Cytokine Expression in the Cervical Stroma of HIV-Positive and HIV-Negative Women with Cervical Intraepithelial Neoplasia

Abstract

Cervical intraepithelial neoplasias (CINs) are a major public health issue. The prevalence of CINs is higher in women with the human immunodeficiency virus (HIV). The objective of this study was to determine whether there are differences in the immune responses in the cervical stroma of HIV-infected and -uninfected women with CIN. The responses were assessed according to the immunohistochemical expression of cytokines interleukin (IL)-4, IL-12, interferon gamma (IFN-γ), and tumor growth factor beta (TGF-β). In addition, we determined whether there were differences in the local immune responses between patients with CIN1 and CIN 2/3. A cross-sectional study was performed using material collected by cervical conization in HIV-infected and -uninfected women with CIN. The conization was performed using loop electrical excision procedure (LEEP) from January 1999 to May 2004. The evaluation of cytokines in the cervical stroma was based on immunohistochemistry. No differences were found between the two groups of women regarding HIV status. However, the associations between IL-12 expression and CIN 2/3 (p=0.016) in HIV-infected women and between IL-4 expression and CIN 1 (p=0.0456) in HIV-infected women were significant when the interaction between HIV infection and lesion grade was assessed. Additionally, a significant association between TGF-β expression and CIN 2/3 in both groups was observed regardless of HIV infection (p=0.000). An interaction between HIV infection and CIN grade was detected because IL-12 and IL-4 expression increased in the presence of HIV infection. Regarding the CIN grade, there was a high prevalence of TGF-β in CIN 2/3 lesions, which reflected the predominance of an immunoregulatory environment.

Introduction

Cervical carcinoma and its precursor lesions, cervical intraepithelial neoplasias (CINs), are a significant public health issue worldwide, and the human papillomavirus (HPV) is the primary etiological agent (22). However, infection with the virus may not lead to cervical cancer. Other factors are involved, such as genetic instability, the virus type, the host immune response, and co-infection with other infectious agents, including the human immunodeficiency virus (HIV) and Chlamydia trachomatis (25,29).

The cell-mediated immune response is the primary host defense mechanism against the development of cervical cancer. The strongest evidence regarding the role of the immune response in controlling HPV-induced lesions was found in studies that demonstrated a high prevalence of CIN in immunosuppressed patients, including patients with HIV and kidney transplant recipients (12,18).

High-grade CIN (CIN 2/3) is considered the precursor lesion to cervical cancer and should be eliminated by a successful immune response. This response is regulated by cytokines that are mainly expressed by T helper (Th) cells. Immunostimulatory cytokines are expressed by Th1 cells and include interferon gamma (IFN-γ), TNF-α, interleukin (IL)-2, and IL-12. Immunoinhibitory cytokines are expressed by Th2 cells and include IL-4, IL-5, IL-6, IL-8, IL-10, and IL-13 (10,26). An additional subtype of T-cells, known as regulatory T-cells (Tregs), secretes TGF-β, which can suppress T lymphocytes and the production of Th1 cytokines. This mechanism alters the immune response (2).

It is unclear whether HIV/HPV co-infection affects cytokine synthesis and whether this co-infection could contribute to the development of squamous intraepithelial lesions. A study by Crowley-Nowick et al. demonstrated that HIV alters the expression of local immune constituents, possibly increasing the virulence of HPV in HIV-infected patients (11).

This study aimed to evaluate the local immune responses in HIV-infected and -uninfected patients with CIN 1 and CIN 2/3 by analyzing the expression of cytokines.

Materials and Methods

Design

This study was a cross-sectional analysis developed from a previous prospective study that evaluated the risk factors for CIN recurrence in HIV-infected and -uninfected women (15). According to the results of previous cervical biopsies, women were submitted to LEEP for the treatment of CIN, and paraffin blocks of the surgical specimens were selected for this analysis.

The main inclusion criterion was the presence of CIN in the paraffin block. Blocks without epithelial representation and poor quality slides (without epithelial or stromal representation or with folded epithelium that precluded a cell count) were excluded.

Two variables were defined for the outcome: HIV and CIN (in the LEEP specimen). The predictor variables were as follows: age, parity, tobacco smoking, age at initiation of sexual activity, cytology, colposcopy, anatomopathological examination of the biopsy sample, and cytokines. Colposcopy was categorized in two grades: minor changes (grade 1), when was present a delicate acetic-white epithelium, a delicate mosaic, or a delicate puncturing; and major changes (grade 2), in the presence of intense acetic-white epithelium, a coarse mosaic, a prominent excretory ducts of glands, a sharp border, or a rapid acetic acid reaction. In addition, CIN was considered an explanatory variable for the HIV outcome, and HIV was considered an explanatory variable for the CIN outcome.

Laboratory methods

The immunohistochemistry technique was used to evaluate the cytokines. For this technique, 4 μm sections were prepared and mounted on silanized slides. Deparaffinization was performed, followed by rehydration in 95%, 70%, and 50% alcohol baths, antigen retrieval using a Pascal Pressure Cooker for 6 minutes at a temperature of 98°C, an endogenous peroxidase block (1.5% hydrogen peroxide solution in methanol), and biotin-avidin block (milk powder-based solution). After the blocking treatment, the slides were incubated overnight at 4°C with the specific primary antibody (NCL TGF-β: Novocastra, Leica Biosystems Ltd., lot# 6001375; IL 12 p70 (14 L7): Santa Cruz Biotechnology, lot# 62409; IL-4: NyRhIL4, sc 73317, monoclonal mouse (IgG2a); IFN-γ: monoclonal mouse, Santa Cruz Biotechnology). The secondary antibody (4 Mouse Probe, UP534 L, Biocare Medical) was applied and incubated at room temperature for 30 minutes. The polymer (Mach 4 HRP Polymer, MRH534 L, Biocare Medical) was applied after washing the slides in Tris-buffered saline with Tween (TBST), and the development was performed using the Diaminobenzidine (DAB) chromogen. Slide counterstaining was performed with hematoxylin solution, followed by dehydration using an alcohol bath and diaphonization in xylene solution and mounting on coverslips. Tonsil samples were used as a positive control. The primary antibody was omitted in the negative controls.

We counted the cells localized in the stroma—just below the epithelium with CIN—that were positive (staining) for the cytokines analyzed. An objective counting was made in three consecutive fields as demonstrated in Figure 1. Only one experienced observer was responsible by this counting to avoid variability. All samples were analyzed without patients' identification. The Olympus BX 41 microscope, with a magnification of 400× (40× objective and 10× optical lens), was used for cell counting. Each field was 0.230 μm×0.310 μm in size.

FIG. 1.

Immunohistochemistry. Stained cells in the stroma, just below the epithelium with cervical intraepithelial neoplasia (CIN), showing the presence of specific cytokines. (A) Tumor growth factor beta (TGF-β); (B) interferon gamma (IFN-γ); (C) interleukin (IL)-12; and (D) IL-4. Olympus BX 41 microscope, with a magnification of 400× (40× objective and 10× optical lens). Color images available online at www.liebertpub.com/vim

The study was approved by the research ethics committee of the Federal University of Minas Gerais, and all participants signed an informed consent form.

Statistical analysis

The statistical analysis was performed using R software v2.13.1. The significance level was set at 5%. Initially, a descriptive analysis of the groups was performed. Next, a univariate analysis was performed using the Mann–Whitney U-test and a chi-square test, and graphs were created to assess the possible interactions between the variables. A logistic regression was used for the multivariate analysis with the variables that were selected as predictors in the univariate analysis (p<0.20), and a 5% significance level was used for the final analysis. The Hosmer–Lemeshow test was performed to assess the goodness of fit of the logistic regressions.

We tested interactions among cytokines in the presence of both HIV and CIN in the final model of analysis in order to investigate if there were any association between these variables that could be related to a different expression of cytokines investigated in this study. Usually, an interaction among variables produces a synergistic effect modification that could potentiate or reduce the isolated association of each one of these variables. It means that in the presence of HIV and a specific lesion grade, the expression of cytokines could be different in comparison with the expression found with every one of those variables singly.

Results

A total of 85 paraffin blocks were included in the study: 47 blocks from HIV-infected women and 38 blocks from uninfected women. Overall, 39 out of 85 blocks were from patients with CIN 1, and 46 blocks were from patients with CIN 2/3. Regarding HIV status, the groups were homogeneous according to age (p=0.70), parity (p=0.21), age at onset of sexual activity (p=0.77), and tobacco smoking (p=0.76). Regarding CIN grade, there were no differences between groups according to age (p=0.42), parity (p=0.25), age at onset of sexual activity (p=0.72), and tobacco smoking (p=0.45).

Table 1 shows that there were no significant differences between the groups according to HIV status, the referral cytology, and the results from the examination of LEEP specimens. However, there was a significant predominance of CIN 2/3 lesions in biopsies before LEEP (81.6%) among the HIV-uninfected women (p=0.000). In addition, worsening prognoses were observed among HIV-infected women.

Table 1.

A Comparison of the Cytology, Colposcopy, and Histopathologic Results from the Previous Biopsies and LEEP According to HIV Status

	HIV infected (n=47)		HIV uninfected (n=38)
Variables	n	%	n	%	p-Value ^*	OR	95% CI
Cytology
Normal	18	38.3	9	23.7	0.27^**	1	0.22–1.52
Altered	29	61.7	25	76.3		0.58
ASCUS			4	10.5
Colposcopy
Grade 1	33	70.2	7	18.4	0.000	1	0.03–0.27
Grade 2	14	29.8	31	81.6		0.10
Previous biopsy
CIN 1	37	78.7	7	18.4	0.000	1	0.02–0.16
CIN 2/3	10	21.3	31	81.6		0.52
Lesion grade at LEEP
CIN 1	25	53.2	14	36.8	0.133	1	0.21–1.23
CIN 2/3	22	46.8	24	63.2		0.51

Pearson's chi-square test.

p-Value, excluding the cases of ASCUS.

LEEP, loop electrical excision procedure; HIV, human immunodeficiency virus; ASCUS, atypical squamous cells of undetermined significance; OR, odds ratio; CI, confidence interval.

No significant associations were found comparing HIV infection and the expression of cytokines IFN-γ (p=0.472), TGF-β (p=0.791), IL-4 (p=0.287), and IL-12 (p=0.842). However, a significant difference was found for TGF-β (p=0.003): there were significantly more TGF-β positive cells in women with CIN 2/3 when the cytokines and lesion grade were compared to the HIV status of patients. There were no significant differences between groups regarding the other cytokines that were tested (Table 2).

Table 2.

Comparison of Median (Range, n) of Local Cytokines (IFN-γ, TGF-β, IL-4, IL-12) and the Degrees of Lesion of HIV-Infected and Uninfected Women

	CIN 1		CIN 2/3
	HIV		HIV
	Infected (+)	Uninfected (−)	Infected (+)	Uninfected (−)
IFN-γ	6.0 (0–25, 22)	4 (0–42, 13)	9.0 (0–43, 23)	4 (0–32, 24)
TGF-β	18.0 (5–39,23)^*	14.5 (10–35, 14)^**	31 (14–112, 23)	32 (13–61, 24)^#
IL-4	1.0 (0–9, 19)	0.0 (0–11, 14)	1.0 (0–23, 22)	1.0 (0–29, 23)
IL-12	0.0 (0–12, 23)^***	0.5 (0–12, 14)	2.0 (0–11, 23)	1.0 (0–11, 23)

Mann–Whitney U-test: some cases were excluded due to inadequate material for analysis.

p<0.001 in comparison with CIN 2/3HIV+; ^** p<0.001 in comparison with CIN 2/3HIV−; ^*** p=0.031 in comparison with CIN 2/3HIV+; ^# p<0.003 in comparison with CIN 1HIV− and with CIN 1HIV+.

The logistic regression model was used in the statistical modeling with the potential predictors from the univariate analysis, and a p-value of<0.20 was used as a selection criterion. In addition, interactions were considered in the assessment of cytokine responses according to lesion grade and HIV status. Regarding the final model of HIV status, significant values for IL-4 and CIN 1 (p=0.0456), IL-12 and CIN 1 (p=0.0192), and IL-12 and CIN 2/3 (p=0.0163) were obtained (Table 3). Regarding the final model of lesion grade, the cytokine TGF-β was the only significant variable (p=0.000; odds ratio [OR]=1.126; confidence interval [CI 1.064–1.192]; data not shown).

Table 3.

A Logistic Regression Analysis of HIV Status: The Final Model

Source	p-Value^*	OR	95% CI
CIN 2/3	0.3652	0.572	0.171–1.917
IL-4^**	0.0456	1.692	1.010–2.834
IL-12^**	0.0192	0.526	0.307–0.900
CIN 2/3×IL-4	0.0515	0.594	0.351–1.003
CIN 2/3×IL-12	0.0163	2.032	1.139–3.624

Hosmer–Lemeshow (p-value=0.598).

Regarding IL-4 and IL-12, the results are related to patients with CIN 1.

Discussion

The immune response to CINs, HPV, and HIV is not fully understood. Studies have demonstrated that even after the introduction of highly active antiretroviral therapy (HAART), which restores CD4-positive T lymphocyte counts to normal values, there was an increased incidence of cervical lesions related to HPV. This finding suggests that other factors affect the action of HPV (8,19,20). Recently, Strickler et al. demonstrated that high Treg levels may be significantly associated with oncHPV persistence in both HIV-positive and HIV-negative women (27).

The findings in the literature regarding the profile of cytokines that are found in cervical lesions are controversial, mainly due to the different types of samples that have been tested (biopsy, lavage cytology, surgical specimen, and blood) and the research techniques that were used (real-time polymerase chain reaction [RT-PCR] and immunohistochemistry). Moreover, blood samples analysis revealed the systemic profile of immune response, which may not match the local microenvironment below the lesion. This result was obtained by Kobayashi et al. when cervical and blood samples from the same patients were assessed (14). These authors found higher proportions of T-cells that expressed IFN-γ and TGF-β in cervical tissue than in blood (14).

We found a significantly increased number of cells positive for TGF-β in cervical tissue samples with CIN 2/3 than in CIN 1 samples in both HIV-infected and -uninfected women (p<0.003). However, we did not find any significant differences between groups when we compared women according to HIV serostatus without considering lesion grade. Kobayashi et al. found increased TGF-β stromal density in infiltrates that were adjacent to CIN 2/3, especially in HIV-uninfected patients, when compared with normal cervix (13). Additionally, they found increased TGF-β density in HIV-infected patients with CIN 2/3, albeit at lower concentrations than in HIV-uninfected patients with CIN 2/3 (13). Increased TGF-β stromal density was observed in HIV-infected women with CIN 2/3 in a recent study by Quaresma et al. (6). In addition, Baritaki et al. found that the increased expression of TGF-β was associated with the progression from low-grade CIN to high-grade CIN in immunocompetent patients, which is in agreement with the results of our study (4). TGF-β belongs to a family of multifunctional growth factors that regulate basic cellular functions, including apoptosis, proliferation, differentiation, and immunosuppression (24). The loss of responsiveness to TGF-β is a special biological characteristic that is associated with tumor progression. According to Massagué, there is often a loss of antiproliferative activity by TGF-β in cancer, and this cytokine may even have proliferative activity (16). The higher TGF-β values that we found in the tissue samples from CIN 2/3 lesions may indicate an increased resistance and/or the desensitization of TGF-β receptors. Additionally, these findings could indicate a greater proportion of regulatory T-cells in this type of injury, which leads to a favorable environment for disease progression.

The expression of Th1-type cytokines may be the key strategy in directing an appropriate immune response in the cervical mucosa, whereas a Th2-type response may be ineffective at preventing the progression to a high-grade lesion (23). The Th1 profile was assessed in this study through the expression of IFN-γ and IL-12.

In this study, no differences were observed in the expression of IFN-γ and IL-12 between groups of HIV-infected and uninfected women, and no differences were observed regarding lesion grade. However, a significant association was found between IL-12, HIV infection, and high-grade CIN (p=0.016) when interaction between CIN, HIV, and local immune responses was evaluated. IL-12 is a key cytokine for the induction of cell-mediated immune responses and is secreted by dendritic cells, macrophages, and B-lymphocytes. IL-12 promotes the differentiation of T-cells into IFN-γ-secreting Th1 cells (5) and induces IFN-γ secretion by NK cells (7). Our finding is in agreement with that of Moscicki et al. who found increased IL-12 levels in HIV-positive patients who developed high-grade lesions (17). In addition, Crowley-Nowick et al. reported that HIV/HPV co-infection was associated with high levels of IL-12-secreting cells, especially in women with a third sexually transmitted disease compared with women infected with HIV or HPV alone, which suggests an interaction (11).

IL-12 inhibits angiogenesis (28). According to Voest et al., the suppression of angiogenesis by IL-12 is dependent on its ability to induce the expression of IFN-γ (28), which was observed in our study. However, the expression was not significant in HIV-infected women with CIN 2/3. The increased levels of IL-12 in our study reflect the role of IL-12 in the inhibition of angiogenesis and epithelial proliferation. Additionally, the increased IL-12 levels indicate a shift to a Th1 response, thereby preventing the progression to cancer.

Th2 profile was evaluated according to IL-4 expression, which suppresses the cellular immune response and preferentially stimulates the humoral response (30). There were no significant differences between IL-4 expression and HIV status in our study. However, we found an association between HIV-infected women with CIN 1 and an increased expression of IL-4 (p=0.0456) in the multivariate analysis, and this group tended to develop a Th2 pattern. This result was not found in women who were uninfected with HIV, which indicates that there is most likely an interaction between HIV and lesion grade. Other authors have evaluated the expression of IL-4 using different techniques and materials and found no differences between the CIN grades (3,7). In contrast, other authors have described an increased expression of IL-4 in patients with CIN 2/3, which indicates an increasingly predominant Th2 response as the lesion worsens (1,9).

The immune response is a complex event that involves several factors. This response may differ even within each group of low- or high-grade lesions depending on which type of HPV is involved. However, HPV type was not assessed in this study. A prospective (cohort) study is needed to assess the immune responses in women with CINs according to cytokine levels and to determine whether there is a correlation between the immune responses and HPV type.

The results of this study suggest that there is an immunoregulatory environment in CIN 2/3 lesions with a predominance of TGF-β. Regarding HIV, we observed an interaction between HIV and CIN grade with a predominance of Th1 responses in women with CIN 2/3 and a predominance of Th2 responses in women with CIN 1. Data in the literature are scarce, especially regarding local immune response in HIV-positive patients. This study is the first to evaluate the local immune responses in a significant number of HIV-positive women in Brazil according to cytokine expression. However, future studies are needed with greater coverage of the local immune response to understand this response better, especially in HIV-infected women.

Footnotes

Acknowledgments

We would like to thank CNPq (National Council of Technological and Scientific Development) and FAPEMIG (Fundação de Amparo à Pesquisa do Estado de Minas Gerais) for the institutional support.

Author Disclosure Statement

No competing financial interests exist.

References

al-Saleh

, Giannini

, Jacobs

, et al. Correlation of T-helper secretory differentiation and types of antigen-presenting cells in squamous intraepithelial lesions of the uterine cervix. J Pathol, 1998; 184:283–290.

Alcocer-González

, Berumen

, Taméz-Guerra

, et al. In vivo expression of immunosuppressive cytokines in Human Papillomavirus-transformed cervical cancer cells. Viral Immunol, 2006; 19:481–491.

Bais

, Beckmann

, Lindemans

, et al. A shift to a peripheral Th2-type cytokine pattern during the carcinogenesis of cervical cancer becomes manifest in CIN III lesions. J Clin Pathol, 2005; 58:1096–1100.

Baritaki

, Sifakis

, Huerta-Yepez

, et al. Overexpression of VEGF and TGF-beta1 mRNA in Pap smears correlates with progression of cervical intraepithelial neoplasia to cancer: implication of YY1 in cervical tumorigenesis and HPV infection. Int J Oncol, 2007; 31:69–79.

Brunda

. Interleukin-12. J Leukoc Biol, 1994; 55:280–288.

Carneiro

, Pacheco

, Xavier

, and Quaresma

. Tissue expression of TGF-β₁ in uterine cervical samples from HIV/AIDS patients. Microb Pathog, 2012; 53:44–48.

Chan

, Perussia

, Gupta

, et al. Induction of interferon gamma production by natural killer cell stimulatory factor: characterization of the responder cells and synergy with other inducers. J Exp Med, 1991; 173:869–879.

Chaturvedi

, Madeleine

, Biggar

, and Engels

. Risk of human papillomavirus-associated cancers among persons with AIDS. J Natl Cancer Inst, 2009; 101:1120–1130.

Clerici

, Merola

, Ferrario

, et al. Cytokine production patterns in cervical intraepithelial neoplasia: association with human papillomavirus infection. J Natl Cancer Inst, 1997; 89:245–250.

10.

Clerici

, Shearer

, and Clerici

. Cytokine dysregulation in invasive cervical carcinoma and other human neoplasias: time to consider the TH1/TH2 paradigm. J Natl Cancer Inst, 1998; 90:261–263.

11.

Crowley-Nowick

, Ellenberg

, Vermund

, Douglas

, Holland

, and Moscicki

. Cytokine profile in genital tract secretions from female adolescents: impact of human immunodeficiency virus, human papillomavirus, and other sexually transmitted pathogens. J Infect Dis, 2000; 181:939–945.

12.

Frisch

, Biggar

, Engels

, and Goedert

. Association of cancer with AIDS-related imunosupression in adults. JAMA, 2001; 285:1736–1745.

13.

Kobayashi

, Greenblatt

, Anastos

, et al. Functional attributes of mucosal immunity in cervical intraepithelial neoplasia and effects of HIV infection. Cancer Res, 2004; 64:6766–6774.

14.

Kobayashi

, Weinberg

, Darragh

, and Smith-McCune

. Evolving immunosupressive miroenvironment during human cervical carcinogenesis. Mucosal Immunol, 2008; 1:412–420.

15.

Lima

, Tafuri

, Araújo

, de Miranda Lima

, and Melo

. Cervical intraepithelial neoplasia recurrence after conization in HIV-positive and HIV-negative women. Int J Gynaecol Obstet, 2009; 104:100–104.

16.

Massagué

. TGF-beta in Cancer. Cell, 2008; 134:215–230.

17.

Moscicki

, Ellenberg

, Crowley-Nowick

, Darragh

, Xu

, and Fahrat

. Risk of high-grade squamous intraephitelial lesion in HIV-infected adolescents. J Infect Dis, 2004; 190:1413–1421.

18.

Moscicki

, Ellenberg

, Vermund

, et al. Prevalence of and risks for cervical and human papillomavirus infection and squamous intraepithelial lesions in adolescent women: impact of infection with human immunodeficiency virus. Arch Pediatr Adolesc Med., 2000; 154:127–134.

19.

Palefsky

. Human papillomavirus-related disease in people with HIV. Curr Opin HIV AIDS, 2009; 4:52–56.

20.

Palefsky

. Cervical human papillomavirus infection and cervical intraepithelial neoplasia in women positive for human immunodeficiency virus in the era of highly active antiretroviral therapy. Curr Opin Oncol, 2003; 15:382–388.

21.

Pardo-Govea

, Callejas

, Núñez-Troconis

, et al. Gamma interferon (IFN-gamma), tumor necrosis factor alpha (TNF-alpha) and interleukins 2, 4 and 6 (IL-2, IL-4, IL-6) in cervical-uterine cells of intraepithelial neoplasia: a preliminary report. Invest Clin, 2005; 46:5–13.

22.

Psyrri

, and DiMaio

. Human papillomavirus in cervical and head-and-neck cancer. Nat Clin Pract Oncol, 2008; 5:24–31.

23.

Scott

, Stites

, Moscicki

. Th1 cytokine patterns in cerviacl human papillomavirus infection. Clin Diagn Lab Immunol, 1999; 6:751–755.

24.

Shi

, Massagué

. Mechanisms of TGF-beta signaling from cell membrane to the nucleus. Cell, 2003; 113:685–700.

25.

Smith

, Bosetti

, Muñoz

, et al. Herpes Herrero R simplex virus-2 as a human papillomavirus cofactor in the etiology of invasive cervical cancer. J Natl Cancer Inst, 2002; 94:1604–1613.

26.

Spellberg

, and Edwards

Jr . Type 1/type 2 immunity in infectious diseases. Clin Infect Dis, 2001; 32:76–102.

27.

Strickler

, Martinson

, Desai

, et al. The relation of plasmacytoid dendritic cells (pDCs) and regulatory T-cells (Tregs) with HPV persistence in HIV-infected and HIV-uninfected women. Viral Immunol, 2014; 27:20–25.

28.

Voest

, Kenyon

, O'Reilly

, Truitt

, D'Amato

, and Folkman

. Inhibition of angiogenesis in vivo by interleukin 12. J Natl Cancer Inst, 1995; 87:581–586.

29.

Wallin

, Wiklund

, Luostarinen

, et al. A population- based prospective study of Chlamydia trachomatis infection and cervical carcinoma. Int J Cancer, 2002; 101:371–374.

30.

Zhu

, and Paul

. CD4 T cells: fates, functions, and faults. Blood, 2008; 112:1557–1569.