Search tips
Search criteria 


Logo of cancimmLink to Publisher's site
Cancer Immun. 2013; 13: 4.
Published online 2013 January 22.
PMCID: PMC3559192

Hsp72 mediates stronger antigen-dependent non-classical MHC class Ib anti-tumor responses than hsc73 in Xenopus laevis


The heat shock proteins (HSPs) gp96 and HSP70 mediate potent antigen-dependent anti-tumor T cell responses in both mammals and Xenopus laevis. We have shown that frogs immunized with total HSP70 generate CD8+ T cell responses against the Xenopus thymic lymphoid tumor 15/0 that expresses several non-classical MHC class Ib (class Ib) genes, but no classical MHC class Ia (class Ia). In the absence of class Ia, we hypothesized that hsp72 can prime class Ib-mediated anti-tumor unconventional CD8+ T cells in an antigen-dependent manner. To test this, we produced Xenopus recombinant HSP70 proteins (both the cognate hsc73 and the inducible hsp72) from stable 15/0 tumor transfectants. We used an in vivo cross-presentation assay to prime animals by adoptive transfer of HSP-pulsed antigen-presenting cells (APCs) and showed that both hsp72-and hsc73-Ag complexes have a similar potential to elicit class Ia-mediated T cell responses against minor histocompatibility (H) Ag skin grafts. In contrast, our in vivo cross-presentation assay revealed that hsp72 was more potent than hsc73 in generating protective immune responses against the class Ia-negative 15/0 tumors in an Ag-dependent and class Ib-mediated manner. These results suggest that hsp72 can stimulate class Ib-mediated immune responses and represents a promising candidate for immunotherapy against malignancies with downregulated class Ia expression.

Keywords: heat shock proteins, tumor immunity, Xenopus laevis, comparative immunology


The cytosolic 70-kDa heat shock proteins, or HSP70, are evolutionarily ancient and highly conserved intracellular molecular chaperones that, like other heat shock proteins (HSPs), help with intracellular transport, folding of newly synthesized proteins, and prevention of protein aggregation (1). Increasing evidence suggests that these proteins are critical players in immune surveillance and tumor immunity (2, 3). The ability of HSP70 to mediate potent anti-tumor responses is due to its intrinsic ability to activate both the adaptive and the innate arms of the immune system (2, 4). HSP70 can elicit antigen (Ag)-dependent responses, in which chaperoned antigenic peptides are cross-presented by antigen-presenting cells (APCs) through classical MHC class Ia (class Ia) molecules for CD8+ T cell recognition and induction of strong cytotoxic T lymphocyte (CTL) responses (5, 6). On the other hand, HSP70 also induces a strong peptide-independent pro-inflammatory response (2). This is due to the interaction of HSP70 with a multitude of signaling receptors such as CD14, Toll-like receptor (TLR) 2, TLR 4, and CD40 that are involved in activation of the innate immune system (79).

Due to these immunological properties, HSP70 is a prime candidate as an anti-tumor vaccine. However, the only HSP that has been evaluated to date in clinical trials is gp96 (10), despite many unknowns associated with this protein. For instance, it is not known how peptide binding occurs in vivo, and it is difficult to associate gp96 with peptides in vitro(11, 12). Conversely, HSP70 represents a potentially more reliable anti-tumor vaccine candidate, since it has defined peptide binding stoichiometry and requirements that are dependent on ATP to ADP hydrolysis (11). Furthermore, anti-tumor immunity generated by tumor-derived HSP70-peptide complexes is evolutionarily conserved since it has been demonstrated in the amphibian Xenopus(13, 14).

There is, however, an unresolved issue: purification of HSP70 from tissues does not distinguish between the inducible hsp72 and the constitutive hsc73, which are the two different cytosolic members of the HSP70 family encoded by different genes with different regulation. Hsc73 is constitutively expressed and poorly stress-inducible, while hsp72 is rapidly induced under stress conditions including oxidative stress, changes in pH, as well as malignancy (1, 15, 16). Therefore, one cannot clearly attribute immune protection to either protein. Interestingly, even though HSP70 (i.e., pool of cytosolic hsp72 and hsc73) is one of the prime candidates for use in tumor immunotherapy, there has been minimal attempt to evaluate the respective contribution of these two HSP70 forms in stimulating anti-tumor responses, especially in the case of tumors that downregulate class Ia expression to escape immune recognition. Some evidence suggests that hsp72 may be the key player in these responses. For example, in a rat carcinoma model, hsp72 expression correlated with tumor immunogenicity (17). Also, hsp72 was found to bind more peptides, more efficiently under oxidative conditions than hsc73 (18). It is noteworthy, however, that these data do not rule out the possible immunogenicity of hsc73.

Although the relationships between class Ia molecules and HSPs, including HSP70, have been studied extensively, there is little known regarding the roles of non-classical MHC class Ib (class Ib) gene products. Class Ib molecules are heterogeneous genes structurally similar to class Ia, but exhibit low polymorphism and limited tissue distribution. Interestingly, many tumors, especially malignant cancers, downregulate class Ia expression to avoid immune detection (19). However, many of these tumors still express class Ib molecules (13). Indeed, class Ib molecules are increasingly implicated in immune surveillance, especially in tumor metastasis where class Ia expression is lost (20). On the other hand, there is strong evidence highlighting the pro-tumorigenic roles of certain class Ib molecules. For instance, patients with malignant breast tumors that have lost class Ia expression, but still express class Ib molecules such as HLA-E and HLA-G, result in more severe clinical outcomes, relapse, and poor prognosis (21). Therefore, the roles of these class Ib molecules in cancer are still largely unclear and warrant further examination.

In the present work, we took advantage of our non-mammalian Xenopus comparative tumor immunity model to test the difference in immunogenicity of hsp72 and hsc73 in class Ib-mediated anti-tumor responses. Xenopus is an attractive animal model for studying HSP-mediated immune responses due to its poor responsiveness to lipopolysaccharide (LPS) (4, 22). For this purpose, we used the transplantable class Ia negative thymic tumor 15/0, which expresses several class Ib genes. Using this model, we were able to show that hsp72, but not hsc73, can prime class Ib-mediated anti-tumor responses in an Ag-dependent manner.


Recombinant protein purification

To evaluate and compare the immunological properties of hsp72 and hsc73, we developed an in vitro system to separately purify the Xenopus recombinant tagged proteins. Importantly, this system provided a simple, more convenient way to obtain large amounts of HSP70, which is advantageous given the poor yield obtained using the conventional ADP/ATP-agarose chromatography (23). Proteins were tagged in Xenopus because we do not have antibodies that distinguish between hsp72 and hsc73.

For this purpose, we generated expression vectors carrying either the Xenopus hsp72 or hsc73 full-length cDNA driven by the Xenopus elongation factor 1α (EF-1α) promoter to provide strong constitutive expression. In addition, these genes were fused with two different tags: 6XHis and MYC for protein purification and Western blot analysis, respectively. Xenopus 15/0 tumor cells were co-transfected with either the hsp72 or the hsc73 vector together with a plasmid carrying a Puromycin resistance gene in order to generate stable transfectants cloned by limiting dilution. Using selected hsp72 or hsc73 stable clones, we were able to produce 15/0 recombinant hsp72 and hsc73 (rec-15/0 hsp72 and rec-15/0 hsc73). It was critical to produce these recombinant proteins in the 15/0 tumor cells to ensure that they will be complexed to 15/0 tumor Ag (as well as LG-15 minor H Ags). Expression of each recombinant HSP was confirmed by RT-PCR and Western blot using anti-His and anti-MYC Abs (data not shown).

Recombinant proteins were purified from large-scale 15/0 tumor cultures (on average 150 μg of proteins from 5x107 cells) and controlled for homogeneity by silver staining (Figure 1A). Western blot analysis confirmed that the purified proteins were the recombinant tagged proteins and not endogenous HSP70, since both were recognized by an anti-MYC antibody (Figure 1B). The identity of the recombinant proteins was further confirmed by their appropriate molecular weights and by Western blot using an anti-HSP70 antibody (Figure 1B). Furthermore, cell lysates from untransfected parental 15/0 cells subjected to the same purification procedure did not result in any detectable endogenous HSP70 by Western blot analysis (Figure 1C). It is, therefore, unlikely that the rec-15/0 hsp72 and hsc73 preparations are contaminated by endogenous HSP70.

Figure 1
Purification of recombinant 15/0 hsp72 and hsc73. (A) Silver stained 8% SDS-PAGE gel loaded with 5 μg of rec-15/0 hsp72 (1) and rec-15/0 hsc73 (2). (B) Western blot of different concentrations of rec-15/0 hsp72 (1) and rec-15/0 hsc73 (2) probed ...

In conclusion, we established a convenient in vitro system to produce pure Xenopus recombinant 15/0 tumor-derived hsp72 and hsc73 proteins.

Class Ia cross-presentation ability of recombinant Xenopus HSP70 proteins

To determine if rec-15/0 hsp72 and hsc73 were biologically active, we took advantage of a well-characterized in vivo Ag cross-presentation assay. This assay is based on minor H mismatched skin graft rejection between LG-6 and LG-15 isogenetic Xenopus clones that display an identical heterozygous (a/c) MHC haplotype, but differ from each other by multiple minor H loci (2427). Skin grafts transplanted between these clones are rejected more slowly than MHC-disparate allografts, unless the recipients are primed against the minor H Ags of the skin donor. Priming is achieved by pulsing peritoneal leukocytes (PLs) that serve as APCs with HSPs bound to minor H Ags, and by adoptively transferring these PLs into naïve recipients. We have shown that such priming with gp96-Ag complexes induces a MHC-restricted minor H Ag-specific T cell response, resulting in a significantly accelerated skin graft rejection (28).

Adoptively transferred LG-6 PLs pulsed with either rec-15/0 hsp72 or hsc73 purified from 15/0 tumor cells derived from the LG-15 background were as potent as PLs pulsed with 15/0-derived gp96 (positive control) to induce an accelerated LG-15 skin graft rejection (Figure 2). In contrast, rec-gp96 purified from E. coli (used as a negative control) did not significantly affect the rejection kinetics as compared to LG-6 frogs adoptively transferred with PLs incubated with amphibian PBS (APBS), which rules out the contribution of pro-inflammatory stimuli such as endotoxins. Furthermore, PLs pulsed with rec-15/0 hsp72 and hsc73 pretreated with ATP to remove bound Ags did not induce accelerated skin graft rejection, which indicates that the responses elicited by these proteins were Ag-dependent (Figure 2).

Figure 2
Recombinant 15/0 hsp72 and hsc73 mediate antigen-dependent accelerated minor H Ag-disparate skin graft rejection. LG-6 PLs were pulsed with APBS, rec-gp96 (E. coli derived), 15/0 gp96, (A) rec-15/0 hsp72, rec-15/0 hsp72 pre-treated with ATP (ATP), (B) ...

Together, our results demonstrate that rec-15/0 hsp72 and rec-15/0 hsc73 derived from 15/0 tumor cells of the LG-15 background (i.e., expressing both 15/0 tumor and LG-15 minor H Ags) are biologically active. Notably, our data indicate that hsp72 and hsc73 have a comparable ability to generate class Ia-mediated Ag-dependent T cell responses against skin minor H Ags.

Class Ia-unrestricted anti-tumor immune responses generated by PLs pulsed with rec-15/0 hsp72

As shown above, our in vivo cross-presentation assay using PLs as APCs is effective for class Ia-restricted responses, such as skin graft rejection. However, this assay has never been used for class Ib-mediated responses such as the ones against the 15/0 tumor (i.e., class Ia-negative but class Ib+). The conventional method for priming animals against the 15/0 tumor has historically been immunization with tumor-derived HSPs by subcutaneous injection (4). We were, therefore, interested to test the feasibility and efficiency of our in vivo cross-presentation assay on class Ib-mediated anti-tumor immunity. LG-15 frogs were primed by adoptive transfer of LG-15 PLs pulsed with either APBS, rec-gp96, 15/0-derived gp96, or rec-15/0 hsp72. Animals were challenged with the 15/0 tumor one week later and the time (in days) for first appearance of solid tumor at the site of injection was monitored. Notably, animals primed either with 15/0 gp96 or rec-15/0 hsp72 had significantly delayed tumor appearance (Figure 3). In fact, most of them did not develop tumors during the time of the experiment (60 days), suggesting that a potent protection was elicited. On the other hand, control LG-15 animals primed by adoptive transfer of PLs pulsed with rec-gp96 derived from E. coli or incubated with APBS alone readily developed solid tumor around day 20 post-challenge (Figure 3). These results indicate that using the cross-presentation assay to prime animals is a highly effective way to elicit HSP-mediated anti-tumor responses against a class Ia-negative tumor. Furthermore, this work demonstrates that the rec-15/0 hsp72 carrying 15/0 tumor Ags is able to elicit protection against the 15/0 tumor through its interaction with APCs. This method of priming animals will allow us to conclusively determine if there are any differences in immunogenicity between hsp72 and hsc73 in anti-tumor responses.

Figure 3
Cross-priming by rec-15/0 hsp72 results in delay of tumor appearance and protection. LG-15 PLs were pulsed with APBS, rec-gp96, 15/0 gp96, or rec-15/0 hsp72, and were adoptively transferred into LG-15 recipients, which were challenged with 5x105 live ...

Comparative ability of rec-15/0 hsp72 and hsc73 to generate class Ia-unrestricted anti-tumor immunity

Having established a reliable priming system to assess in vivo HSPs’ ability to stimulate immune responses against a class Ia-deficient tumor, we sought to investigate potential differences between hsp72 and hsc73.

To test the immunogenicity of hsp72 and hsc73 in vivo, we performed tumor priming/challenge experiments using the in vivo cross-presentation assay described above. Briefly, LG-15 frogs were primed by adoptive transfer of PLs pulsed with rec-15/0 hsp72, or rec-15/0 hsc73, and subsequently challenged with the 15/0 tumor. Data from these experiments confirmed that priming with Xenopus rec-15/0 hsp72 elicits a reproducible and significant delay in tumor appearance (p < 0.01, Figure 4A). As seen previously, most of the animals did not develop tumor even after the completion of the experiments (Figure 4, A and B). In contrast, priming with rec-15/0 hsc73 generated anti-tumor immune responses against 15/0 that were less potent than the ones generated by rec-15/0 hsp72 (Figure 4A). The combined data from three separate experiments plotted in survival Kaplan-Meier curves and analyzed with log-rank test confirmed with high statistical significance the difference in anti-tumor responses of hsp72 and hsc73 (Figure 5). Interestingly, there was a large variation in the onset of tumor appearance following rec-15/0 hsc73 priming. About half of the animals developed tumors early on (20–25 days post-tumor challenge) similar to the control frogs, while the others had delayed tumor growth. In conclusion, these results suggest that Xenopus rec-15/0 hsp72 is more potent than hsc73 for inducing immune responses against the 15/0 tumor that only expresses class Ib molecules.

Figure 4
Recombinant 15/0 hsp72 mediates more potent antigen-dependent anti-tumor responses than rec-15/0 hsc73. (A) LG-15 PLs were pulsed with APBS, rec-15/0 hsp72, rec-15/0 hsc73, as well as ATP-treated rec-15/0 hsp72 and rec-15/0 hsc73, and were adoptively ...
Figure 5
Kaplan-Meier curves for tumor-free survival during different priming conditions. LG-15 PLs were pulsed with APBS, rec-15/0 hsp72, rec-15/0 hsc73, as well as ATP-treated rec-15/0 hsp72 and rec-15/0 hsc73, and were adoptively transferred into LG-15 recipients, ...

Antigen dependency of HSP70-mediated class Ia-unrestricted anti-tumor T cell responses

Based on amino acid sequence analysis of the putative peptide binding domain, we have postulated that some Xenopus class Ib molecules are able to bind and present Ags (29). In addition, previous work has shown that immunization with HSP70 complexed with antigenic peptides from the 15/0 tumor generated anti-tumor responses, whereas Ag-free HSP70 did not (27). To further investigate the Ag dependency of hsp72-mediated responses against 15/0 tumors, and the involvement of class Ib molecules, we took advantage of our in vivo cross-presentation model. We pulsed PLs with rec-15/0 HSP70 protein preparations pretreated with ATP in order to remove bound Ags. Animals primed with rec-15/0 Ag-free hsp72 (ATP) developed tumor as fast as the control animals suggesting that Ags carried by hsp72 are critical for eliciting anti-15/0 responses (Figure 4, A and B, and Figure 5). Interestingly, although the anti-tumor responses elicited by rec-15/0 hsc73 were not as potent as with rec-15/0 hsp72, they were nonetheless Ag-dependent (Figure 4A and Figure 5). Therefore, hsp72 is more potent than hsc73 in inducing protective Ag-dependent class Ia-unrestricted anti-tumor T cell responses in vivo.


The objective of this study was to utilize Xenopus laevis as a non-mammalian comparative model to investigate the respective ability of the stress-inducible hsp72 and the constitutively expressed hsc73 in eliciting anti-tumor immunity. Besides gene regulation, hsp72 and hsc73 have distinct non-overlapping cellular functions where hsc73 mainly acts as a molecular chaperone, while hsp72 provides cytoprotection (1). It is, therefore, possible that hsp72 and hsc73 also display specific immune properties. Our study provides the first evidence that, although hsc73 is as potent as hsp72 to facilitate class Ia-restricted T cell responses, it is less efficient than hsp72 in eliciting class Ia-unrestricted anti-tumor T cell responses that are class Ib-mediated.

Given the high degree of structural conservation between hsp72 and hsc73 in vertebrates, and the remarkable evolutionary conservation of the immune system, including anti-tumor T cell immunity between Xenopus and mammals, these results have fundamental relevance for all vertebrates including humans.

Hsp72 and hsc73 have comparable ability to chaperone minor H Ags and induce Ag-dependent class-Ia restricted T cell responses

Our results provide convincing evidence that rec-15/0 hsc73 is as efficient as rec-15/0 hsp72 in chaperoning minor H Ags and eliciting Ag-dependent class Ia-restricted T cell-mediated accelerated rejection of minor H locus disparate skin grafts. This supports our previous observations using immunization with total HSP70 to elicit Ag-specific class Ia-restricted CD8+ T cells response in vitro and accelerated skin graft rejection in vivo(14).

The in vivo cross-presentation used in this study was initially characterized with gp96. In this case, the cross-presentation of chaperoned minor H Ags by pulsed PLs critically involved the binding and active internalization by the endocytic receptor CD91 (28). Based on those findings, our data suggest that HSP70-mediated responses should also involve active receptor-mediated endocytosis of HSP70-minor H Ag complexes by PLs followed by cross-presentation of the chaperoned minor H Ags to class Ia and activation of specific T cell responses. At this time, it is unknown if HSP70 is internalized by CD91 or other endocytic receptors in Xenopus, but all the data obtained so far collectively support this process.

PLs pulsed with LPS containing recombinant gp96 derived from E. coli do not induce accelerated minor H-disparate skin graft rejection. In addition, adoptive transfer of PLs does not induce any rejection of LG-6 isografts (data not shown). The accelerated skin graft rejection kinetics obtained by priming with hsp72 or hsc73 chaperoning minor H Ags is consistent with a T cell-mediated response. Finally, PLs pulsed with hsp72 or hsc73 stripped of Ags by ATP treatment do not accelerate skin graft rejection.

Therefore, similar to their mammalian counterparts, Xenopus hsp72 and hsc73 proteins can chaperone Ags into APCs via receptor-mediated internalization and mediate Ag-specific CD8+ T cell activation through the cross-presentation pathway.

Hsp72, but not hsc73, can induce potent protective Ag-dependent class-Ib mediated anti-tumor immunity

Interestingly, we found that hsp72 and hsc73 are not equal in eliciting class Ib-mediated responses. Our data demonstrate that hsp72 has the ability to generate protective anti-tumor responses against the 15/0 tumor, while hsc73 is much less potent. Notably, hsp72-mediated class Ib-restricted responses are Ag-dependent. This is reminiscent of the conventional HSP-mediated class Ia-restricted anti-tumor responses seen in mammals. Currently, it is still unclear if any of the class Ib responses in mammals are actually Ag-dependent, especially in tumor immunity. Since human class Ib molecules are not orthologous to Xenopus class Ib, it would be of interest to see if human hsp72 is also capable of eliciting Ag-dependent anti-tumor responses against class Ia-negative tumors.

The distinct capacity of hsp72 and hsc73 to elicit class Ib-mediated anti-tumor responses can come from different potential sources. It is possible, in principle, that the inefficiency of recombinant hsc73 to induce response against the 15/0 tumor results from defects in its Ag binding. However, this deficiency would have to be specifically restricted to the binding of class Ib Ags since our data demonstrated that hsc73 elicits Ag-dependent class Ia-restricted CD8+ T cell responses similar to hsp72. Based on SDS-PAGE migration, Western blot analysis with anti-HSP70 mAbs, and functional assays, we feel confident that hsc73 accurately reflects the immunological properties of its endogenous counterpart. As such, a more probable explanation for the potent ability of hsp72 to elicit immune responses against 15/0 tumors is a better capacity to bind class Ib Ag(s) than hsc73. Currently, we do not know what type of Ags can be bound by Xenopus class Ib molecules, and whether similar Ags would be complexed to hsp72. However, based on the various class Ib ligands that have been characterized in human and mouse (20), we can speculate that Xenonpus non-classical genes (XNCs) could bind modified peptides, lipid, and/or glycolipid Ags. One feature of these types of class Ib Ags is that they are often evolutionarily conserved products of pathogens or byproducts (self-Ag) resulting from stress. One interesting possibility in our case is that some conserved ligands can serve as tumor recognition structures that can be recognized and killed by class Ib-restricted CD8+ T cells. However, it is also possible that hsp72 can chaperone modified peptides specific to the individual tumor (i.e., 15/0 only).

Presumably, the types of Ags carried by HSP70 proteins may have an impact on the effector cells that are generated during these responses. Previous studies showed that gp96 is able to generate unconventional class Ib-mediated CD8+ T cells against the 15/0 tumor (30). Our data further support those findings and suggest that hsp72 also has the ability to activate both conventional and unconventional or potentially innate CD8+ T cells. In contrast, hsc73 is only able to elicit class Ia-mediated responses likely to involve conventional CD8+ T cells, whereas it is unable to participate in immune surveillance against class Ia negative tumors and, therefore, may not be able to stimulate unconventional CD8+ T cells.

Clinical perspective

The above work has several clinical implications, especially for cancer immunotherapy. For the first time, our data revealed that there is indeed a difference in immunogenic potential between hsp72 and hsc73. We showed hsp72 was able to mediate potent Ag-dependent anti-tumor responses resulting in delayed tumor growth and, in most cases, protection against the aggressive 15/0 tumor. In comparison, hsc73 elicited a very weak response against the 15/0 tumor. These data demonstrate that hsp72 is more immunogenic and may be used against tumors with suboptimal class Ia expression. Clinically, this may have relevance in scenarios where HSPs are purified from autologous cancers and used for vaccination. We surmise that the ratio of hsp72 versus hsc73 found in these tumors should have important implications in the individual treatment of patients.

On the other hand, the difference in immunogenicity may be due to the qualitative or quantitative difference of Ags bound to hsp72 and hsc73. It has already been shown that hsp72 can bind more peptides and binds them more efficiently under oxidative stress than hsc73. It is also possible that hsp72 binds antigenic determinants stronger than hsc73, especially since the amino acid sequences of these proteins are most divergent in the peptide binding region. In the present study, we showed that animals primed with hsc73 had a variable response to the 15/0 tumor challenge. Some developed tumor comparable to control animals, while others exhibited delayed growth. In comparison, all animals primed with ATP-treated Ag-free hsc73 developed tumor similarly to control frogs, which suggests that a fraction of hsc73 was complexed to tumor Ags. Elution and sequencing of the specific Ags bound to Xenopus hsp72 and hsc73 will provide more direct evidence of Ag access and binding and may give us an idea if these proteins can activate different effector populations during an immune response.

To date, HSP vaccines alone have not been sufficiently convincing in the treatment of cancer, presumably because they require large amounts of protein and multiple immunizations, and HSPs can only be obtained in limited amounts from patients. Using adoptive transfer, the patient’s own APCs pulsed with HSP may provide a better alternative since much less protein is required. Such an alternative method for generating therapeutic vaccines using HSP70 purified from DC/tumor fusions, using both patient-derived ovarian cancer cells and cultured breast cancer cells, was recently published (31). T cells primed with the HSP70 preparation showed increased expression of IFN-γ and more efficient tumor killing in vitro. Using hsp72 specifically may increase the immunogenicity of these therapeutic vaccines.

Finally, hsp72-mediated anti-tumor response should also be considered in the development of new therapeutic approaches. Notably, as mentioned above, class Ib molecules remain expressed by class Ia-deficient tumors, presumably to prevent killing by NK cells. The roles of these class Ib molecules in cancer biology remains poorly understood (20, 31). While class Ib involvement in tumor immunity awaits further investigation, the adoptive transfer of hsp72-pulsed APCs represents a promising therapeutic approach that may engage both class Iaand class Ib-restricted anti-tumor immunity.


This work was supported by the following grants from the NIH: T32-AI 07285 (H.N.), 1R03-HD061671-01, and R24-AI-059830-06. The expert animal husbandry provided by David Albright and Tina Martin is greatly appreciated. We would also like to thank Drs. Eva-Stina Edholm, Leon Grayfer, and Nikesha Haynes for critically reading the manuscript.


heat shock protein;
heat shock cognate;
peritoneal leukocyte;
Amphibian PBS

Materials and methods

Animals and cell lines

LG-6 and LG-15 clones (32, 33) were obtained from our Xenopus laevis Research Resource for Immunobiology (34) at the University of Rochester. All animals were handled under strict laboratory and UCAR regulations (Approval number 100577/2003-151), minimizing discomfort at all times. The 15/0 lymphoid tumor cell line is derived from a spontaneously arising thymic tumor in a LG-15 frog (24).

Plasmid construction

Full-length Xenopus hsp72 and hsc73 cDNA were obtained from the pSP72-HSP70 and pSP72-hsc70.1 vectors. Hsp72 (2258 bp) was cloned into the PstI and SalI sites of the pXEX vector (35). Hsc73 (2458 bp) was cloned into the XbaI and SacI sites of pXEX. In addition, a tag containing MYC, 6XHIS, STOP from the pAP-tag5 was cloned into pXEX (hsp72 insert was added into the BamHI and SmaI sites, and SacI and ClaI sites were used for hsc73) to get the final vectors: pXEX-hsp72-MYC,6XHIS and pXEX-hsc73-MYC,6XHIS.

Stable transfections

Plasmids were co-transfected with a vector containing a puromycin-resistance gene as a selectable marker into Xenopus 15/0 tumor cells using Lipofectamine 2000 (Invitrogen) according to manufacturer’s directions. Single clones were established by limited dilution cloning and were tested by PCR and RT-PCR for integration and expression of the genes of interest.

Recombinant protein purification

The recombinant hsp72 and hsc73 proteins were purified using Ni-NTA agarose (QIAGEN) according to manufacturer’s instructions (The QIAexpressionist). Washes were performed a minimum of 5x under stringent conditions, which included higher salt concentration (0.5M NaCl), the addition of imidizole (40 mM) which prevents nonspecific binding of non-HIS proteins to the column, and 0.5% Tween. The eluted samples were run through PD10 columns in order to exchange the solution to 20 mM Phosphate buffer. These fractions were subsequently run through a DEAE exchange column and eluted with 20 mM Phosphate buffer + 150 mM NaCl in order to purify HSP70 proteins only. Proteins were quantified by Bradford assay, and the purity was assessed by silver staining.

Western blotting

Purified recombinant proteins were separated on 10% polyacrylamide gels and transferred to a PVDF membrane (Bio-Rad). Membranes were incubated with anti-MYC tag mAb (P/N 46-0603 Invitrogen, 1:5000 dilution) or anti-HSP70/HSC70 mAb (SPA-820 Stressgen, 1:1000 dilution) followed by an HRP-conjugated rabbit anti-mouse Ab (Sigma, 1:5000 dilution) and developed by SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific).

In vivo priming assay

LG clones were primed according to published methods (36). Briefly, PLs were pulsed with 1 μg HSP proteins per 5x105 cells in APBS for 1 h on ice. PLs were then extensively washed (at least 3x) with cold sterile APBS to ensure that there is no residual HSPs left. The PLs were resuspended at a concentration of 5x105 PLs per 300 μl (for skin grafting assays), or 1x106 PLs per 300 μl (for tumor challenge assays), and were adoptively transferred by an IP injection into recipient animals. Frogs were primed for 7 days prior to skin graft or tumor challenge experiments.

Skin grafting

LG-6 clones were grafted according to published methods (36). Briefly, a piece of LG-15 ventral skin (abdominal skin which appears silvery due to the presence of irridophore pigmented cells, 5 mm x 5 mm) was inserted under the skin of the LG-6 recipient clone, silvery side up. Twenty-four hours later the overlaying host skin was removed so that the graft can be freely visualized. Skin graft rejection was determined by the percent of destruction of the irridophores on the grafted skin.

Tumor transplantation

A week before each experiment a fresh batch of 15/0 tumor cells was thawed and expanded. Cloned LG-15 Xenopus frogs were challenged with 5x105 15/0 tumor cells (> 95% viable cells) in 300 μl volume per frog by subcutaneous injection on one side of the dorsal (back side) of the animal (36). The initial tumor appearance was noted and the tumor volume was recorded every 2–3 days. Tumor volume (height x length x width) was measured using calipers. Once the tumors grew out to 3,500 mm3, or the frogs started looking lethargic and not eating, they were euthanized to prevent discomfort.

Statistical analysis

Statistical significance was analyzed using one-way ANOVA and post hoc Tukey multiple comparison tests, and Kaplan-Meier curves analyzed with a log-rank test.


1. Kiang JG, Tsokos GC. Heat shock protein 70 kDa: molecular biology, biochemistry, and physiology. Pharmacol Ther. 1998;80:183–201. [PubMed]
2. Srivastava P. Roles of heat-shock proteins in innate and adaptive immunity. Nat Rev Immunol. 2002;2:185–194. [PubMed]
3. Calderwood SK, Theriault JR, Gong J. Message in a bottle: role of the 70-kDa heat shock protein family in anti-tumor immunity. Eur J Immunol. 2005;35:2518–2527. [PubMed]
4. Robert J. Evolution of heat shock protein and immunity. Dev Comp Immunol. 2003;27:449–464. [PubMed]
5. Castellino F, Boucher PE, Eichelberg K, Mayhew M, Rothman JE, Houghton AN, Germain RN. Receptor-mediated uptake of antigen/heat shock protein complexes results in major histocompatibility complex class I antigen presentation via two distinct processing pathways. J Exp Med. 2000;191:1957–1964. [PMC free article] [PubMed]
6. Bendz H, Ruhland SC, Pandya MJ, Hainzl O, Riegelsberger S, Brauchle C, Mayer MP, Buchner J, Issels RD, Noessner E. Human heat shock protein 70 enhances tumor antigen presentation through complex formation and intracellular antigen delivery without innate immune signaling. J Biol Chem. 2007;282:31688–31702. [PubMed]
7. Asea A, Kraeft SK, Kurt-Jones EA, Stevenson MA, Chen LB, Finberg RW, Koo GC, Calderwood SK. HSP70 stimulates cytokine production through a CD14-dependant pathway, demonstrating its dual role as a chaperone and cytokine. Nat Med. 2000;6:435–442. [PubMed]
8. Asea A, Rehli M, Kabingu E, Boch JA, Bare O, Auron PE, Stevenson MA, Calderwood SK. Novel signal transduction pathway utilized by extracellular HSP70: role of toll-like receptor (TLR) 2 and TLR4. J Biol Chem. 2002;277:15028–15034. [PubMed]
9. Basu S, Binder RJ, Suto R, Anderson KM, Srivastava PK. Necrotic but not apoptotic cell death releases heat shock proteins, which deliver a partial maturation signal to dendritic cells and activate the NF-kappa B pathway. Int Immunol. 2000;12:1539–1546. [PubMed]
10. Srivastava PK. Therapeutic cancer vaccines. Curr Opin Immunol. 2006;18:201–205. [PubMed]
11. Blachere NE, Li Z, Chandawarkar RY, Suto R, Jaikaria NS, Basu S, Udono H, Srivastava PK. Heat shock protein-peptide complexes, reconstituted in vitro, elicit peptide-specific cytotoxic T lymphocyte response and tumor immunity. J Exp Med. 997186:1315–1322. [PMC free article] [PubMed]
12. Nicchitta CV. Re-evaluating the role of heat-shock protein-peptide interactions in tumour immunity. Nat Rev Immunol. 2003;3:427–432. [PubMed]
13. Robert J, Ménoret A, Basu S, Cohen N, Srivastava PR. Phylogenetic conservation of the molecular and immunological properties of the chaperones gp96 and hsp70. Eur J Immunol. 2001;31:186–195. [PubMed]
14. Robert J, Gantress J, Rau L, Bell A, Cohen N. Minor histocompatibility antigen-specific MHC-restricted CD8 T cell responses elicited by heat shock proteins. J Immunol. 2002;168:1697–1703. [PubMed]
15. Mosser DD, Morimoto RI. Molecular chaperones and the stress of oncogenesis. Oncogene. 2004;23:2907–2918. [PubMed]
16. Ciocca DR, Calderwood SK. Heat shock proteins in cancer: diagnostic, prognostic, predictive, and treatment implications. Cell Stress Chaperones. 2005;10:86–103. [PMC free article] [PubMed]
17. Ménoret A, Patry Y, Burg C, Le Pendu J. Co-segregation of tumor immunogenicity with expression of inducible but not constitutive hsp70 in rat colon carcinomas. J Immunol. 1995;155:740–747. [PubMed]
18. Callahan MK, Chaillot D, Jacquin C, Clark PR, Ménoret A. Differential acquisition of antigenic peptides by Hsp70 and Hsc70 under oxidative conditions. J Biol Chem. 2002;277:33604–33609. [PubMed]
19. Zitvogel L, Tesniere A, Kroemer G. Cancer despite immunosurveil-lance: immunoselection and immunosubversion. Nat Rev Immunol. 2006;6:715–727. [PubMed]
20. Gomes AQ, Correia DV, Silva-Santos B. Non-classical major histo-compatibility complex proteins as determinants of tumour immunosurveillance. EMBO Rep. 2007;8:1024–1030. [PubMed]
21. de Kruijf EM, Sajet A, van Nes JG, Natanov R, Putter H, Smit VT, Liefers GJ, van der Elsen PJ, van der Velde CJ, Kuppen PJ. HLA-E and HLA-G expression in classical HLA class I-negative tumors is of prognostic value for clinical outcome of early breast cancer patients. J Immunol. 2010;185:7452–7459. [PubMed]
22. Belicher PA, Rollins-Smith LA, Jacobs DM, Cohen N. Mitogenic responses of frog lymphocytes to crude and purified preparations of bacterial lipopolysaccharide (LPS) Dev Comp Immunol. 1983;7:483–496. [PubMed]
23. Robert J, Gantress J, Cohen N, Maniero GD. Xenopus as an experimental model for studying evolution of hsp—immune system interactions. Methods. 2004;32:42–53. [PubMed]
24. Robert J, Guiet C, Du Pasquier L. Lymphoid tumors of Xenopus laevis with different capacities for growth in larvae and adults. Dev Immunol. 1994;3:297–307. [PMC free article] [PubMed]
25. Arnall JC, Horton JD. Impaired rejection of minor-histocompatibility-antigen-disparate skin grafts and acquisition of tolerance to thymus donor antigens in allothymus-implanted, thymectomized Xenopus. Transplantation. 1986;41:766–776. [PubMed]
26. Flajnik MF, Du Pasquier L, Cohen N. Immune responses of thymus/lymphocyte embryonic chimeras: studies on tolerance and major histocompatibility complex restriction in Xenopus. Eur J Immunol. 1985;15:540–547. [PubMed]
27. Rau L, Cohen N, Robert J. MHC-restricted and -unrestricted CD8 T cells: an evolutionary perspective. Transplantation. 2001;72:1830–1835. [PubMed]
28. Robert J, Ramanayake T, Maniero GD, Morales H, Chida AS. Phylogenetic conservation of glycoprotein 96 ability to interact with CD91 and facilitate antigen cross-presentation. J Immunol. 2008;180:3176–3182. [PubMed]
29. Goyos A, Ohta Y, Guselnikov S, Robert J. Novel nonclassical MHC class Ib genes associated with CD8 T cell development and thymic tumors. Mol Immunol. 2009;46:1775–1786. [PMC free article] [PubMed]
30. Goyos A, Cohen N, Gantress J, Robert J. Anti-tumor MHC class Ia-unrestricted CD8 T cell cytotoxicity elicited by the heat shock protein gp96. Eur J Immunol. 2004;34:2449–2458. [PubMed]
31. Gong J, Zhang Y, Durfee J, Weng D, Liu C, Koido S, Song B, Apostolopoulos V, Calderwood SK. A heat shock protein 70-based vaccine with enhanced immunogenicity for clinical use. J Immunol. 2010;184:488–496. [PMC free article] [PubMed]
32. Kobel HR, Du Pasquier L. Strains and species of Xenopus for immunological research. Amsterdam: Elsevier/North Holland Publishing Co.; 1977.
33. Kobel HR, Du Pasquier L. Production of large clones of histocompatible, fully identical clawed toads (Xenopus) Immunogenetics. 1975;2:87–91.
34. Xenopus laevis Research Resource for Immunobiology. Accessed from:
35. Johnson AD, Krieg PA. pXeX, a vector for efficient expression of cloned sequences in Xenopus embryos. Gene. 1994;147:223–226. [PubMed]
36. Nedelkovska H, Cruz-Luna T, McPherson P, Robert J. Comparative in vivo study of gp96 adjuvanticity in the frog Xenopus laevis. J Vis Exp. 2010;43:2026. [PubMed]

Articles from Cancer Immunity are provided here courtesy of Academy of Cancer Immunology