|
|
Blood, Vol. 113, Issue 19, 4525-4533, May 7, 2009
Antineoplastic activity of lentiviral vectors expressing interferon- in a preclinical model of primary effusion lymphoma
Blood Calabrò et al.
113: 4525
Supplemental materials for: Calabro et al
Proliferation assay and apoptosis analysis. CRO-AP/3 and MBL-2 cells were exposed to increasing doses (100 and 1000 IU/ml) of native or recombinant, human or murine IFN-α, and the antiproliferative activity of IFN-αs was evaluated by 3H-thymidine incorporation analyzed for 4 days. Proliferation was also analyzed in LV-transduced CRO-AP/3 cells and in cells recovered from malignant effusions or from peritoneal washings of treated mice. Briefly, cells were suspended at 0.5–2 × 105/ml, and 24 hours before they had to be harvested, 100 µl of each cell suspension was seeded in triplicate into a 96-well flat-bottomed plate in the presence of 3H-thymidine (1 µCi/well); incorporated 3H-thymidine was quantified by liquid scintillation counting on a β-counter (TopCount, Perkin-Elmer Life Sciences, Cambridge, UK). Apoptosis was analyzed by flow cytometry after staining with annexin V/propidium iodide (PI) using the Annexin-V-FLUOS staining kit (Roche Diagnostic, Indianapolis, IN, USA) according to manufacturer’s instructions. Antibodies and flow cytometry analyses. Cells recovered from ascites, peritoneal washings and tumors were analyzed for the expression of CD138/syndecan-1 using a phycoerythrin (PE)-conjugated mouse monoclonal antibody (mAb) to human CD138 (Caltag Laboratories, Burlingame, CA, USA) using an EPICS XL flow cytometer (Coulter Electronics, Hialeah, FL, USA). To determine the in vivo gene transfer efficiency, PE-conjugated anti-CD138 mAb was used to gate PEL cells recovered from malignant effusions to determine the percentage of EGFP-co-expressing cells. EGFP-LV–transduced cells were pelletted, washed twice with phosphate-buffered saline (PBS, Oxoid, Basingstoke, UK), incubated with anti-CD138 mAb diluted in PBS/2% FCS for 40 min, washed twice with PBS, and then fixed with PBS-1% formaldehyde. Parental CRO-AP/3 cells were used as negative control for EGFP and as positive control for CD138 expression analyses. The expression of the IFN-stimulated Ly-6C marker (Caltag Laboratories) on murine peripheral blood cells was evaluated by labelling with a rat anti-mouse Ly-6C mAb as previously reported.1 Murine natural killer (NK) cells and macrophages were detected and quantified in splenocytes and ascitic fluids from control and treated mice after labelling with PE-conjugated rat anti-mouse pan-NK cells (clone DX5, eBioscience, San Diego, CA, USA) and rat anti-mouse F4/80 (Caltag Laboratories) mAbs, respectively. Isotype-matched irrelevant mAbs were always used in parallel as negative controls. Analysis of the modulation of HHV8 lytic replication program by IFN-αs. To induce HHV8 lytic replication, CRO-AP/3 cells were treated with 1 mM n-butyrate and exposed to 500 IU/ml of human or murine IFN-α 24 hours after lytic cycle induction. Twenty-four and 48 hours after IFN treatment, CRO-AP/3 cells were analyzed for the presence of an HHV8 transcript characteristic of the early lytic phase by RT-PCR using primers specific for ORF50,2 encoding the protein RTA (“replication and transcription activator”) responsible for the switch between the latent and the lytic phase. Moreover, the presence and degree of induction of the viral processivity factor 8 (PF8), characteristic of the delayed-early phase of the lytic cycle and encoded by ORF59, was evaluated 24 hours after IFN treatment using an indirect immunofluorescence assay (IFA), performed as reported previously.2 Stained cells were viewed and counted on a fluorescence microscope (Axiovert 200, Carl Zeiss, Jena, Germany). The positivity rates were determined by counting the positively and negatively stained cells in randomly, blindly selected fields for a total of at least 1,000 cells. Cells were also visualized and images were captured by confocal microscopy using an LSM510 microscope (Carl Zeiss). IFN-α-LV–transduced CRO-AP/3 cells were also analysed for vIL-6 expression using a rabbit polyclonal antibody to vIL-6 (ABI, Advanced Biotechnologies Inc., Columbia, MA) as described previously.2 Molecular analyses. RT-PCR analyses were conducted under previously reported conditions.2 Activation of IFN-inducible pathways was analyzed by RT-PCR at 48 and 72 hours post-treatment by assaying 2′,5′-oligoadenylate synthetase (2′,5′-OAS) expression in CRO-AP/3 cells and murine interferon regulatory factor-7 (mIRF-7) expression in MBL-2 cells. The 2′,5′-OAS–specific transcript was amplified using OAS.S3 (5′-ACTGAGTTCGCTCCAGCTC-3′) and OAS.AS1 (5′-GCAGAGTTGCTGGTAGTTTAT-3′), and the mIRF-7–specific transcript using mIRF-7for2 (5′-ATGGCTGAAGTGAGGGGG-3′) and mIRF-7rev2 (5′-GCAGAACC TGAAGCAAGAGG-3′). After an initial 8-min DNA denaturation and Taq polymerase activation, 30 cycles were run at 94°C for 1 min, 58°C (for 2′,5′-OAS–specific primers) or 60°C (for mIRF-7–specific primers) for 1 min, and 72°C for 1 min; these were followed by a 7-min extension step at 72°C. The murine TRAIL-specific transcript was detected using mTRAILfor (5′-TCACCAACGAGATGAAGCAG-3′) and mTRAILrev (5′-GCCTAAGGTCTTTCCATCC-3′); 35 cycles were run as reported above, using 56°C as annealing temperature. Assessment of human or murine β-actin expression was done in parallel or in multiplex as control in all RT-PCR experiments. Each PCR sample (20 µl) was analyzed by electrophoresis on a 1.6% agarose gel, and the bands were visualized by ethidium bromide staining. Quantitative analysis of systemic distribution of EGFP-LV was performed by real-time PCR, as previously reported,1 using EGFP-specific primers and probe on DNA extracted from peripheral blood cells 10 days after PEL cell injection of EGFP-LV–treated mice, and from different body compartments (ascitic cells, peritoneum, lung, spleen, liver and kidney) obtained when the EGFP-LV–treated animals were culled. Quantitative analyses of interleukin-6 (IL-6), IL-10 and vascular endothelial growth factor (VEGF). ELISA assays were used to measure the amounts of IL-6, IL-10 (Human IL-6 or IL-10 ELISA kits, Diaclone, Besançon, France) and VEGF (Human VEGF ELISA kit, BioSource) in culture supernatants from LV-transduced CRO-AP/3 cells and in cell-free fractions of malignant effusions of treated and untreated mice. Statistical analyses. Curves reporting the percentage of survivors over time were estimated by the Kaplan-Meier method and compared with the log-rank test. Data analysis was performed using the MedCalc statistical software (version 9.3.2.0, Mariakerke, Belgium). Student’s t test was used to determine the statistical significance of the difference between distinct groups of in vitro or in vivo experiments. Significant differences in ascites incidence at one time point were determined by the Fisher’s exact test. All tests were two-sided. P-values < 0.05 were considered significant. Results Assessment of in vitro gene transfer efficiency and expression persistence. We assessed the in vitro gene transfer efficiency mediated by the IFN-α–expressing lentiviral vectors (hIFN-α2b-LV or mIFN-α1-LV) in CRO-AP/3 cells. This cell line was transduced with these LV in parallel with a control EGFP-LV at a final multiplicity of infection (m.o.i.) of 2, 10 and 25; cells were assayed periodically for IFN-α release in culture supernatants by ELISA and for the percentage of EGFP+ cells by flow cytometry. The transduction efficiency of CRO-AP/3 cells was found to be m.o.i.-correlated. Indeed, 5–40% (m.o.i. 2–25) of the EGFP-LV–transduced cells were found to be EGFP+ 3 days after transduction, and EGFP expression was stable over time (Figure S1). hIFN-α2b-LV–transduced CRO-AP/3 cells were found to produce 8–12 ng of hIFN-α2b per 106 cells per ml using an m.o.i. of 2, 35–50 ng using an m.o.i. of 10, and 95–110 ng using an m.o.i. of 25, and these ranges were maintained 30 days after transduction. mIFN-α1-LV–transduced cells were found to stably produce similar, m.o.i.-dependent levels of murine cytokine (10–120 ng mIFN-α1/106 cells/ml). A decreased viability and proliferative activity accompanied by a lowered transgene expression was observed soon after transduction with all LVs in CRO-AP/3 cells (Figure S1). This crisis, likely due to the cytotoxic activity of VSV-G glycoprotein used to pseudotype the lentiviral particles, was generally overcome after 5–7 days, except for hIFN-α2b-LV–transduced cells, that showed a m.o.i.-dependent lowered proliferative capability (Figure 1). On the whole, these data indicate that, in our system, IFN-α– and EGFP-expressing LVs efficiently delivered the transgene to CRO-AP/3 cells and conferred long-term transgene expression. In vivo gene transfer efficiency, expression stability and lentiviral vector biodistribution. The in vivo efficiency of gene transfer in tumor cells was evaluated in flow cytometry by measuring the percentage of CD138+ cells co-expressing EGFP in EGFP-LV–treated mice at the time they were culled. These analyses revealed that 3–16% of the PEL cells were EGFP+ 4–17 days after the last administration of EGFP-LV (data not shown), suggesting efficient transgene delivery to PEL cells in vivo. These results are in line with the range of positivity obtained in the in vitro transductions using an m.o.i. of 2 (Figure S1), which corresponds to that theoretically used in vivo. As murine IFN-α/β were shown to enhance the expression of surface antigens encoded by the Ly-6 locus on murine macrophages and lymphocytes,3,4 we monitored the expression of this marker on murine peripheral blood cells. Ten days after PEL cell injection, Ly-6C was found to be significantly up-regulated in peripheral blood cells from mIFN-α1-LV–treated mice (Student t test, mIFN-α1-LV vs EGFP-LV or medium, P < 0.0001, Figure S2) with respect to control mice. Ly-6C expression remained up-regulated 5 weeks and 8 weeks later, indicating durable in vivo transgene expression. EGFP-LV–treated animals showed Ly-6C up-regulation in 7.4 ± 1.5 % of peripheral blood cells, suggesting that the dose used of lentiviral vector may per se activate an intrinsic, vector-mediated, anti-viral response, leading to a limited induction of endogenous mIFN-α1. Although interferon activity is species restricted, some subtypes or recombinant forms of human IFN-α were shown to be as active on murine cells as on human cells.4,5 Indeed, Ly-6C was found to be significantly up-regulated in 20.41 ± 3.97% of peripheral blood cells of hIFN-α2b-LV–treated mice (Student t test, hIFN-α2b-LV vs EGFP-LV or medium, P < 0.0001) 12 days after PEL cell injection (not shown). The increase in the percentage of Ly-6C–bearing cells was not maintained at the subsequent time points, as the one measured in mIFN-α1–treated animals. Since in vitro and in vivo quantifications showed that the two cytokines (human and murine) were produced in equivalent ranges, this transient effect may be simply linked to a specific activity of human IFN-α in Ly-6C induction lower than that of the mouse cytokine, previously reported for the human recombinant form A/D.4 It is also conceivable that chronic exposure to hIFN-α2b might alter the response of murine target cells, and ongoing experiments are aimed at investigating whether hIFN-α2b may down-regulate the expression of type I IFN receptor on murine cells. Aside from this transient effect, the fact that in our preclinical model hIFN-α2b acted mainly, if not only, on PEL cells is clearly evident from the absence of a significant difference between the two “therapeutic” (murine vs human IFN-α, Figure S3) groups, ruling out the possibility of an additive effect on host cells. On the whole, these data indicate that, in our system, IFN- and EGFP-expressing LVs efficiently delivered the transgene to PEL cells and conferred long-term transgene expression in vivo as well. Vector biodistribution was assessed by real-time PCR using EGFP-specific primers and probe in DNA extracted from peripheral blood cells 10 days after PEL cell injection, and from different body compartments (ascitic cells, peritoneum, lung, spleen, liver, and kidney) recovered when the EGFP-LV–treated animals were sacrificed. As shown in Table S1, the vector infected mainly ascitic cells and cells lining the peritoneal cavity, whereas systemic spreading remained limited. Cumulative analysis of the in vivo anti-neoplastic activity of hIFN-α2b-LV and mIFN-α1-LV. The increase in the survival time of animals treated with IFN-α–expressing LVs remained statistically significant by assembling control groups of all in vivo experiments (Figure S3, log-rank test, hIFN-α2b-LV or mIFN-α1-LV vs EGFP-LV, P < 0.0001). Comparison of the two IFN-α treatments showed no significant differences in survival time (hIFN-α2b-LV vs EGFP-LV: 3.5-fold increase; mIFN-α1-LV vs EGFP-LV: 2.7-fold increase; log-rank test, hIFN-α2b-LV vs mIFN-α1-LV, P = 0.8527) or ascites development (Fisher Exact test, P = 0.5402). Moreover, EGFP-LV treatment led to a significant delay in the development of malignant effusions compared to medium-injected animals (log-rank test, EGPV-LV vs medium, P = 0.0137), as already observed in each in vivo experiment (Figures 3A and 4A). Although additional factors might be considered in vivo, this phenomenon was in line with the in vitro experiments, which showed a reduced viability and transgene expression soon after transduction also in EGFP-LV–transduced cells (Figure S1). As previously reported,6,7 this finding may be linked to cytotoxic activity to most mammalian cells of VSV-G glycoprotein. Accordingly, PEL cell viability was also affected by the transduction of an empty LV (not shown). Moreover, vector ultra-centrifugation was shown to augment intrinsic toxicity to target cells for the parallel concentration of protein debris and dyes derived from vector-producing cell culture media.8 Nevertheless, it is conceivable that the toxicity observed in in vivo experiments affected to the same extent either the EGFP-LV– and the IFN-LV–treated animals, allowing for the comparison of the obtained survival curves. Effects of IFN-αs on HHV8 reactivation. As human IFN-α was shown to inhibit HHV8 lytic program in PEL cell lines,9 we evaluated whether the antiviral status induced by mIFN-α1 could limit HHV8 reactivation from latency. After lytic cycle induction, followed by human or murine IFN-α treatment, we examined the presence of an HHV8 transcript characteristic of the early lytic phase, encoded by ORF50, in parallel with the degree of induction of the viral processivity factor 8 (PF8), a delayed-early lytic protein. As shown in Figure S4, panel A, an ORF50-specific transcript was detected 48 and 72 hours after n-butyrate induction in untreated as well as in mIFN-α1–treated cells, whereas the induction of ORF50 gene expression was strongly reduced in hIFN-α2b–exposed CRO-AP/3 cells (Figure S4, panel B). PF8 expression, analysed by an indirect immunofluorescence assay (IFA), was found in 28 ± 4% of untreated and mIFN-α1–treated CRO-AP/3 cells, whereas it was significantly inhibited in hIFN-α2b–exposed cells (7.6 ± 1.8%, P < 0.0001, panel B2), further indicating that mIFN-α1 does not interfere with the expression of HHV8 lytic proteins and does not play a role in controlling virus replication. Viral interleukin 6 (vIL-6) expression was shown to be activated by IFN-α in PEL cells, and this protects cells from IFN-α–mediated cell cycle arrest and apoptosis.10 We therefore investigated whether IFN-α could lead to vIL-6 up-regulation in hIFN-α-LV–transduced CRO-AP/3 cells by IFA analyses. Low levels of expression (< 2%) were detected in control cells (not transduced and mock-transduced cells), in EGFP-LV– and in mIFN-α1-LV–transduced cells (Figure S4, panel C), reflecting the low percentage of cells entering spontaneously the lytic replication program. On the contrary, CRO-AP/3 cells expressing increasing amounts of the hIFN-α2b showed a m.o.i.-dependent and statistically significant up-regulation of vIL-6 (Student t test, P < 0.01), indicating that also CRO-AP/3 cells respond to hIFN-α2b by up-regulating vIL-6 expression. Evaluation of the contribution of murine NK cells and macrophages to the mIFN-α1 anti-neoplastic activity. As the immunomodulatory activity is one of the multiple effects of type I IFNs,11 we analyzed the possible involvement of murine NK cells and macrophages in the antitumor effects exerted by mIFN-α1. Flow cytometric analyses did not disclose the presence of measurable amounts of murine NK cells in lymphomatous effusions from four mIFN-α1–treated animals or from control and hIFN-α2b-LV–treated animals, using the DX5 monoclonal antibody. However, NK cells were occasionally, but not regularly, detected at higher levels in the spleen of mIFN-α1-LV–treated animals (range 4.7–5.4% in mIFN-α1-LV–treated mice vs 1.2–3% in all other groups). Splenic F4/80+ macrophages showed a significant increase in mIFN-α1-LV–treated animals compared to control and hIFN-α2b-LV–treated mice (range 67.5–75% in mIFN-α1-LV–treated mice vs 11–37.5% in all other groups, P = 0.0288), whereas peritoneal F4/80+ cells were detected in comparable amounts in all groups of animals (not shown). A control vector-induced immune stimulation was never observed in all sets of experiments, since levels of detection of peritoneal or splenic murine NK and macrophages did not significantly differ between medium- and EGFP-LV–injected control groups (not shown). REFERENCES 1. Indraccolo S, Tisato V, Tosello V, et al. Interferon-α gene therapy by lentiviral vectors contrasts ovarian cancer growth through angiogenesis inhibition. Hum Gene Ther. 2005;16:957–970. 2. Gasperini P, Barbierato M, Martinelli C, et al. Use of a BJAB-derived cell line for isolation of human herpesvirus 8. J Clin Microbiol. 2005;43:2866–2875. 3. Dumont FJ, Coker LZ. Interferon-α/β enhances the expression of Ly-6 antigens on T cells in vivo and in vitro. Eur J Immunol. 1986;16:735–740. 4. Dumont FJ. Recombinant human interferon-α A/D enhances the expression of Ly-6A/E, Ly-6C, and TAP antigens on murine T lymphocytes. J Interferon Res. 1988;8:347–356. 5. Kumaran J, Wei L, Kotra LP, Fish EN. A structural basis for interferon-α-receptor interactions. FASEB J. 2007;21:3288–3296. 6. Burns JC, Friedmann T, Driever W, Burrascano M, and Yee JK. Vesicular stomatitis virus G glycoprotein pseudotyped retroviral vectors: concentration to very high titer and efficient gene transfer into mammalian and nonmammalian cells. Proc Natl Acad Sci U S A. 1993;90:8033–8037. 7. Yee JK, Miyanohara A, LaPorte P, Bouic K, Burns JC, and Friedmann T. A general method for the generation of high-titer, pantropic retroviral vectors: highly efficient infection of primary hepatocytes. Proc Natl Acad Sci U S A. 1994;91:9564–9568. 8. Yamada K, McCarty DM, Madden VJ, and Walsh CE. Lentivirus vector purification using anion exchange HPLC leads to improved gene transfer. Biotechniques. 2003;34:1074–1780. 9. Monini P, Carlini F, Stürzl M, et al. Alpha interferon inhibits Human Herpesvirus 8 (HHV-8) reactivation in primary effusion lymphoma cells and reduces HHV-8 load in cultured peripheral blood mononuclear cells. J Virol. 1999;73:4029–4041. 10. Chatterjee M, Osborne J, Bestetti G, Chang Y, Moore PS. Viral IL-6–induced cell proliferation and immune evasion of interferon activity. Science. 2002;298:1432–435. 11. Belardelli F, Ferrantini M, Proietti E, Kirkwood JM. Interferon-α in tumor immunity and immunotherapy. Cytokine Growth Factor Rev. 2002;13:119–134.
Files in this Data Supplement:
- Table S1. Analysis of the systemic distribution of EGFP-LV–injected intraperitoneally (PDF, 14.9 KB)
- Figure S1. Efficiency of EGFP gene transfer (JPG, 61.9 KB)
-
EGFP expression was analyzed by flow cytometry at different time points in CRO-AP/3 cells transduced in vitro with the lentiviral vector EGFP-LV using final multiplicities of infection (m.o.i.) of 2, 10 and 25. The mean and standard deviation values of three independent experiments are reported. The transduction efficiency of CRO-AP/3 cells was found to be m.o.i.-dependent.
- Figure S2. Expression of IFN-α–modulated Ly-6C surface antigen (JPG, 32.3 KB)
-
IFN-α–induced Ly-6C antigen was up-regulated in peripheral blood cells from mIFN-α1-LV–treated mice. Statistically significant differences between mIFN-α1–treated mice and controls (Student t test) are indicated (*).
- Figure S3. Cumulative analysis of the in vivo anti-neoplastic activity of hIFN-α2b-LV and mIFN-α1-LV (JPG, 33.5 KB)
-
Kaplan-Meier survival curves for CRO-AP/3–injected SCID mice treated with medium alone (18 mice), control EGFP-LV (22 mice), hIFN-α2b-LV (12 mice), and mIFN-α1-LV (10 mice). The increase in the survival time of IFN-α-LV-treated mice animals remained statistically significant by assembling all in vivo experiments. Comparison of the two IFN-α treatments showed no significant differences in survival time, indicating that specific targeting of the host peritoneal microenvironment with the mIFN-α1–expressing lentiviral vector may impair PEL cell growth in vivo.
- Figure S4. Antiviral activity of IFN-αs (JPG, 73.7 KB)
-
(A) Analysis of ORF50 gene expression. CRO-AP/3 cells, treated with n-butyrate (n-but) to induce HHV8 lytic cycle for 24 hours and thereafter exposed to human or murine IFN-α, were analyzed for the presence of a viral transcript characteristic of HHV8 early lytic phase by RT-PCR using primers specific for ORF50. Expression of human β-actin is shown below as control. An ORF50-specific transcript was detected 48 and 72 hours after lytic cycle induction in untreated as well as in mIFN-α1–treated cells, whereas the induction of the ORF50-specific transcript was found to be curtailed in CRO-AP/3 cells exposed to hIFN-α2b. (B) Detection of HHV8 processivity factor 8 (PF8). CRO-AP/3 cells, treated as reported above, were analyzed for the induction of viral PF8, characteristic of the delayed-early phase of the lytic cycle, by an indirect immunofluorescence assay, 24 hours after IFN-α treatment. CRO-AP/3 cells were stained with an anti-PF8 monoclonal antibody (red, Alexa 594 label). Untreated (1) and mIFN-α1–treated (3) CRO-AP/3 cells showed similar levels of n-butyrate–induced PF8 expression, whereas hIFN-α2b treatment (2) strongly inhibited PF8 expression, indicating that the human, but not the murine, cytokine is able to control viral reactivation. Original magnifications, ×20. (C) vIL-6 expression in transduced CRO-AP/3 cells. Histograms report mean and standard deviation of percentages of vIL-6–expressing cells analyzed at 3 different time points during transduction experiments. A m.o.i.-dependent and statistically significant (*, Student t test, P < 0.01, each column vs first four groups) up-regulation of vIL-6 was observed in hIFN-α2b-LV–transduced CRO-AP/3 cells.
|
|
