Immune response induced by retrovirus-mediated HSV-tk/GCV pharmacogene therapy in patients with glioblastoma multiforme

Immune response induced by retrovirus-mediated HSV-tk/GCV pharmacogene therapy in patients with glioblastoma multiforme

Horizontal gene transfer from retroviruses to mammals is well documented and extensive, but is rare between unrelated viruses with distinct genome types. Human endogenous retroviruses (HERVs) may be involved in the development of autoimmune disease. Chimeric HSV/AAV products that can mediate transgene integration in human mitotic cells have been constructed, but, to date, genetic modification of dividing cells in animal models using HSV products has not been possible. SAMHD1 was recently identified as a retrovirus restriction factor highly expressed in macrophages. Tumor specimens were obtained at primary or recurrent surgery and at autopsy. These results reveal that a diverse range of ancient sag-containing retroviruses independently donated sag twice from two separate lineages that are distinct from MMTV. Numbers of tumor-infiltrating lymphocytes found weeks after gene therapy were not significantly increased compared with primary tumors.

Mitotic tumor cells were sparse close to the VPC injection sites, but abundant in brain areas somewhat distant from these sites. Depleting SAMHD1 in THP-1 cells enhanced HSV-1 replication, while ectopic overexpression of SAMHD1 in U937 cells repressed HSV-1 replication. Immunophenotyping of PMNC did not show a significant activation of T cells or NK cells during gene therapy. SaHV2 independently acquired these genes after diverging from AtHV3. It demonstrated an antitumor immune response in the gene therapy group, but not in the control group. These findings support the concept of in vivo induction of a systemic immune response by local intracerebral HSV-tk/GCV pharmacogene therapy for primary human GBM. Gene Therapy (2000) 7, 1853-1858.

To date, the tumoricidal effect of RV-mediated HSV-tk gene transfer has been thought to be mediated exclusively by GCV-induced cytotoxicity to transduced tumor cells6,8,9,12 and also to their directly attached nontransduced neighbor cells, a phenomenon which is known as the bystander effect.3,13,14 In addition, it has been shown in animal studies that pharmacological gene therapy with HSV-tk/GCV is also able to induce an effective antitumor immune response involving tumor infiltrating T cells and natural killer (NK) cells.15,16,17,18,19 This immune response is supposed to enhance the therapeutic effects of pharmacological gene therapy beyond local tumor cell killing, thereby contributing towards the eradication of distant tumor cell foci. The histogram shows the number of tBLASTn hits per mammalian genome to any of the four viral sags. All patients signed an informed consent and agreed to participate in the study. The study protocol was approved by the local ethical committee. Patients were randomly assigned to the gene therapy or control arms (Table 1). Tumor extent and localization were assessed by contrast-enhanced cranial MRI. Clade-A shows the rodent ERVs, and RHVP, while clade-B is the predominantly primate clade that also contains AtHV3 and AtHV2.

Glioblastoma was diagnosed at surgery by frozen sections and confirmed later through light-microscopy in paraffin sections. Specimens obtained from three gene therapy patients with recurrent tumors were investigated in the same way. Analysis of primary tumors demonstrated minor to moderate intratumoral T lymphocytic infiltrates with no significant B cell accumulation, and perivascular populations of activated microglial cells (Table 2). Specimens from gene therapy patients obtained at recurrence surgery or autopsy showed varying degrees of tumor necrosis and reactive astrocytic proliferation in and around the former tumor resection cavity, and similar density of tumor-infiltrating lymphocytes compared with the primary tumors (Figure 1). The subsequent lines show the codons for the same region using only clade-A or clade-B sequences, including a larger taxon set compared with the representative subset used for the tree. Interestingly, the MIB1 immunostaining indicative for mitotic activity was lower in the specimens obtained from tumors after gene therapy than at primary tumor surgery (Figure 1). Total counts of PMNC in all patients were markedly decreased before surgery and adjuvant therapy.

Numbers of T cells (CD3+) and CD4+ or CD8+ T cell subsets (helper and cytotoxic lymphocytes, respectively) did not display a significant increase during GCV therapy (days 14-27), and did not differ significantly in either group of patients (Figure 2a-c). Also, absolute numbers of MHC-class I nonrestricted NK cells (CD56+) were low in both groups and did not show significant differences over a 4 week course (Figure 2d). Albeit with lower support for deep nodes, the sag-only tree recovers the distinction between predominantly rodent- and primate-infecting viruses. Similar findings were described with activated NK cells (CD56+/CD69+) in both groups (Figure 2e and f). Serum ELISA measured significantly elevated levels of sFasL in the gene therapy group, especially during GCV therapy (Figure 3a). In the control group, serum sFasL levels were below the detection limit of the assay (5 pg/ml). IL-12 was found in the serum of both gene therapy and control group patients, but there were statistically significant differences in the concentration of this cytokine in both groups (Figure 3b).
Immune response induced by retrovirus-mediated HSV-tk/GCV pharmacogene therapy in patients with glioblastoma multiforme

3 and sag-only tree in the upper panel. T cell reactivity against autologous tumor cells was detected in gene therapy patients on day 7 after VPC injection. GCV administration (days 14-27) was followed by a significantly higher antitumor response on day 28 (Figure 4). The reactivity of control patients’ PMNCs against VPC and autologous tumor was significantly lower compared with those values in gene therapy patients (Figure 4). The synergy of HSV-tk/GCV and other pharmacogene therapy systems with radiation therapy has been established in cell culture and in animal experiments.27,28 However, there are little data in the literature regarding humans. Both trees are shown on the same axis and have been pruned down to the taxa relevant to this study. Stereotactically injected VPC in previously irradiated patients with recurrent GBM caused a strong systemic cellular immune response with major intratumoral infiltrates of immunocompetent cells.18 In tumor specimens taken a week after VPC injection, strong macrophage infiltration and perivascular lymphocytic accumulation were observed.

Destruction of tumor neocapillaries by HSV-tk/GCV therapy caused further intratumoral areas of petechial hemorrhages and ischemic necrosis.13 Kramm et al29 studied the systemic immune response during pharmacogene therapy in a 14-year-old patient with recurrent anaplastic ependymoma. During HSV-tk/GCV pharmacogene therapy, no radiation therapy was applied. The authors demonstrated induction of a systemic immune response involving T cell activation and immunostimulatory cytokine release, which were similar to our findings, but more pronounced. Yellow bars indicate the upper and lower bound age estimates. Therefore, we hypothesized that immunosuppressive medication and adjuvant radiation during locally tumor-destructive gene therapy might inhibit the systemic immune response mounted against tumor antigens. Studies comparing immunological findings in parallel groups of tumor patients receiving the same kind of gene therapy, but differing in the type of medication and the timing of adjuvant treatments should provide a better answer to this question. In conclusion, our findings, as well as previous reports15,16,17,18 on an immunological component of HSV-tk/GCV gene therapy may reveal an additional effect of pharmacogene therapy: destruction of tumor cells by bioactivated drugs provides tumor antigen release without suppression of antitumor immune response, as with conventional chemotherapy or radiation.

Apoptotic tumor cell death with concomitant activation of the immune system may represent an effective antitumor vaccination strategy in vivo. This date (~10 mya) is also similar to the integration date estimate of a closely related sag in the C. This study was supported in part by grants 015VE1997 and 2794A/0087H from the Federal State of Saxony-Anhalt (NGR), grant 0311661/1402 (BioRegio) from the Federal Ministry of Education and Research of Germany (NGR, CMK), grant KR1711/1-1 from the German Research Council, DFG (CMK), grant SFB 503-C6 from the German Research Council, DFG (UB), grants from the Elterninitiative Kinderkrebsklinik Duesseldorf eV (CMK, UB, DK), and by Novartis Pharma GmbH, Nuremberg, Germany. Figure 1 Micrographs demonstrating findings in the primary, untreated (left column) and the treated recurrent tumor (right column, 24 weeks after primary surgery) of a gene therapy patient. (A) hematoxylin and eosin (H&E) staining of the primary GBM: densely cell-rich and well neovascularized tumor; (B) MIB1 immunostaining positive in approximately 10% of the tumor cells (arrows); (C) moderately dense activated microglia (HLA-DR-positive, arrows); (D) this is the invasive front of the recurrent tumor (arrows), H&E stain; (E) MIB1 immunostaining. Note the low density of positive (proliferating) tumor cells ( 0.05). Figure 3 Cytokine ELISA in serum of gene therapy patients and controls.

(a) ELISA for soluble Fas ligand. There is a significant difference between mean values in both groups (P < 0.01, nonpaired Student's t test, Wilcoxon's two-sample test). (b) ELISA for IL-12. There is a significant difference between mean values in both groups (P < 0.01, nonpaired Student's t test, Wilcoxon's two-sample test). Data points are expressed as mean and standard deviation. Serum ELISA for IL-12 and sFasL was performed by using commercial kits (Diaclone, Besancon, France, for IL-12; Coulter Immunotech, Krefeld, Germany, for sFasL) according to the manufacturer's instructions. Figure 4 ELISPOT assay for interferon gamma (IFg) with peripheral mononuclear cells (PMNC) isolated on days 7 and 28 after surgery from gene therapy and control patients. NIH3T3, murine fibroblasts; VPC, virus producer cells; TM, autologous tumor cells; GT, gene therapy group; C, control group. The difference (day 7) between VPC-GT and VPC-C is statistically significant (P < 0.05, Wilcoxon's two-sample test). The differences (day 28) between VPC-GT and VPC-C, between TM-GT and TM-C, and between TM-GT-day 7 and TM-GT-day 28 are statistically significant (P < 0.05, Wilcoxon's two-sample test). Data points are expressed as mean and standard deviation. On days 7 and 28, PMNC from two gene therapy patients and two controls were isolated according to a standard protocol33 and frozen at -80°C. Cell cultures from primary tumors were set up by mechanical suspension of tissue samples in high glucose DMEM with 20% FCS (Sigma, Deisenhofen, Germany) on 10-cm Petri dishes. Every 2-3 weeks, the primary culture was trypsinized and passaged until a sufficient number of tumor cells was obtained. GFAP immunostaining confirmed the presence of GBM cells. Microplate assays of IFg secretion from PMNC34,35 were performed with the ELISPOT kit according to the manufacturer's instructions (AID Diagnostica, Heidelberg, Germany). For these assays, 4 ´ 105 PMNC per well were cocultivated twice in triplicate for 48 h with 1 ´ 105 autologous tumor cells, with 1 ´ 105 VPC, or with 1 ´ 105 NIH3T3 cells (ATCC, Rockville, MD, USA). PMNC and tumor cells from two gene therapy and two control patients were used. Evaluation of IFg secretion (brown spots) was carried out by a computerized microscopic image analysis system (NIH Image Software, Scion Corp., Frederick, MD, USA).

You may also like