Frequent Display of Human Papillomavirus Type 16 E6-specific Memory T-Helper Cells in the Healthy Population as Witness of Previous Viral Encounter¹

Genital infection of both men and women with oncogenic HPV⁴ types is quite common (1, 2, 3), however, only a minor fraction of infected subjects develop progressing epithelial lesions or cancer (4, 5, 6). Most premalignant lesions regress spontaneously. This is likely mediated by cellular immune responses because regressing (genital) warts are infiltrated with CD4+ T cells, CD8+ T cells, and macrophages (7, 8). The role of immune surveillance in controlling infection by various HPV types is additionally indicated by the increased incidence of HPV infections, HPV-associated warts, cervical intraepithelial neoplasia lesions, and cervical carcinoma in immunocompromised subjects (9, 10, 11). This notion is sustained by the observations that papillomavirus-specific T-cell immunity can protect against malignant transformation of papillomas in rabbits (12).

Examination of HPV-specific T-cell immunity in patients with HPV+ cervical lesions has revealed the occasional presence of CTL responses against the E6 and E7 oncoproteins (13, 14, 15, 16, 17, 18). Because an overt correlation between the presence of such CTL immunity and regression or progression of the disease has thus far not been found, the functional significance of these responses is at present a matter of debate. Most likely, not only CTL numbers, but also their activation state needs to be taken into account. Accumulating evidence pertinently shows that proper activation of CTL requires delivery by professional APCs of strong costimulatory signals in addition to the antigenic signal. Especially in the case of antiviral immunity in chronic infections, this costimulatory action of APC requires virus-specific Th cell activity (19, 20, 21, 22, 23). We previously found that E7-specific Th immunity was frequently observed in HPV-16+ patients but only rarely in healthy subjects (24). In contrast, >50% of the healthy subjects tested showed very strong memory-type Th responses against one or more epitopes of HPV-16 E2. It is conceivable that this Th memory is a witness of previous encounter with HPV-16 or possibly other HPV-types encoding cross-reactive epitopes (25). Apparently, the T-cell immune system deals with the E2 and E7 antigens in a completely different manner. Because high expression of E6/E7 and E2 is mutually exclusive during infection and subsequent tumor development, these findings prompted us to now also monitor Th immunity against HPV-16 E6 in patients and healthy subjects. Similar to what was found for E2, memory Th immunity against E6 was frequently and abundantly observed in healthy subjects, suggesting a role for these Th cells in protection against reinfection by high-risk HPV and as such the HPV-mediated induction of anogenital lesions and cervical cancer.

PBMCs of anonymous healthy blood bank donors (D) were obtained after informed consent. Because these donors are anonymous, only a general age and gender distribution can be given. The currently studied healthy individuals displayed a median age of 44 years old (range, 25–68 years) and comprised of 40% women and 60% males. No additional data are available, nor was serum available to determine the HPV infection status by serological analysis. However, donors with a known recent history of infection, including abnormal pap-smear, were as part of normal regulations, discouraged to donate blood.

The study of the subjects (P) with cervical carcinoma in this paper was nested in the CIRCLE study, which investigates cellular immunity against HPV-16-infected cervical lesions. Women presenting with histological proven cervical carcinoma at the department of gynecology of the Leiden University Medical Center were, after informed consent, enrolled in this study. The study design was approved by the Medical Ethical Committee of the Leiden University Medical Center. Blood was drawn at day of treatment before surgery. Subjects with stage IB-IIA were treated by a radical hysterectomy. The age of the patients ranged between 29 and 76 years old, and these women were typed for HPV-16 using HPV-16-specific primers on DNA isolated from paraffin-embedded sections of biopsies or surgical resection specimens (26).

A set of peptides spanning the whole HPV-16 E2 protein, consisting of 22 30-mer peptides with 15 amino acids of overlap and the COOH-terminal peptide with a length of 35 amino acids, was used for the T-cell proliferation assays. They are indicated by the first and last amino acid in the protein (e.g., E2_1–30, residues 1–30, and the last peptide E2_331–365). For the ELISPOT assays, a large set of 22-residue long HPV-16 E6 overlapping (12 residues) peptides was used to obtain a detailed insight in the immunogenic regions of the E6 protein. In addition, 8 HPV-16 E6-derived 32-mer overlapping (14 residues) peptides and 4 HPV-16 E7 35-mer overlapping (14 residues) peptides were used for the T-cell cultures and proliferation assays. These long HPV-16 E2, HPV-16 E6, and HPV-16 E7 peptides were used in the proliferation assay to lower the number of PBMCs required because we wanted to use pools of maximal two peptides.

Pools of four 30-mer peptides, with 15 amino acids overlap, spanning the influenza matrix 1 protein of A/PR/8/34 (M1) were used as positive control in the ELISPOT assay. The peptides were synthesized and dissolved as described previously (27).

TT (1 LF/ml; National Institute of Public Health and the Environment, Bilthoven, the Netherlands) was used as positive control in the short-term T-cell proliferation assays.

Recombinant HPV-16 E6 protein and HPV-16 E7 protein (the latter protein served as control protein in proliferation assays) were produced in recombinant Escherichia coli transformed with Pet-19b-HPV-16 E6 and Pet-19b-HPV-16 E7 (24).

The presence of HPV-16 E6-specific Th cells was analyzed by ELISPOT as described previously (24). Briefly, PBMCs were seeded at a density of 2 × 10⁶ cells/well of a 24-well plate (Costar, Cambridge, MA) in 1 ml of Iscove’s medium (Bio-Whittaker, Verviers, Belgium) enriched with 10% FCS (Greiner, Mannheim, Germany) in the presence or absence of 5 μg/ml of indicated E6-derived 22-mer peptide. As a positive control, PBMCs were cultured in the presence of indicated pools of influenza A/PR/8/34 M1 protein-derived peptides, consisting of four overlapping 30 amino acid long peptides in each pool because previous analyses showed that all healthy individuals displayed reactivity against one or more of these M1 peptide pools (24). After 4 days of incubation at 37°C, PBMCs were harvested, washed, and seeded in 6-replicate wells at a density of 10⁵ cells/well on a Multiscreen 96-well plate (Millipore, Etten-Leur, the Netherlands) coated with an IFN-γ catching antibody (Mabtech AB, Nacha, Sweden). Then, the ELISPOT plates were incubated for 18 h at 37°C to allow catching of the antigen-specific produced IFN-γ. The ELISPOT assay was additionally performed according to the instructions of the manufacturer (Mabtech). The number of spots was analyzed with a fully automated computer-assisted video imaging analysis system (Carl Zeiss Vision). Specific spots were calculated by subtracting the mean number of spots plus 2× SD of the medium only control from the mean number of spots of experimental wells. Antigen-specific T-cell frequencies were considered to be increased compared with nonresponders when T-cell frequencies were ≥1/10⁴ PBMCs (24).

Healthy donor-derived PBMCs were incubated with 12 pools of HPV-16 E2-derived 30-mer peptides, 4 pools of E6 32-mer peptides, and 2 pools of E7 35-mer peptides (each pool containing two overlapping peptides), according to procedures described previously (25, 27). Briefly, freshly isolated PBMCs were seeded at a density of 1.5 × 10⁵ cells/well in a 96-well U-bottomed plate (Costar, Cambridge, MA) in 200 μl of Iscove’s medium (Bio-Whittaker) supplemented with 10% autologous serum. In previous experiments, we observed that proliferative responses to HPV-16 E2 peptides were relatively reduced with the use of (heat-inactivated) human AB serum (Sigma) as compared with autologous serum. HPV-16 E2-, HPV-16 E6- and HPV-16 E7-derived peptides were added at a concentration of 10 μg/ml. Medium alone was taken along as a negative control and TT served as a positive control. For each peptide pool, 8 parallel microcultures were incubated. Peptide-specific proliferation was measured at day 6 by [³H]thymidine incorporation. Cultures were scored positive when the proliferation of ≥75% of the test wells exceeded the mean proliferation plus 3× SD of the medium only control wells, and the SI of the mean test wells over mean medium control wells was ≥3.

PBMCs derived from healthy blood donors were incubated with 8 overlapping 32-mer peptides spanning the entire HPV-16 E6 protein at a concentration of 1 μg/ml in Iscove’s medium (Bio-Whittaker) supplemented with 10% recalcified autologous plasma. One day after stimulation, T-cell growth factor (Biotest, Dreieich, Germany) was added at a final concentration of 5%. At day 10, the cell culture was harvested and incubated with autologous monocytes and E6 peptides (1 μg/ml) at a 10:1 ratio for 16 h, after which, T cells producing IFN-γ in response to the E6 peptides were enriched by a cytokine secretion assay (Miltenyi; Biotech, Bergisch Gladbach, Germany). This assay is based on a bispecific antibody recognizing (a) the IFN-γ molecule and (b) the CD45 molecule. The captured IFN-γ is detected by a phycoerythrin-labeled secondary antibody, and subsequent magnetic labeling allows enrichment of the antigen-specific cells. The isolated cells were cultured for 10–14 days with irradiated autologous PBMCs at 1:10 ratio and 10% T-cell growth factor without any further addition of antigen. The polyclonal T-cell culture was tested for recognition of E6 peptides (5 μg/ml) and protein (10 μg/ml) in a 3-day proliferation assay. Autologous adherent monocytes were used as APCs, and proliferation was measured by [³H]thymidine incorporation.

In addition to E7, the E6 oncoprotein is essential for maintenance of the transformed state in cervical carcinoma. Both proteins are thus interesting targets for antitumor T-cell immunity. We therefore extended our previous analysis of HPV-16 E7-specific Th immunity in HPV-16+ cervical cancer patients (24) to the E6 antigen. Hence, PBMC of 11 HPV-16+ cervical cancer patients were stimulated with a set of overlapping 22-mer peptides covering the complete HPV-16 E6 protein, after which, the number of IFN-γ-producing E6-specific Th cells was determined by ELISPOT. In 3 of 11 HPV-16+ patients (P4, P5, and P8), HPV-16 E6-specific Th frequencies ≥1/10,000 PBMCs were demonstrated. On the basis of our experience, responses of such a magnitude are clearly positive, whereas responses lower than this frequency may fall within background reactivity (24). Patient P4 responded to peptide E6_31–52 and E6_131–152, indicating the presence of memory Th cells specific for two distinct epitopes. The other two patients (P5 and P8) responded to only one E6 peptide, suggesting that in both cases a single E6-epitope was recognized (Table 1). Overall, the frequency and magnitude of Th1-type response against E6 detected in patients is comparable with that previously found for E7 (24).

We subsequently analyzed whether these E6-specific responses are also detectable in healthy donors because this population undoubtedly includes individuals who have effectively dealt with HPV-16 infection (3, 6, 28). Therefore, we assessed the presence of HPV-16 E6-specific Th cells in the blood of a group of 20 healthy individuals. Interestingly, 12 of 20 healthy blood donors showed HPV-16 E6-specific Th type 1 responses upon stimulation with one or more of the HPV-16 E6 22-mer peptides (Table 1). Analysis of the reactivity pattern revealed that most responses detected were directed against peptide sequences of the COOH-terminal half (E6_81–158) of HPV-16 E6 (Table 1). The magnitude of the E6-specific responses was very similar to that found in patients. However, the fraction of healthy donors showing strong IFN-γ Th responses against E6 peptides appears higher than that of the patients (60 versus 27%). Furthermore, some of the healthy subjects reacted even against four or more E6-peptides (range, 1–7), whereas the 3 patients responded to one or two peptides only. It must be noted, however, that also the incidence and breadth of the response against the influenza M1 antigen is higher in healthy subjects (Table 1), indicating an overall reduced immunity in patients.

To further characterize the observed HPV-16 E6-specific IFN-γ-responses in healthy subjects, the E6-induced proliferation capacity of PBMCs was analyzed. Freshly isolated PBMCs of 5 healthy donors were stimulated with a pool of eight of the long 32-mer overlapping HPV-16 E6-peptides for 10 days, and the responding T cells were isolated based on the E6-specific IFN-γ production using cytokine secretion assay and magnetic cell sorting. Flow cytometric analysis of the T cells specifically responding to E6 revealed that these T cells resided in the CD4+ fraction only (Fig. 1). After a 14-day period of nonspecific expansion, the resulting T-cell cultures were tested for their proliferative capacity to both E6 peptides and E6 protein because the latter will indicate their capacity to respond to naturally processed and presented antigen. From 3 of these 5 cultures, we were able to isolate IFN-γ-producing HPV-16 E6-specific Th cells able to proliferate not only upon stimulation with peptide but also to E6 protein-pulsed APCs (Fig. 1). Two subjects (D21 and D23) predominantly responded to E6_73–109 and E6_127–158, whereas the third donor (D22) displayed a broader reaction pattern (Fig. 1). The area covered by E6_73–109 and E6_127–158 was also frequently recognized when the E6 response was analyzed by ELISPOT (Table 1). Taken together, our data demonstrate that the memory T-cell repertoire of the majority of healthy individuals contains proliferative and IFN-γ-producing CD4+ Th-cells capable of reacting against the HPV-16-derived E6 oncoprotein.

We have recently demonstrated frequent HPV-16 E2-specific Th responses in healthy individuals (25), whereas Th immunity against the E7 oncoprotein was rarely found in these subjects (24). This study revealed that the HPV-16 E6 protein is also frequently targeted by Th cells in healthy subjects. To define the relationship between the E2-, E6- and E7-specific Th response in healthy individuals, the presence of Th reactivity against these three antigens was analyzed in parallel. Fresh PBMC samples derived from 12 healthy blood donors were stimulated with 18 pools of two overlapping 30–35-mer peptides that together cover the entire sequences of HPV-16 E2, HPV-16 E6, and HPV-16 E7, and T-cell proliferation was evaluated after 6 days of culture as described previously (25). A SI exceeding 3 was applied as a threshold because this is commonly used for the detection of memory T-cell responses (29, 30). The incidence and specificity of the proliferative responses against HPV-16 E2 and HPV-16 E6 were comparable with our previous ELISPOT experiments (Table 1; Ref. 25). One donor (D33; Fig. 2) displayed responses against all three antigens, and 3 donors [D27 (Fig. 2), D30, and D37] showed specific proliferation against E2 and E6 (Table 2). In three other cases (D26, D29, D34) responses were only found against E2, whereas one PBMC sample (D31) only responded to E6 peptides. Most importantly, a vast majority (8 of 12) of the healthy subjects displayed memory-type Th responses against HPV-16 E2 and/or HPV-16 E6 in their blood, supporting the notion that encounter with high-risk HPV types, including HPV-16, is very common (1, 2, 3). Although in certain cases, this event may result in the prevalence of Th memory against either E2 or E6 rather than against both antigens; such immunity against the E7 antigen is only rarely installed (Table 2, Fig. 2).

Data obtained in the few studies that scrutinized HPV-specific immunity in HPV-16-positive patients suggested that HPV-16 E6-specific T-cell immunity might play a role in the protection against progressive HPV-16 infection (14, 31). Successful control of viral infection by the cellular immune system induces T-cell memory and therefore E6-specific T cells are expected to be present in those individuals who have effectively dealt with infection. On the basis of the cumulative 24-month incidence of infection with HPV-16 of 7% (6), a substantial fraction of the healthy population is expected to harbor such T-cell memory. Indeed, we could detect HPV-16 E6-specific memory Th immunity in more than half (17 of 32) of the healthy subjects by both IFN-γ-ELISPOT and short-term proliferation assays. Isolated E6 peptide-specific Th cells from such healthy subjects specifically reacted upon stimulation with HPV-16 E6 protein showing their capacity to respond not only to peptide but also to naturally processed HPV-16 E6 protein.

Because E6 protein sequences are conserved to considerable extent between HPV types, it is conceivable that infection with a given HPV type induces T cells, which are also able to react with peptides of HPV-16. Amino acid sequence analysis revealed that especially the genital high-risk HPV types displayed strong homology with HPV-16 E6 (data not shown). Therefore, part of the memory Th responses measured in this study might be a reflection of previous infections with high-risk HPV types other than HPV-16. Furthermore, the relatively high incidence and prevalence of (subclinical) infection with high-risk HPV (3, 6, 28) implies that some of these responses reflect current HPV infection.

In PBMCs from patients diagnosed with HPV-16+ cervical carcinoma, E6-specific Th type 1 immunity was detected in ∼30% (3 of 11) of the patients tested. As such, the number of E6-specific responders does not differ from that of HPV-16+ patients with E7-specific Th type 1 reactivity (6 of 16, ∼40%; Ref. 24 and unpublished observations). On the basis of the frequency of E6 responses in healthy individuals, the HPV-16 E6-specific Th cells in patients are likely to represent the remnants of a failing immune response, whereas Th reactivity to E7 may reflect de novo responses after cross-priming of E7 derived from a progressive tumor. Interestingly, a direct comparison of Th responses against HPV-16 E6 and HPV-16 E7 in healthy individuals reveals a different trend of immune activity. Only 1 of 12 healthy blood donors displayed E7-specific Th immunity, whereas 5 of these 12 healthy subjects reacted upon stimulation with HPV-16 E6 peptides. Moreover, we demonstrated that E6-specific Th responses in general coincided with responses against HPV-16 E2. In addition, if one takes into account that probably not all of these anonymous healthy blood donors have encountered HPV-16, whereas the patients are all HPV-16+, the percentage of healthy subjects responding to HPV-16 E6 after successful clearance of the HPV infection is likely to be underestimated. Failure of boosting the E6-specific Th1 type immunity may thus be related to progressive infection, which was also suggested by the work of Kadish et al. (31). To address this, a more comprehensive study of HPV-16-specific immunity in relation to disease and HPV infection status is required.

Collectively, our data suggest that the induction of Th cell reactivity against the HPV-16 E7 protein is suboptimal when compared with that against HPV-16 E2 and HPV-16 E6. Recently, it was shown that CD4+ T-cell reactivity against the plantar wart causing low-risk HPV type 1 was predominantly directed against the E4 protein (50%) and to a lesser extent against E6 (20%). Th-cell reactivity against HPV-1 E7, like our results with HPV-16 E7, could not be demonstrated (32). Presently, it is not clear why Th reactivity is preferentially induced against the HPV-16 E6 protein and not against E7. Because of the nature of HPV infection, induction of T-cell immunity is exclusively dependent on cross-presentation by epithelial resident dendritic cells, the Langerhans cell (33). Possibly, the localization of the antigen in the infected cell may determine the accessibility of these antigens for Langerhans cells. Low-risk and high-risk E7 proteins are expressed in the nucleus of the cell (34), and Th responses against this antigen are hardly detected in healthy subjects (24, 32). The expression of HPV-1 E4 protein is predominantly cytoplasmic, and a large fraction of healthy subjects mount an HPV-1 E4-specific Th response (32). Interestingly, the expression of low-risk HPV E6 protein is also restricted to the nucleus, but E6 of high-risk HPV is expressed in the nucleus and the cytoplasm (34). This difference in localization seems to be reflected by the low Th response rate against HPV type 1 E6 protein and the frequent induction of HPV-16 E6-specific Th reactivity in healthy subjects. Moreover, the similar cytoplasmic and nuclear expression of HPV-16 E2 induces Th memory in healthy individuals (25, 35).

In conclusion, the high rate of spontaneous clearance of high-risk HPV infections suggests that in general, the Langerhans cell-driven antigen presentation machinery is capable of inducing an effective immune defense against HPV. Our study is the first to demonstrate the presence of HPV-16 E6-specific memory Th immunity in the healthy population and shows that HPV infection indeed leads to T-cell immunity against the (immediate) early proteins E2 and E6 expressed during infection. Because the population, in which E2- and E6-specific T-cell immunity frequently is found, consists of healthy subjects, the observed IFN-γ-producing T cells circulating in the peripheral blood may play a role in protection against newly and/or persistent HPV infections and as such the development of cancer.

We thank all healthy blood donors and patients participating in the CIRCLE study. We also thank Willemien Benckhuijsen for synthesis of the peptides, Gemma Ranke for obtaining patient material, and Sandra Uljee for HPV typing.