T- and NK-cell populations with regulatory phenotype and markers of apoptosis in circulating lymphocytes of patients with CIN3 or microcarcinoma of the cervix: evidence for potential mechanisms of immune suppression

According to generally acknowledged conception, under the influence of cancer cells, innate mechanisms that maintain immune cell homeostasis and control the magnitude of immune responses can undergo dramatic dysregulation which leads to suppression of the immune system and enables the tumor to escape from immune attack and to progress. Regulatory pathways that mediate interactions between different functional types of immune cells during tumor development are currently under extensive investigation, and various in vitro and in vivo model systems for studying these interactions at the molecular and cellular levels are employed. Close attention is now being paid to lymphocytes that belong to adaptive or innate immunity and exhibit immunoregulatory properties; these lymphocytes include regulatory CD4/CD8 T cells (Treg), Th17/Th22 cells, γδT cells, regulatory B cells, regulatory NK cells (NKreg), NK-like T cells, and some other cell subsets. Significant progress has been made in the study of phenotypic characteristics of these diverse cell types and their impact on the activity of effector cells using experimental models. On the other hand, this progress highlights the need for additional studies to obtain evidence that carcinogenesis-related mechanisms mediated by these cells are physiologically relevant and can lead to dysregulation of immune homeostasis in cancer patients. Another, no less important issue is the ability of a tumor to exert not only local effects on distribution and activity of regulatory and effector cells throughout its entire development, but systemic effects as well, thus creating favorable conditions for growth both within the tumor microenvironment and at the level of “macroenvironment” [1]. Indeed, numerous studies have revealed dramatic changes in frequencies of the above-mentioned regulatory cells or other suppressor subsets (like M2-like macrophages and myeloid-derived suppressor cells, or MDSC) infiltrating tumor tissue. Altered expression of molecular markers indicating functional impairments in infiltrating cytotoxic cells has also been characterized in many cancer types, including those of epithelial origin. These data allow researchers to reconstruct local regulatory networks formed by immune cells and neoplastic cells at different stages of tumor progression. At the same time, there is a growing body of evidence showing alterations in the number and the phenotype of lymphocytes in blood, lymph nodes, and other metastatic sites in cancer patients, these findings demonstrating that the processes of systemically induced immune suppression can vary depending on tumor etiology, location, histologic type, and stage of development. Of particular interest are the changes in the immune system that occur in early phases of tumor progression, including precancerous lesions, because studying these changes provides an opportunity to evaluate the role of different immune cell populations in creating conditions for transition of premalignant lesions into invasive cancer. However, for many types of cancer such studies may be hampered by inability to diagnose precancerous conditions and early-stage cancer and, therefore, information on specific mechanisms maintaining immunodeficiency in patients is still lacking.

The cross-talk between Tregs and effector T cells has been recognized as a crucial factor in tumor development, with suppressive capacity of Tregs being realized in different ways, including inhibition of proliferation and activity of effector T cells, or induction of T cell apoptosis [2]. Cancer cells have the ability to promote expansion of CD4 Tregs population, which in turn, due to FasL/CD95L expression, are able to stimulate apoptosis in effector T lymphocytes via CD95-dependent pathway [2]. At the same time, Tregs are known to be more resistant to cell death because of elevated expression of endogenous anti-apoptotic molecules. Additionally, tumor cells and, as recently discovered, endothelial cells of tumor microvasculature can upregulate FasL/CD95L, contributing to higher apoptotic rates of activated CD95⁺ Т lymphocytes [3] and exhaustion of effector immune cells [4]. As illustrated by studies in patients with invasive hepatocellular carcinoma [5] and head and neck cancer [2], manifestations of this cross-talk can be detected in populations of peripheral blood lymphocytes, however, little is known about the extent they can reach during early stages of carcinogenesis. CD8 Tregs represent another cell type that has been recently recognized as the crucial component of the population of immunosuppressive Tregs. These cells were demonstrated to have similar capacity to suppress effector functions of T cells as CD4 Tregs, but in contrast to CD4 Tregs, their exact role in the regulation of antitumor immunity (including modulation of apoptosis) remains essentially unstudied. Lastly, the control over the number and the activity of various effector immune cells can be exercised by regulatory cells of innate immunity, as for example cytokine-producing NKregs, which involvement in pathogenesis of oncological diseases, especially at early stages, remains one of the least investigated issues [6, 7].

In light of the problem of immune regulatory mechanisms that maintain early stages of tumor progression, there is growing interest to the study of cervical cancer (CC). In its development, invasive carcinoma of the cervix goes through several morphologically distinct stages – cervical intraepithelial neoplasia (CIN) grade 1, 2, and 3 (including cancer in situ), and microinvasive cancer. The establishment of cervical neoplasia is mostly associated with high-risk papillomaviruses (HPV) that have evolved multiple strategies to evade and/or dysregulate host immunity mechanisms thus preventing clearance of infection [8]; in its turn, a host immune system failure to develop an effective response allows HPV to persist and increases the probability of malignant progression [9]. Pre-invasive and early-stage invasive cancers account for the largest number of cervical cancer cases, thus making possible comparative studies on immune parameters in the continuum of the cancer progression. A considerable amount of data has been amassed to date on molecular and cellular features altering tumor microenvironment upon CIN and СС development, such as dysregulated expression profile of cytokines or other immune-related markers, altered content and distribution of tumor-infiltrating lymphocytes, including cytotoxic cells and Tregs, with high-risk HPV considered to be a major driver of these events. In the case of invasive CC, alterations in the immune cell composition of tumor-positive and tumor-negative lymph nodes have also been reported, indicating early formation of a tolerogenic/immunosuppressive milieu (pre-metastatic niches) that precedes lymphatic spread of tumor cells [10–12]. Furthermore, there are some studies providing evidence that specific changes in immune parameters can occur in the systemic circulation of HPV-positive individuals and patients with CIN1–3 or CC. For example, peripheral blood T lymphocytes obtained from women with CIN1 were found to exhibit impaired ability to become activated in response to ex vivo stimulation with HPV antigens [13]. Several researchers have documented elevated frequency of circulating CD4 lymphocytes expressing Treg markers in blood samples of women with persistent HPV, CIN or CC [14–16]. Other studies have also identified that patients with CIN1/2 and CIN3 display altered numbers of some rare populations of circulating lymphocytes, which exert immunoregulatory/proinflammatory functions, such as Tc17 [17], Th17 and Th22 [18], NK-like CD8 cells [19], and CD4⁺NKG2D⁺ cells [20]. Deregulated expression of several activation markers and co-stimulatory/inhibitory molecules (such as NKG2D, CD28, and PD-1) was also described in peripheral blood lymphocytes from women with invasive cervical carcinoma or preinvasive lesions [21–23]. Additionally to immunoregulatory cell populations, some important changes in the levels of related cytokines (for example, IL-4, −6, −8, −10, −17, −22, −23, TNFα, IFNγ and other) have been revealed in the serum of CIN/CC patients [16, 18, 19], with TGFβ supposed to be one of the key factors operating the mechanisms by which regulatory lymphocytes are induced and can exert their suppressive effect [16, 23].

It is assumed that systemic abnormalities seen in patients with strictly local cervical lesions may arise as a consequence of persistent antigenic stimulation during chronic viral infection and facilitate neoplastic progression [22]; however there are still insufficient data to unequivocally confirm this assumption. Another question is that the observed shifts in the profile of regulatory and effector immune cells can be brought about by activation of apoptosis-related processes with engagement of Tregs as inducers of CD95-mediated lymphocyte death, but physiological significance of this relationship in cervical carcinogenesis is much less understood. In this work we focus on studying preinvasive and microinvasive cervical cancer as the earliest stages in disease development. We analyzed changes in the frequencies of peripheral blood CD4⁺/CD8⁺ regulatory cells and NKregs in parallel with assessment of TGFβ serum concentration, CD95 expression and the level of apoptosis in T cells, with the aim to determine: 1) whether up-regulation of apoptosis-associated markers in major Т-lymphocyte subsets accompanies CC development and correlates with the change in percentages of regulatory cell populations of adaptive and innate branches of immunity, and 2) whether these changes may extend to systemic level of immunity during the initial stages of invasive cervical cancer progression.

Among known populations of suppressor cells, CD4 Tregs and their functional role in carcinogenesis are the most extensively studied; for example in the case of cervical cancer and its precursor lesions, a substantial body of research has convincingly demonstrated accumulation of CD4 Tregs in the tumor site. Based on these studies, mechanisms through which CD4⁺ cells can undergo differentiation into Tregs under the influence of tumor microenvironment, with assistance of dendritic cells and the stromal component, have been proposed. Fewer data are available on whether these processes can spread to systemic level; nevertheless, importantly, most investigators point to the fact that the increase in the level of peripheral blood CD4 Tregs is an early event in cervical cancer development, occurring at the stage of CIN2–3 [15, 16, 30], and is probably linked with the persistence of high-risk HPV infection [14]. Our data showing increased numbers of lymphocytes co-expressing Treg markers in the circulation of women with preinvasive and microinvasive cancer are in agreement with these findings. Therefore, one can expect pathological alterations that reflect the functioning of immunosuppressive mechanisms triggered by expansion of CD4 Treg population would also be detectable in the peripheral circulation during early stages of cervical cancer progression. An example to support this assumption is the study by Visser and co-authors who employed ex vivo system to demonstrate that CD4⁺CD25^high Tregs can inhibit IFNγ production by stimulated HPV16 E6/E7-specific Т cells isolated from peripheral blood of CIN and cervical cancer patients [15]; nonetheless the spectrum of relationships between CD4⁺Tregs and their potential target cell populations formed upon CIN/CC development remains largely uncharacterized.

Apart from inhibiting the functional activity of effector cells, ability to induce lymphocyte apoptosis was recognized as another crucial mechanism whereby tumors can antagonize antitumor immune responses. Through upregulation of CD95L/FasL, peripheral blood CD4 Tregs were also able to trigger apoptotic CD95/Fas-dependent death of autologous CD8⁺ Т effectors in a co-culture system, as was demonstrated in head and neck squamous cell carcinoma [2]. Accordingly, if helper/cytotoxic T lymphocytes upregulate their CD95 membrane expression, then expansion of Tregs in blood of cancer patients might result in enhanced levels of spontaneous apoptosis in T cells. Such relationships have been confirmed in some types of human cancer: for example, a coordinated increase in the number of CD95L⁺ Tregs and the level of Annexin V binding with CD95⁺ cytotoxic Т lymphocytes has been reported in the peripheral blood of hepatocellular carcinoma patients [5]. Another study published by Hoffmann and co-authors revealed significantly elevated frequencies of circulating CD3⁺CD95⁺ T cells that also displayed a higher level of Annexin binding in head and neck cancer (accounting for up to 100% in advanced disease), but a correlation between these markers of apoptosis and Treg numbers was not analyzed [4]. According to our findings, an increase in CD95 expression occurs in peripheral blood T cells in early-stage CC, which may represent a consequence of long-term exposure of T lymphocytes to virus-specific or/and tumor-associated antigens and may also reflect systemic activation status of the immune system over the course of HPV infection and progression of neoplasia. In one third of CC patients examined, the proportion of CD3⁺CD95⁺ cells stained positively for Annexin V exceeded the range of values defined for the healthy control group, suggesting that an elevated level of T-cell apoptosis is likely due to activation of CD95-dependent pathway. Notably, a strong association between the proportions of CD3⁺Anx⁺CD95⁺ cells and FoxP3⁺ lymphocytes was revealed in the CIN/CC patient group. Although the existence of such correlation cannot be appreciated as direct evidence of causal relationships, it nevertheless underpins the possibility that Tregs may execute their suppressive function through activation of T-cell apoptosis in preclinical stages of cervical carcinoma.

From the analysis of data on differential changes in the expression of CD95 and the level of apoptosis in CD4 and CD8 subsets of peripheral blood T cells available from published sources, it can be inferred that these changes exhibit different patterns depending on the type of cancer. For example, Xu and colleagues reported reduced expression of CD95 (as well as some other activation- and cytotoxicity-related markers) by circulating CD3⁺CD8⁺Т lymphocytes from patients with advanced lung cancer and proposed this reduction might be linked to impairment of killer T-cell function [31]. On the contrary, patients with gastric [32] and breast cancer [33] showed up-regulation of CD95 expression in peripheral blood CD8 Т cells, while head and neck cancer patients displayed increased numbers of CD95⁺ cells in both CD4 T helper and CD8 T killer subsets [4]. In blood samples from patients with preinvasive and microinvasive cervical cancer tested our study, we could observe significantly higher proportion of CD4⁺CD95⁺ Т lymphocytes with a concomitant tendency for decreased percentage of CD8⁺CD95⁺ Т lymphocytes compared to healthy controls. Since this decrease was not accompanied by an increase in the amount of CD8⁺CD95⁻ T cells, a plausible explanation for this observation may be the higher rate of apoptosis in CD8⁺ Т lymphocytes resulting in the more rapid elimination. Indeed, as follows from most studies, in which apoptosis-related markers were measured in CD4⁺ and CD8⁺ subsets of circulating T cells of cancer patients, CD8⁺ lymphocytes are more susceptible to apoptosis [2, 4, 32]. Although in the present study we did not compare the level of Annexin V binding between CD4 and CD8 T lymphocytes, we can speculate that the same feature could be responsible for the observed in CIN/CC group correlation between the relative abundance of Anx⁺CD95⁺ T cells and the proportion of CD8⁺CD95⁺ T cells, which appeared to be stronger than between Anx⁺CD95⁺ and CD4⁺CD95⁺ T cells. Regarding CD4+ effector T cells, other, CD95-independent, pathways mediated by Tregs to negatively regulate their activity are supposed to be more important [2], including those mediated by TGFβ, a cytokine secreted both by Tregs and cervical cancer cells [34] (as indicated above, female patients involved in our study had higher level of serum TGFβ than the controls). Interestingly, in CIN/CC patient group, a strong association was seen between the frequencies of CD4⁺CD95⁺ T cells and CD4⁺CD25^high cells (along with a more moderate correlation between CD4⁺CD95⁺ cells and CD4⁺FoxP3⁺ cells, Fig.10) suggesting that these cell populations may be involved in a common mechanism, whereas no such relationship was detected for CD8⁺CD95⁺ T cells.

Recent research on the mechanisms of immunosuppression employed by a developing tumor has provided novel insights on the diversity and function of other cell subsets (besides CD4 Tregs) capable of repressing effector activities of the immune system and thereby facilitating tumor evasion of immune surveillance. These include a relatively poorly studied population of CD8⁺Treg lymphocytes, whose role in tumor development and progression is not yet fully understood. There are only few published data on gynecological cancers, for example a significantly higher frequency of circulating CD8 Tregs has been described in patients with advanced ovarian carcinoma (stage III-IV) compared to patients with early-stage (I-II) or benign tumors of the ovary [35]. In the case of cervical cancer, there is one study by Chen and colleagues who revealed increased levels of circulating CD8⁺CD25⁺CD127^dim/negFoxP3⁺ lymphocytes in women with invasive cervical carcinoma (stage IA-IIB) and CIN3 compared to healthy counterparts, with no significant difference between CIN3 and invasive cancer detected [16]. Results from our study on CIN3 and microcarcinoma presented here are consistent with these observations. Also, several research teams have demonstrated that in spite of the fact that the population of CD8 Tregs is much less numerous in the peripheral circulation than CD4 Tregs, in in vitro assays they are nevertheless able to suppress T-cell proliferation (as effectively as, or even better than, CD4 Tregs), acting through the release of TGFβ or via other mechanisms [25, 35, 36]. Given that in experimental settings CD8⁺ T cells were shown to convert to suppressor CD8 Tregs under the influence of the tumor cell microenvironment [35, 37], it can be assumed that expansion of CD8 Tregs observed in systemic circulation of cervical cancer patients at a pre-metastatic stage might reflect important local level changes in their quantities, that provide a foothold for further tumor dissemination. This assumption was underpinned by Battaglia and co-authors [10] and Heeren and co-authors [11], who showed that CD8⁺FoxP3⁺ Т cells were more frequent in metastatic tumor-draining lymph nodes in early-stage cervical cancer. Apart from oncological diseases, CD8 Tregs are thought to play an important role in establishment of chronic viral infection, including herpesviruses (EBV, CMV), HCV and HIV (discussed in [38]). It is also possible that HPV engages CD8 Tregs for maintaining its long-term persistence, which may underlie an increase in the frequency of CD8 Tregs observed at earliest stages of CC progression.

Diverse types of cells endowed with regulatory properties have been discovered in the innate immune system; among these, NKregs have recently emerged as an important subpopulation whose functions and tissue distribution in different pathological conditions (including cancer) are presently attracting increasing interest. Owing to their high capacity to produce proinflammatory cytokines, NKregs are no longer regarded simply as a minor fraction of peripheral blood NK cells, but there are only a few types of cancer, for which their contribution to antitumor immunity have been described so far. We observed a decrease in the frequency of circulating CD56^bright NK lymphocytes in women diagnosed with cervical cancer, although the overall percentage of a major population of NK cells (CD56^dim) didn’t differ significantly between patients and controls. Similar findings have been made in two other studies on head and neck cancer [39] and prostate cancer [6], whereas conflicting data have been reported for breast cancer (while Bauernhofer and co-authors [40] showed decreased levels of CD56^bright NK cells, Nieto-Velázquez and co-authors [41] revealed no changes in CD56^bright NK-cell numbers relative to healthy individuals). As a consequence, we were able to reveal an increased CD56^dim/CD56^bright NK-cell ratio in the peripheral blood of cervical cancer patients with no difference between patients and controls in the total percentage of NK cells. Similar results have been reported by Bauernhofer et al. [40] and Koo et al. [6]. The most likely explanation for these data is that circulating CD56^bright NKregs are recruited to the lymphoid tissue (regional lymph nodes) and primary tumor site, where they comprise a significant portion of NK lymphocytes and are thought to serve as precursors for cytotoxic/effector CD56^dim NK cells [6]. While the distribution of CD56^bright cells between different tissue compartments in cervical cancer patients remains unclear, flow cytometry analysis of enzymatically dissociated samples of primary high-grade serous ovarian carcinoma confirmed that CD56^bright NK cells were indeed predominant among intratumoral NK lymphocytes [42].

In the present study, we focused on identifying phenotypic changes in circulating lymphocytes of patients with preinvasive and microinvasive carcinoma, since these disease stages represent, on the one hand, the ultimate result of malignant transformation, and, on the other hand, the transition from intraepithelial development to invasive growth, which could be considered as a turning point in tumor progression. In relation to histopathologic and clinical features, preinvasive and microinvasive lesions are very close, that is apparently why we didn’t observe significant differences in most parameters tested between respective subgroups of patients. However, the data obtained indicated that immunological defects seen in the periphery may accumulate during further CC progression, and additional studies that would include more advanced disease stages are needed to clarify this question. In this regard, some valuable results of a flow cytometric study of cell suspensions prepared from tumor-draining lymph nodes of female patients with stage IB1 cervical cancer [11, 12] are worth mentioning: in this study, the authors found altered proportions of CD4⁺ and CD8⁺ Т cells, along with altered expression of T-cell activation and immune checkpoint markers, increased frequencies of CD4⁺ and CD8⁺ Tregs, as well as a decreased CD3⁺CD8⁺/Treg ratio.

It is worth noting, in conclusion, that numerous studies now carried out in the field of cancer immunotherapy and aimed at developing new approaches to treat cancer, such as the use of therapeutic vaccines, targeted drugs (with primary attention to immune checkpoint inhibitors), and cytokine-based therapy, presuppose deep understanding of the processes that underlie both local and systemic dysfunctions of antitumor immunity in different kinds of malignancies and at different stages of disease progression. Immune regulatory networks that maintain cancer onset and progression (including the association between regulatory lymphocytes and the level of T cell apoptosis analyzed in the present work) may serve either as the source of molecular and cellular therapeutic targets or markers for monitoring the efficacy of immunotherapy and its combination with chemo−/radio−/targeted therapy regimens. For example, a vast amount of data has been amassed presenting opportunities for the use of CD4 Tregs as direct targets for new targeted agents with consideration of the features of Treg biology in different human cancers [43]. We have also demonstrated previously the informativity of some apoptosis-related markers in circulating lymphocytes for monitoring the effect of immunomodulatory treatment in patients with CIN3 and CC stage IA [27]. In view of the infectious etiology of cervical cancer, the most promising immune-based strategies for combination therapy of precursor lesions and invasive (especially metastatic) cancer are being discussed as a separate issue [44, 45]. It is plausible that some of these strategies may help disrupt pathologic interplay existing between CD4⁺/CD8⁺ Tregs and populations of cytotoxic lymphocytes and resulting in upregulation of apoptosis in the latter cells, which should be investigated further.

Collectively, analysis of immunophenotypic data on peripheral blood lymphocytes from patients with CIN3/ carcinoma in situ or microinvasive carcinoma of the cervix suggests that the mechanisms, by which Tregs can expand their suppressive influence on antitumor immune responses and impair the balance between effector cell subsets (including via potentiation of CD95-mediated apoptosis), and which were previously established mostly with the use of murine models or other experimental systems, play a relevant role in promoting early cervical cancer progression and spread in humans. These pathological changes may also be accompanied by a regulatory/effector cell disbalance within non-T lymphocytes. Activation of the above mechanisms observed at the level of systemic circulation and coincident with the earliest lesions (virtually confined to epithelial layer) obviously reflects the existence of a complex cross-talk and interdependence between local immune reactions and responses of peripheral blood lymphocytes.