads to different infectivity of GC in non-polarized 
and polarized epithelial cells. 
 
 
117 
 
 
 
 
118 
 
 
 
 
 
 
Figure 25. GC inoculation induces the elevation of intracellular calcium levels in 
polarized but not in non-polarized T84 cells. 
Two-day non-polarized and 10-day polarized T84 cells were incubated apically with 
or without GC Pil+Opa+ or Pil+ΔOpa with or without thapsigargin (10 μM), for 4 h. 
Cells were incubated with the calcium indicator Fluo-4 and the membrane dye 
CellMask and analyzed using 3D-CFM. (A) Representative images of x-z view are 
shown. Scale bar, 20 μm. (B) The level of intracellular calcium was quantified by MFI 
(±SEM) of Fluo-4 in the cytoplasmic region of >50 cells from three independent 
experiments. n=3, *p<0.05; **p< 0.01; ***p<0.001. 
 
 
 
 
 
 
119 
 
3.5 Discussion  
Cell polarity is the key difference in the properties of epithelial cells in the human 
cervix. This study utilized non-polarized and polarized epithelial cell model generated 
from the same cell line to study the impact of cell polarity on GC infectivity and GC-
induced actin cytoskeleton remodeling. I found that GC invade more efficiently into 
non-polarized than polarized epithelial cells, but cell polarity does not significantly 
affect GC of adhesion despite their different surface morphology. GC remodel the 
actin cytoskeleton in non-polarized and polarized epithelial cells in the opposite way, 
accumulating and reducing F-actin underneath GC adherent sites, respectively. The F-
actin accumulation in non-polarized epithelial cells is concurrent with microvilli 
elongation, and the F-actin reduction, which also occurs in GC-infected endocervix, is 
concurrent with microvilli demolishment. These results demonstrate that the epithelial 
cell polarity can significantly impact GC infectivity, and the non-polarized and 
polarized organization of the actin cytoskeleton is a critical underlying factor. 
 
By comparing GC strains expressing phase variable Opa and no Opa, I found that Opa 
expression enhanced the level of adhesion in both non-polarized and polarized 
epithelial cells. This observation is consistent with what I have seen in the human 
cervical tissue explants where expression of CEACAM- binding Opa (OpaCEA) 
enhanced GC colonization in both non-polarized ectocervical and polarized 
endocervical epithelial cells. These findings suggest that epithelial cell polarity does 
not significantly change the regulatory effect of Opa on GC infectivity, and the non-
polarized and polarized epithelial cell model mimics a certain aspect of GC infection 
120 
 
processes in vivo. However, there is an inconsistency in our observation between the 
two models. In the cell line model, GC adhered to non-polarized and polarized 
epithelial cells at a similar level. But GC colonized the ectocervix at a higher level 
than the endocervix. Such differences could be caused by a differential expression 
level of GC host receptors in the ecto and endocervical epithelial cells. Chapter 2 has 
shown a higher level of CEACAM expression in ectocervical than endocervical 
epithelial cells. In contrast, non-polarized and polarized cell model was derived from 
the same cell line and unlikely had different expression level of the host receptors due 
to polarization. Technical differences may also contribute to the inconsistency. In the 
tissue explant model, GC had access to all three cervical regions, but in the cell line 
model, GC were incubated with non-polarized and polarized epithelial cells 
separately. 
 
It is technically difficult to identify GC inside epithelial cells in human cervical tissue 
explants using confocal fluorescence microscopy, thus GC invasion into epithelial 
cells in tissue explants cannot be examined. However, I did not observe significant GC 
staining between the luminal surface and the basal membrane of the cervical 
epithelium. This is not surprising as only 0.1-1% of adhered GC invade into epithelial 
cells using a cell line model and gentamicin resistance assay. The non-polarized and 
polarized epithelial cell line model enabled me to examine the impact of epithelial 
polarity on GC invasion. I found that GC invade more efficiently into the non-
polarized than polarized epithelial cells, which is consistent with the clinical 
observation of GC inside cervical epithelial cells shed off the epithelium [20]. 
121 
 
Furthermore, I found that Opa expression enhanced invasion into both types of 
epithelial cells. Published data from Song lab [95] and the results presented in Chapter 
2 showed that Pil+ΔOpa GC transmigrate across polarized epithelial cells and 
penetrate into the subepithelium of the polarized endocervical epithelial cells more 
efficiently than Pil+Opa+. However, none of the GC strains used penetrate into the 
subepithelium of non-polarized ectocervical epithelial cells. These findings together 
suggest that Opa expression differentially regulates GC infectivity in different types of 
epithelial cells. In non-polarized epithelial cells and the ectocervix, Opa expression 
enhanced GC colonization and invasion, but in polarized epithelial cells and the 
endocervix, inhibited GC penetration. Consequently, expression of Opa drives GC 
infectivity to colonization and invasion but not penetration.  
 
Non-polarized and polarized epithelial cells have different morphology (stratified with 
sparse microvilli or columnar with dense microvilli at the apical membrane) and actin 
organization (non-polarized or apical polarized distribution). Apical polarized F-actin 
generates and maintains the apical microvilli of polarized epithelial cells. It is 
expected that GC can induce different changes in the morphology and actin 
organization in non-polarized and polarized epithelial cells. It is surprising that GC 
modify actin organization in non-polarized and polarized epithelial cells in the 
opposite way, increasing or decreasing F-actin underneath GC adherent sites, which 
may lead to different morphological changes in non-polarized and polarized epithelial 
cells, microvilli elongation and demolishment respectively. Pil+Opa+ GC induce 
microvilli elongation in non-polarized epithelial cells, which may contribute to 
122 
 
enhanced GC invasion by wrapping GC with host cell membranes. The extensive 
membrane protrusion induced by GC Pil+Opa+ is potentially mediated by actin 
polymerization underneath GC microcolonies, demonstrated by the recruitment of F-
actin. In contrast, GC Pil+ΔOpa+ induce microvilli demolishment at the apical surface 
of polarized epithelial cells probably by inhibiting actin polymerization or enhancing 
actin depolymerization. Microvilli demolishment appears to allow GC to interact with 
the apical membrane intimately, probably facilitating GC invasion by the zipper 
mechanism, which requires membrane ruffling [130-132]. However, as Opa 
expression reduces the level of Arp2/3 and F-actin exclusion but enhances GC 
invasion into polarized epithelial cells, the reduction of Arp2/3 and F-actin is not 
essential for GC invasion. These data demonstrate that GC differentially modify the 
morphology and actin organization of non-polarized and polarized epithelial cells to 
promote adhesion and invasion. The mechanism by which GC modulate actin 
organization in epithelial cells is unknown. Unlike those well-studied actin modulating 
pathogens [135], no secreted effector proteins have been found in GC. The potential 
mechanisms include the interactions of GC surface molecules with host receptors and 
the downstream signaling of host receptors. For example, the host receptors for Opa, 
CEACAM, have been shown to direct or indirect interaction with actin regulators [59, 
158]. 
 
GC-induced actin recruitment to GC adherent sites was observed in non-polarized T84 
cells but was not observed in the GC-inoculated ectocervical epithelial cells in tissue 
explants. The ectocervical epithelium has a much more complicated cytoarchitecture 
123 
 
than non-polarized epithelial monolayers. The superficial layers of multiple-layered 
ectocervical epithelial cells are highly differentiated and cornified. Cornified cells 
eventually lose nuclei and die, consequently lacking the actin cytoskeleton response to 
GC infection. However, the inner layers of the ectocervical epithelium, when are 
exposed to GC due to GC-induced exfoliation of the superficial layers, may be able to 
mount actin remodeling response to GC. 
 
The decrease of Arp2/3 in the apical surface of polarized epithelial cells is a potential 
cause of F-actin exclusion underneath GC adherent sites. The exclusion of F-actin 
provides an explanation for the massive microvilli demolishment induced by GC 
Pil+ΔOpa. GC Pil+Opa+ induce the above phenotypes to a significantly less degree, 
suggesting that the Opa expression inhibits GC-induced actin reorganization and 
morphological changes in polarized epithelial cells. In addition, to support microvilli, 
the actin cytoskeleton is a critical support for the apical junction. The perijunctional 
actomyosin ring regulates the stability and permeability of the apical junction [137]. 
Song lab previously showed that GC induce the activation of NMII, which leads to the 
disassembly of the apical junction [95]. Here, I show that inhibition of GC-induced 
NMII activation restores the polarized distribution of Arp2/3 and the actin 
cytoskeleton. My data suggest that GC-induced NMII activation is also involved in 
actin reorganization in response to GC infection. 
 
124 
 
The GC-induced microvilli elongation and F-actin accumulation in non-polarized 
epithelial cells indicate actin polymerization. In contrast, the GC-induced microvilli 
demolishment and F-actin exclusion in polarized epithelial cells indicate actin 
depolymerization. NMII has been shown to play a role in both polymerization and 
depolymerization of the actin filament [156]. The role of Arp2/3 has been implicated 
in the polarized endocervical epithelial cells. Further experiments need to be done on 
the effect of NMII and Arp2/3 inhibitors on the plasma membrane remodeling and 
actin reorganization induced by GC Pil+Opa+ and Pil+ΔOpa in non-polarized and 
polarized epithelial cells. Data presented in Chapter 2 show that Opa expression 
regulates these processes differently in non-polarized and polarized epithelial cells, 
and the role of Opa interaction with its host receptors has been implicated. To test this 
hypothesis, further studies need to done on the plasma membrane remodeling and 
actin reorganization induced by different Opa isoforms in non-polarized and polarized 
epithelial cells.  
 
In summary, the data presented in this chapter demonstrate that GC induced 
differential remodeling of the actin cytoskeleton, which leads to the differential 
membrane remodeling and infectivity in non-polarized and polarized epithelial cells. 
Expression of Opa regulates these processes differently in non-polarized and polarized 
epithelial cells. These data suggest that the polarity of epithelial cells is one of the 
critical factors that regulate GC infectivity.  
 
125 
 
Chapter 4: Conclusions, discussion, and future direction 
 
4.1 Summary  
Gonorrhea, caused by Neisseria gonorrhoeae (GC), is the second most common 
sexually transmitted infection (STI) in the United States. GC is a gram-negative 
bacteria and a human exclusive pathogen. More than 50% GC infection in females is 
asymptomatic, which delays treatments and, leads to severe complications, including 
pelvic inflammatory disease (PID), fallopian tube scarring and blockage with 
subsequent ectopic pregnancy and infertility. The asymptomatic female infection can 
also lead to disseminated gonococcal infection (DGI) and increase the possibility of 
getting coinfection with HIV and other sexually transmitted infections. The continuous 
raising of antibiotic-resistant strains makes gonorrhea public health and women’s 
health crisis. 
 
The cervix is the gate of the female reproductive tract (FRT) and the initiation site of 
GC infection. The epithelium is the first line of defense GC need to overcome to 
establish infection. The cervical epithelium is heterogeneous and has distinct 
properties in different anatomical regions. The ectocervix is line with multiple-layered 
non-polarized epithelial cells; the endocervix, single-layered polarized columnar 
epithelial cells; and the transformation zone (TZ), epithelial cells that gradually 
transformed from multiple-layered to single-layered and from non-polarized to 
126 
 
polarized columnar epithelial cells. However, whether GC infect the heterogeneous 
cervical epithelium in the same way or differently is unknown.  
 
A major barrier to the research of GC infection in women is the lack of experimental 
model mimicking the heterogeneous cervical epithelium. To overcome this barrier, I 
established a human cervical tissue explant model, which maintains the in vivo 
cytoarchitecture of the cervical epithelium. This model enables us to observe that GC 
change their infectivity in different cervical regions based on phase variable virulence 
determinants on the bacterial surface and differential availability of host receptors on 
the cervical epithelial cells. GC colonize the ectocervical and TZ more efficiently than 
the endocervical epithelial cells. In contrast, GC only penetrate into sub-epithelial 
tissue in TZ and the endocervix, with a higher level of penetration in TZ. Pili are 
required for GC colonization in all three cervical regions. Expression of OpaCEA 
enhances colonization in the ecto and endocervix, where CEACAM is abundantly 
expressed at the luminal surface, but only inhibits GC penetration in the endocervix. 
Expression of different Opa proteins does not affect GC infectivity in the TZ, where 
CEACAM is expressed at a low level. The regulatory effect of OpaCEA is dependent on 
the expression of CEACAM1 and the ITIM domain in its cytoplasmic domain, 
potentially also its downstream SHP activation. GC-induced junction disruption and 
epithelial exfoliation in the TZ that expresses low levels of CEACAMs are not 
regulated by OpaCEA expression on GC. The inhibition of GC penetration in the 
endocervix by OpaCEA-CEACAM1 interaction involves SHP activation, which inhibits 
cell-cell junction disruption and subsequent inhibition of epithelial exfoliation. 
127 
 
 
The level of cell polarity is the most prominent difference between the ecto and 
endocervical epithelial cells. The non-polarized and polarized epithelial cell line 
model enables us to examine the impact of epithelial polarity alone on GC infectivity, 
particularly GC invasion that is difficult to study using tissue explants. GC exhibit a 
higher level of invasion into non-polarized than into polarized epithelial cells, but 
similar levels of adhesion to non-polarized cells as to polarized epithelial cells. The 
expression of phase variable Opa enhances both adhesion and invasion in both non-
polarized and polarized epithelial cells, suggesting the cell polarity is the main factor 
influencing GC invasion. GC with phase variable Opa induce microvilli elongation 
exclusively in non-polarized cells, which is mediated by F-actin accumulation to 
facilitate GC adhesion and invasion. In contrast, GC lacking Opa expression induce 
massive disruption of microvilli in the apical surface of polarized epithelial cells, by a 
reduction of the actin cytoskeleton at GC adherent sites, observed in both polarized 
cell line model and endocervical tissue explant model. Such actin cytoskeleton 
exclusion may result from the redistribution of Arp2 and the activation of NMII. 
Furthermore, GC inoculation only induces an increase in intracellular calcium levels 
in polarized, but not in non-polarized epithelial cells. These data suggest that GC 
induces different signaling and actin reorganization in non-polarized and polarized 
epithelial cells, which alter GC infectivity and infection mechanisms. 
 
Taking together, GC modulate its infectivity, colonization, invasion and penetration 
and infection mechanisms, based on both the properties of cervical epithelial cells 
128 
 
(host receptor expression and cell polarity) and the phase variable bacterial surface 
structure. 
 
4.2 Working Model  
 
 
129 
 
Figure 26. Working model for Neisseria gonorrhoeae interaction with the three 
cervical regions, the ectocervix, transformation zone (TZ), and endocervix [159].  
(A) A schematic diagram of GC infectivity in the three cervical regions. The level of 
GC colonization at the luminal surface of the ecto and TZ epithelia is higher than in 
the endocervix. GC can only penetrate into the subepithelial tissue of the TZ and 
endocervix, with a higher level of GC penetration into the TZ epithelium. The OpaCEA 
expressing GC dominate the colonizing GC community at the luminal surface of the 
ecto and endocervical epithelia by interacting with its epithelial receptor CEACAM. In 
the endocervix, OpaHSPG expressing GC penetrate more into the subepithelial tissue by 
disrupting the apical junction. The TZ epithelium expresses CEACAM at a low level, 
thus both the colonizing and penetrating bacterial community is a mixture of GC 
expressing different Opa proteins. 
(B) A schematic diagram of the signaling pathway induced by GC OpaCEA and 
OpaHSPG in the endocervical epithelial cells. GC OpaCEA interact with its receptor 
CEACAM at the apical surface of the endocervical epithelial cells, which potentially 
activates the tyrosine phosphatase SHP, leading to inhibition of the apical junction 
disassembly and subsequent GC penetration. In contrast, GC OpaHSPG infection 
increases intracellular calcium level, leading to NMII activation through MLCK. The 
NMII activation can lead to the apical junction disassembly and consequently GC 
penetration. The NMII activation can also potentially lead to the redistribution of the 
Arp2/3 complex and F-actin from the apical surface of the polarized endocervical 
epithelial cells. 
 
130 
 
My thesis project used the human cervical tissue explant model, non-polarized and 
polarized epithelial cell line model and GC expressing different Opa proteins to study 
the effect of phase variable bacterial surface structures and the heterogeneity of 
cervical epithelium on GC infectivity in the human cervical epithelial cells. Based on 
my results and previously published studies, I propose the following working model. 
 
GC are first deposited onto the vaginal and ectocervical epithelium through sex with 
an infected partner. The vaginal and ectocervical epithelium is colonized by dense 
commensal microbiota creating a unique microenvironment at the mucosal surface 
[11, 160]. To establish infection, successful colonization at the mucosal surface is the 
first step for GC. A higher level of colonization of GC in the vaginal and ectocervical 
epithelium has been reported clinically [15, 20], but GC induced vaginitis is rare [4], 
indicating GC colonization fails to induce immune response in the vagina and 
ectocervix. My results show higher levels of GC colonization in the ectocervix without 
any penetration, consistent with previous clinical observations of no symptomatic 
infection in the vagina and the ectocervix. My finding that pili are essential for GC 
colonization in all anatomic regions of the human cervix indicates that pili-mediated 
interaction of GC with the luminal surface of cervical epithelial cells is the essential 
first step to establish infection in the human cervix. In contrast, Opa expression is not 
essential for GC infection but can regulate GC infectivity. The expression of 
CEACAM-binding Opa (OpaCEA), enhances GC colonization in both non-polarized 
ectocervical epithelial cells and polarized endocervical epithelial cells, through 
interaction with CEACAMs, which is expressed at the luminal surface of the 
131 
 
ectocervical and endocervical but not TZ epithelium. Such Opa-CEACAM interaction 
enhances colonization through the inhibition of GC-induced epithelial shedding. 
OpaCEA inhibit shedding of non-polarized epithelial cells by activation of integrin and 
CD105 showed by Dr. Hauck’s group using cell line and mouse model [61, 62]. In the 
absence of OpaCEA, GC induce higher levels of epithelial shedding, which shed off GC 
colonized epithelial cells resulting in the low level of colonization. Consequently, the 
community of GC colonizing the ectocervical epithelium is dominated by the OpaCEA 
expressing bacteria, given that the OpaHSPG expressing bacteria are shed and the 
majority of 11 Opa proteins interact with CEACAMs. Despite the exfoliation, GC do 
not penetrate into the ectocervical epithelium, may be due to the incapability of GC to 
disrupt the of E-cadherin-based adherens junction in the remaining epithelial layers 
after infection. The failure of GC to penetrate across the ectocervical epithelium 
provides a potential explanation for the localized and asymptomatic vaginal and 
ectocervical infection. GC expressing phase variable Opa adhere and invade more 
efficiently into non-polarized epithelial cells compared to GC expressing no Opa and 
GC invasion into polarized epithelial cells. Concurrent with the enhanced invasion of 
Opa-expressing GC, the F-actin accumulation and microvilli elongation at GC 
adherent sites was observed exclusively in non-polarized epithelial cells. These data 
obtained from the non-polarized epithelial cell line model suggest that GC, 
particularly those expressing phase variable Opa, can invade into non-polarized 
epithelial cells, such as ectocervical epithelial cells and exfoliated cervical epithelial 
cells, by inducing actin cytoskeleton mediated bacterial uptake. 
 
132 
 
The transformation zone (TZ) is a unique area between the ecto and endocervix. The 
low level of CEACAM expression in TZ epithelial cells provides an explanation for 
the similar level of colonization, penetration, and exfoliation induced by GC 
expressing different isoforms of Opa or no Opa. The level of GC colonization in the 
TZ is similar to that in the ectocervix that expresses CEACAMs, suggesting that Opa 
proteins do not play a significant role in GC colonization in TZ. The high level of 
penetration across TZ epithelium is mediated by the disruption of E-cadherin based 
cell-cell adhesion, which also leads to epithelial exfoliation. The multiple-layered 
structure of TZ epithelium indicates that the number of layers is not a major factor that 
deters GC from penetration. The similar infectivity displayed by GC expressing 
OpaCEA and OpaHSPG indicates that the bacteria community infecting the TZ is a 
mixture of GC expressing different Opa proteins. The penetration of GC into 
subepithelial tissue will increase the chance of GC being exposed to immune cells, 
which potentially cause inflammation and symptomatic infection. 
 
The TZ epithelium gradually transforms into the endocervix, which leads to the upper 
genital tract. The level of GC colonization and penetration in the endocervix are 
generally lower than those in the ectocervix (colonization only) and TZ (colonization 
and penetration). However, OpaCEA expression increases the level of colonization in 
the endocervical epithelial cells, which is mediated by CEACAMs expressed on the 
apical surface of the endocervical epithelial cells. Opa-CEACAM interaction inhibits 
the disruption of E-cadherin-based adherens junction, consequently decreasing 
epithelial exfoliation and penetration, and increasing colonization. Consistently, Song 
133 
 
lab has previously shown that OpaCEA expression interferes with GC transmigration 
across polarized T84 epithelial cells, that express CEACAMs [94, 95]. In contrast, in 
the absence of Opa-CEACAM interaction (OpaHSPG GC and TZ epithelial cells), GC 
disrupt the adherens junction, leading to epithelial shedding and GC penetration, and 
colonization reduction. Previously published data from Song lab demonstrate that GC 
disrupt the apical junction through increasing intracellular calcium level and calcium-
dependent activation of NMII [95], providing a mechanistic explanation for the 
distinct behavior of GC in polarized endocervical epithelial cells. Data from both 
epithelial cell line model and cervical tissue explant mode suggest the invasion into 
epithelial cells is not the major mechanism used by GC to infect polarized 
endocervical epithelial cells. GC invade into polarized epithelial cells even less than 
into non-polarized epithelial cells. The invasion of GC into polarized epithelial cells 
involves in the demolishment of microvilli at the apical surface probably by activation 
of NMII and redistribution of the actin nucleator Arp2/3 away from the apical surface 
of polarized epithelial cells. Taking together, in the endocervix, colonizing GC at the 
mucosal surface predominantly express OpaCEA, which can potentially ascend to the 
upper genital tract through the continuity of the epithelium, causing pelvic 
inflammatory disease. Penetrating GC predominantly express OpaHSPG. They penetrate 
through the paracellular space between neighboring epithelial cells by disassembling 
cell-cell junction complex creating breaches in this physical protective barrier. GC 
penetrating into subepithelial tissues potentially induce inflammation and causing 
disseminated gonococcal infection (DGI).  
 
134 
 
In summary, my Ph.D. research focused on the mechanism by which GC cope with the 
heterogeneous mucosal surface of the human cervix to initiate infection. The 
differential infectivity of GC, preferring colonization or penetration, in different 
cervical regions is the first direct evidence for that GC use different mechanisms to 
infect different types of cervical epithelial cells. The infection mechanism changes 
depending on both the bacterial phase variation of bacterial surface structures and the 
distinct properties of cervical epithelial cells.  
 
 
 
 
 
 
 
 
 
 
 
135 
 
4.3 Future Direction  
4.3.1 Highlights  
We are the first to study the mechanism of GC infection in the cervical epithelium 
using human cervical tissue explants. Using this model, I showed that GC switch the 
mode of infectivity, colonization or penetration by Opa phase variation in response to 
the heterogeneous cervical epithelium they encounter. The results from my thesis 
research indicate that OpaCEA expression specifically promotes colonization in the ecto 
and endocervix, which expresses CEACAMs. GC colonization my mainly lead to 
localized, asymptomatic infection. In contrast, the expression of OpaHSPG and low 
level of CEACAM expression all GC penetration into the endocervix and TZ 
respectively, which potentially leads to symptomatic infection and disseminated 
gonococcal infection (DGI).  
 
My thesis research enables us to reconcile conflicting results of previous research 
based on cell line and the mouse model, which do not fully represent the 
characteristics of the human cervix in vivo. No previous model mimics the 
heterogeneity of cervical epithelium, and thus, this important factor was largely 
disregarded previously. The results from my thesis research emphasize the impact of 
the heterogeneous cervical epithelium on the outcomes of GC infection in women and 
raise the importance for the field to take this factor into consideration in GC 
pathogenesis research. 
 
136 
 
4.3.2 Question 1. Does the mucosal microenvironment of the human cervix affect 
GC infectivity and how? 
In my thesis project, I focused on the impact of different types of cervical epithelial 
cells on GC infectivity. However, the environment of the mucosal surface along the 
three cervical regions is another layer of complication that GC have to deal with 
during infection. First, microbiota in the ectocervix, where the Lactobacillus species is 
the major commensal species, creating an acidic environment [11]. Second, the 
endocervix secrets a great amount of mucus, forming mucus plug to block the gate of 
the female reproductive tract [18]. Third, the properties of the cervical epithelial cells, 
including their morphology and mucus secretion, are under the influence of the 
periodic change of the sex hormones progestogen and estrogen [161]. All the above 
environmental factors along the mucosal surface of the cervix can potentially affect 
GC infectivity, but very limited research has been done on these subjects. The human 
cervical tissue explant model provides us a great opportunity to examine the impact of 
these mucosal environment factors on GC infectivity.  
 
 
 
 
 
 
137 
 
4.3.3 Question 2. Does GC infection induce differential immune responses in 
different cervical regions? 
The asymptomatic and symptomatic infection caused by GC is directly linked to the 
ability of GC to evade or activate immune responses. Neutrophils have been detected 
in the cervical papsmear from infected patients. The alternative complement pathway 
has been shown involved in the female gonococcal infection [3]. These data suggest 
that GC infection induces immune responses to a certain level. However, does GC 
infection induce differential immune responses in different cervical regions is 
unknown. How GC manipulate host immune response to suppress symptomatic 
infection in the human cervix is not clear. To test the hypothesis that GC colonization 
in the ectocervix leads to asymptomatic infection and GC penetration into the 
subepithelium of the TZ and endocervix leads to symptomatic infection, the immune 
response induced by GC in different cervical regions should be examined.  
 
 
 
 
 
 
 
 
138 
 
4.3.4 Question 3. How to reconstruct the heterogeneous cervical in vitro? 
Although the human cervix tissue explant model maintains the in vivo cytoarchitecture 
of cervical epithelium, it has some drawbacks. Although the overall structure of the 
cervical epithelium is consistent from one human subject to another, the ethnicity, age, 
menstrual cycle, and overall health condition are highly variable. Human cervical 
tissue explants cannot be manipulated in vitro as easily as cell lines and other models. 
Techniques, such as RNAi and transfection, are not readily available to be performed 
in tissue explants. Thus, we sought to mimic the cervical epithelium using cell lines as 
what I showed in Chapter 3 of my thesis. However, the non-polarized cells do not 
form a multiple-layered structure on transwells, and no method is currently available 
to mimics the unique structure of the TZ epithelium. It is important to develop a 
system to mimic all the three types of cervical epithelial cells, which can be 
manipulated easily in vitro, which will allow us to dissect the molecular and cellular 
mechanisms by which GC infect the human cervix in a more detailed way. 
 
 
 
 
 
 
 
139 
 
Bibliography 
1.Report on global sexually transmitted infection surveillance 2015. 
2.Workowski KA, Bolan GA. Sexually transmitted diseases treatment guidelines, 
2015. MMWR Recomm Rep. 2015;64(RR-03):1-137. PubMed PMID: 26042815. 
3.Edwards JL, Butler EK. The Pathobiology of Neisseria gonorrhoeae Lower Female 
Genital Tract Infection. Front Microbiol. 2011;2:102. doi: 10.3389/fmicb.2011.00102. 
PubMed PMID: 21747805; PubMed Central PMCID: PMCPMC3129011. 
4.Edwards JL, Apicella MA. The molecular mechanisms used by Neisseria 
gonorrhoeae to initiate infection differ between men and women. Clin Microbiol Rev. 
2004;17(4):965-81, table of contents. doi: 10.1128/CMR.17.4.965-981.2004. PubMed 
PMID: 15489357; PubMed Central PMCID: PMCPMC523569. 
5.Rice PA, Shafer WM, Ram S, Jerse AE. Neisseria gonorrhoeae: Drug Resistance, 
Mouse Models, and Vaccine Development. Annu Rev Microbiol. 2017;71:665-86. 
doi: 10.1146/annurev-micro-090816-093530. PubMed PMID: 28886683. 
6.(CDC) CfDCaP. Update to CDC's Sexually transmitted diseases treatment 
guidelines, 2010: oral cephalosporins no longer a recommended treatment for 
gonococcal infections. MMWR Morb Mortal Wkly Rep. 2012;61(31):590-4. PubMed 
PMID: 22874837. 
7.Fairley CK, Hocking JS, Zhang L, Chow EP. Frequent Transmission of Gonorrhea 
in Men Who Have Sex with Men. Emerg Infect Dis. 2017;23(1):102-4. doi: 
10.3201/eid2301.161205. PubMed PMID: 27983487; PubMed Central PMCID: 
PMCPMC5176237. 
8.Fleming DT, Wasserheit JN. From epidemiological synergy to public health policy 
and practice: the contribution of other sexually transmitted diseases to sexual 
transmission of HIV infection. Sex Transm Infect. 1999;75(1):3-17. PubMed PMID: 
10448335; PubMed Central PMCID: PMCPMC1758168. 
9.Draper DL, James JF, Brooks GF, Sweet RL. Comparison of virulence markers of 
peritoneal and fallopian tube isolates with endocervical Neisseria gonorrhoeae isolates 
from women with acute salpingitis. Infect Immun. 1980;27(3):882-8. PubMed PMID: 
6769811; PubMed Central PMCID: PMCPMC550857. 
10.Bhattacharyya MN, Jephcott AE, Morton RS. Diagnosis of gonorrhoea in women: 
comparison of sampling sites. Br Med J. 1973;2(5869):748-50. PubMed PMID: 
4736958; PubMed Central PMCID: PMCPMC1589763. 
11.Ling Z, Liu X, Chen X, Zhu H, Nelson KE, Xia Y, et al. Diversity of 
cervicovaginal microbiota associated with female lower genital tract infections. 
140 
 
Microb Ecol. 2011;61(3):704-14. Epub 2011/02/02. doi: 10.1007/s00248-011-9813-z. 
PubMed PMID: 21287345. 
12.Ghosh M, Shen Z, Fahey JV, Crist SG, Patel M, Smith JM, et al. Pathogen 
recognition in the human female reproductive tract: expression of intracellular 
cytosolic sensors NOD1, NOD2, RIG-1, and MDA5 and response to HIV-1 and 
Neisseria gonorrhea. Am J Reprod Immunol. 2013;69(1):41-51. Epub 2012/09/17. 
doi: 10.1111/aji.12019. PubMed PMID: 22984986; PubMed Central PMCID: 
PMCPMC3518676. 
13.Chen C, Song X, Wei W, Zhong H, Dai J, Lan Z, et al. The microbiota continuum 
along the female reproductive tract and its relation to uterine-related diseases. Nat 
Commun. 2017;8(1):875. Epub 2017/10/17. doi: 10.1038/s41467-017-00901-0. 
PubMed PMID: 29042534; PubMed Central PMCID: PMCPMC5645390. 
14.Mitchell CM, Haick A, Nkwopara E, Garcia R, Rendi M, Agnew K, et al. 
Colonization of the upper genital tract by vaginal bacterial species in nonpregnant 
women. Am J Obstet Gynecol. 2015;212(5):611.e1-9. Epub 2014/12/16. doi: 
10.1016/j.ajog.2014.11.043. PubMed PMID: 25524398; PubMed Central PMCID: 
PMCPMC4754962. 
15.HARKNESS AH. The pathology of gonorrhoea. Br J Vener Dis. 1948;24(4):137-
47. PubMed PMID: 18099876; PubMed Central PMCID: PMCPMC1053606. 
16.Rice PA. Gonococcal arthritis (disseminated gonococcal infection). Infect Dis Clin 
North Am. 2005;19(4):853-61. doi: 10.1016/j.idc.2005.07.003. PubMed PMID: 
16297736. 
17.Wu Z, Xu L, Tu Y, Chen R, Yu Y, Li J, et al. The relationship between the 
symptoms of female gonococcal infections and serum progesterone level and the 
genotypes of Neisseria gonorrhoeae multi-antigen sequence type (NG-MAST) in 
Wuhan, China. Eur J Clin Microbiol Infect Dis. 2011;30(1):113-6. Epub 2010/09/04. 
doi: 10.1007/s10096-010-1040-x. PubMed PMID: 20820835; PubMed Central 
PMCID: PMCPMC2998644. 
18.Gipson IK. Mucins of the human endocervix. Front Biosci. 2001;6:D1245-55. 
Epub 2001/10/01. PubMed PMID: 11578960. 
19.Hladik F, McElrath MJ. Setting the stage: host invasion by HIV. Nat Rev Immunol. 
2008;8(6):447-57. doi: 10.1038/nri2302. PubMed PMID: 18469831; PubMed Central 
PMCID: PMCPMC2587276. 
20.Evans BA. Ultrastructural study of cervical gonorrhea. J Infect Dis. 
1977;136(2):248-55. PubMed PMID: 408425. 
21.Plant L, Jonsson AB. Contacting the host: insights and implications of pathogenic 
Neisseria cell interactions. Scand J Infect Dis. 2003;35(9):608-13. PubMed PMID: 
14620143. 
141 
 
22.Haas R, Meyer TF. The repertoire of silent pilus genes in Neisseria gonorrhoeae: 
evidence for gene conversion. Cell. 1986;44(1):107-15. PubMed PMID: 2866848. 
23.Jonsson AB, Nyberg G, Normark S. Phase variation of gonococcal pili by 
frameshift mutation in pilC, a novel gene for pilus assembly. EMBO J. 
1991;10(2):477-88. PubMed PMID: 1671354; PubMed Central PMCID: 
PMCPMC452669. 
24.Rudel T, Scheurerpflug I, Meyer TF. Neisseria PilC protein identified as type-4 
pilus tip-located adhesin. Nature. 1995;373(6512):357-9. doi: 10.1038/373357a0. 
PubMed PMID: 7830772. 
25.Rudel T, Boxberger HJ, Meyer TF. Pilus biogenesis and epithelial cell adherence 
of Neisseria gonorrhoeae pilC double knock-out mutants. Mol Microbiol. 
1995;17(6):1057-71. PubMed PMID: 8594326. 
26.Rudel T, van Putten JP, Gibbs CP, Haas R, Meyer TF. Interaction of two variable 
proteins (PilE and PilC) required for pilus-mediated adherence of Neisseria 
gonorrhoeae to human epithelial cells. Mol Microbiol. 1992;6(22):3439-50. PubMed 
PMID: 1362447. 
27.Wolfgang M, Park HS, Hayes SF, van Putten JP, Koomey M. Suppression of an 
absolute defect in type IV pilus biogenesis by loss-of-function mutations in pilT, a 
twitching motility gene in Neisseria gonorrhoeae. Proc Natl Acad Sci U S A. 
1998;95(25):14973-8. PubMed PMID: 9844000; PubMed Central PMCID: 
PMCPMC24560. 
28.Dietrich M, Mollenkopf H, So M, Friedrich A. Pilin regulation in the pilT mutant 
of Neisseria gonorrhoeae strain MS11. FEMS Microbiol Lett. 2009;296(2):248-56. 
Epub 2009/05/08. doi: 10.1111/j.1574-6968.2009.01647.x. PubMed PMID: 
19486161; PubMed Central PMCID: PMCPMC4428587. 
29.Edwards JL, Brown EJ, Ault KA, Apicella MA. The role of complement receptor 3 
(CR3) in Neisseria gonorrhoeae infection of human cervical epithelia. Cell Microbiol. 
2001;3(9):611-22. PubMed PMID: 11553013. 
30.Stern A, Brown M, Nickel P, Meyer TF. Opacity genes in Neisseria gonorrhoeae: 
control of phase and antigenic variation. Cell. 1986;47(1):61-71. PubMed PMID: 
3093085. 
31.Bhat KS, Gibbs CP, Barrera O, Morrison SG, Jähnig F, Stern A, et al. The opacity 
proteins of Neisseria gonorrhoeae strain MS11 are encoded by a family of 11 
complete genes. Mol Microbiol. 1991;5(8):1889-901. PubMed PMID: 1815562. 
32.Billker O, Popp A, Gray-Owen SD, Meyer TF. The structural basis of CEACAM-
receptor targeting by neisserial Opa proteins. Trends Microbiol. 2000;8(6):258-60; 
discussion 60-1. PubMed PMID: 10838580. 
142 
 
33.van Putten JP, Paul SM. Binding of syndecan-like cell surface proteoglycan 
receptors is required for Neisseria gonorrhoeae entry into human mucosal cells. 
EMBO J. 1995;14(10):2144-54. PubMed PMID: 7774572; PubMed Central PMCID: 
PMCPMC398320. 
34.Cole JG, Fulcher NB, Jerse AE. Opacity proteins increase Neisseria gonorrhoeae 
fitness in the female genital tract due to a factor under ovarian control. Infect Immun. 
2010;78(4):1629-41. doi: 10.1128/IAI.00996-09. PubMed PMID: 20100859; PubMed 
Central PMCID: PMCPMC2849431. 
35.Yamasaki R, Bacon BE, Nasholds W, Schneider H, Griffiss JM. Structural 
determination of oligosaccharides derived from lipooligosaccharide of Neisseria 
gonorrhoeae F62 by chemical, enzymatic, and two-dimensional NMR methods. 
Biochemistry. 1991;30(43):10566-75. PubMed PMID: 1931980. 
36.Danaher RJ, Levin JC, Arking D, Burch CL, Sandlin R, Stein DC. Genetic basis of 
Neisseria gonorrhoeae lipooligosaccharide antigenic variation. J Bacteriol. 
1995;177(24):7275-9. PubMed PMID: 8522539; PubMed Central PMCID: 
PMCPMC177611. 
37.Banerjee A, Wang R, Uljon SN, Rice PA, Gotschlich EC, Stein DC. Identification 
of the gene (lgtG) encoding the lipooligosaccharide beta chain synthesizing glucosyl 
transferase from Neisseria gonorrhoeae. Proc Natl Acad Sci U S A. 
1998;95(18):10872-7. PubMed PMID: 9724797; PubMed Central PMCID: 
PMCPMC27988. 
38.Harvey HA, Jennings MP, Campbell CA, Williams R, Apicella MA. Receptor-
mediated endocytosis of Neisseria gonorrhoeae into primary human urethral epithelial 
cells: the role of the asialoglycoprotein receptor. Mol Microbiol. 2001;42(3):659-72. 
PubMed PMID: 11722733. 
39.Song W, Ma L, Chen R, Stein DC. Role of lipooligosaccharide in Opa-independent 
invasion of Neisseria gonorrhoeae into human epithelial cells. J Exp Med. 
2000;191(6):949-60. PubMed PMID: 10727457; PubMed Central PMCID: 
PMCPMC2193109. 
40.Weel JF, Hopman CT, van Putten JP. Bacterial entry and intracellular processing of 
Neisseria gonorrhoeae in epithelial cells: immunomorphological evidence for 
alterations in the major outer membrane protein P.IB. J Exp Med. 1991;174(3):705-15. 
PubMed PMID: 1908511; PubMed Central PMCID: PMCPMC2118933. 
41.Deo P, Chow SH, Hay ID, Kleifeld O, Costin A, Elgass KD, et al. Outer membrane 
vesicles from Neisseria gonorrhoeae target PorB to mitochondria and induce 
apoptosis. PLoS Pathog. 2018;14(3):e1006945. Epub 2018/03/30. doi: 
10.1371/journal.ppat.1006945. PubMed PMID: 29601598; PubMed Central PMCID: 
PMCPMC5877877. 
143 
 
42.Mosleh IM, Huber LA, Steinlein P, Pasquali C, Günther D, Meyer TF. Neisseria 
gonorrhoeae porin modulates phagosome maturation. J Biol Chem. 
1998;273(52):35332-8. PubMed PMID: 9857075. 
43.Müller A, Günther D, Brinkmann V, Hurwitz R, Meyer TF, Rudel T. Targeting of 
the pro-apoptotic VDAC-like porin (PorB) of Neisseria gonorrhoeae to mitochondria 
of infected cells. EMBO J. 2000;19(20):5332-43. doi: 10.1093/emboj/19.20.5332. 
PubMed PMID: 11032801; PubMed Central PMCID: PMCPMC314008. 
44.Kozjak-Pavlovic V, Dian-Lothrop EA, Meinecke M, Kepp O, Ross K, Rajalingam 
K, et al. Bacterial porin disrupts mitochondrial membrane potential and sensitizes host 
cells to apoptosis. PLoS Pathog. 2009;5(10):e1000629. Epub 2009/10/23. doi: 
10.1371/journal.ppat.1000629. PubMed PMID: 19851451; PubMed Central PMCID: 
PMCPMC2759283. 
45.Binnicker MJ, Williams RD, Apicella MA. Gonococcal porin IB activates NF-
kappaB in human urethral epithelium and increases the expression of host 
antiapoptotic factors. Infect Immun. 2004;72(11):6408-17. doi: 
10.1128/IAI.72.11.6408-6417.2004. PubMed PMID: 15501771; PubMed Central 
PMCID: PMCPMC523018. 
46.Dupuy AG, Caron E. Integrin-dependent phagocytosis: spreading from 
microadhesion to new concepts. J Cell Sci. 2008;121(11):1773-83. doi: 
10.1242/jcs.018036. PubMed PMID: 18492791. 
47.Edwards JL, Apicella MA. The role of lipooligosaccharide in Neisseria 
gonorrhoeae pathogenesis of cervical epithelia: lipid A serves as a C3 acceptor 
molecule. Cell Microbiol. 2002;4(9):585-98. PubMed PMID: 12390351. 
48.Edwards JL, Brown EJ, Uk-Nham S, Cannon JG, Blake MS, Apicella MA. A co-
operative interaction between Neisseria gonorrhoeae and complement receptor 3 
mediates infection of primary cervical epithelial cells. Cell Microbiol. 2002;4(9):571-
84. PubMed PMID: 12390350. 
49.Edwards JL, Entz DD, Apicella MA. Gonococcal phospholipase d modulates the 
expression and function of complement receptor 3 in primary cervical epithelial cells. 
Infect Immun. 2003;71(11):6381-91. PubMed PMID: 14573659; PubMed Central 
PMCID: PMCPMC219594. 
50.Bos MP, Kuroki M, Krop-Watorek A, Hogan D, Belland RJ. CD66 receptor 
specificity exhibited by neisserial Opa variants is controlled by protein determinants in 
CD66 N-domains. Proc Natl Acad Sci U S A. 1998;95(16):9584-9. PubMed PMID: 
9689124; PubMed Central PMCID: PMCPMC21382. 
51.Beauchemin N, Arabzadeh A. Carcinoembryonic antigen-related cell adhesion 
molecules (CEACAMs) in cancer progression and metastasis. Cancer Metastasis Rev. 
2013;32(3-4):643-71. doi: 10.1007/s10555-013-9444-6. PubMed PMID: 23903773. 
144 
 
52.Gray-Owen SD, Blumberg RS. CEACAM1: contact-dependent control of 
immunity. Nat Rev Immunol. 2006;6(6):433-46. doi: 10.1038/nri1864. PubMed 
PMID: 16724098. 
53.Sadarangani M, Pollard AJ, Gray-Owen SD. Opa proteins and CEACAMs: 
pathways of immune engagement for pathogenic Neisseria. FEMS Microbiol Rev. 
2011;35(3):498-514. doi: 10.1111/j.1574-6976.2010.00260.x. PubMed PMID: 
21204865. 
54.Lee HS, Ostrowski MA, Gray-Owen SD. CEACAM1 dynamics during neisseria 
gonorrhoeae suppression of CD4+ T lymphocyte activation. J Immunol. 
2008;180(10):6827-35. PubMed PMID: 18453603. 
55.Hauck CR, Grassmé H, Bock J, Jendrossek V, Ferlinz K, Meyer TF, et al. Acid 
sphingomyelinase is involved in CEACAM receptor-mediated phagocytosis of 
Neisseria gonorrhoeae. FEBS Lett. 2000;478(3):260-6. PubMed PMID: 10930579. 
56.McCaw SE, Schneider J, Liao EH, Zimmermann W, Gray-Owen SD. 
Immunoreceptor tyrosine-based activation motif phosphorylation during engulfment 
of Neisseria gonorrhoeae by the neutrophil-restricted CEACAM3 (CD66d) receptor. 
Mol Microbiol. 2003;49(3):623-37. PubMed PMID: 12864848. 
57.Booth JW, Telio D, Liao EH, McCaw SE, Matsuo T, Grinstein S, et al. 
Phosphatidylinositol 3-kinases in carcinoembryonic antigen-related cellular adhesion 
molecule-mediated internalization of Neisseria gonorrhoeae. J Biol Chem. 
2003;278(16):14037-45. Epub 2003/02/05. doi: 10.1074/jbc.M211879200. PubMed 
PMID: 12571236. 
58.Sarantis H, Gray-Owen SD. The specific innate immune receptor CEACAM3 
triggers neutrophil bactericidal activities via a Syk kinase-dependent pathway. Cell 
Microbiol. 2007;9(9):2167-80. Epub 2007/05/15. doi: 10.1111/j.1462-
5822.2007.00947.x. PubMed PMID: 17506820. 
59.Billker O, Popp A, Brinkmann V, Wenig G, Schneider J, Caron E, et al. Distinct 
mechanisms of internalization of Neisseria gonorrhoeae by members of the 
CEACAM receptor family involving Rac1- and Cdc42-dependent and -independent 
pathways. EMBO J. 2002;21(4):560-71. PubMed PMID: 11847104; PubMed Central 
PMCID: PMCPMC125849. 
60.Voges M, Bachmann V, Naujoks J, Kopp K, Hauck CR. Extracellular IgC2 
constant domains of CEACAMs mediate PI3K sensitivity during uptake of pathogens. 
PLoS One. 2012;7(6):e39908. doi: 10.1371/journal.pone.0039908. PubMed PMID: 
22768164; PubMed Central PMCID: PMCPMC3386982. 
61.Muenzner P, Bachmann V, Zimmermann W, Hentschel J, Hauck CR. Human-
restricted bacterial pathogens block shedding of epithelial cells by stimulating integrin 
activation. Science. 2010;329(5996):1197-201. doi: 10.1126/science.1190892. 
PubMed PMID: 20813953. 
145 
 
62.Muenzner P, Rohde M, Kneitz S, Hauck CR. CEACAM engagement by human 
pathogens enhances cell adhesion and counteracts bacteria-induced detachment of 
epithelial cells. J Cell Biol. 2005;170(5):825-36. doi: 10.1083/jcb.200412151. 
PubMed PMID: 16115956; PubMed Central PMCID: PMCPMC2171332. 
63.Islam EA, Anipindi VC, Francis I, Shaik-Dasthagirisaheb Y, Xu S, Leung N, et al. 
Specific Binding to Differentially Expressed Human Carcinoembryonic Antigen-
Related Cell Adhesion Molecules Determines the Outcome of Neisseria gonorrhoeae 
Infections along the Female Reproductive Tract. Infect Immun. 2018;86(8). Epub 
2018/07/23. doi: 10.1128/IAI.00092-18. PubMed PMID: 29760215; PubMed Central 
PMCID: PMCPMC6056862. 
64.Dehio C, Gray-Owen SD, Meyer TF. The role of neisserial Opa proteins in 
interactions with host cells. Trends Microbiol. 1998;6(12):489-95. PubMed PMID: 
10036728. 
65.Hauck CR, Meyer TF. 'Small' talk: Opa proteins as mediators of Neisseria-host-cell 
communication. Curr Opin Microbiol. 2003;6(1):43-9. PubMed PMID: 12615218. 
66.van Putten JP, Duensing TD, Cole RL. Entry of OpaA+ gonococci into HEp-2 cells 
requires concerted action of glycosaminoglycans, fibronectin and integrin receptors. 
Mol Microbiol. 1998;29(1):369-79. PubMed PMID: 9701828. 
67.Duensing TD, van Putten JP. Vitronectin mediates internalization of Neisseria 
gonorrhoeae by Chinese hamster ovary cells. Infect Immun. 1997;65(3):964-70. 
PubMed PMID: 9038304; PubMed Central PMCID: PMCPMC175076. 
68.Gómez-Duarte OG, Dehio M, Guzmán CA, Chhatwal GS, Dehio C, Meyer TF. 
Binding of vitronectin to opa-expressing Neisseria gonorrhoeae mediates invasion of 
HeLa cells. Infect Immun. 1997;65(9):3857-66. PubMed PMID: 9284164; PubMed 
Central PMCID: PMCPMC175551. 
69.Grassmé H, Gulbins E, Brenner B, Ferlinz K, Sandhoff K, Harzer K, et al. Acidic 
sphingomyelinase mediates entry of N. gonorrhoeae into nonphagocytic cells. Cell. 
1997;91(5):605-15. PubMed PMID: 9393854. 
70.Freissler E, Meyer auf der Heyde A, David G, Meyer TF, Dehio C. Syndecan-1 and 
syndecan-4 can mediate the invasion of OpaHSPG-expressing Neisseria gonorrhoeae 
into epithelial cells. Cell Microbiol. 2000;2(1):69-82. PubMed PMID: 11207564. 
71.Naumann M, Rudel T, Meyer TF. Host cell interactions and signalling with 
Neisseria gonorrhoeae. Curr Opin Microbiol. 1999;2(1):62-70. PubMed PMID: 
10047561. 
72.Stockert RJ. The asialoglycoprotein receptor: relationships between structure, 
function, and expression. Physiol Rev. 1995;75(3):591-609. doi: 
10.1152/physrev.1995.75.3.591. PubMed PMID: 7624395. 
146 
 
73.Porat N, Apicella MA, Blake MS. Neisseria gonorrhoeae utilizes and enhances the 
biosynthesis of the asialoglycoprotein receptor expressed on the surface of the hepatic 
HepG2 cell line. Infect Immun. 1995;63(4):1498-506. PubMed PMID: 7890416; 
PubMed Central PMCID: PMCPMC173181. 
74.Harvey HA, Ketterer MR, Preston A, Lubaroff D, Williams R, Apicella MA. 
Ultrastructural analysis of primary human urethral epithelial cell cultures infected with 
Neisseria gonorrhoeae. Infect Immun. 1997;65(6):2420-7. PubMed PMID: 9169783; 
PubMed Central PMCID: PMCPMC175335. 
75.Swanson KV, Griffiss JM, Edwards VL, Stein DC, Song W. Neisseria 
gonorrhoeae-induced transactivation of EGFR enhances gonococcal invasion. Cell 
Microbiol. 2011;13(7):1078-90. doi: 10.1111/j.1462-5822.2011.01603.x. PubMed 
PMID: 21501367. 
76.Edwards VL, Wang LC, Dawson V, Stein DC, Song W. Neisseria gonorrhoeae 
breaches the apical junction of polarized epithelial cells for transmigration by 
activating EGFR. Cell Microbiol. 2013;15(6):1042-57. doi: 10.1111/cmi.12099. 
PubMed PMID: 23279089. 
77.Howie HL, Shiflett SL, So M. Extracellular signal-regulated kinase activation by 
Neisseria gonorrhoeae downregulates epithelial cell proapoptotic proteins Bad and 
Bim. Infect Immun. 2008;76(6):2715-21. Epub 2008/04/07. doi: 10.1128/IAI.00153-
08. PubMed PMID: 18391004; PubMed Central PMCID: PMCPMC2423055. 
78.Howie HL, Glogauer M, So M. The N. gonorrhoeae type IV pilus stimulates 
mechanosensitive pathways and cytoprotection through a pilT-dependent mechanism. 
PLoS Biol. 2005;3(4):e100. Epub 2005/03/22. doi: 10.1371/journal.pbio.0030100. 
PubMed PMID: 15769184; PubMed Central PMCID: PMCPMC1065265. 
79.Citri A, Yarden Y. EGF-ERBB signalling: towards the systems level. Nat Rev Mol 
Cell Biol. 2006;7(7):505-16. doi: 10.1038/nrm1962. PubMed PMID: 16829981. 
80.Merz AJ, Enns CA, So M. Type IV pili of pathogenic Neisseriae elicit cortical 
plaque formation in epithelial cells. Mol Microbiol. 1999;32(6):1316-32. PubMed 
PMID: 10383771. 
81.Lee SW, Higashi DL, Snyder A, Merz AJ, Potter L, So M. PilT is required for 
PI(3,4,5)P3-mediated crosstalk between Neisseria gonorrhoeae and epithelial cells. 
Cell Microbiol. 2005;7(9):1271-84. doi: 10.1111/j.1462-5822.2005.00551.x. PubMed 
PMID: 16098215. 
82.Higashi DL, Zhang GH, Biais N, Myers LR, Weyand NJ, Elliott DA, et al. 
Influence of type IV pilus retraction on the architecture of the Neisseria gonorrhoeae-
infected cell cortex. Microbiology. 2009;155(Pt 12):4084-92. Epub 2009/09/17. doi: 
10.1099/mic.0.032656-0. PubMed PMID: 19762436; PubMed Central PMCID: 
PMCPMC2889423. 
147 
 
83.Ison CA. Sexually Transmitted Infections and Sexually Transmitted Diseases. 
Springer-Verlag Berlin Heidelberg2011. 
84.Fichorova RN, Desai PJ, Gibson FC, Genco CA. Distinct proinflammatory host 
responses to Neisseria gonorrhoeae infection in immortalized human cervical and 
vaginal epithelial cells. Infect Immun. 2001;69(9):5840-8. PubMed PMID: 11500462; 
PubMed Central PMCID: PMCPMC98702. 
85.McGee ZA, Johnson AP, Taylor-Robinson D. Pathogenic mechanisms of Neisseria 
gonorrhoeae: observations on damage to human fallopian tubes in organ culture by 
gonococci of colony type 1 or type 4. J Infect Dis. 1981;143(3):413-22. PubMed 
PMID: 6785363. 
86.McCaw SE, Liao EH, Gray-Owen SD. Engulfment of Neisseria gonorrhoeae: 
revealing distinct processes of bacterial entry by individual carcinoembryonic antigen-
related cellular adhesion molecule family receptors. Infect Immun. 2004;72(5):2742-
52. PubMed PMID: 15102784; PubMed Central PMCID: PMCPMC387857. 
87.Dehio M, Gómez-Duarte OG, Dehio C, Meyer TF. Vitronectin-dependent invasion 
of epithelial cells by Neisseria gonorrhoeae involves alpha(v) integrin receptors. 
FEBS Lett. 1998;424(1-2):84-8. PubMed PMID: 9537520. 
88.Lin L, Ayala P, Larson J, Mulks M, Fukuda M, Carlsson SR, et al. The Neisseria 
type 2 IgA1 protease cleaves LAMP1 and promotes survival of bacteria within 
epithelial cells. Mol Microbiol. 1997;24(5):1083-94. PubMed PMID: 9220014. 
89.Edwards JL. Neisseria gonorrhoeae survival during primary human cervical 
epithelial cell infection requires nitric oxide and is augmented by progesterone. Infect 
Immun. 2010;78(3):1202-13. Epub 2010/01/04. doi: 10.1128/IAI.01085-09. PubMed 
PMID: 20048043; PubMed Central PMCID: PMCPMC2825954. 
90.McGee ZA, Stephens DS, Hoffman LH, Schlech WF, Horn RG. Mechanisms of 
mucosal invasion by pathogenic Neisseria. Rev Infect Dis. 1983;5 Suppl 4:S708-14. 
PubMed PMID: 6415784. 
91.Ilver D, Källström H, Normark S, Jonsson AB. Transcellular passage of Neisseria 
gonorrhoeae involves pilus phase variation. Infect Immun. 1998;66(2):469-73. 
PubMed PMID: 9453597; PubMed Central PMCID: PMCPMC107929. 
92.Wang JA, Meyer TF, Rudel T. Cytoskeleton and motor proteins are required for the 
transcytosis of Neisseria gonorrhoeae through polarized epithelial cells. Int J Med 
Microbiol. 2008;298(3-4):209-21. doi: 10.1016/j.ijmm.2007.05.004. PubMed PMID: 
17683982. 
93.Wang J, Gray-Owen SD, Knorre A, Meyer TF, Dehio C. Opa binding to cellular 
CD66 receptors mediates the transcellular traversal of Neisseria gonorrhoeae across 
polarized T84 epithelial cell monolayers. Mol Microbiol. 1998;30(3):657-71. PubMed 
PMID: 9822830. 
148 
 
94.Stein DC, LeVan A, Hardy B, Wang LC, Zimmerman L, Song W. Expression of 
Opacity Proteins Interferes with the Transmigration of Neisseria gonorrhoeae across 
Polarized Epithelial Cells. PLoS One. 2015;10(8):e0134342. Epub 2015/08/05. doi: 
10.1371/journal.pone.0134342. PubMed PMID: 26244560; PubMed Central PMCID: 
PMCPMC4526573. 
95.Wang LC, Yu Q, Edwards V, Lin B, Qiu J, Turner JR, et al. Neisseria gonorrhoeae 
infects the human endocervix by activating non-muscle myosin II-mediated epithelial 
exfoliation. PLoS Pathog. 2017;13(4):e1006269. Epub 2017/04/13. doi: 
10.1371/journal.ppat.1006269. PubMed PMID: 28406994; PubMed Central PMCID: 
PMCPMC5391109. 
96.Ward ME, Watt PJ, Robertson JN. The human fallopian tube: a laboratory model 
for gonococcal infection. J Infect Dis. 1974;129(6):650-9. PubMed PMID: 4209721. 
97.Mosleh IM, Boxberger HJ, Sessler MJ, Meyer TF. Experimental infection of native 
human ureteral tissue with Neisseria gonorrhoeae: adhesion, invasion, intracellular 
fate, exocytosis, and passage through a stratified epithelium. Infect Immun. 
1997;65(8):3391-8. PubMed PMID: 9234803; PubMed Central PMCID: 
PMCPMC175480. 
98.McGee ZA, Jensen RL, Clemens CM, Taylor-Robinson D, Johnson AP, Gregg CR. 
Gonococcal infection of human fallopian tube mucosa in organ culture: relationship of 
mucosal tissue TNF-alpha concentration to sloughing of ciliated cells. Sex Transm 
Dis. 1999;26(3):160-5. PubMed PMID: 10100774. 
99.Kim M, Ashida H, Ogawa M, Yoshikawa Y, Mimuro H, Sasakawa C. Bacterial 
interactions with the host epithelium. Cell Host Microbe. 2010;8(1):20-35. doi: 
10.1016/j.chom.2010.06.006. PubMed PMID: 20638639. 
100.Mulvey MA, Schilling JD, Martinez JJ, Hultgren SJ. Bad bugs and beleaguered 
bladders: interplay between uropathogenic Escherichia coli and innate host defenses. 
Proc Natl Acad Sci U S A. 2000;97(16):8829-35. PubMed PMID: 10922042; PubMed 
Central PMCID: PMCPMC34019. 
101.Mulvey MA, Schilling JD, Hultgren SJ. Establishment of a persistent Escherichia 
coli reservoir during the acute phase of a bladder infection. Infect Immun. 
2001;69(7):4572-9. doi: 10.1128/IAI.69.7.4572-4579.2001. PubMed PMID: 
11402001; PubMed Central PMCID: PMCPMC98534. 
102.Coradini D, Casarsa C, Oriana S. Epithelial cell polarity and tumorigenesis: new 
perspectives for cancer detection and treatment. Acta Pharmacol Sin. 2011;32(5):552-
64. Epub 2011/04/18. doi: 10.1038/aps.2011.20. PubMed PMID: 21499288; PubMed 
Central PMCID: PMCPMC4002515. 
103.Meng W, Takeichi M. Adherens junction: molecular architecture and regulation. 
Cold Spring Harb Perspect Biol. 2009;1(6):a002899. Epub 2009/08/05. doi: 
10.1101/cshperspect.a002899. PubMed PMID: 20457565; PubMed Central PMCID: 
PMCPMC2882120. 
149 
 
104.Hartsock A, Nelson WJ. Adherens and tight junctions: structure, function and 
connections to the actin cytoskeleton. Biochim Biophys Acta. 2008;1778(3):660-9. 
Epub 2007/07/27. doi: 10.1016/j.bbamem.2007.07.012. PubMed PMID: 17854762; 
PubMed Central PMCID: PMCPMC2682436. 
105.Huber AH, Weis WI. The structure of the beta-catenin/E-cadherin complex and 
the molecular basis of diverse ligand recognition by beta-catenin. Cell. 
2001;105(3):391-402. PubMed PMID: 11348595. 
106.Lickert H, Bauer A, Kemler R, Stappert J. Casein kinase II phosphorylation of E-
cadherin increases E-cadherin/beta-catenin interaction and strengthens cell-cell 
adhesion. J Biol Chem. 2000;275(7):5090-5. PubMed PMID: 10671552. 
107.Nelson WJ. Regulation of cell-cell adhesion by the cadherin-catenin complex. 
Biochem Soc Trans. 2008;36(Pt 2):149-55. doi: 10.1042/BST0360149. PubMed 
PMID: 18363555; PubMed Central PMCID: PMCPMC3368607. 
108.Rhee J, Mahfooz NS, Arregui C, Lilien J, Balsamo J, VanBerkum MF. Activation 
of the repulsive receptor Roundabout inhibits N-cadherin-mediated cell adhesion. Nat 
Cell Biol. 2002;4(10):798-805. doi: 10.1038/ncb858. PubMed PMID: 12360290. 
109.Roura S, Miravet S, Piedra J, García de Herreros A, Duñach M. Regulation of E-
cadherin/Catenin association by tyrosine phosphorylation. J Biol Chem. 
1999;274(51):36734-40. PubMed PMID: 10593980. 
110.Piedra J, Miravet S, Castaño J, Pálmer HG, Heisterkamp N, García de Herreros A, 
et al. p120 Catenin-associated Fer and Fyn tyrosine kinases regulate beta-catenin Tyr-
142 phosphorylation and beta-catenin-alpha-catenin Interaction. Mol Cell Biol. 
2003;23(7):2287-97. PubMed PMID: 12640114; PubMed Central PMCID: 
PMCPMC150740. 
111.Balsamo J, Leung T, Ernst H, Zanin MK, Hoffman S, Lilien J. Regulated binding 
of PTP1B-like phosphatase to N-cadherin: control of cadherin-mediated adhesion by 
dephosphorylation of beta-catenin. J Cell Biol. 1996;134(3):801-13. PubMed PMID: 
8707857; PubMed Central PMCID: PMCPMC2120944. 
112.Winter MC, Shasby S, Shasby DM. Compromised E-cadherin adhesion and 
epithelial barrier function with activation of G protein-coupled receptors is rescued by 
Y-to-F mutations in beta-catenin. Am J Physiol Lung Cell Mol Physiol. 
2008;294(3):L442-8. Epub 2007/12/14. doi: 10.1152/ajplung.00404.2007. PubMed 
PMID: 18083766. 
113.Timmerman I, Hoogenboezem M, Bennett AM, Geerts D, Hordijk PL, van Buul 
JD. The tyrosine phosphatase SHP2 regulates recovery of endothelial adherens 
junctions through control of β-catenin phosphorylation. Mol Biol Cell. 
2012;23(21):4212-25. Epub 2012/09/05. doi: 10.1091/mbc.E12-01-0038. PubMed 
PMID: 22956765; PubMed Central PMCID: PMCPMC3484100. 
150 
 
114.Huber AH, Stewart DB, Laurents DV, Nelson WJ, Weis WI. The cadherin 
cytoplasmic domain is unstructured in the absence of beta-catenin. A possible 
mechanism for regulating cadherin turnover. J Biol Chem. 2001;276(15):12301-9. 
Epub 2000/12/19. doi: 10.1074/jbc.M010377200. PubMed PMID: 11121423. 
115.Van Itallie CM, Anderson JM. Architecture of tight junctions and principles of 
molecular composition. Semin Cell Dev Biol. 2014;36:157-65. Epub 2014/08/27. doi: 
10.1016/j.semcdb.2014.08.011. PubMed PMID: 25171873; PubMed Central PMCID: 
PMCPMC4254347. 
116.Harhaj NS, Antonetti DA. Regulation of tight junctions and loss of barrier 
function in pathophysiology. Int J Biochem Cell Biol. 2004;36(7):1206-37. doi: 
10.1016/j.biocel.2003.08.007. PubMed PMID: 15109567. 
117.Ulluwishewa D, Anderson RC, McNabb WC, Moughan PJ, Wells JM, Roy NC. 
Regulation of tight junction permeability by intestinal bacteria and dietary 
components. J Nutr. 2011;141(5):769-76. Epub 2011/03/23. doi: 
10.3945/jn.110.135657. PubMed PMID: 21430248. 
118.Tanos B, Rodriguez-Boulan E. The epithelial polarity program: machineries 
involved and their hijacking by cancer. Oncogene. 2008;27(55):6939-57. doi: 
10.1038/onc.2008.345. PubMed PMID: 19029936. 
119.Capaldo CT, Macara IG. Depletion of E-cadherin disrupts establishment but not 
maintenance of cell junctions in Madin-Darby canine kidney epithelial cells. Mol Biol 
Cell. 2007;18(1):189-200. Epub 2006/11/08. doi: 10.1091/mbc.e06-05-0471. PubMed 
PMID: 17093058; PubMed Central PMCID: PMCPMC1751327. 
120.Fukuyama T, Ogita H, Kawakatsu T, Inagaki M, Takai Y. Activation of Rac by 
cadherin through the c-Src-Rap1-phosphatidylinositol 3-kinase-Vav2 pathway. 
Oncogene. 2006;25(1):8-19. doi: 10.1038/sj.onc.1209010. PubMed PMID: 16170364. 
121.Miyoshi J, Takai Y. Structural and functional associations of apical junctions with 
cytoskeleton. Biochim Biophys Acta. 2008;1778(3):670-91. doi: 
10.1016/j.bbamem.2007.12.014. PubMed PMID: 18201548. 
122.Baum B, Georgiou M. Dynamics of adherens junctions in epithelial 
establishment, maintenance, and remodeling. J Cell Biol. 2011;192(6):907-17. doi: 
10.1083/jcb.201009141. PubMed PMID: 21422226; PubMed Central PMCID: 
PMCPMC3063136. 
123.Blaskewicz CD, Pudney J, Anderson DJ. Structure and function of intercellular 
junctions in human cervical and vaginal mucosal epithelia. Biol Reprod. 
2011;85(1):97-104. Epub 2011/04/06. doi: 10.1095/biolreprod.110.090423. PubMed 
PMID: 21471299; PubMed Central PMCID: PMCPMC3123383. 
124.Draper DL, Donegan EA, James JF, Sweet RL, Brooks GF. Scanning electron 
microscopy of attachment of Neisseria gonorrhoeae colony phenotypes to surfaces of 
151 
 
human genital epithelia. Am J Obstet Gynecol. 1980;138(7 Pt 1):818-26. PubMed 
PMID: 6108716. 
125.Sobel G, Szabó I, Páska C, Kiss A, Kovalszky I, Kádár A, et al. Changes of cell 
adhesion and extracellular matrix (ECM) components in cervical intraepithelial 
neoplasia. Pathol Oncol Res. 2005;11(1):26-31. Epub 2005/03/31. doi: 
PAOR.2005.11.1.0026. PubMed PMID: 15800679. 
126.Ponce de León-Rodríguez MDC, Guyot JP, Laurent-Babot C. Intestinal in vitro 
cell culture models and their potential to study the effect of food components on 
intestinal inflammation. Crit Rev Food Sci Nutr. 2018:1-19. Epub 2018/10/02. doi: 
10.1080/10408398.2018.1506734. PubMed PMID: 30277794. 
127.Dean P, Kenny B. Intestinal barrier dysfunction by enteropathogenic Escherichia 
coli is mediated by two effector molecules and a bacterial surface protein. Mol 
Microbiol. 2004;54(3):665-75. doi: 10.1111/j.1365-2958.2004.04308.x. PubMed 
PMID: 15491358. 
128.Weflen AW, Alto NM, Hecht GA. Tight junctions and enteropathogenic E. coli. 
Ann N Y Acad Sci. 2009;1165:169-74. doi: 10.1111/j.1749-6632.2009.04060.x. 
PubMed PMID: 19538303; PubMed Central PMCID: PMCPMC3325541. 
129.Simovitch M, Sason H, Cohen S, Zahavi EE, Melamed-Book N, Weiss A, et al. 
EspM inhibits pedestal formation by enterohaemorrhagic Escherichia coli and 
enteropathogenic E. coli and disrupts the architecture of a polarized epithelial 
monolayer. Cell Microbiol. 2010;12(4):489-505. Epub 2009/11/12. doi: 
10.1111/j.1462-5822.2009.01410.x. PubMed PMID: 19912240. 
130.Edwards JL, Shao JQ, Ault KA, Apicella MA. Neisseria gonorrhoeae elicits 
membrane ruffling and cytoskeletal rearrangements upon infection of primary human 
endocervical and ectocervical cells. Infect Immun. 2000;68(9):5354-63. PubMed 
PMID: 10948165; PubMed Central PMCID: PMCPMC101799. 
131.Griffiss JM, Lammel CJ, Wang J, Dekker NP, Brooks GF. Neisseria gonorrhoeae 
coordinately uses Pili and Opa to activate HEC-1-B cell microvilli, which causes 
engulfment of the gonococci. Infect Immun. 1999;67(7):3469-80. PubMed PMID: 
10377128; PubMed Central PMCID: PMCPMC116533. 
132.Grassmé HU, Ireland RM, van Putten JP. Gonococcal opacity protein promotes 
bacterial entry-associated rearrangements of the epithelial cell actin cytoskeleton. 
Infect Immun. 1996;64(5):1621-30. PubMed PMID: 8613370; PubMed Central 
PMCID: PMCPMC173971. 
133.Higashi DL, Lee SW, Snyder A, Weyand NJ, Bakke A, So M. Dynamics of 
Neisseria gonorrhoeae attachment: microcolony development, cortical plaque 
formation, and cytoprotection. Infect Immun. 2007;75(10):4743-53. doi: 
10.1128/IAI.00687-07. PubMed PMID: 17682045; PubMed Central PMCID: 
PMCPMC2044525. 
152 
 
134.Merz AJ, So M. Attachment of piliated, Opa- and Opc- gonococci and 
meningococci to epithelial cells elicits cortical actin rearrangements and clustering of 
tyrosine-phosphorylated proteins. Infect Immun. 1997;65(10):4341-9. PubMed PMID: 
9317047; PubMed Central PMCID: PMCPMC175623. 
135.Haglund CM, Welch MD. Pathogens and polymers: microbe-host interactions 
illuminate the cytoskeleton. J Cell Biol. 2011;195(1):7-17. doi: 
10.1083/jcb.201103148. PubMed PMID: 21969466; PubMed Central PMCID: 
PMCPMC3187711. 
136.Goley ED, Welch MD. The ARP2/3 complex: an actin nucleator comes of age. 
Nat Rev Mol Cell Biol. 2006;7(10):713-26. doi: 10.1038/nrm2026. PubMed PMID: 
16990851. 
137.Ivanov AI. Actin motors that drive formation and disassembly of epithelial apical 
junctions. Front Biosci. 2008;13:6662-81. Epub 2008/05/01. PubMed PMID: 
18508686. 
138.Georgiou M, Marinari E, Burden J, Baum B. Cdc42, Par6, and aPKC regulate 
Arp2/3-mediated endocytosis to control local adherens junction stability. Curr Biol. 
2008;18(21):1631-8. Epub 2008/10/30. doi: 10.1016/j.cub.2008.09.029. PubMed 
PMID: 18976918. 
139.Ivanov AI, Parkos CA, Nusrat A. Cytoskeletal regulation of epithelial barrier 
function during inflammation. Am J Pathol. 2010;177(2):512-24. Epub 2010/06/25. 
doi: 10.2353/ajpath.2010.100168. PubMed PMID: 20581053; PubMed Central 
PMCID: PMCPMC2913378. 
140.Fedwick JP, Lapointe TK, Meddings JB, Sherman PM, Buret AG. Helicobacter 
pylori activates myosin light-chain kinase to disrupt claudin-4 and claudin-5 and 
increase epithelial permeability. Infect Immun. 2005;73(12):7844-52. doi: 
10.1128/IAI.73.12.7844-7852.2005. PubMed PMID: 16299274; PubMed Central 
PMCID: PMCPMC1307049. 
141.Crawley SW, Mooseker MS, Tyska MJ. Shaping the intestinal brush border. J 
Cell Biol. 2014;207(4):441-51. doi: 10.1083/jcb.201407015. PubMed PMID: 
25422372; PubMed Central PMCID: PMCPMC4242837. 
142.Iizumi Y, Sagara H, Kabe Y, Azuma M, Kume K, Ogawa M, et al. The 
enteropathogenic E. coli effector EspB facilitates microvillus effacing and 
antiphagocytosis by inhibiting myosin function. Cell Host Microbe. 2007;2(6):383-92. 
doi: 10.1016/j.chom.2007.09.012. PubMed PMID: 18078690. 
143.McGee ZA, Johnson AP, Taylor-Robinson D. Human fallopian tubes in organ 
culture: preparation, maintenance, and quantitation of damage by pathogenic 
microorganisms. Infect Immun. 1976;13(2):608-18. PubMed PMID: 816745; PubMed 
Central PMCID: PMCPMC420652. 
153 
 
144.Jerse AE. Experimental gonococcal genital tract infection and opacity protein 
expression in estradiol-treated mice. Infect Immun. 1999;67(11):5699-708. PubMed 
PMID: 10531218; PubMed Central PMCID: PMCPMC96944. 
145.Eades-Perner AM, van der Putten H, Hirth A, Thompson J, Neumaier M, von 
Kleist S, et al. Mice transgenic for the human carcinoembryonic antigen gene maintain 
its spatiotemporal expression pattern. Cancer Res. 1994;54(15):4169-76. PubMed 
PMID: 8033149. 
146.Apicella MA, Shero M, Jarvis GA, Griffiss JM, Mandrell RE, Schneider H. 
Phenotypic variation in epitope expression of the Neisseria gonorrhoeae 
lipooligosaccharide. Infect Immun. 1987;55(8):1755-61. PubMed PMID: 2440807; 
PubMed Central PMCID: PMCPMC260597. 
147.Swanson J. Studies on gonococcus infection. IV. Pili: their role in attachment of 
gonococci to tissue culture cells. J Exp Med. 1973;137(3):571-89. PubMed PMID: 
4631989; PubMed Central PMCID: PMCPMC2139381. 
148.Chen T, Grunert F, Medina-Marino A, Gotschlich EC. Several carcinoembryonic 
antigens (CD66) serve as receptors for gonococcal opacity proteins. J Exp Med. 
1997;185(9):1557-64. PubMed PMID: 9151893; PubMed Central PMCID: 
PMCPMC2196295. 
149.LeVan A, Zimmerman LI, Mahle AC, Swanson KV, DeShong P, Park J, et al. 
Construction and characterization of a derivative of Neisseria gonorrhoeae strain 
MS11 devoid of all opa genes. J Bacteriol. 2012;194(23):6468-78. doi: 
10.1128/JB.00969-12. PubMed PMID: 23002223; PubMed Central PMCID: 
PMCPMC3497525. 
150.White LA, Kellogg DS. An improved fermentation medium for Neisseria 
gonorrhoeae and other Neisseria. Health Lab Sci. 1965;2(4):238-41. PubMed PMID: 
4953822. 
151.Schürch W, McDowell EM, Trump BF. Long-term organ culture of human 
uterine endocervix. Cancer Res. 1978;38(11 Pt 1):3723-33. PubMed PMID: 698932. 
152.Wang LC, Yu Q, Stein DC, Song W. Immunofluorescence Analysis of Human 
Endocervical Tissue Explants Infected with. Bio Protoc. 2018;8(3). doi: 
10.21769/BioProtoc.2720. PubMed PMID: 29780854; PubMed Central PMCID: 
PMCPMC5959038. 
153.Swanson KV, Jarvis GA, Brooks GF, Barham BJ, Cooper MD, Griffiss JM. 
CEACAM is not necessary for Neisseria gonorrhoeae to adhere to and invade female 
genital epithelial cells. Cell Microbiol. 2001;3(10):681-91. PubMed PMID: 11580753. 
154.Muenzner P, Bachmann V, Kuespert K, Hauck CR. The CEACAM1 
transmembrane domain, but not the cytoplasmic domain, directs internalization of 
human pathogens via membrane microdomains. Cell Microbiol. 2008;10(5):1074-92. 
doi: 10.1111/j.1462-5822.2007.01106.x. PubMed PMID: 18081725. 
154 
 
155.Chen L, Sung SS, Yip ML, Lawrence HR, Ren Y, Guida WC, et al. Discovery of 
a novel shp2 protein tyrosine phosphatase inhibitor. Mol Pharmacol. 2006;70(2):562-
70. Epub 2006/05/22. doi: 10.1124/mol.106.025536. PubMed PMID: 16717135. 
156.Vicente-Manzanares M, Ma X, Adelstein RS, Horwitz AR. Non-muscle myosin II 
takes centre stage in cell adhesion and migration. Nat Rev Mol Cell Biol. 
2009;10(11):778-90. doi: 10.1038/nrm2786. PubMed PMID: 19851336; PubMed 
Central PMCID: PMCPMC2834236. 
157.Day RM, Mitchell TJ, Knight SC, Forbes A. Regulation of epithelial syndecan-1 
expression by inflammatory cytokines. Cytokine. 2003;21(5):224-33. PubMed PMID: 
12824007. 
158.Pils S, Kopp K, Peterson L, Delgado Tascón J, Nyffenegger-Jann NJ, Hauck CR. 
The adaptor molecule Nck localizes the WAVE complex to promote actin 
polymerization during CEACAM3-mediated phagocytosis of bacteria. PLoS One. 
2012;7(3):e32808. doi: 10.1371/journal.pone.0032808. PubMed PMID: 22448228; 
PubMed Central PMCID: PMCPMC3308951. 
159.Wira CR, Rodriguez-Garcia M, Patel MV. The role of sex hormones in immune 
protection of the female reproductive tract. Nat Rev Immunol. 2015;15(4):217-30. doi: 
10.1038/nri3819. PubMed PMID: 25743222. 
160.Kyrgiou M, Mitra A, Moscicki AB. Does the vaginal microbiota play a role in the 
development of cervical cancer? Transl Res. 2017;179:168-82. Epub 2016/07/15. doi: 
10.1016/j.trsl.2016.07.004. PubMed PMID: 27477083; PubMed Central PMCID: 
PMCPMC5164950. 
161.Van Kooij RJ, Roelofs HJ, Kathmann GA, Kramer MF. Human cervical mucus 
and its mucous glycoprotein during the menstrual cycle. Fertil Steril. 1980;34(3):226-
33. PubMed PMID: 7409244. 
 
155