Catherine A. King

DNA vaccines with single-chain Fv fused to fragment C of tetanus toxin induce protective immunity against lymphoma and myeloma

Vaccination with idiotypic protein protects against B-cell lymphoma, mainly through anti-idiotypic antibody. For use in patients, DNA vaccines containing single-chain Fv derived from tumor provide a convenient alternative vaccine delivery system. However, single-chain Fv sequence alone induces low anti-idiotypic response and poor protection against lymphoma. Fusion of the gene encoding fragment C of tetanus toxin to single-chain Fv substantially promotes the anti-idiotypic response and induces strong protection against B-cell lymphoma. The same fusion design also induces protective immunity against a surface Ig-negative myeloma. These findings indicate that fusion to a pathogen sequence allows a tumor antigen to engage diverse immune mechanisms that suppress growth. This fusion design has the added advantage of overcoming potential tolerance to tumor that may exist in patients.

Idiotypic DNA vaccines against B-cell lymphoma.

Idiotypic antigens are clearly defined tumor-associated protein antigens, which can induce protective immunity against lymphoma. Because each patient requires an individual vaccine, idiotypic antigens also provide ideal candidates for exploring the feasibility of replacing protein antigens by DNA vaccines. Component idiotypic variable region genes can be identified in patients tumor biopsies and rapidly assembled as scFv sequences. These can be used to produce recombinant scFv protein in bacteria, or as direct naked DNA vaccines. A preliminary small trial of DNA vaccines for chemotherapy-resistant patients with lymphoma has begun. Intramuscular idiotypic DNA vaccination in a mouse model induces low levels of anti-idiotypic antibody in serum. Levels can be increased dramatically by coinjection of DNA plasmids encoding either IL-2 or GM-CSF, and specific proliferative anti-idiotypic T cells are induced. However protective immunity remains to be demonstrated, and a possible reason for this may lie in the continued secretion of idiotypic scFv antigen which blocks antibody activity by formation of immune complexes. Methods for regulating secretion of antigen are required before this category of tumor antigen can be fully exploited as a vaccine. The power of DNA technology should allow analysis and manipulation of pathways of antigen presentation to induce maximal therapeutic attack on neoplastic B cells. In addition, lymphoma presents a model for application of DNA technology to the wide range of human tumors known to harbor potential tumor antigens.

Electroporation as a “Prime/Boost” Strategy for Naked DNA Vaccination against a Tumor Antigen

We have developed novel DNA fusion vaccines encoding tumor Ags fused to pathogen-derived sequences. This strategy activates linked T cell help and, using fragment C of tetnus toxin, amplification of anti-tumor Ab, CD4+, and CD8+ T cell responses is achievable in mice. However, there is concern that simple DNA vaccine injection may produce inadequate responses in larger humans. To overcome this, we tested electroporation as a method to increase the transfection efficiency and immune responses by these tumor vaccines in vivo in mice. Using a DNA vaccine expressing the CTL epitope AH1 from colon carcinoma CT26, we confirmed that effective priming and tumor protection in mice are highly dependent on vaccine dose and volume. However, suboptimal vaccination was rendered effective by electroporation, priming higher levels of AH1-specific CD8+ T cells able to protect mice from tumor growth. Electroporation during priming with our optimal vaccination protocol did not improve CD8+ T cell responses. In contrast, electroporation during boosting strikingly improved vaccine performance. The prime/boost strategy was also effective if electroporation was used at both priming and boosting. For Ab induction, DNA vaccination is generally less effective than protein. However, prime/boost with naked DNA followed by electroporation dramatically increased Ab levels. Thus, the priming qualities of DNA fusion vaccines, integrated with the improved Ag expression offered by electroporation, can be combined in a novel homologous prime/boost approach, to generate superior antitumor immune responses. Therefore, boosting may not require viral vectors, but simply a physical change in delivery, facilitating application to the cancer clinic.

Manipulation of pathogen-derived genes to influence antigen presentation via DNA vaccines

To gain insight into the routes of presentation of pathogen sequences via DNA vaccines, we have compared the abilities of sequences encoding fragment C of tetanus toxin (FrC) and influenza A virus nucleoprotein (NP) to induce antibody or cytotoxic T-cell (CTL) responses in vivo. Strong antibody and CTL responses were induced against FrC targeted to the endoplasmic reticulum (ER) and both were reduced by removal of the leader sequence. In contrast, targeting of NP to the ER generated only a modest antibody response, likely due to misfolding in this site. Removal of the leader sequence led to anti-NP antibodies via cross-priming. For NP, induction of CTLs was not influenced by the leader sequence. Exogenous FrC or NP delivered as proteins were unable to induce CTLs. Routes to induction of optimal immune responses via DNA evidently differ according to the nature of the encoded pathogen sequence. Understanding processing pathways for pathogen sequences should assist rational design of DNA vaccines.

Immunoglobulin VH gene sequence analysis of spontaneous murine immunoglobulin-secreting B-cell tumours with clinical features of human disease

The 5T series of multiple myelomas (MM) and Waldenstrsöm’s macroglobulinaemia‐like lymphomas (WM), which developed spontaneously in ageing mice of the C57BL/KaLwRij strain, shows clinical and biological features that closely resemble their corresponding human diseases. In order to compare the patterns of somatic mutation in VH genes of mouse tumours with those of human counterparts, we have determined and analysed sequences of immunoglobulin VH genes in five cases of murine MM, two of WM and one of biclonal benign monoclonal gammopathy (BMG). Four of five MM and 2/2 WM cases used VH genes of the large J558 family; one MM used a gene of the VGAM3.8 family, and both clones of the BMG used genes of the 36‐60 family. N‐region insertions were observed in all cases, but D‐segment genes were only identified in 6/9 cases, which were all from the D‐SP family and translated in reading frame 3. Compared with human MM, in which the VH genes have been found to be consistently hypermutated (mean%±SD=8·8±3·2), the degree of somatic mutation in the murine tumours was significantly lower (mean%±SD=2·9±2·3). There was no significant evidence of clustering of replacement mutations in complementarity determining regions (CDR), a feature considered to be characteristic of antigen‐selected sequences. However, one clone of the biclonal BMG case showed intraclonal variation, a feature described in some cases of human BMG. These results indicate that murine VH genes in mature tumours differ from human counterparts in the level and distribution of somatic mutations, but support the concept that BMG may be distinct from MM.

A genetic approach to idiotypic vaccination for B cell lymphoma.

Idiotypic immunoglobulin expressed by a B cell tumor presents a clear tumor antigen which could be attacked by vaccination of the host. Vaccination with idiotypic protein has been shown to induce protective immunity against lymphoma, but application to patients is limited by the requirement of personal vaccines for each patient. A genetic approach enables V-region sequences encoding idiotypic antigen to be rescued from tumor biopsies, and to be assembled as scFv fragments. These can be expressed in bacteria to produce recombinant protein, or used directly as naked DNA vaccines. Intramuscular injection of idiotypic DNA from a mouse B cell lymphoma induces low levels of syngeneic anti-idiotypic antibody in serum. Response can be stimulated by co-injection of DNA plasmids encoding either IL-2 or GM-CSF, and T cells which proliferate in response to idiotypic IgM are generated. However, protection against tumor appears to be blocked by continuing secretion of idiotypic antigen from the persisting vaccine vector, which forms immune complexes with serum antibody. Methods for regulating the level of scFv to engage the immune system, but not to block the effector arm are being investigated. Similar control will be applicable to the cytokine vectors, which can deliver encoded cytokines designed to activate immune pathways for tumor destruction. Experience gained in lymphoma may be extended to other tumors with defined tumor antigens.

Insight into the potential for DNA idiotypic fusion vaccines designed for patients by analysing xenogeneic anti-idiotypic antibody responses.

DNA vaccines induce immune responses against encoded proteins, and have clear potential for cancer vaccines. For B‐cell tumours, idiotypic (Id) immunoglobulin encoded by the variable region genes provides a target antigen. When assembled as single chain Fv (scFv), and fused to an immunoenhancing sequence from tetanus toxin (TT), DNA fusion vaccines induce anti‐Id antibodies. In lymphoma models, these antibodies have a critical role in mediating protection. For application to patients with lymphoma, two questions arise: first, whether pre‐existing antibody against TT affects induction of anti‐scFv antibodies; second, whether individual human scFv fusion sequences are able to fold consistently to generate antibodies able to recognize private conformational Id determinants expressed by tumour cells. Using xenogeneic vaccination with scFv sequences from four patients, we have shown that pre‐existing anti‐TT immunity slows, but does not prevent, anti‐Id antibody responses. To determine folding, we have monitored the ability of nine DNAscFv–FrC patients vaccines to induce xenogeneic anti‐Id antibodies. Antibodies were induced in all cases, and were strikingly specific for each patients immunoglobulin with little cross‐reactivity between patients, even when similar VH or VL genes were involved. Blocking experiments with human serum confirmed reactivity against private determinants in 26–97% of total antibody. Both immunoglobulin G1 (IgG1) and IgG2a subclasses were present at 1·3u200a:u200a1–15u200a:u200a1 consistent with a T helper 2‐dominated response. Xenogeneic vaccination provides a simple route for testing individual patients DNAscFv–FrC fusion vaccines, and offers a strategy for production of anti‐Id antibodies. The findings underpin the approach of DNA idiotypic fusion vaccination for patients with B‐cell tumours.

DNA vaccination against cancer antigens

Rational approaches aimed at manipulating the immune system to act against cancer cells are now becoming feasible. There are two main reasons for this, both arising from developments in molecular genetics: first, there is a greater understanding of the tumor-associated changes occurring in cancer cells, some of which will generate candidate target antigens. Second, there is increasing knowledge of the processes involved in inducing an effective immune response. In attempting to activate the immune system against cancer, we need to remember that the major task of the immune response is to control or eliminate pathogens. Development of vaccines against cancer will rely on lessons from infectious diseases, and is bringing back together the fields of microbiology and immunology which had become separated, perhaps partly due to the effectiveness of antibiotics.

DNA fusion vaccines against B-cell tumors.

The ability of naked DNA to induce immune responses against encoded antigen has been clearly demonstrated for infectious diseases (1). In many cases, the induced immunity is able to protect against infection, and can approach the efficacy of exogenous antigen (2).