Volume 8, Issue 3, Pages 412-424 (September 2003) Novel immunocompetent murine tumor models for the assessment of replication- competent oncolytic adenovirus efficacy  Gunnel Halldén, Richard Hill, Yaohe Wang, Arthi Anand, Ta-Chiang Liu, Nick R Lemoine, Jennelle Francis, Lynda Hawkins, David Kirn  Molecular Therapy  Volume 8, Issue 3, Pages 412-424 (September 2003) DOI: 10.1016/S1525-0016(03)00199-0 Copyright © 2003 The American Society of Gene Therapy Terms and Conditions

FIG. 1. In vitro assessment of Ad5 uptake, replication, and cytopathic effects in murine and human carcinoma cell lines. (A) Cells were infected at 10 (black bars) or 100 (gray bars) particles/cell (ppc) with Ad-GFP and harvested 24 h postinfection to determine infectivity by FACS analysis. The data are from three experiments, each done in triplicate (±SE). (B) Subconfluent murine and human cell lines were infected with Ad5 wild-type virus at 1000 ppc for 1 h. Cells were harvested 48 and 96 h postinfection, and Western blots were carried out for E1A and late coat proteins (hexon, fiber) as described under Methods. All viral proteins expressed in the murine cell lines were of the same molecular weight as those expressed in the human cell lines. (C) Cells were infected at 100 ppc with Ad5 and harvested 96 h later and infectious virus units were quantitated by limiting dilution assay on HEK293 cells (results in triplicate, repeated twice, ±SE); A549 burst performed at 48 h due to induction of severe cytopathic effects. (D) Results from representative cytopathic effect assay showing Ad5 and dl312 (E1A-negative) activity on two human (H460, A549) and nine murine cancer lines. Cells were infected with either Ad5 or dl312 at the ppc shown and stained with crystal violet 5 days later. Molecular Therapy 2003 8, 412-424DOI: (10.1016/S1525-0016(03)00199-0) Copyright © 2003 The American Society of Gene Therapy Terms and Conditions

FIG. 2. In vivo screening for E1A and coat protein expression in immunocompetent murine tumor xenografts following intratumoral injection with Ad5. Murine tumors were grown subcutaneously in the flanks of immunocompetent mice and were subsequently injected directly with Ad5 on days 1–3. On day 6 ± 1 posttreatment tumors were harvested and processed and immunohistochemistry (IHC) was performed for Ad E1A or Ad coat proteins as described under Methods. Representative cross sections of JC (permissive) and CMT-tk (nonpermissive) tumors are shown. Brown-staining cells were positive for protein expression (Note. For JC tumors only, negative controls are not shown but they included tumors injected with UV-inactivated dl312, and for IHC the primary antibody was omitted). Cytopathic effects (C.P.E.) are shown on H and E staining of JC tumors only (cell rounding, eosinophilic cytoplasm, nuclear inclusions). Molecular Therapy 2003 8, 412-424DOI: (10.1016/S1525-0016(03)00199-0) Copyright © 2003 The American Society of Gene Therapy Terms and Conditions

FIG. 3. Evidence for intratumoral viral replication in JC and CMT-64 tumors growing in immunocompetent mice. CMT-93 (white bar), JC (black bar), and CMT-64 (hatched bar) tumors were grown subcutaneously in the flanks of immunocompetent mice and were subsequently injected directly with Ad5 or nonreplicating control adenovirus (PUV-dl312; not detectable). On days 1, 5, or 15 posttreatment tumors were harvested and processed and infectious units were quantitated as described under Methods. The mean values (±SE) are shown for six tumors per time point. *Undetectable. Molecular Therapy 2003 8, 412-424DOI: (10.1016/S1525-0016(03)00199-0) Copyright © 2003 The American Society of Gene Therapy Terms and Conditions

FIG. 4. Local cellular immune response following injection with Ad5 versus control adenovirus or PBS in immunocompetent murine tumor xenografts. JC or CMT-93 tumors were grown subcutaneously in the flanks of immunocompetent mice and were subsequently injected directly with Ad5, PBS, or replication-attenuated adenovirus (dl312) on days 1–3. On days 5, 15, and 22 posttreatment tumors were harvested, processed, and sectioned and H and E staining was performed to assess cellular immune infiltration, including eosinophils and PMNs. Immunohistochemistry (IHC) was performed for macrophage and CD8(+) cell infiltration between days 5 and 22. (A) Representative cross sections of CMT-93 and JC tumors demonstrating Ad5-associated eosinophilic infiltration (CMT-93) and macrophage infiltration (both models). (B) Representative IHC staining of CD8-positive cell infiltration in CMT-93 tumors on day 15 posttreatment. (C) Semiquantitative histopathology scoring of eosinophil, macrophage, and CD8(+) cell infiltration at the indicated time points in CMT-93 tumors (scored as described under Methods). Molecular Therapy 2003 8, 412-424DOI: (10.1016/S1525-0016(03)00199-0) Copyright © 2003 The American Society of Gene Therapy Terms and Conditions

FIG. 5. Antitumoral efficacy following treatment with Ad5 versus nonreplicating control adenovirus (PUV-inactivated dl312) in immunocompetent versus athymic murine tumor xenografts. (A) Kaplan–Meier curve for CMT-64 groups. CMT-64 tumors were injected as described with Ad5 (black circle), replication-deficient adenovirus (PUV-dl312; open circle), or PBS (open square) and were followed for progression (tumor volume >400 μl; P = 0.04 and 0.01, respectively). Kaplan–Meier curves for JC treatment groups in (B) immunocompetent BalbC mice or (C) athymic nu/nu mice are also shown. JC tumor cells were preinfected (5%) as described with Ad5 (black) or replication-incompetent adenovirus (PUV-dl312; open) prior to inoculation subcutaneously in the flanks of immunocompetent mice. The percentage of mice alive over time is shown (P = 0.04 for Ad5 versus control). CMT-93 tumors were grown subcutaneously in the flanks of (D) C57B/6 immunocompetent or (E) athymic (nu/nu) mice and were subsequently injected directly with PBS (open square), Ad5 (black square), adenovirus control particle (open triangle; PUV-dl312), or dl312 (open circle; E1A-negative control). The percentage of mice free from progression at each time point was estimated using the Kaplan–Meier method (±SE) (P = 0.004 for Ad5 vs control particle, P < 0.001 vs PBS, P = 0.03 vs dl312). Molecular Therapy 2003 8, 412-424DOI: (10.1016/S1525-0016(03)00199-0) Copyright © 2003 The American Society of Gene Therapy Terms and Conditions