British Journal of Dermatology (1992) 127, 551-559.
Lymphoproliferative responses to human papillomaviruses in patients with cutaneous warts F.C.CHARLESON, M.NORVAL. E.C.BENTON* AND J.A.A.HUNTER* Department of Medical Microbiology. University of Edinburgh Medical School, Edinburgh, U.K, 'University of Edinburgh Department of Dermatology, Royal Infirmary of Edinburgh. U.K. Accepted for publication 2 i June 1992
Summary
In vitro lymphoproliferative responses of peripheral blood mononuclear cells (PBMC) from patients with cutaneous warts, caused by infection with human papillomavirus type 1 (HPV-1) or type 2 (HPV-2), were assayed during the course of treatment. Purified HPV-1 and HPV-2 were used as antigens, as well as herpes simplex virus (HSV) and concanavalin A (Con A). All patients had normal percentages of subsets within the PBMC population and normal iymphoproliferative responses to Con A, and those with a clinical history of HSV infections had positive lympboproliferative responses to HSV. Responses to both HPV antigens were poor. Only 10 of 100 assays of PBMC from 26 patients showed a stimulation index greater than 2. Addition of interleukin 2 made little difference in most cases. No correlation of clinical status of warts, i.e. improving, unchanged or resolved, with lymphoproliferation was found. When the PBMC were depleted of plastic-adherent cells and enriched for T cells, some samples which had not shown a lymphoproliferative response to HPV-1 or HPV-2 became positive: this response was abolished when the adherent cells were re-added. Thus it is possible that the adherent cell population from a proportion of patients contains cells which suppress lymphoproliferation, or that an immunoregulatory network is present so that iymphoproliferation does not take place in vitro without prior activation and cloning of T cells.
Warts caused by infection with human papillomavirus (HPV) types are common, affecting 10% ofthe population at any one time. The lesions are predominantly persistent in nature and last for several months before resolving spontaneously. About 35% of skin warts resolve in 6 months but. in some apparently healthy individuals, warts may be extensive, recalcitrant to treatment, and persist for many years.' It appears that the immune response generated against HPV infection is important in the control of the extent and duration of warts. Clinically this is apparent in patients who are immunosuppressed by disease or drugs, as they will often develop extensive, disabling and persistent warts, particularly when cell-mediated immunity is depressed, rather than humoral immunity. The following clinical examples illustrate this point: (i) pregnant women may develop florid warts which often regress spontaneously following delivery; (ii) patients immunosuppressed by diseases such as lymphomas or leukaemia, or those
Correspondenee: IJr M.Norval, Department of Medieal Microbiology. University of Edinburgh Medical School, Teviot Place. Edinburgh EH8 9AG. U,K,
infected with human Immunodeficiency virus, may develop extensive cutaneous and mucocutaneous warts: (iii) patients with epidermodysplasia verruciformis, an autosomal recessive disorder with an associated defect in cell-mediated immunity, have extensive infection with HPV of unusual types:^ and, (iv) patients immunosuppressed therapeutically following renal or cardiac allografting frequently develop extensive and persistent warts. ^^ Histological evidence also points to a role for the immune response in resolution of viral warts. For example, in regressing plane warts caused by HPV-3. an infiltrate of T cells and mononuclear phagocytes is found adjacent to damaged keratinocytes. indicating classical cell-mediated cytotoxicity.'' On the other hand, such an infiammatory cellular infiltrate is infrequent in other skin warts which are regressing, and is seldom seen in mucocutaneous warts. Furthermore, in clinical terms, the immune response to HPV does not seem to be very effective or long-lasting, with no obvious protection from subsequent episodes of infection. Assays to demonstrate specific immune responses against HPV have proved difficult, and have often
551
552
F.C.CHARLESON et al.
yielded inconsistent results, with the patient's previous history of HPV infection being largely unknown. In general, no correlation has been found between antibody titre to HPV and a past or present history of cutaneous warts'' or cervical lesions.'" Two studies have used in vitro lymphoproliferative responses to HPV to assess T-ceil immunity in patients with cutaneous warts. In the first, a small response to semipurified HPV was noted in patients with active warts, and this increased after resolution." In the second, patients with warts of short duration (less than 1 year) showed small responses, although these diminished in patients with warts of longer duration.*^ In addition, some controls who thought they had never had warts were positive. It is difficult, therefore, to correlate accurately HPV history and lymphoproliferation. Similar lymphoproliferation results have been reported in patients with cervical HPV lesions and dysplasia, using a variety of HPV antigen preparations, including several fusion proteins representing early regions ofthe HPV genome. No correlation was found between the clinical diagnosis of cervieal HPV infection and positive responses.'"' In the present study, patients with either simple plantar warts caused by HPV-1 or common hand and plantar mosaic warts caused by HPV-2 were followed during the course of treatment. Peripheral blood mononuclear cells (PBMC) were assayed for their ability to respond to purified HPV in vitro. As several patients showed improvement or resolution of their warts during the study, correlations between lymphoproliferative responses and clinical status could be drawn, PBMC suspensions were depleted of antigen-presenting cells by plastic adherence, and enriched for T cells by adsorption on nylon wool columns. The responses of the T cells, depleted of, and reconstituted with plastic adherent cells, to HPV antigens were ascertained.
Methods Clinical samples
The patients studied were attending the dermatology out-patient department at the Royal Infirmary of Edinburgh for treatment of cutaneous warts, and were otherwise healthy. The type of wart present was determined by clinical examination. Two groups of patients are reported—those with simple plantar warts, caused by HPV-1. and those with common hand warts or mosaic warts, caused by HPV-2, The patients were seen and assessed clinically, usually at monthly intervals, and underwent a variety of treatments for their warts such as freezing with liquid nitrogen, application of salicylic
acid, or intralesional injection of bleomycin. Age- and sex-matched healthy volunteers, with no known cutaneous HPV infection at the time of sampling, acted as controls. Approximately 50 ml of venous blood was collected in preservative-free heparin at each visit, and processed immediately. Antigens
Pooled parings from either simple plantar warts (HPV-1) or plantar mosaic warts (HPV-2) acted as a source of HPV antigen. Purified whole virion samples were prepared by caesium chloride density gradient centrifugation' and stored in phosphate-buffered saline (PBS) at — 20°C. Total protein content was determined by the comparison of virion samples with serial dilutions of bovine serum albumin on silver-stained polyacrylamide gels under reducing conditions. The presence of HPV-1 proteins was confirmed by Western blotting of virion sampies using a monoclonal antibody for the LI protein of HPV-1 (Sera-lab. U.K.), and any HPV-2 samples which contained HPV-1 proteins were discarded. Herpes simplex virus (HSV) antigen was prepared by infecting Vero cells with a clinical isolate of HSV type 1 for 24-48 h at 37°C; the number of plaque forming units (p.f.u.)/ml was assayed by titration in Vero cells before inactivation with ultraviolet B irradiation of 300 m]/cm^ (HSV titre before inactivation was 10" p.f.u./ml)," The lysate from uninfected Vero cells acted as a negative control.'^ Concanavalin A (Con A) was obtained from Sigma. Separation of PBMC
PBMC were separated by centrifugation on 'Lymphopaque' (Nyegaard. Oslo) and washed three times in sterile PBS containing 5 U/ml of heparin before resuspension in assay medium (see below). Plastic-adherent cells were isolated from PBMC by incubation in "Falcon' tissue-culture fiasks (Becton Dickinson, U.K.) for 1 h at 37°C." Non-adherent cells were further purified by incubation on nylon wool (Fenwal. Illinois) for 2 h at 37°C to provide a sample which was enriched for T lymphocytes (T-enriched cells).'- All cell samples were resuspended at 10^/ml in assay medium whieh consisted of RPMI 1 640 (Northumbria Biologicals. U.K.) with 1 5% autologous plasma plus 5 U/ml heparin, 2 mM glutamine, 0-05 mM 2-mercaptoethanol, 100 lU/penicillin, 200 /ig/ml streptomycin, 1()() fig/ml gentamicin and 2 /ig/ml Fungizone®. The ceil populations were monitored at each stage by fiow cytometry using monoclonal antibodies to various cell surface markers (see below).
LYMPHOPROLIFERATIVE RESPONSES TO HPV
Lymphoproliferation
assay
Phenotyping of cell populations
Cells were seeded into round-bottomed 96-well 'Cellcult' plates (Sterilin, U.K.) in 200 ^1 volumes (2 x 10^ cells) in the presence or absence of antigen. Con A was added at 10 /^g/ml and HSV at 1 p.f.u./cell, with the equivalent amount of uninfected Vero cell iysate as a negative controi. Approximately 1 /ig of HPV-1 or HPV2 in 20 /d PBS/ml of cell suspension was used, with 20 /il/ml sterile PBS as a negative control. PBMC were assayed in quintuplicate for each antigen, whiist Tenriched cells were assayed in tripiicate to conserve cell yieids. A reconstituted ceii sampie consisting of Tenriched cells plus plastic-adherent cells (added to form 10% of the total cell number within the same sample volume) was also assayed in triplicate. Cells were incubated in 5% CO2 at i7°C for 6 days when cultured with Con A or HSV, or 7 days when cultured with HPV antigens. Tritiated thymidine (0 75 /iCi in 10 fi\ PBS/ well) was added for the final 18-24 h of eulture. Plates were harvested on to 'Titertek' filter paper (Flow Laboratories) using an 'Automash 2000' cell harvester (Dynatech) and discs were counted for 1 min in a 'Tri-Carb' scintillation counter (Canberra Packard), Some results were expressed as a stimulation index (SI) where:
SI-
mean count per minute (c.p,m.) in presence of antigen mean c.p.m. in absence of antigen An SI value of 2 or more was taken as positive.
Table 1. Monoclonal antibodies used in phcnntypiiii' by flow cytometry
553
Cell surface marker CD 3 CD4 CD8 CDla MHC class 11
CD57 CDIlb
Cells were incubated with various mouse monoclonal antibodies, followed by sheep anti-mouse Ig labelled with fiuorescein isothiocyanate (Sigma), according to the method of Neill and Miller,'' except that the first step was performed at 4°C overnight. The monoclonal antibodies which were used are shown in Tabie 1. Cells were washed twice, resuspended at 5 x 10^ cells/mi in coid PBS, fixed with 1% formaldehyde, and approximately 10* cells analysed on a Coulter EPICS 'C fiow cytometer (output 100 mW at 488 nm. fiow rate 400-500 cells/s).
Results Clinical status
The clinical details of the patients and healthy volunteers studied are shown in Table 2. The results for patients with common hand warts and plantar mosaic warts are reported as a single group because both are caused by the same HPV type, and no differences were observed between them. In some cases results have been divided into three groups so that immune responses could be correlated with changes in clinical status. These were: (a) no improvement: (b) improvement (i.e, a deerease in the number of warts or degree of hyperkeratosis) or, (c) complete resolution ('resolved'). Patients were monitored over varying lengths of time according to changes in their clinical status.
Target cell T lymphocytes T-helper lymphocytes T-supprcssor/cytotoxic lymphocytes Thymocytes Dendritk" ceils Monocytes/macTophages B lymphocytes Activated T lymphocytes LGL* NK cellsj: K cells t Monocytes T-suppressor lymphocytes LGL/NK cells
Monoclonal antibody
Source
DAKO-Ti DAK0T4 DAK0-T8 BAK0-T6
Dakopatts Dakopatts Dakopatts Dakopatts
DAf).2il
Dr K,Guy§
Uu 7
Becton Dickinson
OK Ml
Becton Dickinson
* LCI., large granular lymphocytes, t NK, natural killer cells, t K, killer cells. (j Kind gift from Dr Keitb Cuy. Western General Hospitai. Edinburgh.
554
F.C.CHARLESON et al
Table 2. Clinical history of study groups Duration of wart(s) (years) HPV type
Number in group (male, female)
Mean age
Simple plantar
HPV-1
9 f5M, 4F)
Common hand warts
HPV-2
11 (9M. 2F)
Plantar mosaic
HPV-2
6(3M. 3F)
27-8 (range 17-46) 38-9 (range 22-70) 39 3 (range 20-62)
—
11 (SM, 6F)
379
Diagnosis
Healthy controls
(years)
1-5
5-10
>10
2
3
3
1
4
2
4
1
0
5
0
1
—
—
—
—
(range 24-62)
Table 3. Phenotype of peripheral blood mononuclear cells in HPV patients and controls
Number of
Clinica] status
samples
HPV-1 warts No improvement Impmvement Resolved
8 7 3
HPV-2 warts No improvement Improvement Resolved
9 16 5
Normal controls
9
MHC CD3
CD4
CD8
CDla
aass II
45 + 13 34 + 12 2 1 + 6 1-2 + 0-4 19 + 7 53 + 15 37 + 6 24 + 6 3 + 2 20 + 6 49 + 9 4 2 + 5 21+2 1+0 4 21 + 1
CDS 7
CDllb
n + 7 27+10 18 + 6 32 + 10 11+8 23 + 0-5
48+10 35+11 20 + 8 16 + 1 2 9 + 1 3 13 + 8 30+13 44 + 8 36 + 10 21+7 1-9 + 0-9 26 + 8 16 + 8 32 + 10 54+12 37 + 8 25 + 2 2-8 + 1-9 16 + 9 14 + 8 27+17 45 + 7
33 + 11 24+7
3 + 1-9 2 1 + 6
16 + 9 29 + 9
Values represent the mean % positive + lSD,
Phenotyping of PBMC Table 3 shows the expression of various cell surface markers on PBMC from patients in the different clinical groups, expressed as a percentage of the total number of cells analysed. There was no significant difference in the PBMC phenotype of any group at a particular clinical stage. The phenotype of individual patients did vary in some cases, but this was observed both for patients who showed clinical progress as well as those who did not. and there was no correlation between changes in phenotype and change in clinical status {data not shown).
Lymphoproliferation assays The in vitro mitogenic response to Con A was consistently high in all PBMC samples. In those patients and control subjects who reported a clinical history of HSV infections (14 of 37), the proliferative response to HSV
was high, the stimulation index varying from 2 4 to 26-8 with a mean of 7-4. The proliferative responses to HPV antigens of PBMC from patients with warts caused by either HPV-1 or HPV-2 are shown in Figures 1 and 2, respectively. In both cases, most patients showed no responses to HPV antigens at any stage of their wart infection, nor did any ofthe controls. Indeed only 10 samples out ofa total of 100 assays showed a stimulation index greater than 2. Patients who showed a positive response to at least one HPV antigen sample had had their warts for at least 1 year; in one case more than 6 years and in another more than 10. The patients who showed no response to HPV antigens had wart infections for less than 1 year to more than 10 years. Assays were also carried out using the E4 proteins of HPV-1 and HPV-2 expressed as fusion products with ^galactosidase of E. coli as antigen.^'* They proved unsuccessful because, in almost all subjects, there was a strong lymphoproliferative response to the ^-galacto-
LYMPHOPROLIFERATIVE RESPONSES TO HPV
No improvtmmt -
• Improvemsnt .^——» Resolved
Normal controls
CLINICAL STATUS
HaimprovemBfit.
ImprDVBnwnt
». Resolvsd
555
Normal controd
CLINICAL STATUS
Figure 1. hi vitro proliferative responses of PBMC from patients with HPV-1 type warts to whole virion samples of (a) HPV-1 and (b) HPV 2 during changes in clinical status ofthe wart infection. Points joined by a line indicate successive samples from the same patient as the wart infection progresses from no improvement, through improvement, to resolved, ln the normal controls' column, points represent samples from individual members of the control group.
Figure 2. In vitro proliferative responses of PBMC from patients with HPV--2 type warts to whole virion samples of (a) HPV-1 and (b) HPV-2 during changes in clinical status of tbe wart infection. Points joined by a line indicate successive samples from the same patient as the wart infection progresses from no improvement to resolved. The clinical status of one patient changed from improvement, to no improvement, before improving again ( • •).
sidase protein itself which was ofthe same magnitude as the response to the fusion protein. In four cases, exogenous interleukin 2 (IL-2) was added to lymphoproliferation cultures to see if this would augment responses to HPV antigens. IL-^2 was added on day 4 of culture at 10 U/well in 10 ^1 sterile PBS. This had no effect on the negative response of a healthy control, nor on that of a patient with warts caused by HPV-2 which were not improving. In another patient with unimproved warts caused hy HPV-2. a strong response to HPV-2 virions was seen only after the addition of IL-2, and in a third patient whose HPV-2 type warts were improving, a small positive response to HPV1 virions became negative after the addition of IL-2.
tion of plastic-adherent cells and a population enriched for T lymphocytes (T-enriched cells). The phenotype of these populations in the different clinical groups is shown in Figure 3. All T-enriched cell samples showed an increase of approximately 20% for CD3-positive T lymphocytes, compared with the whole PBMC populations, with very few CD3-positive cells remaining in the adherent cell samples. Class II MHC-positive and CD 11 bpositive cells were enriched in the adherent cell samples hy between 20 and 40% compared with the whole PBMC samples, whilst cells hearing these markers were depleted in the T-enriched samples, although not to absolute levels. No differences were apparent between the percentages of positive cells in the two patient groups and the control subjects.
Studies of separated PBMC populations
PBMC samples were further purified to obtain a popula-
Table 4 shows examples ofthe responses to HPV-2 by whole PBMC populations, and reconstituted populations where adherent cells have been re-added to T-enriched
556
F.C.CHARLESON et al.
(b)
those patients without a clinical history of HSV infection, the response of all cell populations to HSV antigen was negative. In those patients with a ciinicai history of HSV infection, the responses of the PBMC and the reconstituted samples were positive while the response to the Tenriched samples was diminished, often to negative values (for example: PBMC 22,936± 7640 c.p.m. S I - 4 - 3 : T-enriched 521 5 ± 9 4 6 c.p.m., SI = 2 0 ) . Thus there was no evidence that the PBMC. or the T-cell population reconstituted with the adherent cell population, of patients with cutaneous warts, were unable to respond to a viral antigen, other than HPV.
100-| 80&0-
to20HPVl Worts
HPV2 Warts
Normal Controls
Discussion mi Worts
Normal Controls
The 26 patients assayed had cutaneous warts which were static, improving or had resolved during the study and, in several cases, more than one specimen of blood was obtained from each patient during the course of treatment. Despite the fact that all patients, with one exception, showed at least improvement of their warts, with a smaller number showing resolution, the in vitro iymphoproliferative responses to HPV-1 or HPV-2 as antigens were predominantly negative. In contrast, the patients responded well to the mitogen Con A. and those with a clinical history of HSV infections also responded to HSV in an equivalent test. Thus, their T cells' responses were not suppressed generally, Phenotyping of their PBMC showed normal percentage ranges for CD4. CDS. MHC class II. CDla. CDllb and C57 positive cells.
Figure 3. Phenotype of separated PBMC populations by flow cytometry using monoclonal antibodies specific for (a) CUl. (b) class II MHC and (c) CDl lb. • , PBMC: • . T-enriched cells; 0, plastic-adherent cells.
cells to form 10% ofthe total cell numbers and sample volume. None ofthe populations in patient 44 A shows a positive response to HPV-2, The PBMC populations in patient 34B show a positive response which is depleted in the T-enriched population and subsequently restored when adherent cells are re-added. Conversely, the PBMC sample in patient 46B shows no response, whereas the T-enriched sample does, although this response is abolished when adherent cells are re-added. These lymphoproliferative patterns of PBMC and T-enriched populations were observed in all clinical groups in response to both HPV-1 and HPV-2 antigens (Table 5). On the other hand, the response to HSV was different. In
44A Patient PBMC alone PBMC + HPV-2 T-enriched cells alone T-enriched cells + HPV-2 Reconstituted cells alone Reconstituted cells + HPV 2 Clinical status
In the first place, it is pertinent to ask whether a positive response to HPV might be expected in such a group of patients. This depends, principally, on viral
Table 4. Lymphoproliferative responses to HPV-2 in separated PBMC populations of three patients
46B
}4B
c.p.m,+ SE
SI
c.p.m.+ SF.
SI
c.p.m.+SE
SI
2940+351 2587 + 459
0-9
1009-1-97 5OJ7 + 395
5-0
2883 + 504 2933 + 392
1-0
310 + 68 173 + 65
0-6
305+14 JS4-I-81
\2
1194+109 8902 + 2866
7-4
2919 + 973 4846+1237
1-7
2248 + 863 691 J+396
3-1
3510+348 2927+K)4
0'5
HPV2 Improving
c.p.m.+SE, mean count per minute + standard error. SI. stimulation index.
HPV2 Resolved
HPVl Improving
LYMPHOPROLIFERATIVE RESPONSES TO HPV
557
Table S. Summary of lymphoproliferative responses to HPV in separated PBMC populations
Number of patients in group
Numher of patients showing no response hy PBMC, or T-enriched cells
Number of patients .showing response by PBMC. no response by T-enriched cells
Number of patients showing no response by PBMC. response by T-enriched cells
Clinical group
Antigen
HPV-1 warts
HPV-1 HPV-2 HSV
8
7 3
4 2 2
1 1 1
3 4 0
HPV-1 HPV-2 HSV
14 17 14
11 9 3
1 1 11
2 7 0
HPV-1 HPV-2 HSV
6 7 4
3 2 2
0 0 2
3 5 0
HPV-2 warts
Normal controls
,
antigens being expressed in the wart, and on the initiation of an immune response, probably by Langerhans cells interacting with T cells. It is generally believed that HPV infects basal cells of the epidermis following abrasion of the skin, and that expression of early viral genes promotes proliferation and delays maturation of superficial keratinocytes, leading to the characteristic aeanthosis and parakeratosis of warts. The control of late gene expression is tightly linked to the state of differentiation of the squamous cells, so that the structural proteins of the virus are only found in the upper layers of the epidermis, the stratum granulosum and stratum corneum. In both HPV-1 and HPV-2 lesions, particularly in the former, many virions are seen, sometimes in crystalline arrays within the nuclei of degenerating keratinocytes in the superficial layers of the stratum corneum.'' Although the cell bodies of most Langerhans cells tend to be located near the basement membrane. Langerhans cells have long dendritic processes which form a network throughout the epidermis, and play a crucial role in immunosurveillance.'^ Thus, despite most virus particles being produced in the upper epidermis, it is probable that the Langerhans cells are still capable of monitoring their presence and initiating appropriate immune responses. However, two studies have indicated that Langerhans cell numbers are reduced in skin warts,^^'** and there is a similar picture in cervical HPV lesions and dysplasia. where Langerhans cell numbers are reduced.'''^" and T-cell numbers are also reduced, with altered subset ratios.^' The reduction in Langerhans ceil numbers may result in ineffective local antigen presentation. HPV may infect Langerhans cells and be cytolytic for them, although this
seems unlikely, as HPV particles have never been seen inside Langerhans cells. It is more likely that migration or recruitment of Langerhans cells may be affected, and the interaction between Langerhans cells, HPV-infected keratinocytes and effector T cells within the lesion is unknown. It would be of particular relevance to assay cytokine levels, which may affect migration and function of particular populations of cells, and also adhesion molecule expression. One of the major difficulties in the study of immune responses in HPV infections has been the lack of an in vitro culture system for HPV which would enable preparation of appropriate antigens and target cells. We used viral particles purified from pooled wart parings which were checked by their density on CsCl gradient centrifugation, by electron microscopy and, in the case of HPV-1, by Western blotting. It is probable that some contamination by host cell proteins remains, to which lymphocytes from some subjects may respond. In addition, oniy the late structural proteins of the virus are present in such a preparation, and the immune response may be generated preferentially to the early proteins which are synthesized at various stages during viral replication. The protein derived from the E4 open reading frame may be particularly important in this context, as it accumulates in the cytoplasm of terminally differentiated keratinocytes where it interacts with the cytokeratin matrix, and can comprise 20-30% of the total cell protein.-^^ However, using a fusion protein composed of E4 of HPV-l or HPV-2 and /i-glactosidase expressed in E. coli. led to disappointing results, as lymphocytes from most subjects responded to the Pgalactosidase itself, to a level indistinguishable from that
558
F.C.CHARLESON et ai
of the fusion protein. In addition, some of the threedimensional epitopes ofthe native protein may be lost on expression as a fusion protein, and these may be crucial for stimulating T-cell responses. An alternative to fusion proteins is short synthetic peptides, which have been used to test T-cell responses to HPV in subjects without known HPV infections.^^ It was found that no single peptide, representing determinants in HPV-16 LI and E6. was recognized by all donors and no individual responded to all the peptides. Because in this study stimulation with IL-2 in the presence of the peptides and feeder cells took place before measuring lymphoproliferative responses to individual peptides, it is possible that primary responses may have been generated in vitro. This approach has not been used, thus far, in patients with cutaneous warts. Although only four patients out of 22 in the group where separated PBMC populations were assayed showed a iymphoproliferative response of PBMC to HPV, 16 became positive if the PBMC were first depleted by plastic adherence, and then T cell-enriched by nylon wool adsorption. Five out of seven ofthe normal control subjects behaved similarly. Equivaient studies using HSV as antigen did not follow this unexpected pattern. All cell populations were negative in subjects without a clinical history of HSV infection. In subjects with a clinical history of HSV infection, the PBMC showed positive lymphoproliferation to HSV, as did the reconstituted cell population of T-enriched cells plus plastic adherent cells, whereas the T-enriched population alone was either negative or much reduced in its lymphoproliferative response. Thus, it is possible that the adherent cell population contains cells which suppress the lymphoproliferative response to HPV. or that the nylon wool column removes some cell population which inhibits T-cell stimulation hy HPV. As can be seen from Figure 3, MHC class II and CDllbpositive cells are not entirely depleted from the Tenriched population by plastic adherence and nylon wool. An alternative method using L-leucine methyl ester had no apparent effect in depleting antigen presenting cells {data not shown).^'* It is possibie that T cells may be subject to an immunoregulatory network in HPV infection, as has been descrihed recently in the T-celi response to human tumours.^^-^^ It was shown, for paraganglioma and melanoma, that fresh peripheral blood lymphocytes were not cytotoxic against autologous tumour ceiis. However, if the iymphocytes were first activated in vitro by co-culture with tumour ceiis in the presence of IT.,-2 and then cloned, the T ceils couid be functionaiiy
characterized for cytotoxic and reguiatory functions. Some T-ceii ciones ampiified cytotoxic T-cell responses, whilst others down-reguiated them. Thus, it may be necessary to activate and cione T cells from patients with warts in order to demonstrate the functionai properties of the T ceiis in the response to HPV infection.
Acknowledgment This work was funded by a grant from the Scottish Home and Heaith Department.
References 1 Bunney MH. Benton EC, Cubie HA. Viral Warts: Biology and Treatment. 2nd edn. Oxford: Oxford University Press. 1992. 2 Orth G, Favre M. Breitburd F et al, Epidermodysplasia verruciformis: a model for the role of papillomaviruses in human cancer. Cold Spring Harbor Conf Cell Prolif 19«0: 7: 259-82, 3 Spencer ES. Andersen HK. Clinically evident, non-terminal infections with herpes viruses and the wart virus in immunosuppressed renal allograft recipients. Br Med} 1970; iii: 251-4. 4 Rlidlinger R. Smith IW, Bunney MH, Hunter lAA, Human papillomavirus infections in a group of renal transplant recipients. Br I Dermatol \98&. 115:681-92. 5 Iwatsuki K, Tagami H. Takigawa M, Yamada M. Plane warts under spontaneous regression, Immunopathologic study on cellular constituents leading to the inflammatory reaction. Arch Dermato! 198(i: 122: 6 T S - 9 .
6 Cubie HA, Serological studies in a student population prone to infection with human papillomavirus. / Hi/g Camfc 1972; 70: 67790. 7 Cuhie HA. Norva! M, Humoral and cellular immunity to papiilomavirus in patients with cervical dysplasia, ) Med Virol 1988; 24: 85-95, 8 Ivanyi L, Morison WL. In vitro lymphocyte stimulation by wart antigen in man. Br / Dermatol 1976; 94: 523-7, 9 Lee AKY, Eisinger M. Cell-mediated immunity (CMI) to human wart virus and wart-associated tissue antigens, Ciin Exp Immunol 1976:26:419-24. 10 Cubie HA. Norval M. Crawford L et al. Lymphoproliferative responses to fusion proteins of human papillomaviruses in patients with cervical intraepithelial neoplasia. Epidemiol Infect 1989; 103: 625-32. 11 Vestey IP, Norval M, Howie SEM et al. Antigen presentation in patients with recrudescent orofacial herpes simplex virus infections. Rr I Dermatol 1990: 122: 33-42, 12 Julius MH. Simpson E, Herzenberg LA, A rapid method for the isolation of functional thymus-derived niurine lymphocytes. Eur / /mmufio/1973: 3:645-9. 1 3 Neill W A. Miller EP, Generation of cytotoxic T cells to cytomegalovirus in patients with cervical intraepithelial neoplasia, Med Lab Sci 1987:44: 141-9. 14 Doorbar |, Campbell D. Grant RJA. Gallimore PH. Identification of thehumanpapilloma virus-la E4 gene products. EMBO] 1986; 5: 355-62. 15 Ptister H, Biology and biochemistry of papillomaviruses. Rev Physiol Biochem Pharmacol 1984:99: 111-81.
LYMPHOPROLIFERATIVE RESPONSES TO HPV
16 Streilein JW, Circuits and signals ofthe skin-associated lymphoid tissues (SALT), } Invest Dermatol 1985: 85: 10s-13s. 17 Bhawan 1. Dayal Y, Bhan AK. Langerhans cells in molluscum contagiosum, verruca vulgaris, plantar wart, and condyloma acuminatum. I Am Acad Dermatol 1986; IS: 645-9. 18 Chardonnet Y. Viac ], Thivolet J. Langerhans cells in human warts. BrIDermatol 1986; 115: 669-75. 19 Tay SK, Jenkins D, Maddox P fl aL Subpopulations of Langerhans cells in cervical neoplasia. Br / Obstet Gynaeco! 1987; 94: 10-1 5, 20 Hughes RG. Norval M. Howie SEM, Expression of major histocompatibility class II antigens by Langerhans cells in cervical intraepithelial neoplasia. / Clin Pathol 1988: 41: 253-9, 21 TaySK.JenkinsD, Maddox P, Singer A. Lymphocyte phenotypes in cervical intraepithelia! neoplasia and human papillomavirus infection. Br } Obstet Cynaecol 1987:94; 16-21. 22 Doorbar J. Ely S. Sterling J et al. Specific interaction hetween HPV-
23
24
25
26
559
16 E1-E4 and cytokeratins results in collapse ofthe epithelial cell intermediate filament network. Nature 1991; 352: 824-7. Strang G. Hickiing JK. Mclndoe GAJ et al. Human T cell responses to human papillomavirus type 16 LI and F6 synthetic peptides: identification of T cell determinants. HLA-DR restriction and virus type specificity. / O n Virol 1990; 7 1 : 4 2 3 - 3 1 , Thiele DL, Kurosaka M, Lipsky PE. Phenotype of the accessory cell necessary for mitogen-stimulated T and B cell responses in human peripheral blood: delineation by its sensitivity to the lysosmotropic agent. L-leucine methyl ester, / Immunol 1983; 1 31: 2282-90. Mukherji B, Guha A. Loomis R, Ergin MT. Cell-mediated amplification and down regulation of cytotoxic immune response against autologous human cancer. //mmnfio/ 1987; 138: 1987-91. Mukherji B, Guha A. Chakraborty NG et al, Clonal analysis of cytotoxic and regulatory T cell responses against human melanoma. / Exp Med 1989: 169: 1961-76,
Lymphoproliferative responses to human papillomaviruses in patients with cutaneous warts F.C.CHARLESON, M.NORVAL. E.C.BENTON* AND J.A.A.HUNTER* Department of Medical Microbiology. University of Edinburgh Medical School, Edinburgh, U.K, 'University of Edinburgh Department of Dermatology, Royal Infirmary of Edinburgh. U.K. Accepted for publication 2 i June 1992
Summary
In vitro lymphoproliferative responses of peripheral blood mononuclear cells (PBMC) from patients with cutaneous warts, caused by infection with human papillomavirus type 1 (HPV-1) or type 2 (HPV-2), were assayed during the course of treatment. Purified HPV-1 and HPV-2 were used as antigens, as well as herpes simplex virus (HSV) and concanavalin A (Con A). All patients had normal percentages of subsets within the PBMC population and normal iymphoproliferative responses to Con A, and those with a clinical history of HSV infections had positive lympboproliferative responses to HSV. Responses to both HPV antigens were poor. Only 10 of 100 assays of PBMC from 26 patients showed a stimulation index greater than 2. Addition of interleukin 2 made little difference in most cases. No correlation of clinical status of warts, i.e. improving, unchanged or resolved, with lymphoproliferation was found. When the PBMC were depleted of plastic-adherent cells and enriched for T cells, some samples which had not shown a lymphoproliferative response to HPV-1 or HPV-2 became positive: this response was abolished when the adherent cells were re-added. Thus it is possible that the adherent cell population from a proportion of patients contains cells which suppress lymphoproliferation, or that an immunoregulatory network is present so that iymphoproliferation does not take place in vitro without prior activation and cloning of T cells.
Warts caused by infection with human papillomavirus (HPV) types are common, affecting 10% ofthe population at any one time. The lesions are predominantly persistent in nature and last for several months before resolving spontaneously. About 35% of skin warts resolve in 6 months but. in some apparently healthy individuals, warts may be extensive, recalcitrant to treatment, and persist for many years.' It appears that the immune response generated against HPV infection is important in the control of the extent and duration of warts. Clinically this is apparent in patients who are immunosuppressed by disease or drugs, as they will often develop extensive, disabling and persistent warts, particularly when cell-mediated immunity is depressed, rather than humoral immunity. The following clinical examples illustrate this point: (i) pregnant women may develop florid warts which often regress spontaneously following delivery; (ii) patients immunosuppressed by diseases such as lymphomas or leukaemia, or those
Correspondenee: IJr M.Norval, Department of Medieal Microbiology. University of Edinburgh Medical School, Teviot Place. Edinburgh EH8 9AG. U,K,
infected with human Immunodeficiency virus, may develop extensive cutaneous and mucocutaneous warts: (iii) patients with epidermodysplasia verruciformis, an autosomal recessive disorder with an associated defect in cell-mediated immunity, have extensive infection with HPV of unusual types:^ and, (iv) patients immunosuppressed therapeutically following renal or cardiac allografting frequently develop extensive and persistent warts. ^^ Histological evidence also points to a role for the immune response in resolution of viral warts. For example, in regressing plane warts caused by HPV-3. an infiltrate of T cells and mononuclear phagocytes is found adjacent to damaged keratinocytes. indicating classical cell-mediated cytotoxicity.'' On the other hand, such an infiammatory cellular infiltrate is infrequent in other skin warts which are regressing, and is seldom seen in mucocutaneous warts. Furthermore, in clinical terms, the immune response to HPV does not seem to be very effective or long-lasting, with no obvious protection from subsequent episodes of infection. Assays to demonstrate specific immune responses against HPV have proved difficult, and have often
551
552
F.C.CHARLESON et al.
yielded inconsistent results, with the patient's previous history of HPV infection being largely unknown. In general, no correlation has been found between antibody titre to HPV and a past or present history of cutaneous warts'' or cervical lesions.'" Two studies have used in vitro lymphoproliferative responses to HPV to assess T-ceil immunity in patients with cutaneous warts. In the first, a small response to semipurified HPV was noted in patients with active warts, and this increased after resolution." In the second, patients with warts of short duration (less than 1 year) showed small responses, although these diminished in patients with warts of longer duration.*^ In addition, some controls who thought they had never had warts were positive. It is difficult, therefore, to correlate accurately HPV history and lymphoproliferation. Similar lymphoproliferation results have been reported in patients with cervical HPV lesions and dysplasia, using a variety of HPV antigen preparations, including several fusion proteins representing early regions ofthe HPV genome. No correlation was found between the clinical diagnosis of cervieal HPV infection and positive responses.'"' In the present study, patients with either simple plantar warts caused by HPV-1 or common hand and plantar mosaic warts caused by HPV-2 were followed during the course of treatment. Peripheral blood mononuclear cells (PBMC) were assayed for their ability to respond to purified HPV in vitro. As several patients showed improvement or resolution of their warts during the study, correlations between lymphoproliferative responses and clinical status could be drawn, PBMC suspensions were depleted of antigen-presenting cells by plastic adherence, and enriched for T cells by adsorption on nylon wool columns. The responses of the T cells, depleted of, and reconstituted with plastic adherent cells, to HPV antigens were ascertained.
Methods Clinical samples
The patients studied were attending the dermatology out-patient department at the Royal Infirmary of Edinburgh for treatment of cutaneous warts, and were otherwise healthy. The type of wart present was determined by clinical examination. Two groups of patients are reported—those with simple plantar warts, caused by HPV-1. and those with common hand warts or mosaic warts, caused by HPV-2, The patients were seen and assessed clinically, usually at monthly intervals, and underwent a variety of treatments for their warts such as freezing with liquid nitrogen, application of salicylic
acid, or intralesional injection of bleomycin. Age- and sex-matched healthy volunteers, with no known cutaneous HPV infection at the time of sampling, acted as controls. Approximately 50 ml of venous blood was collected in preservative-free heparin at each visit, and processed immediately. Antigens
Pooled parings from either simple plantar warts (HPV-1) or plantar mosaic warts (HPV-2) acted as a source of HPV antigen. Purified whole virion samples were prepared by caesium chloride density gradient centrifugation' and stored in phosphate-buffered saline (PBS) at — 20°C. Total protein content was determined by the comparison of virion samples with serial dilutions of bovine serum albumin on silver-stained polyacrylamide gels under reducing conditions. The presence of HPV-1 proteins was confirmed by Western blotting of virion sampies using a monoclonal antibody for the LI protein of HPV-1 (Sera-lab. U.K.), and any HPV-2 samples which contained HPV-1 proteins were discarded. Herpes simplex virus (HSV) antigen was prepared by infecting Vero cells with a clinical isolate of HSV type 1 for 24-48 h at 37°C; the number of plaque forming units (p.f.u.)/ml was assayed by titration in Vero cells before inactivation with ultraviolet B irradiation of 300 m]/cm^ (HSV titre before inactivation was 10" p.f.u./ml)," The lysate from uninfected Vero cells acted as a negative control.'^ Concanavalin A (Con A) was obtained from Sigma. Separation of PBMC
PBMC were separated by centrifugation on 'Lymphopaque' (Nyegaard. Oslo) and washed three times in sterile PBS containing 5 U/ml of heparin before resuspension in assay medium (see below). Plastic-adherent cells were isolated from PBMC by incubation in "Falcon' tissue-culture fiasks (Becton Dickinson, U.K.) for 1 h at 37°C." Non-adherent cells were further purified by incubation on nylon wool (Fenwal. Illinois) for 2 h at 37°C to provide a sample which was enriched for T lymphocytes (T-enriched cells).'- All cell samples were resuspended at 10^/ml in assay medium whieh consisted of RPMI 1 640 (Northumbria Biologicals. U.K.) with 1 5% autologous plasma plus 5 U/ml heparin, 2 mM glutamine, 0-05 mM 2-mercaptoethanol, 100 lU/penicillin, 200 /ig/ml streptomycin, 1()() fig/ml gentamicin and 2 /ig/ml Fungizone®. The ceil populations were monitored at each stage by fiow cytometry using monoclonal antibodies to various cell surface markers (see below).
LYMPHOPROLIFERATIVE RESPONSES TO HPV
Lymphoproliferation
assay
Phenotyping of cell populations
Cells were seeded into round-bottomed 96-well 'Cellcult' plates (Sterilin, U.K.) in 200 ^1 volumes (2 x 10^ cells) in the presence or absence of antigen. Con A was added at 10 /^g/ml and HSV at 1 p.f.u./cell, with the equivalent amount of uninfected Vero cell iysate as a negative controi. Approximately 1 /ig of HPV-1 or HPV2 in 20 /d PBS/ml of cell suspension was used, with 20 /il/ml sterile PBS as a negative control. PBMC were assayed in quintuplicate for each antigen, whiist Tenriched cells were assayed in tripiicate to conserve cell yieids. A reconstituted ceii sampie consisting of Tenriched cells plus plastic-adherent cells (added to form 10% of the total cell number within the same sample volume) was also assayed in triplicate. Cells were incubated in 5% CO2 at i7°C for 6 days when cultured with Con A or HSV, or 7 days when cultured with HPV antigens. Tritiated thymidine (0 75 /iCi in 10 fi\ PBS/ well) was added for the final 18-24 h of eulture. Plates were harvested on to 'Titertek' filter paper (Flow Laboratories) using an 'Automash 2000' cell harvester (Dynatech) and discs were counted for 1 min in a 'Tri-Carb' scintillation counter (Canberra Packard), Some results were expressed as a stimulation index (SI) where:
SI-
mean count per minute (c.p,m.) in presence of antigen mean c.p.m. in absence of antigen An SI value of 2 or more was taken as positive.
Table 1. Monoclonal antibodies used in phcnntypiiii' by flow cytometry
553
Cell surface marker CD 3 CD4 CD8 CDla MHC class 11
CD57 CDIlb
Cells were incubated with various mouse monoclonal antibodies, followed by sheep anti-mouse Ig labelled with fiuorescein isothiocyanate (Sigma), according to the method of Neill and Miller,'' except that the first step was performed at 4°C overnight. The monoclonal antibodies which were used are shown in Tabie 1. Cells were washed twice, resuspended at 5 x 10^ cells/mi in coid PBS, fixed with 1% formaldehyde, and approximately 10* cells analysed on a Coulter EPICS 'C fiow cytometer (output 100 mW at 488 nm. fiow rate 400-500 cells/s).
Results Clinical status
The clinical details of the patients and healthy volunteers studied are shown in Table 2. The results for patients with common hand warts and plantar mosaic warts are reported as a single group because both are caused by the same HPV type, and no differences were observed between them. In some cases results have been divided into three groups so that immune responses could be correlated with changes in clinical status. These were: (a) no improvement: (b) improvement (i.e, a deerease in the number of warts or degree of hyperkeratosis) or, (c) complete resolution ('resolved'). Patients were monitored over varying lengths of time according to changes in their clinical status.
Target cell T lymphocytes T-helper lymphocytes T-supprcssor/cytotoxic lymphocytes Thymocytes Dendritk" ceils Monocytes/macTophages B lymphocytes Activated T lymphocytes LGL* NK cellsj: K cells t Monocytes T-suppressor lymphocytes LGL/NK cells
Monoclonal antibody
Source
DAKO-Ti DAK0T4 DAK0-T8 BAK0-T6
Dakopatts Dakopatts Dakopatts Dakopatts
DAf).2il
Dr K,Guy§
Uu 7
Becton Dickinson
OK Ml
Becton Dickinson
* LCI., large granular lymphocytes, t NK, natural killer cells, t K, killer cells. (j Kind gift from Dr Keitb Cuy. Western General Hospitai. Edinburgh.
554
F.C.CHARLESON et al
Table 2. Clinical history of study groups Duration of wart(s) (years) HPV type
Number in group (male, female)
Mean age
Simple plantar
HPV-1
9 f5M, 4F)
Common hand warts
HPV-2
11 (9M. 2F)
Plantar mosaic
HPV-2
6(3M. 3F)
27-8 (range 17-46) 38-9 (range 22-70) 39 3 (range 20-62)
—
11 (SM, 6F)
379
Diagnosis
Healthy controls
(years)
1-5
5-10
>10
2
3
3
1
4
2
4
1
0
5
0
1
—
—
—
—
(range 24-62)
Table 3. Phenotype of peripheral blood mononuclear cells in HPV patients and controls
Number of
Clinica] status
samples
HPV-1 warts No improvement Impmvement Resolved
8 7 3
HPV-2 warts No improvement Improvement Resolved
9 16 5
Normal controls
9
MHC CD3
CD4
CD8
CDla
aass II
45 + 13 34 + 12 2 1 + 6 1-2 + 0-4 19 + 7 53 + 15 37 + 6 24 + 6 3 + 2 20 + 6 49 + 9 4 2 + 5 21+2 1+0 4 21 + 1
CDS 7
CDllb
n + 7 27+10 18 + 6 32 + 10 11+8 23 + 0-5
48+10 35+11 20 + 8 16 + 1 2 9 + 1 3 13 + 8 30+13 44 + 8 36 + 10 21+7 1-9 + 0-9 26 + 8 16 + 8 32 + 10 54+12 37 + 8 25 + 2 2-8 + 1-9 16 + 9 14 + 8 27+17 45 + 7
33 + 11 24+7
3 + 1-9 2 1 + 6
16 + 9 29 + 9
Values represent the mean % positive + lSD,
Phenotyping of PBMC Table 3 shows the expression of various cell surface markers on PBMC from patients in the different clinical groups, expressed as a percentage of the total number of cells analysed. There was no significant difference in the PBMC phenotype of any group at a particular clinical stage. The phenotype of individual patients did vary in some cases, but this was observed both for patients who showed clinical progress as well as those who did not. and there was no correlation between changes in phenotype and change in clinical status {data not shown).
Lymphoproliferation assays The in vitro mitogenic response to Con A was consistently high in all PBMC samples. In those patients and control subjects who reported a clinical history of HSV infections (14 of 37), the proliferative response to HSV
was high, the stimulation index varying from 2 4 to 26-8 with a mean of 7-4. The proliferative responses to HPV antigens of PBMC from patients with warts caused by either HPV-1 or HPV-2 are shown in Figures 1 and 2, respectively. In both cases, most patients showed no responses to HPV antigens at any stage of their wart infection, nor did any ofthe controls. Indeed only 10 samples out ofa total of 100 assays showed a stimulation index greater than 2. Patients who showed a positive response to at least one HPV antigen sample had had their warts for at least 1 year; in one case more than 6 years and in another more than 10. The patients who showed no response to HPV antigens had wart infections for less than 1 year to more than 10 years. Assays were also carried out using the E4 proteins of HPV-1 and HPV-2 expressed as fusion products with ^galactosidase of E. coli as antigen.^'* They proved unsuccessful because, in almost all subjects, there was a strong lymphoproliferative response to the ^-galacto-
LYMPHOPROLIFERATIVE RESPONSES TO HPV
No improvtmmt -
• Improvemsnt .^——» Resolved
Normal controls
CLINICAL STATUS
HaimprovemBfit.
ImprDVBnwnt
». Resolvsd
555
Normal controd
CLINICAL STATUS
Figure 1. hi vitro proliferative responses of PBMC from patients with HPV-1 type warts to whole virion samples of (a) HPV-1 and (b) HPV 2 during changes in clinical status ofthe wart infection. Points joined by a line indicate successive samples from the same patient as the wart infection progresses from no improvement, through improvement, to resolved, ln the normal controls' column, points represent samples from individual members of the control group.
Figure 2. In vitro proliferative responses of PBMC from patients with HPV--2 type warts to whole virion samples of (a) HPV-1 and (b) HPV-2 during changes in clinical status of tbe wart infection. Points joined by a line indicate successive samples from the same patient as the wart infection progresses from no improvement to resolved. The clinical status of one patient changed from improvement, to no improvement, before improving again ( • •).
sidase protein itself which was ofthe same magnitude as the response to the fusion protein. In four cases, exogenous interleukin 2 (IL-2) was added to lymphoproliferation cultures to see if this would augment responses to HPV antigens. IL-^2 was added on day 4 of culture at 10 U/well in 10 ^1 sterile PBS. This had no effect on the negative response of a healthy control, nor on that of a patient with warts caused by HPV-2 which were not improving. In another patient with unimproved warts caused hy HPV-2. a strong response to HPV-2 virions was seen only after the addition of IL-2, and in a third patient whose HPV-2 type warts were improving, a small positive response to HPV1 virions became negative after the addition of IL-2.
tion of plastic-adherent cells and a population enriched for T lymphocytes (T-enriched cells). The phenotype of these populations in the different clinical groups is shown in Figure 3. All T-enriched cell samples showed an increase of approximately 20% for CD3-positive T lymphocytes, compared with the whole PBMC populations, with very few CD3-positive cells remaining in the adherent cell samples. Class II MHC-positive and CD 11 bpositive cells were enriched in the adherent cell samples hy between 20 and 40% compared with the whole PBMC samples, whilst cells hearing these markers were depleted in the T-enriched samples, although not to absolute levels. No differences were apparent between the percentages of positive cells in the two patient groups and the control subjects.
Studies of separated PBMC populations
PBMC samples were further purified to obtain a popula-
Table 4 shows examples ofthe responses to HPV-2 by whole PBMC populations, and reconstituted populations where adherent cells have been re-added to T-enriched
556
F.C.CHARLESON et al.
(b)
those patients without a clinical history of HSV infection, the response of all cell populations to HSV antigen was negative. In those patients with a ciinicai history of HSV infection, the responses of the PBMC and the reconstituted samples were positive while the response to the Tenriched samples was diminished, often to negative values (for example: PBMC 22,936± 7640 c.p.m. S I - 4 - 3 : T-enriched 521 5 ± 9 4 6 c.p.m., SI = 2 0 ) . Thus there was no evidence that the PBMC. or the T-cell population reconstituted with the adherent cell population, of patients with cutaneous warts, were unable to respond to a viral antigen, other than HPV.
100-| 80&0-
to20HPVl Worts
HPV2 Warts
Normal Controls
Discussion mi Worts
Normal Controls
The 26 patients assayed had cutaneous warts which were static, improving or had resolved during the study and, in several cases, more than one specimen of blood was obtained from each patient during the course of treatment. Despite the fact that all patients, with one exception, showed at least improvement of their warts, with a smaller number showing resolution, the in vitro iymphoproliferative responses to HPV-1 or HPV-2 as antigens were predominantly negative. In contrast, the patients responded well to the mitogen Con A. and those with a clinical history of HSV infections also responded to HSV in an equivalent test. Thus, their T cells' responses were not suppressed generally, Phenotyping of their PBMC showed normal percentage ranges for CD4. CDS. MHC class II. CDla. CDllb and C57 positive cells.
Figure 3. Phenotype of separated PBMC populations by flow cytometry using monoclonal antibodies specific for (a) CUl. (b) class II MHC and (c) CDl lb. • , PBMC: • . T-enriched cells; 0, plastic-adherent cells.
cells to form 10% ofthe total cell numbers and sample volume. None ofthe populations in patient 44 A shows a positive response to HPV-2, The PBMC populations in patient 34B show a positive response which is depleted in the T-enriched population and subsequently restored when adherent cells are re-added. Conversely, the PBMC sample in patient 46B shows no response, whereas the T-enriched sample does, although this response is abolished when adherent cells are re-added. These lymphoproliferative patterns of PBMC and T-enriched populations were observed in all clinical groups in response to both HPV-1 and HPV-2 antigens (Table 5). On the other hand, the response to HSV was different. In
44A Patient PBMC alone PBMC + HPV-2 T-enriched cells alone T-enriched cells + HPV-2 Reconstituted cells alone Reconstituted cells + HPV 2 Clinical status
In the first place, it is pertinent to ask whether a positive response to HPV might be expected in such a group of patients. This depends, principally, on viral
Table 4. Lymphoproliferative responses to HPV-2 in separated PBMC populations of three patients
46B
}4B
c.p.m,+ SE
SI
c.p.m.+ SF.
SI
c.p.m.+SE
SI
2940+351 2587 + 459
0-9
1009-1-97 5OJ7 + 395
5-0
2883 + 504 2933 + 392
1-0
310 + 68 173 + 65
0-6
305+14 JS4-I-81
\2
1194+109 8902 + 2866
7-4
2919 + 973 4846+1237
1-7
2248 + 863 691 J+396
3-1
3510+348 2927+K)4
0'5
HPV2 Improving
c.p.m.+SE, mean count per minute + standard error. SI. stimulation index.
HPV2 Resolved
HPVl Improving
LYMPHOPROLIFERATIVE RESPONSES TO HPV
557
Table S. Summary of lymphoproliferative responses to HPV in separated PBMC populations
Number of patients in group
Numher of patients showing no response hy PBMC, or T-enriched cells
Number of patients .showing response by PBMC. no response by T-enriched cells
Number of patients showing no response by PBMC. response by T-enriched cells
Clinical group
Antigen
HPV-1 warts
HPV-1 HPV-2 HSV
8
7 3
4 2 2
1 1 1
3 4 0
HPV-1 HPV-2 HSV
14 17 14
11 9 3
1 1 11
2 7 0
HPV-1 HPV-2 HSV
6 7 4
3 2 2
0 0 2
3 5 0
HPV-2 warts
Normal controls
,
antigens being expressed in the wart, and on the initiation of an immune response, probably by Langerhans cells interacting with T cells. It is generally believed that HPV infects basal cells of the epidermis following abrasion of the skin, and that expression of early viral genes promotes proliferation and delays maturation of superficial keratinocytes, leading to the characteristic aeanthosis and parakeratosis of warts. The control of late gene expression is tightly linked to the state of differentiation of the squamous cells, so that the structural proteins of the virus are only found in the upper layers of the epidermis, the stratum granulosum and stratum corneum. In both HPV-1 and HPV-2 lesions, particularly in the former, many virions are seen, sometimes in crystalline arrays within the nuclei of degenerating keratinocytes in the superficial layers of the stratum corneum.'' Although the cell bodies of most Langerhans cells tend to be located near the basement membrane. Langerhans cells have long dendritic processes which form a network throughout the epidermis, and play a crucial role in immunosurveillance.'^ Thus, despite most virus particles being produced in the upper epidermis, it is probable that the Langerhans cells are still capable of monitoring their presence and initiating appropriate immune responses. However, two studies have indicated that Langerhans cell numbers are reduced in skin warts,^^'** and there is a similar picture in cervical HPV lesions and dysplasia. where Langerhans cell numbers are reduced.'''^" and T-cell numbers are also reduced, with altered subset ratios.^' The reduction in Langerhans ceil numbers may result in ineffective local antigen presentation. HPV may infect Langerhans cells and be cytolytic for them, although this
seems unlikely, as HPV particles have never been seen inside Langerhans cells. It is more likely that migration or recruitment of Langerhans cells may be affected, and the interaction between Langerhans cells, HPV-infected keratinocytes and effector T cells within the lesion is unknown. It would be of particular relevance to assay cytokine levels, which may affect migration and function of particular populations of cells, and also adhesion molecule expression. One of the major difficulties in the study of immune responses in HPV infections has been the lack of an in vitro culture system for HPV which would enable preparation of appropriate antigens and target cells. We used viral particles purified from pooled wart parings which were checked by their density on CsCl gradient centrifugation, by electron microscopy and, in the case of HPV-1, by Western blotting. It is probable that some contamination by host cell proteins remains, to which lymphocytes from some subjects may respond. In addition, oniy the late structural proteins of the virus are present in such a preparation, and the immune response may be generated preferentially to the early proteins which are synthesized at various stages during viral replication. The protein derived from the E4 open reading frame may be particularly important in this context, as it accumulates in the cytoplasm of terminally differentiated keratinocytes where it interacts with the cytokeratin matrix, and can comprise 20-30% of the total cell protein.-^^ However, using a fusion protein composed of E4 of HPV-l or HPV-2 and /i-glactosidase expressed in E. coli. led to disappointing results, as lymphocytes from most subjects responded to the Pgalactosidase itself, to a level indistinguishable from that
558
F.C.CHARLESON et ai
of the fusion protein. In addition, some of the threedimensional epitopes ofthe native protein may be lost on expression as a fusion protein, and these may be crucial for stimulating T-cell responses. An alternative to fusion proteins is short synthetic peptides, which have been used to test T-cell responses to HPV in subjects without known HPV infections.^^ It was found that no single peptide, representing determinants in HPV-16 LI and E6. was recognized by all donors and no individual responded to all the peptides. Because in this study stimulation with IL-2 in the presence of the peptides and feeder cells took place before measuring lymphoproliferative responses to individual peptides, it is possible that primary responses may have been generated in vitro. This approach has not been used, thus far, in patients with cutaneous warts. Although only four patients out of 22 in the group where separated PBMC populations were assayed showed a iymphoproliferative response of PBMC to HPV, 16 became positive if the PBMC were first depleted by plastic adherence, and then T cell-enriched by nylon wool adsorption. Five out of seven ofthe normal control subjects behaved similarly. Equivaient studies using HSV as antigen did not follow this unexpected pattern. All cell populations were negative in subjects without a clinical history of HSV infection. In subjects with a clinical history of HSV infection, the PBMC showed positive lymphoproliferation to HSV, as did the reconstituted cell population of T-enriched cells plus plastic adherent cells, whereas the T-enriched population alone was either negative or much reduced in its lymphoproliferative response. Thus, it is possible that the adherent cell population contains cells which suppress the lymphoproliferative response to HPV. or that the nylon wool column removes some cell population which inhibits T-cell stimulation hy HPV. As can be seen from Figure 3, MHC class II and CDllbpositive cells are not entirely depleted from the Tenriched population by plastic adherence and nylon wool. An alternative method using L-leucine methyl ester had no apparent effect in depleting antigen presenting cells {data not shown).^'* It is possibie that T cells may be subject to an immunoregulatory network in HPV infection, as has been descrihed recently in the T-celi response to human tumours.^^-^^ It was shown, for paraganglioma and melanoma, that fresh peripheral blood lymphocytes were not cytotoxic against autologous tumour ceiis. However, if the iymphocytes were first activated in vitro by co-culture with tumour ceiis in the presence of IT.,-2 and then cloned, the T ceils couid be functionaiiy
characterized for cytotoxic and reguiatory functions. Some T-ceii ciones ampiified cytotoxic T-cell responses, whilst others down-reguiated them. Thus, it may be necessary to activate and cione T cells from patients with warts in order to demonstrate the functionai properties of the T ceiis in the response to HPV infection.
Acknowledgment This work was funded by a grant from the Scottish Home and Heaith Department.
References 1 Bunney MH. Benton EC, Cubie HA. Viral Warts: Biology and Treatment. 2nd edn. Oxford: Oxford University Press. 1992. 2 Orth G, Favre M. Breitburd F et al, Epidermodysplasia verruciformis: a model for the role of papillomaviruses in human cancer. Cold Spring Harbor Conf Cell Prolif 19«0: 7: 259-82, 3 Spencer ES. Andersen HK. Clinically evident, non-terminal infections with herpes viruses and the wart virus in immunosuppressed renal allograft recipients. Br Med} 1970; iii: 251-4. 4 Rlidlinger R. Smith IW, Bunney MH, Hunter lAA, Human papillomavirus infections in a group of renal transplant recipients. Br I Dermatol \98&. 115:681-92. 5 Iwatsuki K, Tagami H. Takigawa M, Yamada M. Plane warts under spontaneous regression, Immunopathologic study on cellular constituents leading to the inflammatory reaction. Arch Dermato! 198(i: 122: 6 T S - 9 .
6 Cubie HA, Serological studies in a student population prone to infection with human papillomavirus. / Hi/g Camfc 1972; 70: 67790. 7 Cuhie HA. Norva! M, Humoral and cellular immunity to papiilomavirus in patients with cervical dysplasia, ) Med Virol 1988; 24: 85-95, 8 Ivanyi L, Morison WL. In vitro lymphocyte stimulation by wart antigen in man. Br / Dermatol 1976; 94: 523-7, 9 Lee AKY, Eisinger M. Cell-mediated immunity (CMI) to human wart virus and wart-associated tissue antigens, Ciin Exp Immunol 1976:26:419-24. 10 Cubie HA. Norval M. Crawford L et al. Lymphoproliferative responses to fusion proteins of human papillomaviruses in patients with cervical intraepithelial neoplasia. Epidemiol Infect 1989; 103: 625-32. 11 Vestey IP, Norval M, Howie SEM et al. Antigen presentation in patients with recrudescent orofacial herpes simplex virus infections. Rr I Dermatol 1990: 122: 33-42, 12 Julius MH. Simpson E, Herzenberg LA, A rapid method for the isolation of functional thymus-derived niurine lymphocytes. Eur / /mmufio/1973: 3:645-9. 1 3 Neill W A. Miller EP, Generation of cytotoxic T cells to cytomegalovirus in patients with cervical intraepithelial neoplasia, Med Lab Sci 1987:44: 141-9. 14 Doorbar |, Campbell D. Grant RJA. Gallimore PH. Identification of thehumanpapilloma virus-la E4 gene products. EMBO] 1986; 5: 355-62. 15 Ptister H, Biology and biochemistry of papillomaviruses. Rev Physiol Biochem Pharmacol 1984:99: 111-81.
LYMPHOPROLIFERATIVE RESPONSES TO HPV
16 Streilein JW, Circuits and signals ofthe skin-associated lymphoid tissues (SALT), } Invest Dermatol 1985: 85: 10s-13s. 17 Bhawan 1. Dayal Y, Bhan AK. Langerhans cells in molluscum contagiosum, verruca vulgaris, plantar wart, and condyloma acuminatum. I Am Acad Dermatol 1986; IS: 645-9. 18 Chardonnet Y. Viac ], Thivolet J. Langerhans cells in human warts. BrIDermatol 1986; 115: 669-75. 19 Tay SK, Jenkins D, Maddox P fl aL Subpopulations of Langerhans cells in cervical neoplasia. Br / Obstet Gynaeco! 1987; 94: 10-1 5, 20 Hughes RG. Norval M. Howie SEM, Expression of major histocompatibility class II antigens by Langerhans cells in cervical intraepithelial neoplasia. / Clin Pathol 1988: 41: 253-9, 21 TaySK.JenkinsD, Maddox P, Singer A. Lymphocyte phenotypes in cervical intraepithelia! neoplasia and human papillomavirus infection. Br } Obstet Cynaecol 1987:94; 16-21. 22 Doorbar J. Ely S. Sterling J et al. Specific interaction hetween HPV-
23
24
25
26
559
16 E1-E4 and cytokeratins results in collapse ofthe epithelial cell intermediate filament network. Nature 1991; 352: 824-7. Strang G. Hickiing JK. Mclndoe GAJ et al. Human T cell responses to human papillomavirus type 16 LI and F6 synthetic peptides: identification of T cell determinants. HLA-DR restriction and virus type specificity. / O n Virol 1990; 7 1 : 4 2 3 - 3 1 , Thiele DL, Kurosaka M, Lipsky PE. Phenotype of the accessory cell necessary for mitogen-stimulated T and B cell responses in human peripheral blood: delineation by its sensitivity to the lysosmotropic agent. L-leucine methyl ester, / Immunol 1983; 1 31: 2282-90. Mukherji B, Guha A. Loomis R, Ergin MT. Cell-mediated amplification and down regulation of cytotoxic immune response against autologous human cancer. //mmnfio/ 1987; 138: 1987-91. Mukherji B, Guha A. Chakraborty NG et al, Clonal analysis of cytotoxic and regulatory T cell responses against human melanoma. / Exp Med 1989: 169: 1961-76,
