The Gonococcal Transcriptome during Infection of the Lower Genital Tract in Women

Gonorrhea is a highly prevalent disease resulting in significant morbidity worldwide, with an estimated 106 cases reported annually. Neisseria gonorrhoeae, the causative agent of gonorrhea, colonizes and infects the human genital tract and often evades host immune mechanisms until successful antibiotic treatment is used. The alarming increase in antibiotic-resistant strains of N. gonorrhoeae, the often asymptomatic nature of this disease in women and the lack of a vaccine directed at crucial virulence determinants have prompted us to perform transcriptome analysis to understand gonococcal gene expression patterns during natural infection. We sequenced RNA extracted from cervico-vaginal lavage samples collected from women recently exposed to infected male partners and determined the complete N. gonorrhoeae transcriptome during infection of the lower genital tract in women. On average, 3.19% of total RNA isolated from female samples aligned to the N. gonorrhoeae NCCP11945 genome and 1750 gonococcal ORFs (65% of all protein-coding genes) were transcribed. High expression in vivo was observed in genes encoding antimicrobial efflux pumps, iron response, phage production, pilin structure, outer membrane structures and hypothetical proteins. A parallel analysis was performed using the same strains grown in vitro in a chemically defined media (CDM). A total of 140 genes were increased in expression during natural infection compared to growth in CDM, and 165 genes were decreased in expression. Large differences were found in gene expression profiles under each condition, particularly with genes involved in DNA and RNA processing, iron, transposase, pilin and lipoproteins. We specifically interrogated genes encoding DNA binding regulators and iron-scavenging proteins, and identified increased expression of several iron-regulated genes, including tbpAB and fbpAB, during infection in women as compared to growth in vitro, suggesting that during infection of the genital tract in women, the gonococcus is exposed to an iron deplete environment. Collectively, we demonstrate that a large portion of the gonococcal genome is expressed and regulated during mucosal infection including genes involved in regulatory functions and iron scavenging.


Introduction
The sexually transmitted infection (STI) gonorrhea, caused by the Gram-negative bacterium Neisseria gonorrhoeae, is the second most common reportable disease in the U.S. In 2014, there were 331,000 reported cases and the total annual number is estimated to be 800,000 due to underreporting (Centers for Disease Control and Prevention). Gonorrhea is also prevalent worldwide with an estimated 106 million cases annually [1]. The relatively straightforward therapeutic approaches to treat gonococcal infections have been complicated recently by the emergence of antibiotic resistance [2,3]. In 2013, the CDC reported that antibiotic-resistant N. gonorrhoeae was a major cause for concern in the US and abroad, and a recent study has described cephalosporin-resistant N. gonorrhoeae in the United States [4].
Like many other human pathogens, N. gonorrhoeae must adapt to the environment encountered during infection. Control of gene expression results from a number of distinct mechanisms, including DNA binding proteins and regulatory sRNAs that bind to target mRNAs to control their translation and stability [5][6][7][8]. DNA binding proteins that are important in infection include sigma factors, such as RpoH, shown to increase transcription of genes involved in stress response and adherence to epithelial cells [9,10], and OxyR, which regulates expression of genes such as catalase, involved in resistance to bactericidal reactive oxygen species [11]. Another well-known factor crucial for a successful host infection is Fur, which regulates intracellular iron levels [12,13]. Targets of Fur include the ferric-binding protein (FbpA), transferrin binding proteins (TbpAB) and lactoferrin binding proteins (LbpAB).
Despite numerous studies that have examined gonococcal gene expression in vitro, little is known about how this pathogen regulates gene expression during natural infection. In the female genital tract, N. gonorrhoeae is exposed to unique conditions, such as low pH [14,15], varying oxygen and iron levels [16][17][18][19], and the presence of additional microbes and host cells [20]. In this environment, free iron is scarce and is usually complexed to host iron binding proteins, including transferrin and lactoferrin, thus making the female genital tract an iron-deplete environment [17]. However, during menses iron levels can rise and N. gonorrhoeae responds to these changes via expression of the Fur protein [17]. Previous work from our group has shown that gonococcal genes encoding the tbpAB genes and the fur gene itself are differentially regulated during in vivo infection [16,17]. However, these studies were carried out using a combination of subset microarrays and qRT-PCR and thus were limited in their sensitivity and detection.
Here, we present the first study based on an RNA deep sequencing (RNA-seq) analysis of N. gonorrhoeae RNA samples derived directly from cervico-vaginal lavage specimens of infected women. The gonococcal strains isolated from these samples were also grown in vitro and the resulting transcriptomes were compared to those expressed during natural infection to define infection-specific gene expression profiles. This analysis has identified gonococcal genes that are highly expressed during cervical infection in women and other genes that are differentially regulated during infection compared to growth in vitro.

Global analysis of gonococcal gene expression during natural mucosal infection in women
Following RNA extraction and sequencing from four cervico-vaginal lavage samples, cDNA reads were aligned to the three fully sequenced strains of N. gonorrhoeae in the NCBI database: N. gonorrhoeae FA1090, NCCP11945 and TCDC-NGO8107, using Rockhopper software [21]. Gonococcal RNA isolated from all subjects aligned most closely to the NCCP11945 strain, a strain isolated from a vaginal specimen in 2008 in Korea [22]. On average, 3.19% of the total RNA from each sample aligned to the N. gonorrhoeae NCCP11945 genome and~45% aligned to the human genome. To be considered expressed, a gene had to have a reads per kilobase per million reads (RPKM) value of at least 10. Our analysis detected expression of 1750 gonococcal genes, representing~65% of the gonococcal genome. The level of expression of all genes, regardless of RPKM level in each subject and infecting strain, is summarized in S1 Table. High expression of several similar genes was detected in each of the vaginal samples (Table 1). These similarities in gene expression occurred despite temporal differences in the exposure of female subjects to the infected partner(s), menstrual cycle and degree of cervico-vaginal neutrophilia ( Table 2). This core set of similar genes included those encoding outer membrane (omp3, NGK_0749) housekeeping (elongation factor P, Tu), stress response (groES), phage associated (NGK_1145) and hypothetical proteins (NGK_0472, NGK_2043) ( Table 1). In addition to ORFs, 75 short non-coding transcripts representing regulatory small RNAs (described in detail below), were also expressed. It is important to note that several of these highly expressed genes showed similar high expression in vitro during growth in CDM. Specific analysis of genes showing changes in expression as a result of infection is discussed below.
To gain information on general trends of gonococcal responses to infection in women, we categorized and organized the top 100 gonococcal genes that were most highly expressed when RPKM values from all subjects were averaged together. These genes were categorized based on their gene annotation according to strain NCCP11945, which was chosen because sequences aligned most closely to that strain [22]. To determine the ratio of enrichment, these highly expressed genes were first categorized by function. Then, the percentage of genes in each functional category was compared to the percentage of genes within the same functional category in the whole gonococcal genome. The resulting ratio was defined as the "ratio of enrichment". The most commonly represented enriched categories that were highly expressed included genes associated with pilin, membrane and stress response as well as traditional housekeeping genes such as those involved in transcription, replication and energy metabolism (Fig 1). Expression of these genes showed variable levels in each of the cervico-vaginal samples, but all were expressed at high levels and, as a whole, reflected the profile of the most highly expressed gonococcal genes during natural infection. As above, several of these genes were also highly expressed in vitro. A specific analysis of genes that showed changes in expression relative to in vitro growth is presented below.
Expression of N. gonorrhoeae antimicrobial resistance genes during natural mucosal infection in women N. gonorrhoeae strains isolated recently worldwide have shown increased resistance to antibiotics [2][3][4]. The infecting strains in our study also were resistant to several antibiotics (S2 Table). Thus, we examined expression of genes encoding the mtrCDE efflux pump [23][24][25][26], a system that exports toxic, host-derived compounds and hydrophobic antibiotics from gonococci. MtrR, the regulatory protein which represses transcription of mtrCDE genes [27], was also examined. Expression of all genes within the mtrCDE locus was found in all 4 cervico-vaginal lavage samples (Fig 2). We observed lower levels of expression of the regulatory mtrR gene as compared to the efflux pump components for all 4 infecting strains. In Subjects 2, 3 and 4, the mtrR gene was not detected at an RPKM level of at least 10. Expression of mtrR in Subject 1 was higher than in other subjects, corresponding to a higher expression of the mtrCDE genes. Expression of the mtrA gene, which activates the mtrCDE locus [28], was also detected in Subjects 1, 2 and 4 with RPKM values of 67, 31 and 167 respectively (S1 Table-Line 262). It should be noted that many of the genes in the mtr locus also showed high expression in vitro, and that mtr genes were not specifically regulated by infection but rather showed a pattern of expression in vivo consistent with low levels of the repressor and high levels of the efflux pump genes (Fig 2).

Gonococcal gene expression during natural mucosal infection relative to expression during growth in CDM
The majority of studies aimed at characterizing the response of the gonococcus to different environmental stimuli have examined N. gonorrhoeae cultured in vitro (i.e. during incubation in media or with host cells) conditions that do not completely replicate the environment encountered in the human host during infection. Therefore, gonococcal genes that are expressed during natural infection may differ markedly from those expressed in vitro. Thus, we cultured gonococcal isolates from Subjects 1, 2, and 3 (Subject 4 was diagnosed by PCR only; no isolate was recovered) in a chemically defined medium (CDM) with added ferric nitrate (100 μM final concentration) and compared the resulting transcriptome to that expressed during infection in the female genital tract. We did not observe differences in the growth rates between the 3 strains in CDM plus ferric nitrate (data not shown). Expression levels for all genes detected under in vitro conditions are shown in S1 Table. Regulated genes were those that showed a difference in expression of at least 3-fold (up or down) in natural infection compared to growth in CDM. On average, 397 gonococcal genes, representing 20% of the genome, were regulated in each subject. Functional enrichment analysis, performed on genes showing differences in expression during natural infection compared to growth in CDM, showed enrichment of genes encoding for tRNA, iron, membrane, transposase and lipoproteins among genes whose expression was increased in vivo compared to growth in CDM. In contrast, metabolism, growth and housekeeping functions were enriched in genes showing decreased expression in vivo compared to CDM (Fig 3). Subject-specific gene expression patterns appeared overall remarkably similar, although subtle differences were also noted between subjects. For example, in the comparison between in vivo gene expression and growth in CDM, pilin-related genes were enriched among genes showing increased expression in Subject 2, enriched among genes showing decreased expression in Subject 3 and approximately   equal Subject 1. A similar variability was observed for stress response genes, where enrichment among genes with increased expression was observed in Subject 3 but not in Subject 1 or 2. Finally, genes encoding lipoproteins were enriched among genes with increased expression in Subject 1 but not in Subjects 2 or 3. Overall, we observed general trends and a core set of genes differentially expressed in all subjects regardless of the specifics of infection. However, the subtle differences in regulation and expression of gonococcal genes, for example those involved in pilin and stress response, indicate that the genital tract of each subject was unique.
Regulation of N. gonorrhoeae genes encoding DNA binding proteins during natural mucosal infection relative to growth in CDM Regulatory proteins that bind DNA appear to play a prominent role in N. gonorrhoeae infection of the genital tract in women. Among the 3 subjects, each examined individually, 14 DNA-binding regulators were differentially expressed during natural infection as compared to growth in CDM, with increased expression of 6 and decreased expression of 8 regulatory protein genes. Certain genes that encode DNA binding proteins were regulated in multiple subjects, for example an AraC-like regulator (NGK_2588) was increased in Subjects 1 and 2, a two-component regulator (NGK_2403) was decreased in Subjects 2 and 3, and a CysB-like regulator decreased (NGK_1878) in Subjects 1 and 2 ( Table 3). The fur gene also showed lower levels in RNA isolated from Subjects 2 and 3 compared to growth in CDM.

Composite analysis of gonococcal gene expression during natural mucosal infection relative to expression during growth in CDM
Several gonococcal genes were regulated in a similar fashion in each subject, suggesting that a common response of the gonococcus may be dependent on intrinsic features of the human female genital tract. Thus, we next examined the combined results from the 4 cervico-lavage samples to define an 'average' expression of each gonococcal gene during infection (S3 Table). As a comparison, we also combined the data resulting from the in vitro growth of the 3 available infecting strains, and lastly, used these averages to compare in vivo and in vitro gene expression. We identified 305 genes with statistically significant changes in expression, defined as a q-value of <0.05 and > than a 2-fold change in expression. A total of 140 genes showed increased expression and 165 decreased expression during in vivo vs. in vitro growth (S4 Table; and Fig 4A). Functional enrichment analysis was performed to define how the gonococcus alters gene expression during infection (Fig 4B and 4C). We observed that genes involved in membrane composition/function, iron scavenging, transposases and tRNAs were strongly enriched among genes with increased expression in vivo compared to growth in CDM. In contrast, several housekeeping genes were enriched among genes showing decreased expression, suggesting changes in growth and replication rates of N. gonorrhoeae during infection compared to growth in CDM (Fig 4B and 4C). Other categories such as pilin and lipoprotein genes also were enriched among genes showing decreased expression despite the fact that when examining subjects individually these genes were sometimes found to be enriched among genes showing increased expression. Interestingly, 20 genes encoding phage-associated proteins showed statistically significant regulation, with seven of these genes showing increased expression during infection as compared to growth in CDM (S4 Table). Phage-associated genes were also slightly enriched among genes showing decreased expression in vivo. The expression levels of all genes from cervico-vaginal lavage samples compared to strains grown in CDM are shown in S3 Table. A small number of genes that encoded DNA binding regulatory proteins also exhibited statistically significant regulation when infection in vivo vs. growth in vitro in CDM was compared. Some were the same regulatory proteins that showed changes in expression when individual subjects were examined; for example an ArsC transcriptional regulator (NGK_2588) was expressed approximately 4-fold higher during in vivo infection vs. growth in vitro and was also regulated in Subjects 1 and 2 individually (Table 3).
We also identified 75 non-coding RNAs between 30-250 nucleotides in length in the composite analysis that possibly represented sRNAs (S3 Table). Expression of putative sRNAs was observed both as antisense transcripts opposite known ORFs and from intergenic regions. Ten of these sRNAs showed increased expression during natural infection compared to growth in CDM; 16 sRNAs showed decreased expression (S4 Table). Several of the sRNAs expressed during infection have previously been identified in N. gonorrhoeae under a variety of in vitro conditions [8,29]. For example, the sRNA NrrF exhibited statistically significant increased expression in vivo as compared to in vitro. We also detected transcription of the sRNAs 4.5S and M1 [30]. Other sRNAs, including a total of~50% of sRNAs that had been experimentally verified in a previous study [8] were also detected, including one at genomic coordinates 1201200-1201287,which showed decreased expression in vivo compared to in vitro (S4 Table) [8]. We did not detect expression of the pilin antigenic variation small RNA [5]. A large number of transfer RNAs (tRNAs) also exhibited statistically significant changes in expression (S4   Table). Most tRNAs were detected at higher levels during natural infection compared to growth in CDM. Of 33 tRNAs that showed significant regulation, only one was expressed at higher levels during growth in CDM compared to natural infection.

Expression and regulation of gonococcal genes that encode for iron scavenging proteins during natural mucosal infection relative to growth in CDM
Previously, we reported that the gonococcal iron and Fur-regulated genes were expressed in urethral and cervico-vaginal lavage samples obtained from men and women with gonococcal infections [16,17]. While these studies confirmed expression of tbp and fbp genes during infection, they did not address differential expression relative to corresponding infecting strains grown in vitro. In the current study we examined expression of tbp and fbp genes in the three culture-positive subject samples and compared this expression to the corresponding strains grown in vitro. Predominantly, increased expression of iron regulated tbpA and tbpB genes was seen in samples from each subject as compared to expression in vitro, although expression in individual samples varied. We also observed increased expression of fbpABC genes in vivo as compared to expression in vitro (Fig 5). Analysis of average expression of tbpAB and fbpAC genes from all cervico-vaginal lavage samples also revealed statistically significant increased levels compared to organisms growth in CDM with ferric nitrate (q < 0.0001; Fig 5). Consistent with these results, the repressor of these genes, Fur, was expressed at lower levels during natural infection compared to growth in CDM with ferric nitrate (Table 3 and S3 Table). Other iron-regulated genes also showed regulation; both hpuA and hpuB were highly expressed in vivo as compared to in vitro, and two out of four heme utilization proteins also showed a similar pattern (although not statistically significant). In contrast, lactoferrin binding proteins showed higher expression in vitro compared to in vivo (Fig 5). Because in vitro growth was performed under iron-replete conditions (when iron scavenging genes exhibit low expression), the high in vivo expression of tbp, fbp, and hpu genes, as well as NrrF, combined with the low expression of Fur suggests that gonococci encounter an iron-deplete environment in the female genital tract.

Discussion
In this study we report the complete gonococcal transcriptome expressed during infection in women. Our results show that more than 65% of the genome was transcribed and detected. Genes that were highly expressed across all subjects during infection included genes associated with stress response, outer membrane and hypothetical proteins, and a high number of phage genes which previously have not been well characterized in N. gonorrhoeae. Phage associated genes have been shown previously to be intimately involved in infection including in the closely related human pathogen N. meningitidis [31][32][33]. While prior studies have reported the presence of at least 9 complete phage genomes within the genome of N. gonorrhoeae, the function of these phage genomes has not been investigated [34,35]. A recent study reported on the possible function of these phage proteins in the context of infection [36]. Prior studies from our group identified a phage-associated DNA binding regulatory protein that controls genes required for gonococcal infection of human epithelial cells and colonization in a mouse model [37]. Our analysis showed that 20 phage-associated genes were regulated during infection as compared to growth in vitro and further analysis of these genes may provide new information on the function of these genes during human gonococcal infection. Of particular relevance in light of an alarming increase in antibiotic-resistant strains of N. gonorrhoeae were the high levels of expression of the multiple transferable resistance (mtr)  (100 μM final concentration). These genes were expressed at significantly higher levels and were placed in a separate graph for ease of viewing. (C) Average expression levels of fbpC, fbpB, lbpB, lbpA, hbpA, hbpB, hbpC, tbpA, tbpB, hpuB, and hpuA gene expression in RNA isolated from 4 cervico-lavage samples (red bars) and average expression CDE locus genes and a corresponding low expression of the mtrR repressor that controls expression of the operon in the gonococcal strains and samples obtained directly from subjects. Strains isolated in this study were not resistant to ceftriaxone (the recommended treatment in the United States), but other gonococcal strains with decreased susceptibility to ceftriaxone have been isolated recently in China and in the United States [4]. The mtrCDE locus confers resistance to several host-derived antimicrobials, including cationic antimicrobial peptides (eg. LL-37), long chained fatty acids and other hydrophobic and artificial compounds (i.e. protegrin-1, progesterone and Triton X-100). Exposure of N. gonorrhoeae to macrolides (azithromycin), β-lactam antimicrobials (penicillin) ciprofloxacin, tetracycline, rifampicin and extended spectrum cephalosporins, results in increased MICs to these antimicrobials [23][24][25][26]38]. The mtrCDE locus is under the control of two classical DNA binding proteins: the Mtr repressor (MtrR) protein, which down regulates expression of the mtrC transcript and presumably other proteins in the operon [27], and the Mtr activator (MtrA) [28], which enhances expression of mtrCDE. Several studies have shown that mutations of MtrR lead to enhanced fitness of N. gonorrhoeae in a mouse model of infection, suggesting that this system may also be crucial for survival of the organism in the human host [38,39].
A parallel comparative analysis of the transcriptome of N. gonorrhoeae in individual subjects and corresponding infecting isolates grown in vitro in a CDM identified 140 genes with increased expression and 165 genes with decreased expression. Predictably, there were subjectspecific changes in the transcriptomes of N. gonorrhoeae isolates, due to the unique genital tract environment of individual subjects, which lead to tailored and specific N. gonorrhoeae responses during infection. These specific responses are likely mediated, in part, by DNA-binding transcription factors, several of which were found to be expressed and regulated during infection, including Fur and MtrR, and an ArsR-like regulator. We have shown that the latter is crucial to N. gonorrhoeae survival during infection of female endocervical cells, and that this regulator is also regulated by Fur (C. A. Genco, unpublished data). We found that numerous other regulators showed variable expression during infection, particularly those involved in nutrient acquisition and metabolism (NGK_1662, GntR, BasS). This likely reflects the large differences in metabolite composition in the female genital tract as compared to growth in CDM. In addition to well-known transcription factors, others such as NGK_1277, NGK_1662, and NGK_1878 are relatively uncharacterized and may be involved in regulation of gene products that are crucial to gonococcal infection.
Increased expression of the iron regulated tbpAB and fbpABC genes during infection was also observed as compared to in vitro growth. It is important to note that our studies only examined gonococcal strains grown under iron-replete conditions in vitro, and that a comparison of the gonococcal transcriptome expressed during infection with other in vitro conditions will likely identify additional genes that are differentially expressed during infection. We have previously reported that tbpAB and fbpA are expressed in vivo [17]. High expression and regulation of these proteins in uncomplicated cervical infection strongly suggests that the genital tract in women is an iron-deplete environment.
In addition, a large number of non-coding RNAs, including putative sRNAs and tRNAs, exhibited differential expression during infection. Relatively little is known regarding gonococcal sRNAs and studies on these regulators in N. gonorrhoeae are scarce. Work from our group and others have described the repertoire of N. gonorrhoeae sRNAs expressed during in vitro growth [5][6][7][8]. It is possible that post-transcriptional regulation by sRNAs plays an important role in N. gonorrhoeae gene expression during infection because of the relatively small number of DNA binding proteins expressed by gonococci compared to other Gram-negative bacteria.
In Streptococcus pneumoniae, Staphylococcus aureus, Enterococcus faecalis and Pseudomonas aeruginosa, expression of tRNAs have been reported to play a role in the development of antibiotic resistance through changes in membrane permeability [40][41][42]. Thus the increased expression of tRNAs observed during natural gonococcal infection relative to growth in vitro may represent an infection specific response to resist host-derived antimicrobial components. The current study lacks a time course analysis of gonococcal gene expression during of natural infection; it seems Iikely that the gonococcal transcriptome changes as disease progresses from initial attachment and invasion of host cells and the recruitment of immune cells. Future studies will focus on such an analysis.
A number of bacterial factors expressed during mucosal infection likely contribute to development of gonococcal disease, to the persistent nature of the infection and to disparate symptomatic responses observed in men and women. However, the majority of these microbial factors have been identified through experimental conditions that may not faithfully replicate those encountered by bacteria in their specific host environment. It is likely that many pathogenicity-associated factors expressed during natural infection are still unknown or undetected. Our studies using RNA-seq analysis are the first to define the gonococcal global transcriptome during natural mucosal infection. Collectively, we demonstrate that a large portion of the gonococcal genome is expressed during mucosal infection and that the gonococcus responds specifically to the anti-microbial and iron-deplete nature of the lower genital tract in women. Future studies should focus on examining highly expressed genes for both therapeutic and vaccine studies, as well as genes specifically regulated as a function of infection to increase our knowledge of the molecular mechanisms of infection of this pathogen.

Ethics statement
All subjects provided informed written consent. The BUMC Institutional Review Boards (IRB) at the University of Massachusetts Medical School, Boston University Medical Campus (protocol number H-28858), and at the Institute of Dermatology, Chinese Academy of Medical Sciences Nanjing approved the study including the protocol described in this manuscript and have determined that the study meets the requirements set forth by the IRB.

Subject samples
Cervico-vaginal lavage samples (for RNA isolation) and swabs (for isolation of infecting strains) were obtained from four subjects attending the National Center for Sexually Transmitted Diseases (NCSTD) in Nanjing, China. Women were the exclusive sex-contacts of male subjects who had presented to the clinic with symptomatic urethritis; men were diagnosed with gonorrhea by Gram's stain and culture (see below). Women were screened for N. gonorrhoeae infection by culture and PCR (see below), and tested positive by one or both methods. Cervicovaginal lavage specimens were obtained according to standard collection methods [43], also demonstrated in the Microbicide Trials Network (MTN) training video used for NIAID/NIH sponsored clinical trials. Lavage fluid was stored in RLT Buffer (Qiagen) at -80°C and shipped to the United States on dry ice. Additional specimens were used for culturing and isolation/ identification of N. gonorrhoeae as described below.
Isolation, culture and polymerase chain reaction (PCR) to identify N. gonorrhoeae in cervical swab specimens Cervical swab specimens were taken and inoculated onto Thayer-Martin agar plates and cultured in candle jars at 36°C for 24-48 hr. Following growth on plates, bacteria were examined by colony morphology, Gram's stain and oxidase testing. Colonies identified as N. gonorrhoeae by these criteria were isolated and sub-cultured onto chocolate agar plates overnight. The next day, colonies were resuspended in Trypticase Soy Broth/Glycerol and frozen at -70°C. PCR testing to identify N. gonorrhoeae was performed utilizing primers directed against the gonococcal porA pseudogene [44] and touchdown PCR for amplification [45]. Mean inhibitory concentrations (MICs) of individual strains against: penicillin, tetracycline, spectinomycin, ceftriaxone and ciprofloxacin were determined by the agar dilution method [46].
Cervical swab specimens were also tested for Chlamydia trachomatis and Mycoplasma genitalium by PCR, and Ureaplasma urealyticum and Mycoplasma hominis by culture. Vaginal swabs were collected and tested for BV, candidiasis and trichomoniasis by wet mount microscopy. Wet mount slides were examined by trained medical technicians.
For analysis of the transcriptome of N. gonorrhoeae in vitro, individual strains were grown on gonococcal base (GCB) agar plates for 16-18 hr at 37°C in 5% CO 2 before they were resuspended in Chemically Defined Media (CDM) [47] containing ferric nitrate (100 μM final concentration). Cultures were started at an O.D. 600 of 0.1 and grown for 3 hrs before 2 ml cultures were centrifuged at 2000 rpm for 5 min and the resulting pellet used for RNA extraction.

RNA isolation
RNA was isolated from cervico-vaginal lavage samples or from in vitro-grown organisms using identical protocols to avoid method-based biases. TRIzol (Invitrogen) was used to isolate RNA directly from cervico-vaginal samples or from in vitro-grown N. gonorrhoeae pellets. RNA was washed twice with 70% ethanol and treated with TURBO DNase (Ambion) per the manufacturer's instructions. Eukaryotic rRNA was depleted from samples using the RiboMinus Kit (Invitrogen). To avoid biases, in vitro-grown N. gonorrhoeae was also treated with the eukaryotic rRNA depletion kit. Prokaryotic rRNA was depleted using the Microbe EXPRESS Kit (Ambion). Extractions were repeated twice to ensure sufficient depletion of rRNA. At no time was EDTA used. All aqueous solutions used for RNA isolation were in Diethylpyrocarbonate (DEPC)-treated water.

RNA sequencing and analysis
Representative samples of RNA were prepared for sequencing using BioChain's High Yield Directional mRNA Sample Prep Kit per the manufacturer's instructions. cDNA libraries were electrophoresed on 8% Tris/Borate/EDTA (TBE) gels for 60 min at 140 V. cDNA fragments between 200-500 base pairs were extracted from the gel in TE buffer with agitation and libraries were analyzed on an Agilent Technologies 2100 Bioanalyzer to confirm cDNA library size and quantity. Libraries were sequenced using 72 and 100 single-end base pair reads on an Illumina GAIIx or High-Seq machine respectively. All analysis of RNA-seq results was carried out using the Rockhopper program [21] with expression of genes defined in reads per kilobase per million reads (RPKM). All reads from each subject's sample were aligned to the three fully sequenced strains of N. gonorrhoeae in NCBI (NCCP11945, TCDCNGO8107 and FA1090) and the strain with the closest alignment (NCCP11945) used for downstream analysis.
To define gonococcal genes that were regulated by infection, comparisons were made between individual subject samples that yielded an isolate (total of 3; the 4 th was positive by PCR only) and their corresponding infecting strains (total of 3). This analysis considered genes to be regulated if they showed a fold change of > 3 between natural infection and growth in vitro. Not having an isolate available, Subject 4 was not used for this analysis. An additional analysis was made between a composite average of all 4 cervico-lavage samples compared to a composite average of 3 gonococcal strains grown in vitro. In this second analysis, availability of more than one sample permitted statistical analysis; genes were considered to be differentially expressed between two samples if the q value was < 0.05 and the fold-change was > 2. Categorization of genes was carried out using the NCBI RefSeq public database. Gene information for the NCCP11945 strain was used, the same strain used for sequence alignment. Functional enrichment was carried out after assigning all genes in the N. gonorrhoeae genome to one of 21 different categories. Functional enrichment ratios were derived through determining the percentage of genes in a given set (for example, increased expression in vivo compared to growth in CDM) that were assigned to a particular function and dividing this number by the percentage of genes assigned to this function in the genome as a whole. This ratio was used to interpret how enriched a given gene set was for a particular function.
Supporting Information S1 Table. Alignment of RNA from all Subjects and from infecting strains of Subjects 1-3 to the NCCP11945 strain of N. gonorrhoeae. Ã Expression is defined as the RPKM level when aligning RNA isolated from Subjects 1-4 to the NCCP11945 strain of N. gonorrhoeae. ÃÃ Expression is defined as the RPKM level when aligning RNA isolated from the infecting strain of Subjects 1-3 after 3 hours of growth in CDM to the NCCP11945 strain of N. gonorrhoeae. Subject 4 was negative for N. gonorrhoeae by plating and so an infecting strain was not isolated. (XLSX) S2 Table. Antibiotic Resistance and Porin Serovar (Type) of infecting N. gonorrhoeae strains. Ã All MICs displayed as μg/mL. (XLSX) S3 Table. Composite view of alignment of RNA from all Subjects and from infecting strains of Subjects 1-3 to the NCCP11945 strain of N. gonorrhoeae. Ã Expression is defined as the RPKM level when aligning RNA isolated from Subjects 1-4 to the NCCP11945 strain of N. gonorrhoeae. RNA was aligned as biological replicates of a single experiment to generate a composite alignment combining all Subjects. ÃÃ Expression is defined as the RPKM level when aligning RNA isolated from the infecting strain of Subjects 1-3 after 3 hours of growth in CDM to the NCCP11945 strain of N. gonorrhoeae. RNA was aligned as biological replicates of a single experiment to generate a composite alignment combining all in vitro analyses. Subject 4 was negative for N. gonorrhoeae by plating and so an infecting strain was not isolated. (XLSX) S4 Table. Regulated Genes When Examining Composite view of alignment of RNA from all Subjects and from infecting strains of Subjects 1-3 to the NCCP11945 strain of N. gonorrhoeae. Ã Expression is defined as the RPKM level when aligning RNA isolated from Subjects 1-4 to the NCCP11945 strain of N. gonorrhoeae. RNA was aligned as biological replicates of a single experiment to generate a composite alignment combining all Subjects. ÃÃ Expression is defined as the RPKM level when aligning RNA isolated from the infecting strain of Subjects 1-3 after 3 hours of growth in CDM to the NCCP11945 strain of N. gonorrhoeae. RNA was aligned as biological replicates of a single experiment to generate a composite alignment combining all in vitro analyses. Subject 4 was negative for N. gonorrhoeae by plating and so an infecting strain was not isolated. (XLSX)