Toll-like Receptors: Important Immune Checkpoints in the Regression of Cervical Intra-epithelial Neoplasia 2

Abstract

Toll-like receptors (TLRs) are innate immune defenders thought to be critical for the clearance of Human Papillomavirus (HPV) infections hence preventing the development of HPV-associated high-grade cervical intra-epithelial neoplasia (CIN2 or 3), a potential cervical cancer precursor. However, the role of TLRs in the regression of established cervical lesions, such as CIN2, is hindered by a lack of prospective design studies. Using SYBR green real-time PCR assays, we have examined the gene expression of TLR2, TLR3, TLR7, TLR8, and TLR9, in cytobrush collected endocervical cells of 63 women diagnosed with CIN2 at study entry (baseline) and followed over a 3-year period. Wilcoxon rank-sum test was used to examine the association between TLR expression levels, measured at baseline, and CIN2 outcome (regression versus persistence/progression) over time. HPV genotyping was performed using Roche Linear Array Assay detecting 37 HPV types. Women with CIN2 regression showed significantly higher baseline levels of TLR2 (p=0.006) and TLR7 (p=0.007), as well as a non-significant trend for a higher TLR8 expression (p=0.053) compared to women with CIN2 persistence/progression. Six women with CIN2 regression, who presented with an HR-HPV DNA-negative CIN2 lesion at study entry, had significantly higher baseline levels of TLR2 (p=0.005), TLR7 (p=0.013), and TLR8 (p=0.012), compared to women with CIN2 persistence/progression, suggesting their role in clearance of HPV prior to clearance of the lesion. Our results confirm a key role of TLRs in regression of CIN2 and support the potential use of TLR-agonists for treatment of these lesions.

INTRODUCTION

Persistent infection with one of the 12 carcinogenic or high-risk human papillomavirus (HR-HPV) types, is a prerequisite for the development of cervical cancer ¹. Cervical cancer develops through a series of molecular events which are reflected by histologic changes. True pre-cancers are considered to be represented by histologic alterations referred to as cervical intra-epithelial neoplasia (CIN) grades 2 and 3 whereas CIN grade 1 is considered a benign reversible lesion, with regression rate of 90% among adolescents and young women ^{2, 3}. CIN2 lesions can also spontaneously regress though at the lower rate ^4–6. Ours and other studies have shown that 70% of CIN2 cases in young women regress within three years without any intervention ^{3, 6, 7}. Underlying immune mechanisms critical for regression of such established lesions have not been well defined. Therefore, identification of targets for treatment of CIN2 and/or novel biomarkers to predict lesion progression, are continuously being sought out in order to prevent overtreatment and provide more specific/personalized medical care.

The host immune responses to cervical HPV are considered instrumental in shaping the natural history of HPV infection. Key components of that response include the initial pathogen recognition by the innate immune system, cytokines that bridge the innate and adaptive responses and mediate the latter, cytotoxic effector mechanisms, and regulatory cells and mechanisms that modulate the strength and duration of the response ^8–13. Among these, innate immune cells and receptors, such are toll-like receptors (TLRs) have been garnering increasing attention as therapeutics for treatment of HPV-associated diseases ¹⁴. However, the roles of different TLRs in the natural history of CIN have not been studied in detail.

TLRs are central molecules of the innate immune system engaged in the continuous detection of highly conserved, non-self, structural motifs known as pathogen-associated microbial patterns (PAMPs), exclusively expressed by microbial pathogens. Besides PAMPs, TLRs can also sense endogenous molecules released from necrotic or dying cells known as danger-associated molecular patterns (DAMPs). Of the ten TLRs characterized to date in humans, several are particularly important in viral infections and are also expressed in keratinocytes within the female genital tract (the cells HPV infects) ¹⁵. Intracellularly localized TLRs recognize nucleic acid motifs and include TLR3 (dsRNA), TLR7 (ssRNA), TLR8 (ssRNA), and TLR9 (CpG DNA); while cell surface-localized TLRs, such are heterodimers of TLR2, are important bacterial lipoprotein and viral protein sensors ^16–18. It has been described that viral dsDNA, or viral dsRNA (potential intermediate during viral replication), as well as viral proteins such as HPV L1 and L2 capsid proteins, can act as potential PAMPs which trigger TLR signaling ^{13, 19, 20}.

We have previously demonstrated that high cervical mucosal expression levels of TLR2, TLR3, TLR7, TLR8 and TLR9 upon incident HPV16 infection were critical for HPV16 clearance by the following (4-month) visit ^{13, 21}, suggesting both a role for TLRs in the successful host response to HPV and a mechanism by which HPV16 can evade immune recognition, persist, and drive lesion progression. In the present study, we explored the role of these TLRs in the regression of an established cervical neoplastic lesion, CIN2. We have measured, in a cross-sectional design, cervical mRNA expression of TLR2, TLR3, TLR7, TLR8, and TLR9 in adolescent and young women with biopsy-confirmed CIN2 at the study entry (baseline), and followed over a 3-year period. We hypothesized that TLR baseline levels would be the highest in women whose CIN2 lesions regressed and who at the same time cleared their baseline HR-HPV infection.

MATERIALS AND METHODS

Study design.

The recruitment of women for the parental study on the Natural History of CIN2 took place from 2002 – 2007 in Northern California, USA ^{6, 22}. The study was approved by the institutional review boards of the University of California, San Francisco (UCSF) and Kaiser Permanente, Northern California (KPNC), respectively. Women aged 13 to 25 years referred for abnormal cytology while visiting one of the 12 KPNC participating clinics, with CIN2 diagnosed on H&E histology, were recruited and followed at 4 – 6 month intervals for on an average of 3 years. Exclusion criteria included previous treatment for CIN, immunosuppression, pregnancy, or planning to leave the area within 3 years. Analysis included histologically confirmed CIN2 and at least one follow-up visit resulting in the participation of 95 women. Details of the study design have been previously published ⁶. Biological samples obtained at each of the visits included a cytobrush collected endocervical cells stored in RNAlater for RNA analysis. Cervical biopsies were collected at the study entry (for all 95 women) and the study exit (for 68% of 95 participating women), and at the intermediate visits if the colposcopist was concerned about progression, or if the cytology on any visit suggested progression ⁶. Two pathologists reviewed each cervical biopsy collected at the baseline, while CIN3 diagnosis, if established, was confirmed by a 3rd pathologist and women were exited for treatment. All study cytology and histology samples during follow-up were sent to the centralized laboratory to be reviewed by a single pathologist (TMD). During the interval of our study p16INK4a immunostaining was not a recommended/standard practice for confirmation of CIN2 diagnosis. The study entry biopsies could not be retrieved from the original centers for p16INK4a immunostaining after The Lower Anogenital Squamous Terminology (LAST) Standardization Project for HPV-Associated Lesions was introduced in 2012 ²³.

HPV testing.

Samples for cytology and HPV were immediately placed into liquid based media (PreservCyt; Hologic Corp, Marlborough, MA) and were sent to the UCSF laboratory where 9 mL was removed for amplification and genotyping as described ^{6, 22}. HPV testing was performed using the Roche Linear Array Assay (Roche Molecular Diagnostics, In, Alameda, CA) which detects 37 different HPV types: 12 HR-HPV (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 59), nine probable/possible (p)HR-HPV (26, 53, 66, 67, 68, 69, 70, 73, and 82), and 16 low-risk (LR-)HPV types (6, 11, 40, 42, 54, 55, 61, 62, 64, 71, 72, 81, 83, 84, IS39, and CP6108) ¹.

Definition of CIN2 lesion regression, persistence, or progression.

Definitions of CIN2 regression, persistence or progression for this cohort have been previously described ⁶. In brief, the definition of regression was based on having three consecutive visits with negative cytology, and negative biopsy on any of these visits (CIN0), if available. If there was an insufficient follow-up (i.e., only a single visit with negative cytology), the analysis was censored at the last visit with abnormal cytology or histology. Women who continued to have a low grade (L) or high grade (H) squamous intraepithelial lesion (SIL) on cytology, or CIN1 or 2 on histology at the end of follow-up, were considered non-regressors, i.e. having persistent CIN2. The definition of progression required a biopsy-proven CIN3 or greater at any visit after the baseline diagnosis of CIN2. Time to CIN2 regression or progression was defined as the time between the time point of CIN2 detection and the first of the three consecutive visits with normal cytology/histology (CIN0/LSIL), or the first CIN3 diagnosis ⁶.

RNA Extraction.

Of 95 women who fulfilled requirements for participation in the study on Natural History of CIN2 ^{6, 22}, 63 women had available endocervical cell specimens from the baseline visit for TLRs mRNA analysis. Specimen collection by cytology brush and subsequent RNA extraction using TRI Reagent (Molecular Research Center, Cincinnati, Ohio) have been previously described ²⁴. One µg total mRNA was DNase-treated using TURBO DNA-free Kit (Ambion, Austin, TX, USA), following the manufacturer’s instructions. Reverse transcription was performed using 0.5 µg DNase-treated RNA in a 20 µl reaction volume with random hexamer (37 ng/µl final concentration, Promega, Billerica, Massachusetts) and random nonamer (25 µM, Gene Link, Hawthorne, New York) priming, RNasin RNase inhibitor (10 U/reaction, Promega), and OmniScript RT (4 U/reaction, Qiagen, Valencia, California) at 37°C for 60 min, following the manufacturer’s instructions.

TLR Testing by Quantitative Reverse-Transcription PCR.

Quantitative PCR was performed in triplicate in a 384-well plate format on an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems, Foster City, California) using Power SYBR Green Master Mix (Applied Biosystems), forward and reverse primers (Integrated DNA Technologies, Coralville, Iowa) at 150 nM final concentration, and either cervical unknown cDNA template (12.5 ng/reaction), standard, or water for the no-template control. Primer sequences for TLRs selection and assay validation have been previously described ²⁴. Standard curves for each plate were prepared from Stratagene qPCR Human Reference Total RNA (Stratagene, La Jolla, California), reverse transcribed identically to the cervical RNA samples, and diluted in 4-fold steps from 80 ng/reaction (final concentration) to 0.078 ng/reaction. The amplification program was run as described ²⁴. Quantitative (threshold cycle) PCR values were obtained during exponential amplification. TLR expression, in relative units calculated by the efficiency-corrected method using amplification efficiencies determined from standard-curve slopes ²⁵, was normalized to the geometric mean ²⁶ of two reference genes previously validated for cervical samples in HPV studies; Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH), and Ribosomal Protein Lateral stalk subunit P0 (RPLP0) ²⁴.

Data Analysis.

Because normalized TLR expression levels were not normally distributed, Wilcoxon rank-sum test was used to examine associations between follow-up (independent variable) and TLR expression (dependent variable). GraphPad Prism 7.0 software was used for statistical calculations and preparation of figures. Fisher exact probability test was used to examine the significance of the contingency between different groups (CIN2 regression vs. CIN2 persistence/progression).

RESULTS

Study population

This study included 63 young women (average age 20.9, SD 2.3 years; range 16.3 – 25.0 years), with histologically confirmed CIN2 diagnosis at study entry (baseline). Women were followed on average for 19.2 months (range 4.9 – 34.4 months). At the end of the follow-up, 67% (42/63) of women showed CIN2 regression (CIN2 → CIN0), 20% (13/63) showed lesion persistence (CIN2 → CIN1 or 2) and 13% (8/63) presented with CIN2 progression (CIN2 → CIN3), respectively.

There were no differences in any of characteristics examined between women with CIN2 lesion persistence/progression vs. CIN2 regression (Table 1).

Table 1.

Selected characteristics of the study population

	All women (n = 63)	CIN2 persistence/progression (n = 21)	CIN2 regression (n = 42)	p-valuea
	N (%)	N (%)	N (%)
Race
White, non-Hispanic	27 (43)	5 (24)	22 (52)	p=0.057b
Black, non-Hispanic	16 (25)	6 (28)	10 (24)
Latin/Hispanic	13 (21)	8 (38)	5 (12)
Other race	7 (11)	2 (10)	5 (12)
Age
16 - 20	34 (54)	13 (62)	22 (52)	p=0.593
21 - 25	29 (46)	8 (38)	20 (48)
Alcohol consumption
ever	60 (95)	21 (100)	39 (93)	p=0.544
never	3 (5)	0 (0)	3 (7)
Drug usec
ever	53 (84)	16 (76)	37 (88)	p=0.279
never	10 (16)	5 (24)	5 (12)
Pregnant
ever	23 (37)	9 (43)	14 (33)	p=0.580
never	40 (63)	12 (57)	28 (67)
Current smoker
yes	17 (27)	5 (24)	12 (29)	p=0.770
no	46 (73)	16 (76)	30 (71)
Current use of hormonal contraception
yes	30 (48)	11 (52)	19 (45)	p=0.789
no	33 (52)	10 (48)	23 (55)
Reported sexually transmitted diseasesd
yes	20 (32)	9 (45)	11 (55)	p=0.251
no	43 (68)	12 (28)	31 (72)

^ap-value <0.05 (Fisher exact probability two-tailed test) indicated no significant differences for any of the variables listed between CIN2 persistence/progression compared to CIN2 regression group.

^bThere was a non-significant trend for higher number of non-Caucasian women in CIN2 persistence/progression group (p=0.057).

^cDrug use recorded for pot, hallucinogen or other drugs, with pot use being most common (96%), followed by pot/other drugs (49%), and pot/hallucinogen (32%), respectively.

^dSTI testing included detection of: Chlamydia trachomatis, Neisseria gonorrhoeae, Trichomonas vaginalis, Treponema pallidum, and Herpes simplex virus, respectively.

HPV Type(s) at the Baseline and Relation with CIN2 Regression vs. CIN2 Persistence/Progression

At baseline, 90% (57/63) of women were positive for HPV DNA of one or more HR-HPV types. Six (10%) women with CIN2 had no detectable HR-HPV type at the baseline or at any other CIN2 visit. Among CIN2 lesions positive for a single HR-HPV type (n=31), the most common type was HPV16 (n=12), followed by HPV31 (n=4), HPV18 (n=3), and HPV52 (n=3). Infections with pHR- or LR-HPV types occurred in 43% (26/61) of all HPV DNA+ cases.

HR-HPV persistence was more common in women with CIN2 persistence/progression compared to those with CIN2 regression (43% vs. 17%; p=0.027). HPV16 persistence was also more common among women with persistence/progression vs. those with regression (64% vs. 27%; p=0.099). All six women who were negative for any of the 12 HR-HPV types at the baseline had CIN2 regression. These results are summarized in Table 2.

Table 2.

HPV types in CIN2 persistence/progression vs. CIN2 regression group

HR-HPV type at the baseline	#CIN2 persistence/progression			#CIN2 regression			p-valuea
	All	HPVpersisted	HPVcleared	All	HPVpersisted	HPVcleared
HPV16	11	7 (64%)	4 (36%)	11	3 (27%)	8 (73%)	0.099
other HR-HPV	10	2 (20%)	8 (80%)	25	4 (16%)	21 (84%)	0.563
no HR-HPV	0	0	0	6	0	6 (100%)	NA
Total #cases	21	9 (43%)	12 (57%)	42	7 (17%)	35 (83%)	0.027

^aFisher exact probability test

TLR Expression in Women with CIN2 Regression vs. CIN2 Persistence/Progression

Women with CIN2 regression had significantly higher levels of TLR2 (p=0.006) and TLR7 (p=0.007) and showed a non-significant trend for a higher expression of TLR8 (p=0.053) compared to women with CIN2 persistence/progression (Figure 1A). We separated out the six women who were negative for HR-HPV DNA at their baseline visit. Each of these six women showed CIN2 regression and presented with significantly higher levels of TLR2 (p=0.005), TLR7 (p=0.013), and TLR8 (p=0.012), and a trend for a higher expression of TLR9 (p=0.110) compared to women with CIN2 persistence/progression (Figure 1B). TLR2 (p=0.018) and TLR7 (p=0.017) levels remained significantly higher among CIN2 regressors who were positive for HR-HPV at baseline compared to women with CIN2 persistence/progression (Figure 1B).

Figure 1.

A Expression of TLRs in CIN2 persistence/progression vs. regression group

B Expression of TLRs in CIN2 persistence/progression vs. regression group according to HR-HPV DNA genotyping results

DISCUSSION

Established cervical cancer precursors such as CIN2 have a regression rate of up to 70%, especially in young women ^{4, 27, 28}. Since the precise immune mechanisms by which such HPV-associated pre-cancers resolve are not well understood, we evaluated the role of TLRs (TLR2, 3, 7, 8 and 9) in the regression of CIN2. Our data suggest that high levels of TLR2, TLR7, and potentially TLR8 in cervical mucosa are important for CIN2 regression. As this study examined a single, baseline time point, it was not possible to define if it was the TLRs upregulation that lead to lesion regression or, TLRs downregulation that lead to lesion progression/persistence. The levels of TLRs 2, 7, 8, as well as TLR9, were even higher in the small subset of six women who were regressors, had CIN2 at the H&E histology, but were negative for any of the 12 HR-HPV types ¹ at the time point of CIN2 diagnosis. This finding underscores the importance of TLRs in viral clearance as suggested before ²¹, and necessity of viral recognition for the upregulation of these TLRs.

We previously showed that mRNA expression of protein-sensing TLR2, along with viral nucleic acid-sensing TLR3, TLR8, and TLR9 increases upon incident HPV infection (over a pre-infection baseline measurement) in women who show HPV16 clearance on 4-months follow-up compared with those who do not ²¹. This finding significantly correlated with increases in interferon-alpha2 in cervical lavage specimens, suggesting a key role for TLRs and downstream cytokines in viral clearance ²¹. As an extension of that incident study, we also examined the association of TLR expression in the clearance of HPV16 persistent infections and we found similar associations with elevated TLRs 3, 7, 8 and 9 but only in presence of measurable cell-mediated immune response to HPV16 E6 in the peripheral blood ¹³. This finding suggested an important link between innate and adaptive immunity in the control of HPV infections following periods of persistence ¹³. The important role of TLRs in HPV persistence and lesion progression is supported by in vitro studies demonstrating direct downregulation of TLRs by HPV types, specifically downregulation of TLR9 by HPV16 ²⁹ as well as studies suggesting that specific TLR polymorphisms may be genetic factors contributing to cervical cancer development ^30–33.

Mechanisms involving TLR expression and regression of CIN2 are likely complex. TLR activation results in expression of co-stimulatory molecules on dendritic cells and production of inflammatory cytokines that polarize T-helper (Th)1 immune responses to viral infections ³⁴. Our laboratory has previously demonstrated that a Th1 pattern (expression of interferon-gamma (IFN-gamma) mRNA and absence of IL-4 mRNA) precedes clearance of HPV infection in young women without an HPV-associated lesion supporting our in vivo findings ³⁵. The finding associated with TLR2 was of particular interest since its thought to be primarily involved in recognition of bacterial cell wall components. TLR2 has a critical role in expansion of regulatory T cells (Tregs) ^36–39 and previous studies from our group showed that CIN2/3 lesions had higher levels of mucosal Forkhead box P3 (Foxp3), a transcription regulator of Tregs, than samples with normal histology or CIN1 ⁴⁰. Therefore, in vivo studies incorporating other immune markers, including known inhibitory molecules such are Foxp3, PD-1 and PD-L1, would be informative.

Due to their anti-tumor potential, many TLR agonists have been studied in in vitro and animal studies, and developed either as monotherapy, as adjuvants to vaccination, or other therapeutic modalities ^{41, 42}. In the context of HPV, synthetic TLR7-agonist imiquimod is of particular interest for the treatment of HPV-related disease ¹⁴. The successful use of imiquimod in the treatment of genital warts was first demonstrated two decades ago ⁴³. Since then, application of imiquimod was also shown to be effective in the treatment of cervical, vaginal and vulvar intra-epithelial neoplasia as well ¹⁴.

However, TLRs seem to have complex roles in development of different cancer types including both tumor-suppressive and tumor-promoting effects ^{44, 45}. As mentioned, continuous upregulation of TLRs might result in the activation of several inflammatory and tumor-promoting pathways via NF-κB pathway ³⁴, including upregulation of Tregs, or PD-L1; immunosuppressants which are known to be upregulated in cervical pre-cancer and cancer ^{46, 47}. These concepts are supported by multiple studies demonstrating significantly higher expression of TLRs in cervical cancer cell lines and invasive cervical cancer biopsies compared to cervical dysplasia and/or normal cervix, respectively ⁴⁸. Interestingly, DeCarlo and colleagues have demonstrated that the upregulation of TLR2, 7, 8, and 9 in cervical cancer seem to be associated with stromal and not epithelial compartment, while in cervical precancerous lesions both stromal and epithelial compartments exhibited upregulation of most TLRs ⁴⁹.

Our study has few limitations. One of them is sample size. Because most women ≤25 years old are expected to clear their CIN2 lesion within 12 – 36 months observation period ^{4, 27, 28, 50} the number of women in the persistence/progression group was relatively small. Also, at the time of this study, p16INK4a immunostaining was not a standard practice for confirmation of CIN2 diagnosis on H&E, which is a less reproducible stand-alone diagnosis than CIN3 ⁵¹. The Lower Anogenital Squamous Terminology (LAST) Standardization Project for HPV-Associated Lesions, which recommends when p16INK4a immunostaining should be used in combination with histology to describe HPV-associated precancerous lesions, was only introduced in 2012 ²³. In addition, the source of TLRs (i.e. macrophages, dendritic, epithelial or other cell types) remains unknown in our study and it is possible that the proportion of dysplastic to normal cells might influence TLR expression. Unfortunately, identification of specific cell types was not possible. Therefore, an important step for future studies would be to define the source of TLR mRNA and whether the epithelial make-up of the samples influences TLR expression. Another limitation was the lack of suitable specimens in the parent study design for studying downstream cytokine protein levels (e.g., interferon alpha2) to document activation of these TLRs ²¹. Finally, this study examined a single baseline time point with many women requiring several years before lesion regression. Therefore, studying the expression of TLRs closer to the time point of lesion regression may be more informative.

To our knowledge, ours is the first study reporting on the association between cervical TLRs expression and CIN2 regression. Combined with our previous findings demonstrating the importance of TLRs in the clearance of incident HPV infections (specifically HPV16) in women without an established cervical disease, this study highlights the important protective role of the innate immune system at different stages of HPV infection and sequelae. Detailed understanding of host immune responses, as well as viral factors that influence whether a lesion progresses or regresses, is important to make the right choice of today’s available immunotherapeutics for HPV-associated disease ^{14, 52}. In addition, cytology and HPV testing are more powerful biomarkers of cervical lesion progression or regression when combined than as stand-alone biomarkers ^{53, 54} and a recent study suggested that measuring multiple clinical biomarkers might be helpful in accurately predicting regression of cervical lesions ⁵⁵. Our data suggest that measurement of TLR levels in cervical mucosa of women with HR-HPV DNA+ CIN2 lesions might be suitable in helping differentiate women who require only careful monitoring over time vs. those for whom excisional procedures would be the most beneficial early option, and support the potential use of TLR-agonists for treatment of these lesions.

Novelty: We found high levels of viral protein-sensing TLR2, and viral nucleic acid-sensing TLR7 and TLR8, measured at the time point of CIN2 diagnosis, to be associated with CIN2 regression over time. The highest levels of these TLRs were observed in cervical mucosa of women whose CIN2 lesions did not contain measurable HR-HPV infection. Our data support the use of TLR agonists for treatment of CIN2.

Abbreviations:

HPV

Human Papillomavirus

TLR

Toll-like receptor

CIN

Cervical Intra-epithelial Neoplasia