Tumor immunogenicity is determined by the mechanism of cell death via induction of heat shock protein expression.

In situ killing of tumor cells using suicide gene transfer to generate death by a non-apoptotic pathway was associated with high immunogenicity and induction of heat shock protein (hsp) expression. In contrast, a syngeneic colorectal tumor line, CMT93, killed predominantly by apoptosis, showed low levels of hsp expression and less immunogenicity. When apoptosis was inhibited in CMT93 cells by overexpression of bcl-2, hsp was also induced. Furthermore, when cDNA encoding hsp70 was stably transfected into B16 and CMT93 cells, its expression significantly enhanced the immunogenicity of both tumors. Increased levels of hsp, induced by non-apoptotic cell killing, may provide an immunostimulatory signal in vivo which helps break tolerance to tumor antigens. These findings have important implications for the development of novel anti-cancer therapies aimed at promoting patients' immune responses to their own tumors.

A synthetic peptide adhesion epitope as a novel antimicrobial agent.

The earliest step in microbial infection is adherence by specific microbial adhesins to the mucosa of the oro-intestinal, nasorespiratory, or genitourinary tract. We inhibited binding of a cell surface adhesin of Streptococcus mutans to salivary receptors in vitro, as measured by surface plasmon resonance, using a synthetic peptide (p1025) corresponding to residues 1025-1044 of the adhesin. Two residues within p1025 that contribute to binding (Q1025, E1037) were identified by site-directed mutagenesis. In an in vivo human streptococcal adhesion model, direct application of p1025 to the teeth prevented recolonization of S. mutans but not Actinomyces, as compared with a control peptide or saline. This novel antimicrobial strategy, applying competitive peptide inhibitors of adhesion, may be used against other microorganisms in which adhesins mediate colonization of mucosal surfaces.

Adoptive transfer of immature dendritic cells with autologous or allogeneic tumor cells generates systemic antitumor immunity.

Dendritic cells (DCs) are potent antigen-presenting cells that are capable of priming systemic antitumor immune responses in animal tumor models. However, many of the model tumor systems tested need definition of the specific tumor antigens involved. To use DCs in situations that are more relevant to the majority of human cancers, where the antigens are unknown, we have tested the adoptive transfer of immature DCs in mouse colorectal and melanoma models of varying immunogenicity but with undefined antigens. When DCs admixed with a syngeneic primary tumor inoculum were seeded s.c., the growth of the primary tumor was unchanged; however, if the primary tumor was then surgically excised and the animal was rechallenged with the same tumor, significant protection (75%) was generated when DCs were present in the original primary inoculum of a moderately immunogenic colorectal model (CMT93tk). This effect was not observed when a nonimmunogenic melanoma (B16) was tested in an identical protocol. Next, DCs were injected directly into 6-9-mm established tumors; again, protection (55%) was achieved against a secondary tumor challenge following excision of the primary, but only in the CMT93tk model of moderate immunogenicity. To increase the clinical relevance of this approach still further, we tested irradiated allogeneic K1735 melanoma cells mixed with syngeneic DCs as a vaccine against subsequent challenge with the poorly immunogenic syngeneic melanoma B16. The allogeneic vaccine alone was ineffective, but when admixed with DCs, a significant number of animals rejected a subsequent B16 challenge, suggesting that DCs are able to prime an immune response against melanoma antigens shared between K1735 and B16. The generation of systemic antitumor immunity by adoptive transfer of DCs has significant clinical potential because it is technically straightforward and does not require the definition of specific tumor antigens.

Preclinical evaluation of "whole" cell vaccines for prophylaxis and therapy using a disabled infectious single cycle-herpes simplex virus vector to transduce cytokine genes.

The development of genetically modified "whole" tumor cell vaccines for cancer therapy relies on the efficient transduction and expression of genes by vectors. In the present study, we have used a disabled infectious single cycle-herpes simplex virus 2 (DISC-HSV-2) vector constructed to express cytokine or marker genes upon infection. DISC-HSV-2 is able to infect a wide range of tumor cells and efficiently express the beta-galactosidase reporter gene, granulocyte-macrophage colony-stimulating factor (GM-CSF), or IL-2 genes. Gene expression occurred rapidly after infection of tumor cells, and the level of production of the gene product (beta-galactosidase, GM-CSF, or IL-2) was shown to be both time-and dose-dependent. Vaccination with irradiated DISC-mGM-CSF or DISC-hIL-2-infected murine tumor cells resulted in greatly enhanced immunity to tumor challenge with live parental tumor cells compared with control vaccines. When used therapeutically to treat existing tumors, vaccination with irradiated DISC-mGM-CSF-infected tumor cells significantly reduced the incidence and growth rates of tumors when administered locally adjacent to the tumor site, providing up to 90% protection. The prophylactic and therapeutic efficacy of DISC-mGM-CSF-infected cells was shown initially using a murine renal cell carcinoma model (RENCA), and the results were confirmed in two additional murine tumor models: the M3 melanoma and 302R sarcoma. Therapy with DISC-infected RENCA "whole" cell vaccines failed to reduce the incidence or growth of tumor in congenitally T-cell deficient (Nu+/Nu+) mice or mice depleted of CD4+ and/or CD8+ T-lymphocytes, confirming that both T-helper and T-cytotoxic effector arms of the immune response are required to promote tumor rejection. These preclinical results suggest that this "novel" DISC-HSV vector may prove to be efficacious in developing genetically modified whole-cell vaccines for clinical use.

Apoptosis or necrosis for tumor immunotherapy: what's in a name?

Here we discuss how the mechanisms by which tumor cells are killed in vivo by gene transfer affects their immunogenicity. Our own work has shown that necrotic cell death induces immunological activation signals which recruit, load, activate and mature appropriate subsets of antigen-presenting cells. In contrast, for apoptotic cell death to be immunogenic, signals additional to cell death alone must be provided within the milieu of the dying tumor. Our conclusion is that the immunogenicity of tumor killing is determined by a combination of factors, including the mechanism of killing, the levels of cell death, the local environment that exists within the dying tumor and, as a result, the nature of the immune/scavenger cells which are present at the time of antigen release. Knowledge of how these factors can influence the immune system and lead to the breaking of tolerance to tumor-associated antigens, can potentially be exploited in the design of effective immunotherapies for cancer using gene transfer.

Disabled infectious single-cycle herpes simplex virus as an oncolytic vector for immunotherapy of colorectal cancer.

New modalities of treatment for colorectal cancer are required to support and improve those currently available. One such approach is immunotherapy by transfer of immunostimulatory genes to tumor cells. Here, we report the use of a herpes simplex virus (HSV) vector that is capable of a single round of infection (disabled infectious single-cycle [DISC]-HSV) as a gene transfer vehicle for colorectal cancer. This vector has potential advantages over other vectors for cancer immunotherapy in that it lyses infected tumor cells. Infection with DISC-HSV inhibited tumor cell growth both in vitro and in vivo. In addition, DISC-HSV-mediated cell killing occurs by both apoptotic and necrotic mechanisms. A range of colorectal tumor cell lines could be rapidly transduced with DISC-HSV/lacZ (14-90% in 4 hr). Both tumor prevention and tumor therapy protocols showed clear antitumor effects with DISC-HSV/mGM-CSF. In the prophylactic approach, an infected/irradiated whole cell vaccine protected up to 80% of mice from rechallenge. In addition, intratumoral injection of established tumors with DISC-HSV/GM-CSF caused rejection in 40% of mice and generated some protection from subsequent rechallenge. In both cases, however, it is clear that a dominant therapeutic effect of the DISC-HSV vector derives from its oncolytic properties, irrespective of the transduced gene. As a prelude to taking these studies forward to human clinical trials, we demonstrate that tumor cells could be successfully grown from freshly obtained human colorectal cancer resections (within 1 week of surgery), were transduced with DISC-HSV/hGM-CSF, and secreted the cytokine. This study provides the preclinical basis for trials of immunotherapy of colorectal cancer using DISC-HSV.

CD40 ligation for immunotherapy of solid tumours.

Tumour vaccines provide an important focus of current cancer research and are often based on the premise that although T-cells do respond naturally to certain tumours, this is usually weak and therefore ineffective at controlling disease. An integral and necessary part of a T-cell immune response involves triggering of CD40 on antigen-presenting cells (APC) by its ligand, CD154, on responding T helper (Th) cells. Furthermore, cytotoxic responses to tumours may fail because the Th-cell response is inadequate and unable to provide CD40 stimulation of APC. Growing evidence shows that stimulating APC with soluble CD40L or an agonistic anti-CD40 mAb can, at least in part, replace the need for Th cells and generate APC that are capable of priming cytotoxic T lymphocytes (CTL). The aim of this study was to investigate whether a range of solid tumours (CD40(-)) could be treated with anti-CD40 mAb. It was found that this treatment was effective, and correlated with the intrinsic immunogenicity and aggressiveness of the tumours. The mAb could be delivered locally or at a distal site, but increased antigen load provided by irradiated tumour cells added little to the effectiveness of the treatment. T-cells were required since cytokine (interferon-gamma) and CTL activity were demonstrated following treatment and the therapeutic efficacy was lost in nude mice. In addition, depletion of CD8(+) cells abrogated protection whilst depletion of CD4(+) cells had no effect. This study demonstrates that solid CD40(-) tumours are sensitive to anti-CD40 mAb therapy and that the response bypasses the need for Th cells.

Cell death associated with genetic prodrug activation therapy of colorectal cancer.

Genetic prodrug activation therapy (GPAT) is a form of cancer gene therapy that has potential use against tumours such as colorectal malignancy. The characterization of such therapies using laboratory models provides a basis for clinical trials. In this study the gene encoding Herpes Simplex Virus thymidine kinase (HSVtk) was delivered to colorectal tumour cells using an Adenoviral (Ad) vector in vitro. In this way the cells were made susceptible to killing with the prodrug ganciclovir to various degrees depending on cell infectability with Ad. Bystander killing effect appeared minimal both in vitro and when transduced cells were injected in vivo. Mechanisms of cell death, measured in vitro using anti-BrDU (DNA-break labelling) and propidium iodide staining variously showed a combination of apoptosis in the G1 cell cycle phase and late apoptotic or necrotic sub-G1 DNA fragmentation, depending on the tumour cell line. These findings suggest that gene therapy of colorectal cancer by GPAT gives rise to therapeutic forms of direct cell death, but requires improvements in transduction, and possibly immune augmentation.

Coxsackie B and adenovirus receptor, integrin and major histocompatibility complex class I expression in human prostate cancer cell lines: implications for gene therapy strategies.

Gene therapy strategies based on modifying tumour cells using high efficiency adenoviral vectors have shown promise in the clinic. Recently the Coxsackie and adenovirus receptor (CAR) has been shown to mediate adenoviral entry into tumour cells, although previous studies also suggested a role for MHC class I heavy chain. Detailed evaluation of the expression of both CAR and MHC class I in prostate cancer cell lines would have important implications for therapeutic strategies. We have found that, unlike cell lines derived from other malignancies, in human and murine prostate cancer loss of CAR expression appears to be relatively infrequent and does not correlate with loss of MHC class I expression. These findings, together with the demonstration of appreciable levels of cell-surface expression of integrins, suggest that cancer vaccine strategies based on modifying whole prostate cancer cells should be feasible using the current generation of recombinant adenoviral vectors, without deleterious effects on either the virus vector or the target cell.

A novel murine model of allogeneic vaccination against prostate cancer.

Prostate cancer continues to be a major cause of death in men. Surgical and medical treatments of the disease have improved, but metastasic disease remains a significant clinical problem. Novel therapies such as whole cell vaccination offer the potential of treating disease by stimulating the immune system. To study the efficacy of a whole cell vaccine in prostate cancer two strains of mice were used: C57BL/6 (H-2Kb) and C3H/HeJ (H-2K(k)) in combination with four different cell lines. Thus, a model was constructed of allogeneic and syngeneic vaccine, as well as a challenge tumour for each strain. Two novel cell lines were developed during this study. Firstly, the non tumourigeneic PMC-1 was derived from a normal mouse prostate and immortalized with HPV16. Secondly, the tumourigeneic PMC-1 C6ras1p1 was transformed with human ras gene which formed tumours in both SCID and C3H/HeJ mice. Protection, and the nature of the immune response to syngeneic and allogeneic vaccine, in males and females was examined in both strains. Vaccination with both syngeneic and allogeneic irradiated whole cell vaccines induced protection from syngeneic challenge in females. However, no protection was observed when allogeneic vaccine was given to male mice. This correlated with the immune response. Two types of cellular immune responses were generated in females. A NK-mediated response was observed in C57BL/6 mice, whilst C3H/HeJ mice developed a CTL response. Little or no cellular immune response was observed in males. The cytokine profile in C3H/HeJ females was a mixture of Th1 and Th2 whilst a mainly Th1 profile was observed in C57BL/6 mice. Male mice showed a diminished cytokine secretion compared to females which was further depressed after challenge. The difference in immunity was largely as expected, since tolerance to prostate antigens should not normally develop in female mice. However, this makes this model particularly relevant clinically since it directly mimics the human situation and thus may accelerate the development of whole cell vaccines for clinical use.

Safety and immunogenicity of the malaria candidate vaccines FP9 CS and MVA CS in adult Gambian men.

We assessed the safety and immunogenicity of prime-boost vectors encoding the Plasmodium falciparum circumsporozoite (CS) protein expressed either in the attenuated fowl-pox virus (FP9) or modified vaccinia virus Ankara (MVA). Thirty-two adult Gambians in groups of four to eight received one, two or three doses of FP9 CS and/or MVA CS. No serious adverse event was observed following vaccination. The most immunogenic regimen was two doses of FP9 followed by a single dose of MVA 4 weeks later (an average of 1000 IFN-gamma spot forming units/million PBMCs). This level of effector T-cell responses appears higher than that seen in previously reported studies of CS-based candidate malaria vaccines.

Effect of route of immunisation and adjuvant on T and B cell epitope recognition within a streptococcal antigen.

Oral immunisation may elicit both systemic and mucosal immunity. Antibodies directed to a portion (residues 816-1213) of a cell surface adhesin termed streptococcal antigen I/II (SA I/II) of Streptococcus mutans prevent colonisation of this bacterium in vivo. This polypeptide is highly immunogenic in mice and is immunodominant in naturally sensitised humans. In this study, the effects of immunisation by different routes and of adjuvant on T and B cell epitope recognition were investigated. The recombinant polypeptide comprising residues 816-1213 of SA I/II was administered to groups of SJL mice intraperitoneally, subcutaneously or orally. For systemic immunisation, incomplete Freund's adjuvant was used, whereas for oral immunisation the antigen was coupled to the cholera toxin B subunit. The hierarchy of T and B cell epitope recognition differed significantly following different routes of immunisation. These differences in T and B cell responses may be accounted for by extracellular protease activity and processing by antigen presenting cells at the sites of immunisation. Furthermore, epitope recognition may be critical if the immune response elicited by a vaccine must be directed specifically to functional determinants within an antigen. This study emphasises the importance of route of immunisation in vaccine development.

The role of herpes simplex virus thymidine kinase in the treatment of solid tumours.

Suicide gene therapy is the approach whereby the genetic alteration of a cell renders it susceptible to an otherwise non-toxic prodrug. Suicide gene therapy for solid tumours has progressed rapidly since the concept was originally described: nearly all tumour types have been explored, with some, such as glioma, melanoma and colon cancer frequently used experimentally. The exciting aspect of suicide gene therapy is the bystander effect, the phenomenon whereby there is extended tumour death when only a small fraction is transfected with the suicide gene. This phenomenon implies that there is a reduced need to target specifically all tumour cells, as the effect mechanism itself carries out this function. The bystander effect mode of action has not yet been fully characterised, but the role of gap junctions and the immune system are implicated as the main instruments in its potentiation. This approach is also amenable to pharmacological intervention, which may help to optimise parameters prior to commencing suicide gene therapy. Clinical trails have already commenced using this form of treatment and results are eagerly awaited.

A protein fragment of streptococcal cell surface antigen I/II which prevents adhesion of Streptococcus mutans.

Attachment of Streptococcus mutans to the tooth surface involves a cell surface protein with an M(r) of 185,000, termed streptococcal antigen (SA) I/II. Four overlapping fragments of the gene encoding SA I/II were amplified by polymerase chain reaction, cloned, and expressed in Escherichia coli. The recombinant polypeptides were assayed for adhesion-binding activity to salivary receptors and for recognition by a panel of monoclonal antibodies (MAbs) raised against SA I/II. Two of the MAbs which are known to prevent colonization of S. mutans in vivo bound the recombinant polypeptide comprising residues 816 to 1161. In vitro adhesion of S. mutans to saliva-coated hydroxyapatite beads was also inhibited specifically by a polypeptide (residues 816 to 1213) encompassing the same region. The evidence from the MAbs preventing colonization of S. mutans and the adherence inhibition assay suggests that an adhesion-binding activity resides within the portion of SA I/II comprising residues 816 to 1213, which is highly conserved among oral streptococcal species.

T-cell, adhesion, and B-cell epitopes of the cell surface Streptococcus mutans protein antigen I/II.

The T-cell and antibody responses to a cell surface streptococcal antigen (SA I/II) were investigated in naturally sensitized humans. Serum antibody responses were directed predominantly to the N-terminal (residues 39 to 481) and central (residues 816 to 1213) regions of SA I/II which may be involved in bacterial adhesion to salivary receptors. T-cell responses were also directed predominantly towards the central region. The linear peptide relationship of the immunodominant and minor T- and B-cell as well as adhesion epitopes was mapped within residues 816 to 1213. Immunodominant T-cell and B-cell epitopes were identified within residues 803 to 853, which were separated in linear sequence from the adhesion epitopes (residues 1005 to 1044). Adhesion epitopes overlapped with minor B- and T-cell epitopes (residues 1005 to 1054 and 1085 to 1134). An immunodominant promiscuous T-cell epitope (residues 985 to 1004) was adjacent to an adhesion epitope (residues 1005 to 1024). The limited B-cell response to adhesion epitopes is consistent with the success of Streptococcus mutans in colonizing the oral cavity. The strategy of T-cell, adhesion, and B-cell epitope mapping has revealed a general approach for identifying components of subunit vaccines which may focus responses to critical functional determinants. Such epitopes of SA I/II may constitute the components of a subunit vaccine against dental caries.

Can immunotherapy by gene transfer tip the balance against colorectal cancer?

Gene therapy, in particular the transfer of genes encoding immunostimulatory molecules (cytokines and costimulatory molecules) as well as selectively cytotoxic enzymes and DNA vaccination, has the potential of enhancing cell mediated immune responses against tumours including those of colorectal origin. Genes can be transferred using viral vectors either to cultured tumour cells in vitro that can be returned to the patient as a "cancer vaccine", or directly to tumour cells in vivo. Vaccination with DNA constructs expressing specific tumour antigens characteristic of colorectal neoplasia can trigger immune recognition and destruction of tumour cells. The aim is to tip the balance from protumour to antitumour mechanisms by generating a local immune response and systemic antitumour immune memory to destroy metastases. Studies in murine models, combined with human studies, show that such approaches could become an adjunct to current treatments for human colorectal cancer in the near future.

Tumour cell expression of B7 costimulatory molecules and interleukin-12 or granulocyte-macrophage colony-stimulating factor induces a local antitumour response and may generate systemic protective immunity.

Previously, we showed that expression of B7-1 in CMT93 murine colorectal tumour cells inhibited their growth in immunocompetent animals. However, this did not result in any significant increase in systemic protective immunity, relative to that elicited by the parental tumour. To potentiate the effects of B7-1 on systemic immunity. Interleukin-12 (IL-12) or granulocyte-macrophage colony-stimulating factor (GM-CSF) was co-expressed with this molecule. These combinations of immunostimulatory molecules were effective in eliciting systemic immunity. We also show that expression of B7-2 led to a local antitumour response as well as significantly raised systemic immunity. In another tumour model. K1735 minutes melanoma, which is moderately immunogenic, tumours secreting GM-CSF alone were as effective as the parental tumours in generating protective immunity. Previously, we described the deleterious effect of B7-1 expression on protective immunity. Co-expression of GM-CSF did not counteract this consequence of B7-1 expression. Expression of IL-12 was extremely effective in causing rejection of inoculated tumour cells, but evoked only minimal protective systemic immunity. These results suggest that combing costimulatory molecules and cytokines may be a useful therapeutic approach in some, but not all, tumours.

Induction of immune responses to functional determinants of a cell surface streptococcal antigen.

Antibodies directed against the cell surface adhesin, termed streptococcal antigen I/II of Streptococcus mutans can protect against dental caries. Streptococcal antigen I/II (SA I/II) interacts with salivary glycoproteins and promotes adhesion to the tooth surface. Topical application of monoclonal antibodies which recognize a domain within residues 816-1213 (fragment 3) prevents colonization by S. mutans in primates. In this study the immunogenicity and antigenicity of fragment 3 was investigated in five strains of mice. Fragment 3 induced an immune response following immunization with whole cells of S. mutans in all strains of mice. Immunization with recombinant fragment 3 also induced T-cell proliferative and antibody responses both to fragment 3 and to the SA I/II. Antibody responses to the previously defined adhesion determinants (residues 1005-1044) were weak or undetectable. Immunization of three representative strains of mice with a recombinant polypeptide (residues 975-1044) comprising this adhesion epitope and an adjacent T-cell epitope (residues 975-1004) elicited both T- and B-cell responses to the polypeptide and to native SA I/II. The B-cell epitopes overlapped with the adhesion determinant. These findings provide a means of directing immune responses to functional determinants of SA I/II.

The immunotherapy of prostate cancer.

Advanced prostate cancer remains incurable with standard treatment options. Immunotherapy may be a realistic alternative given the growing evidence that the immune response can affect the growth of other solid tumours and the regulation of both specific and shared prostate cancer antigens. Early studies suggest that both non-specific and specific vaccines can effect relevant animal models and clinical trials based on these observations are now in progress. A number of other approaches including gene therapy with HSVtk are already undergoing clinical studies (Herman et al. Hum Gene Ther 1999; 10: 1239-1249). Prostate Cancer and Prostatic Diseases (2000) 3, 303-307