High-sensitivity array analysis of gene expression for the early detection of disseminated breast tumor cells in peripheral blood

Early detection is an effective means of reducing cancer mortality. Here, we describe a highly sensitive high-throughput screen that can identify panels of markers for the early detection of solid tumor cells disseminated in peripheral blood. The method is a two-step combination of differential display and high-sensitivity cDNA arrays. In a primary screen, differential display identified 170 candidate marker genes differentially expressed between breast tumor cells and normal breast epithelial cells. In a secondary screen, high-sensitivity arrays assessed expression levels of these genes in 48 blood samples, 22 from healthy volunteers and 26 from breast cancer patients. Cluster analysis identified a group of 12 genes that were elevated in the blood of cancer patients. Permutation analysis of individual genes defined five core genes (P ≤ 0.05, permax test). As a group, the 12 genes generally distinguished accurately between healthy volunteers and patients with breast cancer. Mean expression levels of the 12 genes were elevated in 77% (10 of 13) untreated invasive cancer patients, whereas cluster analysis correctly classified volunteers and patients (P = 0.0022, Fisher's exact test). Quantitative real-time PCR confirmed array results and indicated that the sensitivity of the assay (1:2 × 108 transcripts) was sufficient to detect disseminated solid tumor cells in blood. Expression-based blood assays developed with the screening approach described here have the potential to detect and classify solid tumor cells originating from virtually any primary site in the body.

Subcutaneous vaccination with irradiated, cytokine-producing tumor cells stimulates CD8+ cell-mediated immunity against tumors located in the "immunologically privileged" central nervous system.

Vaccination with cytokine-producing tumor cells generates potent immune responses against tumors outside the central nervous system (CNS). The CNS, however, is a barrier to allograft and xenograft rejection, and established tumors within the CNS have failed to respond to other forms of systemic immunotherapy. To determine what barriers the "immunologically privileged" CNS would pose to cytokine-assisted tumor vaccines and what cytokines would be most efficacious against tumors within the CNS, we irradiated B16 murine melanoma cells producing murine interleukin 2 (IL-2), IL-3, IL-4, IL-6, gamma-interferon, or granulocyte-macrophage colony stimulating factor (GM-CSF) and used these cells as subcutaneous vaccines against tumors within the brain. Under conditions where untransfected B16 cells had no effect, cells producing IL-3, IL-6, or GM-CSF increased the survival of mice challenged with viable B16 cells in the brain. Vaccination with B16 cells producing IL-4 or gamma-interferon had no effect, and vaccination with B16 cells producing IL-2 decreased survival time. GM-CSF-producing vaccines were also able to increase survival in mice with pre-established tumors. The response elicited by GM-CSF-producing vaccines was found to be specific to tumor type and to be abrogated by depletion of CD8+ cells. Unlike the immunity generated against subcutaneous tumors by GM-CSF, however, the effector responses generated against tumors in the CNS were not dependent on CD4+ cells. These data suggest that cytokine-producing tumor cells are very potent stimulators of immunity against tumors within the CNS, but effector responses in the CNS may be different from those obtained against subcutaneous tumors.

Transloading of tumor cells with foreign major histocompatibility complex class I peptide ligand: a novel general strategy for the generation of potent cancer vaccines.

The major hurdle to be cleared in active immunotherapy of cancer is the poor immunogenicity of cancer cells. In previous attempts to overcome this problem, whole tumor cells have been used as vaccines, either admixed with adjuvant(s) or genetically engineered to express nonself proteins or immunomodulatory factors before application. We have developed a novel approach to generate an immunogeneic, highly effective vaccine: major histocompatibility complex (MHC) class I-positive cancer cells are administered together with MHC class I-matched peptide ligands of foreign, nonself origin, generated by a procedure we term transloading. Murine tumor lines of the H2-Kd or the H2-Db haplotype, melanoma M-3 and B16-F10, respectively, as well as colon carcinoma CT-26 (H2-Kd), were transloaded with MHC-matched influenza virus-derived peptides and applied as irradiated vaccines. Mice bearing a deposit of live M-3 melanoma cells were efficiently cured by this treatment. In the CT-26 colon carcinoma and the B16-F10 melanoma, high efficacies were obtained against tumor challenge, suggesting the universal applicability of this new type of vaccine. With foreign peptide ligands adapted to the requirements of a desired MHC class I haplotype, this concept may be used for the treatment of human cancers.

Expression of hyaluronidase by tumor cells induces angiogenesis in vivo.

Hyaluronic acid is a proteoglycan present in the extracellular matrix and is important for the maintenance of tissue architecture. Depolymerization of hyaluronic acid may facilitate tumor invasion. In addition, oligosaccharides of hyaluronic acid have been reported to induce angiogenesis. We report here that a hyaluronidase similar to the one on human sperm is expressed by metastatic human melanoma, colon carcinoma, and glioblastoma cell lines and by tumor biopsies from patients with colorectal carcinomas, but not by tissues from normal colon. Moreover, angiogenesis is induced by hyaluronidase+ tumor cells but not hyaluronidase- tumor cells and can be blocked by an inhibitor of hyaluronidase. Tumor cells thus use hyaluronidase as one of the "molecular saboteurs" to depolymerize hyaluronic acid to facilitate invasion. As a consequence, breakdown products of hyaluronic acid can further promote tumor establishment by inducing angiogenesis. Hyaluronidase on tumor cells may provide a target for anti-neoplastic drugs.

The activation of human gene MAGE-1 in tumor cells is correlated with genome-wide demethylation.

Human gene MAGE-1 encodes tumor-specific antigens that are recognized on melanoma cells by autologous cytolytic T lymphocytes. This gene is expressed in a significant proportion of tumors of various histological types, but not in normal tissues except male germ-line cells. We reported previously that reporter genes driven by the MAGE-1 promoter are active not only in the tumor cell lines that express MAGE-1 but also in those that do not. This suggests that the critical factor causing the activation of MAGE-1 in certain tumors is not the presence of the appropriate transcription factors. The two major MAGE-1 promoter elements have an Ets binding site, which contains a CpG dinucleotide. We report here that these CpG are demethylated in the tumor cell lines that express MAGE-1, and are methylated in those that do not express the gene. Methylation of these CpG inhibits the binding of transcription factors, as seen by mobility shift assay. Treatment with the demethylating agent 5-aza-2'-deoxycytidine activated gene MAGE-1 not only in tumor cell lines but also in primary fibroblasts. Finally, the overall level of CpG methylation was evaluated in 20 different tumor cell lines. It was inversely correlated with the expression of MAGE-1. We conclude that the activation of MAGE-1 in cancer cells is due to the demethylation of the promoter. This appears to be a consequence of a genome-wide demethylation process that occurs in many cancers and is correlated with tumor progression.

Cell cycle-dependent modulation of telomerase activity in tumor cells.

Telomerase is a ribonucleoprotein complex that is thought to add telomeric repeats onto the ends of chromosomes during the replicative phase of the cell cycle. We tested this hypothesis by arresting human tumor cell lines at different stages of the cell cycle. Induction of quiescence by serum deprivation did not affect telomerase activity. Cells arrested at the G1/S phase of the cell cycle showed similar levels of telomerase to asynchronous cultures; progression through the S phase was associated with increased telomerase activity. The highest level of telomerase activity was detected in S-phase cells. In contrast, cells arrested at G2/M phase of the cell cycle were almost devoid of telomerase activity. Diverse cell cycle blockers, including transforming growth factor beta1 and cytotoxic agents, also caused inhibition of telomerase activity. These results establish a direct link between telomerase activity and progression through the cell cycle.

Conjugates of folate and anti-T-cell-receptor antibodies specifically target folate-receptor-positive tumor cells for lysis.

High-affinity folate receptors (FRs) are expressed at elevated levels on many human tumors. Bispecific antibodies that bind the FR and the T-cell receptor (TCR) mediate lysis of these tumor cells by cytotoxic T lymphocytes. In this report, conjugates that consist of folate covalently linked to anti-TCR antibodies are shown to be potent in mediating lysis of tumor cells that express either the alpha or beta isoform of the FR. Intact antibodies with an average of five folate per molecule exhibited high affinity for FR+ tumor cells but did not bind to FR- tumor cells. Lysis of FR+ cell lines could be detected at concentrations as low as 1 pM (approximately 0.1 ng/ml), which was 1/1000th the concentration required to detect binding to the FR+ cells. Various FR+ mouse tumor cell lines could be targeted with each of three different anti-TCR antibodies that were tested as conjugates. The antibodies included 1B2, a clonotypic antibody specific for the cytotoxic T cell clone 2C; KJ16, an anti-V beta 8 antibody; and 2C11, an anti-CD3 antibody. These antibodies differ in affinities by up to 100-fold, yet the cytolytic capabilities of the folate/antibody conjugates differed by no more than 10-fold. The reduced size (in comparison with bispecific antibodies) and high affinity of folate conjugates suggest that they may be useful as immunotherapeutic agents in targeting tumors that express folate receptors.

Construction, purification, and functional incorporation on tumor cells of glycolipid-anchored human B7-1 (CD80).

To generate a potent cell-mediated immune response, at least two signals are required by T cells. One is engagement of the T-cell receptor with peptide-bearing major histocompatibility complex molecules. The other signal can be delivered by various molecules on the antigen-presenting cell, such as B7-1 (CD80). Many tumor cells escape immune recognition by failing to express these costimulatory molecules. Transfection of the B7 gene into some murine tumor cells allows for immune recognition and subsequent rejection of the parental tumor. We have studied an alternative approach for the introduction of B7-1 onto the surface of tumor cells. This method involves purified glycosyl-phosphatidylinositol (GPI)-anchored proteins which can spontaneously incorporate their lipid tail into cell membranes. We have created and purified a GPI-anchored B7-1 molecule (called GPI-B7) which is able to bind its cognate ligand, CD28, and incorporate itself into tumor cell membranes after a short incubation. Tumor cells that have been reconstituted with GPI-B7 can provide the costimulatory signal needed to stimulate T cells. These findings suggest an approach for the introduction of new proteins onto cell membranes to create an effective tumor vaccine for potential use in human immunotherapy.

The bacterial colicin active against tumor cells in vitro and in vivo is verotoxin 1.

We have identified verotoxin 1 (VT1) as the active component within an antineoplastic bacteriocin preparation from Escherichia coli HSC10 studied over two decades. Recombinant VT1 can simulate the toxicity of anticancer proteins (ACP), and the antineoplastic activity of ACP (and VT1) was abrogated by treatment with anti-VT1 antibody. Similarly, VT1 mimics the protective effect of ACP in a murine metastatic fibrosarcoma model. Prior immunization with VT1 B subunit prevents the effect of VT1 or ACP in this model. The activity of ACP against a variety of human ovarian cell lines was mimicked by VT1, and multidrug-resistant variants were significantly hypersensitive. Primary ovarian tumors and metastases contain elevated levels of globotriaosylceramide compared with normal ovaries, and overlay of frozen tumor sections showed selective VT binding to tumor tissue and the lumen of invading blood vessels. Our contention that VT1 could provide an additional approach to the management of certain human neoplasms is discussed.

Anti-melanoma antibodies from melanoma patients immunized with genetically modified autologous tumor cells: selection of specific antibodies from single-chain Fv fusion phage libraries.

Fusion phage libraries expressing single-chain Fv antibodies were constructed from the peripheral blood lymphocytes of two melanoma patients who had been immunized with autologous melanoma cells transduced the gamma-interferon gene to enhance immunogenicity, in a trial conducted at another institution. Anti-melanoma antibodies were selected from each library by panning the phage against live cultures of the autologous tumor. After two or three rounds of panning, clones of the phage were tested by ELISA for binding to the autologous tumor cells; > 90% of the clones tested showed a strong ELISA reaction, demonstrating the effectiveness of the panning procedure for selecting antimelanoma antibodies. The panned phage population was extensively absorbed against normal melanocytes to enrich for antibodies that react with melanoma cells but not with melanocytes. The unabsorbed phage were cloned, and the specificities of the expressed antibodies were individually tested by ELISA with a panel of cultured human cells. The first tests were done with normal endothelial and fibroblast cells to identify antibodies that do not react, or react weakly, with two normal cell types, indicating some degree of specificity for melanoma cells. The proportion of phage clones expressing such antibodies was approximately 1%. Those phage were further tested by ELISA with melanocytes, several melanoma lines, and eight other tumor lines, including a glioma line derived from glial cells that share a common lineage with melanocytes. The ELISA tests identified three classes of anti-melanoma antibodies, as follows: (i) a melanoma-specific class that reacts almost exclusively with the melanoma lines; (ii) a tumor-specific class that reacts with melanoma and other tumor lines but does not react with the normal melanocyte, endothelial and fibroblast cells; and (iii) a lineage-specific class that reacts with the melanoma lines, melanocytes, and the glioma line but does not react with the other lines. These are rare classes from the immunized patients' repertoires of anti-melanoma antibodies, most of which are relatively nonspecific anti-self antibodies. The melanoma-specific class was isolated from one patient, and the lineage-specific class was isolated from the other patient, indicating that different patients can have markedly different responses to the same immunization protocol. The procedures described here can be used to screen the antibody repertoire of any person with cancer, providing access to an enormous untapped pool of human monoclonal anti-tumor antibodies with clinical and research potential.

Priming of tumor-specific T cells in the draining lymph nodes after immunization with interleukin 2-secreting tumor cells: three consecutive stages may be required for successful tumor vaccination.

Although both CD4+ and CD8+ T cells are clearly required to generate long-lasting anti-tumor immunity induced by s.c. vaccination with interleukin 2 (IL-2)-transfected, irradiated M-3 clone murine melanoma cells, some controversy continues about the site and mode of T-cell activation in this system. Macrophages, granulocytes, and natural killer cells infiltrate the vaccination site early after injection into either syngeneic euthymic DBA/2 mice or athymic nude mice and eliminate the inoculum within 48 hr. We could not find T cells at the vaccination site, which argues against the concept that T-cell priming by the IL-2-secreting cancer cells occurs directly at that location. However, reverse transcription-PCR revealed transcripts indicative of T-cell activation and expansion in the draining lymph nodes of mice immunized with the IL-2-secreting vaccine but not in mice vaccinated with untransfected, irradiated M-3 cells. We therefore propose that the antigen-presenting cells, which invade the vaccination site, process tumor-derived antigens and, subsequently, initiate priming of tumor-specific T lymphocytes in lymphoid organs. These findings suggest a three-stage process for the generation of effector T cells after vaccination with IL-2-secreting tumor cells: (i) tumor-antigen uptake and processing at the site of injection by antigen-presenting cells, (ii) migration of antigen-presenting cells into the regional draining lymph nodes, where T-cell priming occurs, and (iii) circulation of activated T cells that either perform or initiate effector mechanisms leading to tumor cell destruction.

Species-specific metastasis of human tumor cells in the severe combined immunodeficiency mouse engrafted with human tissue.

We have attempted to model human metastatic disease by implanting human target organs into the immunodeficient C.B-17 scid/scid (severe combined immunodeficiency; SCID) mouse, creating SCID-hu mice. Preferential metastasis to implants of human fetal lung and human fetal bone marrow occurred after i.v. injection of human small cell lung cancer (SCLC) cells into SCID-hu mice; the homologous mouse organs were spared. Clinically more aggressive variant SCLC cells metastasized more efficiently to human fetal lung implants than did cells from classic SCLC. Metastasis of variant SCLC to human fetal bone marrow was enhanced in SCID-hu mice exposed to gamma-irradiation or to interleukin 1 alpha. These data indicate that the SCID-hu mice may provide a model in which to study species- and tissue-specific steps of the human metastatic process.