Breast Disease 20 (2004) 311 IOS Press

Peptide-Based Vaccines in Breast Cancer

Mary L. Disis , Lupe G. Salazar and Keith L. Knutson
Tumor Vaccine Group, Oncology, University of Washington, Seattle, WA, USA

Abstract. Human tumors are immunogenic and tumor-associated proteins that generate immunity in cancer patients have been dened. Many of these proteins are involved in the malignant transformation and play a role in either initiating or maintaining the malignant phenotype. Furthermore, due to technical advances in basic immunology over the last decade we have a better understanding of the immune effector cell phenotypes that are potentially involved in tumor eradication and have developed methods to quantitate and characterize these immune effectors. Breast cancer is an intriguing model tumor to target with active immunization. Dozens of breast cancer antigens have been dened [1]. Although many patients with breast cancer can be rendered free of disease with standard therapy such as surgery, radiation, and chemotherapy, some patients will have their disease recur. However, relapse may not occur for many months to years after denitive treatment giving an extended period of micrometastatic disease that may be amenable to immune eradication or modulation. Peptide based vaccines are one of the most commonly studied vaccine strategies targeting breast cancer. Abbreviations: APC: antigen presenting cell; CEA: carcinoembryonic antigen; CTL: cytotoxic T cell; DC: dendritic cell; ECD: extracellular domain; GM-CSF: granulocyte macrophage colony stimulating factor; HLA: human leukocyte antigen; IFA: incomplete Freunds adjuvant; IFN: interferon; IL: interleukin; KLH: keyhole limpit hemocyanigen; LC: Langerhans cell; MHC: major histocompatibility antigen; PBMC: peripheral blood mononuclear cells; TCR: T cell receptor; Th: T helper cell

THE RATIONALE SUPPORTING PEPTIDE BASED VACCINES Vaccines have been one of the most successful clinical interventions for the prevention of human disease. Prospective immunization against a variety of pathogens has resulted in protection from the development of life-threatening infection. A recent review cited data derived from records kept by the Centers of Disease Control since 1912 and outlined the decrease in infectious disease since vaccines targeting specic pathogens have become widely available [2]. The incidence of indigenous poliomyelitis has decreased 100%, diphtheria, measles and rubella have decreased by 99%, and whooping cough has had a 97% decrease in incidence associated with vaccination. Most cases of vaccine failure can be linked to insufcient immune responses generated after active immunization [2]. The
author: Mary L. Disis, Box 356527, Oncology, University of Washington, Seattle, WA 98195-6527, USA. Tel.: +1 206 616 1823; Fax: +1 206 685 3128; E-mail: ndisis@ u.washington.edu.
Corresponding

purpose of a vaccine is to induce an antigen-specic immune response that will result in disease prevention. The identication of specic tumor antigens has signicantly advanced the eld of tumor immunology, in particular, the development of breast cancer vaccines. Improved understanding of the molecular basis of antigen recognition has resulted in the development of rationally designed breast cancer antigen-specic vaccines based on motifs predicted to bind to human class I or class II major histocompatibility molecules (MHC). The use of synthetic peptides in breast cancer vaccines offers practical advantages such as relative ease of construction and production, chemical stability, and a lack of infectious or oncogenic potential. Peptides may also allow better manipulation of the immune response through the use of epitopes designed for stimulating particular subsets of T cells (e.g. helper and cytotoxic T cells). In animal models, peptide vaccines are particularly effective in generating immune responses to self-proteins [3]. Theoretically, immunization to foreign proteins normally elicits immunity to only a subset of potential epitopes, dominant epitopes, whereas other potentially immunogenic epitopes, subdominant

0888-6008/04/$17.00 2004 IOS Press and the authors. All rights reserved

M.L. Disis et al. / Peptide-Based Vaccines in Breast Cancer

epitopes, are ignored. There is evidence that T cells are tolerant to the dominant epitopes of self-proteins, but may respond to the subdominant epitopes [4,5]. As many newly dened tumor antigens are self-proteins, peptide immunization may play a key role in the ability to elicit an immune response to such antigens. Peptide based vaccines can be designed to represent subdominant epitopes, thus, elicit immunity. The development of a rational approach to breast cancer vaccination must be based on how the immune system is stimulated to respond to breast cancer associated proteins. The goal of many infectious disease vaccines is to stimulate a cellular immune response, i.e. T cell response that would result in the development of long-term memory T cells that would be capable of responding to the pathogen when a patient is exposed. Generation of T cell immunity would be the goal for a cancer vaccine developed to prevent disease in patients at high risk for relapse. T cells comprise the majority of circulating peripheral blood lymphocytes. T cell activation depends on recognition of MHC-antigenic peptide complexes on the surface of antigen presenting cells (APC). Tumors themselves could present antigenic peptides in MHC molecules expressed on cell surface. Vaccines composed of both class I and class II peptides derived from breast cancer antigens have been studied in human clinical trials [6,7]. Peptide-based vaccines have advanced from preclinical to human clinical studies over the last decade and several important issues have been elucidated during this clinical progression. First, investigators have developed powerful paradigms to choose peptide(s) for immunization and established the pre-clinical evaluation needed to develop a peptide-based tumor vaccine. Secondly, the importance of class II peptides in active immunization is being increasingly dened. The generation of an endogenous CD4+ T cell response will improve the magnitude and maintenance of CD8+ immunity generated with class I binding epitopes. Finally, methods have been developed to modify peptides to improve the immunogenicity of epitope-specic vaccines. Tumor antigen-specic peptide-based vaccines have been an excellent model to test active immunization in breast cancer patients over the last several years. CD8+ cytotoxic T cells (CTL) can kill cells expressing an antigen. The action of CD8+ T cells requires recognition of antigenic proteins that have been processed and presented as peptide fragments in the context of MHC class I. Epitopes recognized by CD8+ T cells are, in general, peptides of 810 amino acids in length anchored at each end within the major groove of

MHC class I molecules. As a consequence, MHC class I molecules exhibit more amino acid sequence specicity based on these anchoring residues, than MHC class II molecules which may be more promiscuous in the binding of peptides. A potential problem in the development of CTL epitope-based vaccines is the large degree of MHC polymorphism and the need for knowledge of MHC restrictions in the population to be vaccinated. However, it is now known that that MHC class I molecules can be divided into several families or supertypes based on similar peptide-binding repertoires [8]. For example, the A2 supertype consists of at least eight related molecules and of these, the most frequently observed are HLA-A*0201, A*0202, A*0203, A*0206, and A*6802. Many peptides that bind A*0201 also exhibit degenerate binding, i.e. binding to multiple alleles, thus, an A2 supertype multi-epitope vaccine could be designed to provide broad population coverage [8]. Initial studies of CTL epitopes were based on the elution of naturally processed peptides from MHC class I molecules on tumor cells. The identication of sequence motifs from eluted peptides, as well as crystallographic data of peptide binding to specic MHC class I molecules, has allowed predicted binding of specic peptides. Recent investigations demonstrate that computer algorithms designed for predicting class I binding can be good predictors of peptides that are naturally-processed epitopes of tumor antigens. Lu and colleagues used two computerbased prediction algorithms that are readily available on the Internet; syfpeithi.bmi-heidelberg.com and bimas.dcrt.nih.gov/molbio/hla bind/hla motif search info.html. The algorithms were able to identify three candidate peptides of CEA specic for HLA-B7. After constructing the peptides based on motif identication, investigators demonstrated one of the peptides generated CTL capable of lysing HLA-matched CEA expressing tumors [9]. Similar strategies have been extremely useful for identifying class I peptide epitopes from a variety of tumor antigens. For example, Rongcun and colleagues identied four HER-2/neu-specic HLA-A2.1-restricted CTL epitopes which were able to elicit CTL that specically killed peptide-sensitized target cells, and most, importantly, a HER-2/neutransfected cell line and autologous tumor cells. In addition, CTL clones specic for particular HER-2/neu epitopes were isolated from tumor-specic CTL lines, further demonstrating the immunogenicity of these epitopes as candidates for a potential breast cancer vaccine specically designed to develop a tumor-specic CTL response [10]. Thus, peptide based vaccines can

M.L. Disis et al. / Peptide-Based Vaccines in Breast Cancer

be designed for eliciting CTL immunity and may even be able to immunize across multiple class I haplotypes. CD4+ T helper cells (Th) augment the antigenspecic immune response via cytokine secretion. Th1 subsets of CD4+ T cells secrete cytokines such as interleukin (IL)-2 and interferon (IFN)- that will stimulate the proliferation and activity of CTL. Th2 subsets of CD4+ T cells secrete cytokines such as IL-4, -5, -6, and -10 that will result in production of high afnity effective antibody response. The action of CD4+ T cells requires recognition of antigenic proteins that have been processed and presented as peptide fragments in the context of MHC class II. Unlike MHC class I molecules, which typically bind peptides of eight to ten amino acids in length, MHC class II molecules are generally believed to be more permissive in length and amino acid sequence of bound peptide [11]. For example, in studies dening the T helper epitopes of tetanus toxoid, panels of overlapping peptides were screened for their ability to stimulate T cell proliferation in lymphocytes derived from tetanus-immune subjects. Peptides of greater than 12 amino acids and less than 31 amino acids and typically in the range of 14 to 16 amino acids were found to be most efcient in dening recognized epitopes [12]. Moreover, T helper peptides can be universal or promiscuous and bind to multiple alleles. Promiscuous class II helper epitopes have been identied for the HER-2/neu [13], NY-ESO [14], and human telomerase reverse transcriptase antigens [15], three common immunogenic proteins associated with breast cancer. CD4+ T cells play an important role in the generation of an anti-tumor immune response. CD4+ helper T cells initiate and maintain CD8+ T cell responses [16, 17]. Peptide-based cancer vaccines must consider the important and potentially critical role of CD4+ T cells. Similar to the identication of MHC class I binding motifs, computer algorithms have played a major role in the identication of MHC class II binding motifs. MHC class II antigenic epitopes tend to share a common secondary structure, namely amphipathic helices [18] suggesting antigenic epitopes recognized by T cells can be predicted from the primary structure of a protein based on predicted secondary structure [19, 20]. There are now several amphipathicity and hydrophobicity algorithms that have been demonstrated to be useful in predicting T cell presented sequences within potentially antigenic proteins [18,19,21]. These algorithms have been used to identify potential MHC class II-binding epitopes from several tumor-associated antigens. In vitro generation of peptide-specic T cells

from patients with cancer, similar to those used with MHC class I peptides, is the endpoint of the identication of epitopes appropriate for use in immunization. This strategy was used to dene putative class II HER-2/neu binding epitopes for use in breast cancer vaccines [19]. Subsequently, when these peptides were tested for binding to HLA-DR molecules in vitro, the peptides which were the most immunogenic in vivo were also those peptides that had the highest binding afnity to HLA-DR [22].

BREAST CANCER VACCINES COMPOSED OF CLASS I PEPTIDES AND DESIGNED TO GENERATE A CD8+ CYTOTOXIC T CELL (CTL) RESPONSE In an initial clinical study, HLA-A2 positive patients with metastatic HER-2/neu overexpressing breast, ovarian, or colorectal carcinomas were immunized with p369-377, a MHC class I peptide dened from the HER-2/neu protein sequence, admixed in incomplete Freunds adjuvant (IFA) every three weeks [23]. Peptide-specic CTL were isolated and expanded from the peripheral blood of patients after two or four immunizations. The CTL could lyse MHC-matched peptidepulsed target cells but could not lyse MHC-matched tumors expressing the HER-2/neu protein. Even when tumors were treated with IFN- to upregulate MHC class I, the CTL lines generated from the patients would not respond to the peptide presented endogenously on tumor cells. A similar study performed with the same MHC class I peptide in patients with metastatic breast and ovarian cancer demonstrated peptide and tumorspecic responses in some patients [24]. Another group of investigators, immunizing patients with p369-377, used GM-CSF as an adjuvant [6]. GM-CSF is a recruitment and maturation factor for skin dendritic cells (DC), Langerhans cells (LC) and theoretically may allow more efcient presentation of peptide epitopes than standard adjuvants such as IFA. Six HLA-A2 patients with HER-2/neu-overexpressing cancers received six monthly vaccinations with HER-2/neu peptide, p369377, admixed with GM-CSF. The patients had either stage III or IV breast or ovarian cancer. Immune responses to the p369-377 were examined using an IFNELIspot assay. Following vaccination, HER-2/neu peptide-specic precursors developed to p369-377 in just two of four evaluable subjects. The responses were short-lived and not detectable at ve months after the nal vaccination. Immunocompetence was evident as

M.L. Disis et al. / Peptide-Based Vaccines in Breast Cancer

patients had detectable T cell responses to tetanus toxoid and inuenza. Another vaccine strategy commonly used in breast cancer vaccination is the delivery of MHC class I peptides loaded on DC and injected subcutaneously or intradermally. Brossart and colleagues reported a vaccine trial immunizing breast and ovarian cancer patients with autologous DC pulsed with MUC-1 or HER-2/neu class I binding peptides [25]. In this trial, 10 patients were immunized. Half the patients developed CD8+ T cell precursors to their immunizing peptides. Moreover, some patients developed new immunity to other tumor antigens expressed in their cancers such as CEA and MAGE-3, i.e. epitope spreading. Epitope, or determinant spreading, is a phenomenon rst described in autoimmune disease [26] and represents the generation of an immune response to a particular portion of an immunogenic protein and then the natural spread of that immunity to other areas of the protein or even to other antigens present in the environment. Epitope spreading may represent the ability of the initial immune response to create a microenvironment at the site of the tumor that enhances endogenous local immunity [27]. These phase I clinical trials demonstrate that patients with even advanced stage breast cancer can be immunized against antigen expressed by their tumors. Although early trials have focused on only one or two dened breast cancer antigens, many class I peptide epitopes are being dened for novel immunogenic proteins expressed in breast cancer. For example, investigators have identied several HLA-A3 binding epitopes of mammaglobin-A, a protein highly expressed on about 80% of breast tumors [28]. Similarly, MHC class I binding epitopes have been dened for the near universal antigen, survivin [29]. The identication of an armamentarium of MHC class I binding peptides derived from multiple immunogenic breast cancer proteins will lay the foundation of multi-antigen approaches to prevent breast cancer relapse in antigenically diverse tumors across a broad patient population.

BREAST CANCER VACCINES COMPOSED OF CLASS II PEPTIDES AND DESIGNED TO GENERATE A CD4+ T HELPER CELL (TH) RESPONSE Potential pitfalls of the use of single MHC class I binding peptides are highlighted in the trials described above; immunity is short lived and the peptide-specic T cells may not recognize endogenously presented

antigen. A successful vaccine strategy for generating peptide-specic CTL capable of lysing tumor expressing HER-2/neu and resulting in durable immunity involved immunizing patients with putative T-helper epitopes of HER-2/neu which had, embedded in the natural sequence, HLA-A2 binding motifs of HER-2/neu. Thus, both CD4+ T cell helper and CD8+ specic epitopes were encompassed in the same vaccine. HLAA2 patients with HER-2/neu-overexpressing cancers received a vaccine preparation consisting of putative HER-2/neu helper peptides [6]. Contained within these sequences were the HLA-A2 binding motifs. Patients developed both HER-2/neu-specic CD4+ and CD8+ T cell responses. The level of HER-2/neu immunity was similar to viral and tetanus immunity. In addition, the peptide-specic T cells were able to lyse tumor, demonstrating generation of T cells that recognized naturally processed antigen. The responses were long-lived and detectable for greater than one year after the nal vaccination in some patients. These results demonstrate that HER-2/neu MHC class II epitopes containing encompassed MHC class I epitopes are able to induce long-lasting HER-2/neu-specic IFN-producing CD8 T cells. Stimulating an effective T helper (Th) response, even without concomitant CD8+ peptide vaccination, is a way to boost antigen-specic immunity as CD4+ T cells generate the cytokine environment required to support an evolving immune response. In one recent study, patients with advanced stage HER-2/neu overexpressing breast, ovarian, and non-small cell lung cancer were enrolled on a trial of a multi-peptide HER-2/neu Th vaccine [7]. Over 90% of patients developed T cell immunity to HER-2/neu Th peptides and over 60% to a HER-2/neu protein domain. Thus, immunization with peptides resulted in the generation of T cells that could respond to protein processed by APC. Furthermore, at 1-year follow-up, immunity to the HER-2/neu protein persisted in over a third of patients. Immunity elicited by active immunization with CD4+ T helper epitopes was durable. An additional nding of this study was that epitope spreading was observed in the majority of patients and signicantly correlated with the generation of HER-2/neu protein-specic T cell immunity. Vaccines targeting HER-2/neu are among the most commonly tested immunotherapeutic strategies being evaluated in breast cancer patients. However, vaccination with longer peptides derived from the MUC-1 breast cancer tumor antigen has undergone extensive clinical study. MUC-1 is a membrane-bound glycoprotein and is a primary component of the ductal cell

M.L. Disis et al. / Peptide-Based Vaccines in Breast Cancer

surface of most glandular cells. When a normal cell becomes cancerous, MUC-1 is both overexpressed and aberrantly glycosolated. Overexpression of MUC-1 in cancer cells interferes with cell adhesion and may facilitate the development of metastasis. Several peptide based MUC-1 vaccines have been evaluated in clinical trials using it as a model antigen for eliciting both humoral and cellular immune responses. Early trials using MUC-1 with BCG as adjuvant induced DTH responses, but only in a minority of patients [30]. Formulating MUC-1 as a mannan-MUC-1 fusion protein and immunizing patients multiple times resulted in augmentation of MUC-1 specic antibodies in about 50% of patients, with titers higher than 1/20,000. However, patient numbers were small [31]. A more recent study describes vaccination of breast cancer patients with MUC-1/KLH along with QS-21, an adjuvant specifically developed to stimulate cellular immunity [32]. While antibody titers were elevated, and remained so over 100 days after vaccination, no antigen-specic T cell responses were detected. MUC-1 specic antibodies generated after active immunization have, however, been demonstrated to bind to and lyse tumor in vitro [33].

PEPTIDE-SPECIFIC T CELLS ELICITED BY ACTIVE IMMUNIZATION AND DIRECTED AGAINST A SINGLE BREAST CANCER ANTIGENIC EPITOPE CAN BE REMARKABLY DIVERSE IN PHENOTYPE AND FUNCTION A theoretical drawback of the use of peptide epitopes as immunogens is the risk of generating very restricted immune responses, e.g. T cells of a single phenotype directed against the epitope. A strongly restricted response may limit potential therapeutic efcacy. Dissection of the peptide-specic immune response after immunization has demonstrated a marked phenotypic diversity of the vaccinated response [34]. T cell clones specic for HER-2/neu HLA-A2 peptide p369-377 were isolated from a patient who had been vaccinated with HER-2/neu helper epitopes that contained HLA-A2-binding CTL epitopes within their sequences. Throughout the course of immunization, PBMC from this patient showed strong proliferative responses to the HER-2/neu helper epitope p369-384. Following vaccination, T cell clones specic for p369377, the HLA-A2 binding peptide were isolated by limiting dilution and then characterized. The responding T cell repertoire generated was both phenotypically and

functionally diverse. A total of 21 p369-377 T-cell clones were isolated from this patient; peptide-specic CD8+, CD4+ and T cell clones. Of note, despite being generated with an HLA-A2-binding peptide, CD4+ T cells were generated. Although CD4+ T cells are predominantly associated with responding to peptides associated with MHC class II, at lower frequencies CD4 T cells play a role in regulating MHC class I-restricted T cells, particularly T cells associated with cancers such as melanoma, colon, and pancreatic cancer [35,36]. The majority of the clones expressed the T cell receptor (TCR), however, two clones expressed the TCR. Clones could lyse HLAA2-transfected HER-2/neu overexpressing tumor cells as well as peptide loaded MHC-matched cells. In addition to their lytic capabilities these clones could be induced to produce IFN- specically in response to p369-377 peptide stimulation. The TCR clones expressed CD8 and lysed HLA-A2 HER-2/neu positive tumor cells, but not HLA-A2 negative HER-2/neu overexpressing tumor cells. T cells are involved in a wide range of immune responses to infectious and non-infectious diseases, including malaria, mycobacterial infections, cancers, as well as autoimmune disorders such as multiple sclerosis [37]. Often, T cells clones are isolated from the tumor-inltrating lymphocyte population of many tumors including dysgerminoma [38], seminoma [38], renal carcinoma [39], lung [40], colorectal [41], and melanoma [42]. The recruitment to and role of these unique cells in mediating anti-tumor immunity is unknown. One hypothesis is that, given their broad range of regulation by multiple mechanisms of antigen presentation and natural localization to epithelial tissue, TCR T cells are sentinels for the immune system and are capable of alerting the immune system to the presence of danger (e.g. infection, tumors, etc.). These results suggest that a tumor antigen-specic T cell response can be markedly polyclonal at multiple levels including T cell subset and TCR even if peptide epitopes are used as the immunizing antigens.

BREAST CANCER VACCINES COMPOSED OF PEPTIDES AND DESIGNED TO GENERATE BREAST CANCER SPECIFIC ANTIBODY IMMUNITY Another aspect of peptide epitope prediction would be to identify peptide portions of a breast cancer antigen, such as the HER-2/neu extracellular domain

M.L. Disis et al. / Peptide-Based Vaccines in Breast Cancer

(ECD), appropriate to target with an antibody response. Several monoclonal antibodies against the HER-2/neu ECD have been isolated and one such antibody, trastuzumab, has demonstrated clinical efcacy in the treatment of metastatic breast cancer [43] . Although many HER-2/neu-specic antibodies inhibit the growth of cancer cells, some antibodies have no effect on cell growth, while others even actively stimulate cancer growth [44]. This wide range of biological effects is thought to be related to the epitope specicity of the antibodies and to consequent changes in receptor signaling [44]. An alternative to the use of passive antibody therapy would be active immunization against the HER-2/neu ECD. However, inappropriately induced immune responses could have untoward effects on cancer growth. Therefore, it is crucial to identify epitopes on HER-2/neu that are targeted by stimulatory and inhibitory antibodies in order to ensure the induction of a benecial endogenous antibody response. In a recent study, investigators constructed HER2/neu gene fragment phage display libraries to epitopemap a number of HER-2/neu-specic antibodies with different biological effects on tumor cell growth [44]. Regions responsible for opposing effects of antibodies were identied and then used to immunize mice. Mice immunized with the peptides developed antibodies that recognized both the peptide and native HER2/neu. More importantly, HER-2/neu-specic antibodies puried from mouse sera were able to inhibit up to 85% of tumor cell proliferation in vitro. Furthermore, using peptide regions that contain multiple inhibitory B cell epitopes is likely to be superior to the use of single epitope immunogens [45].

THE NEXT GENERATION PEPTIDE BASED VACCINES FOR USE IN BREAST CANCER THERAPY Initial clinical trials of peptide-based breast cancer vaccines indicated that patients can be immunized, but immune responses were generally not to the levels of a vaccinated antigen such as tetanus toxoid. Therefore, more recent investigations are focused on improving the immunogenicity of peptide immunization. The choice of adjuvant is an important consideration in peptide vaccine trials, particularly as peptides themselves are typically only weakly immunogenic. Due to recent renewed enthusiasm for peptide-based tumor vaccines coupled with the search for adjuvants

that promote cellular as well as antibody responses, the list of adjuvants being studied continues to grow. The choice of adjuvant has largely been determined by pre-clinical models. Similar to alum, oil-based incomplete Freunds-type adjuvants, such as Montanide ISA51 (Seppic, Paris, France) and TiterMax (CytRx Corp., Norcross, GA), have been used in human peptide vaccine studies [46,47], with their adjuvant effect likely mediated by a depot effect,increasing the half life of the peptide antigen at the site of immunization. Other proinammatory biologic adjuvants such as BCG, Detox (mycobacterial cell-wall skeleton plus a lipid moiety) (Corixa Corp., Seattle, WA), and inuenza virosomes have been used as adjuvants in peptide- and proteinbased human vaccine studies [4850]. One study compared a variety of adjuvants in a rodent model targeting human MUC-1 peptide antigens conjugated to keyhole limpet hemocyanin (KLH), and found the saponin adjuvant QS-21 (Aquilla Biopharmaceuticals, Worcester, MA) to be the most effective in promoting an antigenspecic antibody responses [51]. Subsequently, QS-21 has been used as an adjuvant in human peptide cancer vaccine trials [52,53], with the hope that the use of a saponin may allow entry of peptides directly into MHC complexes to stimulate a T cell response [54]. Another type of adjuvant that has been studied in animal models is polymer microspheres that encapsulate the peptide antigen, permitting slow release over variable lengths of time [55]. Potential advantages to this type of adjuvant includes the ability to use an oral route of delivery [56], continuous release of antigen obviating the need for booster immunizations, and the ability to incorporate other biologic adjuvants or cytokines within the polymer [57]. Finally, the discovery of various cytokines participating in the initiation of immune responses has suggested that cytokines themselves may be useful immunologic adjuvants. As an example, soluble GM-CSF is a vaccine adjuvant, and has the ability to induce the differentiation of dendritic cells and act as a chemoattractant for various immune cell effectors [58,59]. The relationship between MHC class I afnity and tumor antigen epitope immunogenicity is of importance because tissue-specic and developmental tumor antigens, such as those that are breast cancer antigens, are expressed on normal tissues at some point in time at some location within the body. T cells specic for these self antigens could be functionally inactivated by T cell tolerance. However, CTL responses to tumor epitopes in both volunteer donors and breast cancer patients indicate that tolerance to these tumor antigens, if it ex-

M.L. Disis et al. / Peptide-Based Vaccines in Breast Cancer

ists at all, is incomplete [8,60]. It is unknown whether T cells recognizing high afnity epitopes have been selectively eliminated, leaving a repertoire capable of recognizing only low afnity epitopes. One investigation evaluated several peptides derived from four different breast cancer tumor antigens, p53, HER-2/neu, CEA, and MAGE proteins, for their capacity to induce CTL in vitro capable of recognizing tumor target lines [8]. In order to increase the likelihood of overcoming tolerance, xed anchor analogues that demonstrate improved HLA-*0201 afnity and binding, were used along with wild type peptides. A total of 20/20 wild-type and 9/9 single amino acid substitution analogues were found to be immunogenic in primary in vitro CTL induction assays, using normal PBMCs and monocyte derived dendritic cells as APC. Cytotoxic T cells generated by 13/20 of the wild type epitopes and 6/9 of single substitution analogues tested recognized HLA-matched antigen bearing cancer cell lines. Further analysis suggested that high binding afnity epitopes are frequently generated and can be recognized as a result of natural antigen processing. Indeed, the immunogenicity of CTL epitopes derived from human tumor antigens such as CEA can be improved by altering the peptide motif to improve binding to MHC [61]. Studies such as these demonstrate that recognition of self-tumor antigens is within the realm of the T cell repertoire and that binding afnity may be an important criterion for epitope selection.

ACKNOWLEDGEMENTS This work is supported for MLD by grants from the NIH, NCI (R01 CA75163 and CA85374, and K24 CA85218) as well as funding from the Cancer Research Treatment Foundation, for LGS by NIH training grant T32 (HL07093), and for KLK by a fellowship from the Department of Defense Breast Cancer Program (DAMD17-00-1-0492). REFERENCES
[1] [2] [3] M. Disis, Immunologic targets for breast cancer therapy, Breast Disease 15 (2003), 8390. G. Ada, Vaccines and vaccination, N Engl J Med 345 (2001), 10421053. M.L. Disis, J.R. Gralow and H. Bernhard et al., Peptide-based, but not whole protein, vaccines elicit immunity to HER-2/neu, an oncogenic self-protein, J Immunol 156 (1996), 31513158. R. Cibotti, J.M. Kanellopoulos and J.P. Cabaniols et al., Tolerance to a self-protein involves its immunodominant but does not involve its subdominant determinants, Proc Natl Acad Sci USA 89 (1992), 416420. E.E. Sercarz, P.V. Lehmann and A. Ametani et al., Dominance and crypticity of T cell antigenic determinants, Annu Rev Immunol 11 (1993), 729766. K.L. Knutson, K. Schiffman and M.A. Cheever et al., Immunization of cancer patients with a HER-2/neu, HLA-A2 peptide, p369-377, results in short-lived peptide-specic immunity, Clin Cancer Res 8 (2002), 10141018. M.L. Disis, T.A. Gooley and K. Rinn et al., Generation of Tcell immunity to the HER-2/neu protein after active immunization with HER-2/neu Peptide-based vaccines, J Clin Oncol 20 (2002), 26242632. E. Keogh, J. Fikes and S. Southwood et al., Identication of new epitopes from four different tumor-associated antigens: recognition of naturally processed epitopes correlates with hlaa(*)0201-binding afnity, J Immunol 167 (2001), 787796. J. Lu and E. Celis, Use of two predictive algorithms of the world wide web for the identication of tumor-reactive T-cell epitopes, Cancer Res 60 (2000), 52235227. Y. Rongcun, F. Salazar-Onfray and J. Charo et al., Identication of new HER2/neu-derived peptide epitopes that can elicit specic CTL against autologous and allogeneic carcinomas and melanomas, J Immunol 163 (1999), 10371044. A.M. Livingstone and C.G. Fathman, The structure of T-cell epitopes, Annu Rev Immunol 5 (1987), 477501. J.C. Reece, D.L. McGregor and H.M. Geysen et al., Scanning for T helper epitopes with human PBMC using pools of short synthetic peptides, J Immunol Methods 172 (1994), 241254. H. Kobayashi, M. Wood and Y. Song et al., Dening promiscuous MHC class II helper T-cell epitopes for the HER2/neu tumor antigen, 60 (2000), 52285236. H.M. Zarour, B. Maillere and V. Brusic et al., NY-ESO-1 119143 is a promiscuous major histocompatibility complex class II T-helper epitope recognized by Th1- and Th2-type tumorreactive CD4+ T cells, Cancer Res 62 (2002), 213218. R. Schroers, X.F. Huang and J. Hammer et al., Identication of HLA DR7-restricted epitopes from human telomerase reverse transcriptase recognized by CD4+ T-helper cells, Cancer Res 62 (2002), 26002605.

[4]

[5]

[6]

[7]

[8]

[9]

CONCLUSION
[10]

The denition of multiple breast cancer antigenspecic peptides, both class I and class II derived epitopes are rapidly being translated into human clinical studies. Investigations over the last decade have allowed the development of technology such as computer algorithms and high throughput peptide binding assays to rapidly screen vaccine candidates. Clinical trials of single peptides have demonstrated that cancer patients can be immunized against self-tumor antigens and some studies have shown early positive clinical results. Current effort in the eld is focused on improving the immunogenicity of individual MHC binding peptides as well as developing multiple peptide vaccines for the prevention and treatment of breast cancer.

[11] [12]

[13]

[14]

[15]

10 [16]

M.L. Disis et al. / Peptide-Based Vaccines in Breast Cancer K. Hung, R. Hayashi and A. Lafond-Walker et al., The central role of CD4(+) T cells in the antitumor immune response, J Exp Med 188 (1998), 23572368. S.A. Kalams and B.D. Walker, The critical need for CD4 help in maintaining effective cytotoxic T lymphocyte responses [comment], J Exp Med 188 (1998), 21992204. C. DeLisi BJ, T-cell antigenic sites tend to be amphipathic structures, Proc. Natl. Acad. Sci. USA 82 (1985), 70487052. D.C. Feller, dlCV. Identifying antigenic T-cell sites, Nature 349 (1991), 720721. A. OGarra and K. Murphy, T-cell subsets in autoimmunity, Curr Opin Immunol 5 (1993), 880886. C.G. Ioannides and T.L. Whiteside, T cell recognition of human tumors: implications for molecular immunotherapy of cancer, Clin Immunol Immunopathol 66 (1993), 91106. L. Salazar, J. Fikes and S. Southwood et al., Immunization of cancer patients with HER-2/neu derived peptides demonstrating high afnity binding to multiple class II alleles, Clin Ca Res (in press 2003). T.Z. Zaks and S.A. Rosenberg, Immunization with a peptide epitope (p369-377) from HER-2/neu leads to peptide-specic cytotoxic T lymphocytes that fail to recognize HER- 2/neu+ tumors, Cancer Res 58 (1998), 49024908. J.L. Murray, M.E. Gillogly and D. Przepiorka et al., Toxicity, immunogenicity, and induction of E75-specic tumorlytic CTLs by HER-2 peptide E75 (369-377) combined with granulocyte macrophage colony-stimulating factor in HLAA2+ patients with metastatic breast and ovarian cancer, Clin Cancer Res 8 (2002), 34073418. P. Brossart, S. Wirths and G. Stuhler et al., Induction of cytotoxic T-lymphocyte responses in vivo after vaccinations with peptide-pulsed dendritic cells, Blood 96 (2000), 31023108. P.V. Lehmann, T. Forsthuber and A. Miller et al., Spreading of T-cell autoimmunity to cryptic determinants of an autoantigen, Nature 358 (1992), 155157. E.E. Sercarz, Driver clones and determinant spreading, J Autoimmun 14 (2000), 275277. A. Jaramillo, K. Majumder and P.P. Manna et al., Identication of HLA-A3-restricted CD8+ T cell epitopes derived from mammaglobin-A, a tumor-associated antigen of human breast cancer, Int J Cancer 102 (2002), 499506. M.H. Andersen, L.O. Pedersen and B. Capeller et al., Spontaneous cytotoxic T-cell responses against survivin-derived MHC class I-restricted T-cell epitopes in situ as well as ex vivo in cancer patients, Cancer Res 61 (2001), 59645968. M. Reddish, G.D. MacLean and R.R. Koganty et al., AntiMUC1 class I restricted CTLs in metastatic breast cancer patients immunized with a synthetic MUC1 peptide, Int J Cancer 76 (1998), 817823. V. Karanikas, G. Thynne and P. Mitchell et al., Mannan mucin1 peptide immunization: inuence of cyclophosphamide and the route of injection, J Immunother 24 (2001), 172183. T. Gilewski, S. Adluri and G. Ragupathi et al., Vaccination of high-risk breast cancer patients with mucin-1 (MUC1) keyhole limpet hemocyanin conjugate plus QS-21, Clin Cancer Res 6 (2000), 16931701. F.G. Snijdewint, S. von Mensdorff-Pouilly and A.H. KaruntuWanamarta et al., Antibody-dependent cell-mediated cytotoxicity can be induced by MUC1 peptide vaccination of breast cancer patients, Int J Cancer 93 (2001), 97106. K. Knutson and M. Disis, Clonal diversity of the T cell repertoire responding to a dominant HLA-A2 eptiope of HER-2/neu after active immunization, Hum Immunol 63 (2002), 547557. [35] J.E. de Vries and H. Spits, Cloned human cytotoxic T lymphocyte (CTL) lines reactive with autologous melanoma cells. I. In vitro generation, isolation, and analysis to phenotype and specicity, J Immunol 132 (1984), 510519. R. Somasundaram, P. Robbins and D. Moonka et al., CD4(+), HLA class I-restricted, cytolytic T-lymphocyte clone against primary malignant melanoma cells, Int J Cancer 85 (2000), 253259. R. Boismenu and W.L. Havran, Gammadelta T cells in host defense and epithelial cell biology, Clin Immunol Immunopathol 86 (1998), 121133. X. Zhao, Y.Q. Wei and Y. Kariya et al., Accumulation of gamma/delta T cells in human dysgerminoma and seminoma: roles in autologous tumor killing and granuloma formation, Immunol Invest 24 (1995), 607618. A. Choudhary, F. Davodeau and A. Moreau et al., Selective lysis of autologous tumor cells by recurrent gamma delta tumorinltrating lymphocytes from renal carcinoma, J Immunol 154 (1995), 39323940. D. Yu, A. Matin and W. Xia et al., Liposome-mediated in vivo E1A gene transfer suppressed dissemination of ovarian cancer cells that overexpress HER-2/neu, Oncogene 11 (1995), 13831388. N. Watanabe, A. Hizuta and N. Tanaka et al., Localization of T cell receptor (TCR)-gamma delta + T cells into human colorectal cancer: ow cytometric analysis of TCR-gamma delta expression in tumour-inltrating lymphocytes, Clin Exp Immunol 102 (1995), 167173. H. Bachelez, B. Flageul and L. Degos et al., TCR gamma delta bearing T lymphocytes inltrating human primary cutaneous melanomas, J Invest Dermatol 98 (1992), 369374. C. Vogel, M.A. Cobleigh and D. Tripathy et al., First-line, single-agent Herceptin(R) (trastuzumab) in metastatic breast cancer. a preliminary report, Eur J Cancer 37 (Suppl 1) (2001), 2529. Y.L. Yip, G. Smith and J. Koch et al., Identication of epitope regions recognized by tumor inhibitory and stimulatory anti-ErbB-2 monoclonal antibodies: implications for vaccine design, J Immunol 166 (2001), 52715278. N.K. Dakappagari, D.B. Douglas and P.L. Triozzi et al., Prevention of mammary tumors with a chimeric HER-2 B-cell epitope peptide vaccine, Cancer Res 60 (2000), 37823789. R.M.W.J. van Driel, G.G. Kenter, R.M. Brandt, E.J. Krul, A.B. van Rossum, E. Schuuring, R. Offringa, T. Bauknecht, A. Tamm-Hermelink, P.A. van Dam, G.J. Fleuren, W.M. Kast, C.J. Melief and J.B. Trimbos, Vaccination with HPV16 peptides of patients with advanced cervical carcinoma: clinical evaluation of a phase I-II trial, Eur J Cancer 35 (1999), 946 952. F.B.E.Wang, C. Kuniyoshi, L. Spears, G. Jeffery, V. Marty, S. Groshen and J. Weber, Phase I trial of a MART-1 peptide vaccine with incomplete Freunds adjuvant for resected highrisk melanoma, Clin. Cancer Res. 5 (1999), 27562765. J.S.E.E. Goydos, T.L. Whiteside, O.J. Finn and M.T. Lotze, A phase I trial of a synthetic mucin peptide vaccine. Induction of specic immune reactivity in patients with adenocarcinoma, J. Surg. Res. 63 (1996), 298304. N. Schultz, R. Oratz, D. Chen, A. Zeleniuch-Jacquotte, G. Abeles and J.C. Bystryn, Effect of DETOX as an adjuvant for melanoma vaccine, Vaccine 13 (1995), 503508. R. Gluck, Immunopotentiating reconstituted inuenza virosomes (IRIVs) and other adjuvants for improved presentation of small antigens, Vaccine 10 (1992), 915919.

[17]

[36]

[18] [19] [20] [21]

[37]

[38]

[22]

[39]

[23]

[40]

[24]

[41]

[42]

[25]

[43]

[26]

[27] [28]

[44]

[45]

[29]

[46]

[30]

[31]

[47]

[32]

[48]

[33]

[49]

[34]

[50]

M.L. Disis et al. / Peptide-Based Vaccines in Breast Cancer [51] S.K.R.G. Kim, C. Musselli, S.J. Choi, Y.S. Park and P.O. Livingston, Comparison of the effect of different immunological adjuvants on the antibody and T-cell response to immunization with MUC1-KLH and GD3-KLH conjugate cancer vaccines, Vaccine 18 (1999), 597603. T.A.S. Gilewski, G. Ragupathi, S. Zhang, T.J. Yao, K. Panageas, M. Moynahan, A. Houghton, L. Norton and P.O. Livingston, Vaccination of high-risk breast cancer patients with mucin-1 (MUC1) keyhole limpet hemocyanin conjugate plus QS-21, Clin. Cancer Res. 6 (2000), 16931701. J. Pinilla-Ibarz, K. Cathcart and T. Korontsvit et al., Vaccination of patients with chronic myelogenous leukemia with bcrabl oncogene breakpoint fusion peptides generates specic immune responses, Blood 95 (2000), 17811787. M.J.W.J. Newman, B.H. Gardner, K.J. Munroe, D. Leombruno, J. Recchia, C.R. Kensil and R.T. Coughlin, Saponin adjuvant induction of ovalbumin-specic CD8+ cytotoxic T lymphocyte responses, J. Immunol. 148 (1992), 23572362. Y.G.B. Men, H.P. Merkle and G. Corradin, Induction of sustained and elevated immune responses to weakly immunogenic synthetic malarial peptides by encapsulation in biodegradable polymer microspheres, Vaccine 14 (1996), 14421450. C.O.R.R. Tacket, E.C. Boedeker, G. Losonsky, J.P. Nataro,

11

[57]

[52]

[58]

[53]

[59]

[54]

[60]

[55]

[61]

H. Bhagat and R. Edelman, Enteral immunization and challenge of volunteers given enterotoxigenic E. coli CFA/II encapsulated in biodegradable microspheres, Vaccine 12 (1994), 12701274. R.C.J. Langer and J. Hanes, New advances in microspherebased single-dose vaccines, Adv. Drug Deliv. Rev. 28 (1997), 97119. S.E.E. Markowicz, Granulocyte-macrophage colonystimulating factor promotes differentiation and survival of human peripheral blood dendritic cells in vitro, J. Clin. Invest. 85 (1990), 955961. R.H.G.D. Weisbart, S.C. Clark, G.G. Wong and J.C. Gasson, Human granulocyte-macrophage colony stimulating factor is a neutrophil activator, Nature 314 (1985), 361363. I. Kawashima, S.J. Hudson and V. Tsai et al., The multiepitope approach for immunotherapy for cancer: identication of several CTL epitopes from various tumor-associated antigens expressed on solid epithelial tumors, Hum. Immunol. 59 (1998), 114. E. Huarte, P. Sarobe and J. Lu et al., Enhancing Immunogenicity of a CTL Epitope from Carcinoembryonic Antigen by Selective Amino Acid Replacements, Clin Cancer Res 8 (2002), 23362344.

[56]