© 2002 American Thoracic Society
Peripheral Blood Mononuclear Cell Proliferation to Heat Shock Protein–70 Derived from Autologous Lung CarcinomaChest Department, Department of Pathology, IRIBHN Statistical Unit, and Department of Thoracic Surgery, Erasme University Hospital, CUB Erasme, Brussels, Belgium Correspondence and requests for reprints should be addressed to Alain Michils, M.D., Chest Department, Erasme University Hospital, CUB Erasme, 808 Route de Lennik, B-1070 Brussels, Belgium. E-mail: amichils{at}ulb.ac.be
In animals, tumor-derived heat shock proteins (HSP) induce immune-mediated protection against autologous cancer. We investigated whether HSP-70 derived from human lung carcinoma are also complexed to tumor-specific antigens. Peripheral blood mononuclear cells collected 10 days after surgery from patients with lung cancer were stimulated with HSP-70 purified from autologous and heterologous tumors. The stimulation index (SI) obtained when stimulating cells with autologous tumor-derived HSP-70 averaged 3.07 ± 0.75 in patients with lung cancer and 1.57 ± 0.33 in control subjects (p < 0.001 by analysis of variance). No significant stimulation was observed with HSP-70 derived either from the majority of heterologous tumors or from autologous tumor-free lung tissue. SI decreased from 3.59 ± 0.65 to 1.65 ± 0.38 in six patients tested again 3 months after surgery (p = 0.02 by Wilcoxon test for paired data). HSP-70 derived from lung carcinoma are shown to be associated with T cell antigens. The T cell reactivity appears transient and restricted to antigens complexed to HSP-70 derived from autologous tumors only. This suggests that the antigenicity of human lung tumors is unique, which may be crucial for the design of new vaccines.
Key Words: lung cancer • heat shock proteins (HSP-70) • lymphoblast transformation test • cancer immunotherapy
Heat shock proteins (HSP) derived from animals with cancers elicit protective immunity specific for the cancer from which the HSP are purified (1–9). The mechanism of this autologous restricted protection has been clarified: HSP chaperone, a broad spectrum of peptides generated within the cells from which they have been purified, including tumor rejection antigens (10). Upon injection, peptides chaperoned by HSP are channeled into the major histocompatibility complex class I processing pathway in antigen-presenting cells and elicit protective immunity against the autologous tumor (11–17), boosted by the immunoadjuvant function of HSP (5, 18). This strategy, currently tested in patients suffering from several advanced cancers (19), could also be used in the adjuvant setting after surgical excision of lung carcinoma in patients with a high risk of relapse. However, there has been no definite evidence so far that HSP derived from human lung carcinoma are complexed with T cell antigens. Therefore, we studied whether HSP-70, expressed in large amounts by human lung carcinoma (20–24), is associated with antigens that are able to induce T cell proliferation in vitro and whether this T cell reactivity would display the autologous restricted specificity documented in animals.
Tissue Samples Tumor and corresponding nontumor lung samples were obtained from 11 patients (age range 51–72, 8 with squamous cell carcinoma and 3 with adenocarcinoma) at the time of open thoracotomy for a previously untreated resectable lung carcinoma. Tissue samples were immediately frozen and stored at -70°C until use. Tumors were staged at surgery according to the International Staging System for Lung Cancer (25) and histologically classified according to the World Health Organization study (26). The study was approved by the local institutional ethics committee.
Fragments of tissue, 0.3 g, were homogenized at 4°C in Tris–HCl (20 mM, pH 7.6) containing 1 mM ethylenediamine-tetraacetic acid and 1 UI/ml aprotinin (Trasylol; Bayer, Leverkusen, Germany). The homogenates were centrifuged (2,300 x g, 20 minutes) at 4°C, and 0.3% wt/vol gelatin (Hemacel; Hoechst, Brussels, Belgium) was added to the supernatants; aliquots were stored at -70°C until use.
HSP-70 was purified from cell lysates by adenosine diphosphate affinity chromatography (27, 28). Purified HSP-70 extracts were characterized by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and Immunoblot.
Heparinized blood samples collected from 11 patients (10 days after surgery) and 11 control subjects were diluted with an equal volume of Hanks' solution free of calcium and magnesium. Peripheral blood mononuclear cells (PBMC) were isolated by density centrifugation on Ficoll–Hypaque (Nycomed Pharma AS, Oslo, Norway, d = 1.077 µg/ml), washed three times in Hanks' solution, and resuspended in culture medium (Roswell Park Memorial Institute; Biowithaker, Verviers, Belgium) supplemented with heat-inactivated human serum (10% vol/vol), bicarbonate (1 g), gentamycine (40 µg/ml), and L-glutamine (2 mM) (pH 7.2). PBMC (3 x 106/ml) were stimulated in 96-well flat-bottom trays in a volume of 200 µl, with the following extracts: human recombinant HSP-70 (Stress-Gen Biotechnology Corporation, Victoria, BC, Canada), cell lysates from tumoral and nontumoral lung tissue, HSP-70 derived from autologous tumors with or without previous exposure to adenosine triphosphate (ATP) (3 mM final dilution, 2 hours), HSP-70 derived from tumor-free lung tissue, and HSP-70 derived from heterologous lung tumors (n = 6) (100, 10, and 1 ng/ml final dilution). Positive parallel controls were performed with candidin (500 µg/ml final dilution) and varidase (100 µg/ml final dilution). Cultures were set up in triplicate and kept at 37°C in 5% carbon dioxide /95% air. Cell proliferation was evaluated by measuring DNA synthesis using an enzyme-linked immunosorbent assay (ELISA) system (Biothrak-Cell Proliferation ELISA system; Amersham International, Buckinghamshire, UK) (29). Proliferation was expressed as a stimulation index (SI), which was calculated as the ratio of mean optical density of stimulated to unstimulated cultures. In accordance with the literature (30, 31), a proliferative response with an SI > 2 was regarded as positive. To record the time evolution of cell proliferation to autologous tumor-derived HSP-70, PBMC collected from six patients who did show a significant immediate response were restimulated with the autologous extract 3 months later.
Endotoxin levels were determined in the extracts used for stimulation by a quantitative chromogenic assay (limulus amoebocyte lysate coatest endotoxin; Chromogenics, Mölndal, Sweden) (32).
SDS-PAGE and Immunoblotting Coomassie staining of SDS-PAGE followed by immunoblotting using antibody against human HSP-70 shows that the protein preparations consist mainly of HSP-70 molecules (Figure 1) . However, in three preparations HSP-70 was not identified by this technique, although detectable levels were documented using competitive ELISA (24). Furthermore, 54-kD molecules were detected in some extracts. It is currently not known whether these proteins result from HSP-70 denaturation.
Autologous Stimulations Figure 2 shows the SI obtained on stimulating cells from patients with lung cancer and healthy control subjects, with optimal dilution of the various extracts. The SI obtained with varidase (positive control) averaged 9.06 ± 2.75 in patients with cancer and 10.13 ± 1.94 in control subjects (p > 0.1 by one-way analysis of variance [ANOVA], data not shown). No significant proliferation was observed with human recombinant HSP-70 (mean SI 1.27 ± 0.35 versus 1.03 ± 0.31; p > 0.1 by one-way ANOVA), lysates from nontumoral tissue (mean SI 1.24 ± 0.21 versus 1.19 ± 0.24; p > 0.1 by one-way ANOVA), HSP-70 purified from nontumoral lung tissue (mean SI 1.29 ± 0.17 versus 1.09 ± 0.16; p > 0.1 by one-way ANOVA), or lysates from lung carcinoma tissue (mean SI 1.39 ± 0.15 versus 1.28 ± 0.09; p > 0.1 by one-way ANOVA). The SI obtained when stimulating cells with HSP-70 purified from autologous tumor averaged 3.07 ± 0.75 in patients with lung cancer, compared with 1.57 ± 0.33 in healthy control subjects (p < 0.001 by one-way ANOVA). This was also significantly different from the SI observed when stimulating cells from patients with lung cancer with all the extracts mentioned previously (p < 0.001 by ANOVA with one within-subject factor).
Figure 3 shows the effect of ATP exposure on tumor-derived HSP-70 stimulating capacities. Mean SI decreased from 3.01 ± 0.28 to 2.04 ± 0.63 after exposure to ATP (p < 0.02 by Wilcoxon test for paired data). However, in two cases, the SI were not significantly influenced by HSP-70 incubation with ATP (mean SI 3.02 versus 2.94 with and without ATP, respectively).
Figure 4 shows the time evolution of cell proliferation induced by autologous tumor-derived HSP-70. The SI decreased from 3.58 ± 0.65 to 1.65 ± 0.38 when cells from six patients were stimulated 10 days and 3 months after surgery, respectively (p = 0.028 by Wilcoxon test for paired data). The SI of the latter did not differ significantly from that obtained when stimulating cells from healthy control subjects (p > 0.05 by nonparametric Mann–Whitney test).
Heterologous Stimulations Figure 5 shows the distributions of SI obtained when PBMC from six patients were stimulated 10 days after surgery with HSP-70 derived from autologous and heterologous (n = 6) tumors, respectively. Significant stimulation occurred with autologous HSP-70, with mean SI reaching 3.76 ± 0.64, whereas no proliferative response toward heterologous HSP was observed in most cases (mean SI reaching 1.40 ± 1.13; p = 0.04 by Friedman rank test). In addition, no significant proliferation was detected when stimulating cells from control subjects with all the extracts mentioned previously (mean SI 1.15 ± 0.15; data not shown, p < 0.002 by nonparametric test of Mann–Whitney). Considering SI greater than 2 as significant, a proliferative response to HSP-70 derived from heterologous tumors occurred in five cases only. SI obtained with autologous extracts averaged 3.76 ± 0.64, compared with 1.97 ± 0.15, 1.31 ± 0.19, 1.86 ± 0.31, 1.21 ± 0.09, 1.30 ± 0.12, and 2.04 ± 1.16 with the six heterologous extracts (HE) (p < 0.05 by Wilcoxon test for paired data, except for HE6 because of a high level of T cell response from Patient 4).
Endotoxin Levels Endotoxin levels averaged 180 ± 113 and 216 ± 91 pg/ml in HSP-70 purified from tumoral and nontumoral tissues, respectively, and 229 ± 116 pg/ml in cell lysates from tumoral tissues (p > 0.1 by one-way ANOVA).
Our data document a transient in vitro proliferative response of PBMC toward HSP-70 purified from autologous tumors from patients treated surgically for lung cancer. Although antimelanoma T cell lines (33) can proliferate in vitro in response to autologous tumor-derived HSP, as do tumor-infiltrating lymphocytes after immunotherapy with dendritic cells (34), this is, as far as we are aware, the first report of a natural cellular response to cancer-derived HSP–peptide complex in humans. From a technical point of view, a nonspecific stimulation induced by endotoxin detected at significant levels in our extracts (35) is unlikely to explain the results. Indeed, tumor-derived HSP-70 induced proliferation of PBMC from patients with cancer but not from healthy control subjects under similar conditions. Furthermore, no proliferation was observed with lysates from whole tumoral tissue, despite endotoxin levels similar to those detected in tumor-derived HSP-70.
The cell response of patients with lung cancer appears to be restricted to peptides complexed with HSP-70 derived from tumoral tissue, i.e., tumor-specific antigens. Indeed, no stimulation could be detected when adding HSP-70 derived from autologous tumor-free lung tissue to the cultures. A nonspecific HSP-70 effect (36) was reasonably excluded. Indeed, no proliferation was observed when stimulating cells with human recombinant HSP-70 alone or HSP-70 derived from autologous tumor-free lung tissue. In addition, depletion of tumor-derived HSP-70 of its peptide cargo by exposure to ATP (27, 28) abolished or significantly reduced proliferative responses in most cases. Interestingly, cell lysates from autologous tumors obviously containing tumoral HSP-70 failed to induce any proliferation. This could be attributed to the release of endogenous proteases denaturing the HSP–peptide complex or to the interference of soluble factors (e.g., transforming growth factor-ß) present in tumor lysates (37, 38), which are known to prevent in vitro T cell proliferation (39–41). Such factors, as well as tumor antigen–induced suppressor cells that can be reactivated in cultures (42), are thought to account for the depressed T cell proliferation to mitogens and alloantigens that are frequently associated with cancers in humans (43–46). Therefore, in this rather unfavorable context, the relatively low SI observed when stimulating PBMC with HSP-70 quantities as small as 1–10 ng should be regarded as reflecting a significant immune reactivity. However, it is currently unknown whether this in vitro response will be exploitable for cancer immunotherapy, especially if one considers that important functions of proliferating cells such as Finally, in most cases we did not observe PBMC proliferation induced by HSP-70 derived from heterologous lung carcinoma. One may emphasize that human leukocyte antigen typing was not performed. However, PBMC from every patient were stimulated with the eight HE (i.e., the same human leukocyte antigen type for the eight HE). Under this condition, the inability of peptides issued from the different extracts to bind to the major histocompatibility complex groove of antigen-presenting cells that would result in the absence of cell proliferation, indicates differences in peptides structure. The antigenicity associated with HSP-70 derived from lung tumors appears to be mostly private, sharing only a few specificities, apart from some exceptions (e.g., HE1 carries T cell antigens that are common to several tumors; Figure 6). This absence of extensive cross-reactivity does suggest that human lung carcinoma, like experimental cancers in animals (49, 50) and some other cancers in humans (51, 52), displays individually distinct antigenicity at least to some extent (i.e., HSP-70–associated antigenicity). As tumor antigens associated with tumor-derived HSP-70 include protective antigens in animals (8–16), the absence of cross-reactivity may be crucial for the design of new vaccines. In conclusion, HSP-70 derived from human lung carcinoma is shown to chaperone antigens that are specific to the tumor from which HSP-70 was purified and that they are responsible for a transient cellular immune response. HSP-70 isolated from lung carcinoma may therefore be a reliable source of tumor-specific antigens for clinical application such as immunotherapy. HSP-based cancer vaccines could be best used in the adjuvant setting after surgery, when adequate quantities of tumor-derived HSP-70 would be available, the tumor burden and its associated immunosuppression minimal, and primed T cell still present.
The authors are grateful to Professor J. Duchateau (CHU Brugmann, Department of Immunology) for allowing them to perform HSP-70 purifications and proliferation assays in his department in collaboration with Isa Sprangers (B.S.) and Henri Collet (B.S.). SDS-PAGE and immunoblot were performed by W. Burny (Ph.D.) from Cypro SA (Brussels, Belgium). The authors thank Drs. P. de Francquen and M. Cappello from the Department of Thoracic Surgery (CUB Erasme), David Young from Lewis Gace Bozell, and Alain Van Muylem from the Chest Department (CUB Erasme) for help during the study, as well as C. Piesen for typing the manuscript.
This article has an online data supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Received in original form February 19, 2002; accepted in final form May 23, 2002
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
-interferon secretion were not studied here. Preliminary data show that, in most cases, this reactivity disappears within 3 months after surgery, well before the denaturation of HSP-70 stored at -70°C usually occurs (
