An Oncolytic Adenovirus Enhanced for Toll-like Receptor 9 Stimulation Increases Antitumor Immune Responses and Tumor Clearance

original article

© The American Society of Gene & Cell Therapy

An Oncolytic Adenovirus Enhanced for Toll-like Receptor 9 Stimulation Increases Antitumor Immune Responses and Tumor Clearance Vincenzo Cerullo1,2, Iulia Diaconu1, Valentina Romano1, Mari Hirvinen1, Matteo Ugolini1, Sophie Escutenaire1, Sirkka-Liisa Holm1, Anja Kipar3,4, Anna Kanerva1,5 and Akseli Hemminki1 Cancer Gene Therapy Group, Molecular Cancer Biology Research Program, Transplantation Laboratory, Haartman Institute and Finnish Institute of Molecular Medicine, University of Helsinki, Helsinki, Finland;; 2Laboratory of Immunovirotherapy, Molecular Cancer Biology Research Program, University of Helsinki, Helsinki, Finland; 3Finnish Centre for Laboratory Animal Pathology, Section of Veterinary Pathology and Parasitology, Faculty of Veterinary Medicine, University of Helsinki, Helsinki, Finland; 4Department of Infection Biology, Institute of Global Health, School of Veterinary Science, University of Liverpool, Liverpool, UK; 5Department of Obstetrics and Gynecology, Helsinki University Central Hospital, Helsinki, Finland 1

Oncolytic viruses represent a multifaceted tool for ­cancer treatment. In addition to specific killing of cancer cells (oncolysis), these agents also provide danger signals prompting the immune system to stimulate an antitumor immune response. To increase adenovirus adjuvancy, we engineered the genome of Ad5D24 by inserting 18 immunostimulatory islands (Ad5D24-CpG). The toxicity and immunogenicity profile of Ad5D24-CpG showed that the safety of the maternal virus was retained. The efficacy of the CpG-enriched virus was assessed in a xenograft model of lung cancer where a significant increase in antitumor effect was seen in comparison with controls. When the experiment was repeated in animal depleted of natural killer (NK) cells, Ad5D24-CpG lost its advantage. The same was seen when Toll-like receptor (TLR)9 was blocked systemically. In a syngeneic model of melanoma (B16-OVA), we observed a significant increase of OVA-specific T cells and a decrease of activation of myeloid-derived suppressor cells in Ad5D24CpG–treated mice. In conclusion, we have generated the first genetically modified oncolytic adenovirus backbone able to enhance TLR9-stimulation for increased antitumor activity. Received 10 February 2012; accepted 17 June 2012; advance online publication 24 July 2012. doi:10.1038/mt.2012.137

Introduction

Multiple identified mechanisms by which tumors can evade control by the immune system illustrate the challenges in generating effective and specific antitumor immunity for therapeutic benefit.1 Gene therapy with adenoviruses is a promising modality for the treatment of advanced cancer refractory to the other therapies. Importantly, the safety and the efficacy of this approach have been recently validated in several studies.2–8 Classically, the immune system is thought to limit the efficacy of oncolytic adenoviruses, leading to their clearance. However,

recent preclinical and clinical data suggest that in some cases virotherapy may in fact act as cancer vaccine stimulating tumorspecific immune response.5,8,9 The rationale behind these discoveries is that adenovirus can activate antigen-presenting cells, such as dendritic cells (DCs), via stimulation of pattern recognition receptors. Once activated, DCs can present tumor antigen epitopes to CD8+ and CD4+ T cells. Unfortunately, the frequency of activated CD8+ T cells is often too low to mount an effective response against the tumor,10 especially when immune suppressive mechanisms are present as in advanced human tumors11 hence it is important to investigate new strategies to augment DC stimulation and migration. All the biologic responses crucial for DC functioning are influenced by signaling through Toll-like receptors (TLRs).12 TLRs are a group of receptors expressed primarily by cells of the innate immune system, and are activated by pathogen-associated molecular patterns. TLR9, 1 of the 11 TLRs identified in humans, is localized in late endosomes or lysosomes where it detects unmethylated CpG motifs in double-stranded DNA (CpG-DNA) from bacterial, viral, and human sources.13–15 Upon activation, TLR9 induces DCs to secrete high levels of type I interferons (IFNs) and interleukin-12 (IL-12), leading to rapid activation of natural killer (NK) cell cytotoxicity and IFN-γ production12 for a potent antitumor response.16,17 Agonists of TLR9 have demonstrated potential for the treatment of cancer. They directly induce activation and maturation of DCs, enhance differentiation of B cells into antibody-secreting plasma cells, activate T cells,16,17 and reduce inhibition by myeloidderived suppressor cells (MDSCs).18,19 MDSCs are a population of immature myeloid cells present in high numbers in a wide range of cancers.11,20 MDSCs accumulate due to a tumor-induced maturity block preventing differentiation of myeloid precursors into antigen-presenting cells.21 They inhibit effector T-cell proliferation in vitro and are thought to contribute to tumor-related immune suppression in vivo by inducing T-cell and NK-cell dysfunction.22,23 Murine MDSC are defined by the coexpression of the granulocyte differentiation antigen Gr1

Correspondence: Vincenzo Cerullo, Cancer Gene Therapy Group, Molecular Cancer Biology Research Program, Transplantation Laboratory, University of Helsinki, Helsinki 00330, Finland. E-mail: [email protected]

2076

www.moleculartherapy.org vol. 20 no. 11, 2076–2086 nov. 2012

© The American Society of Gene & Cell Therapy

Despite evolutionary reduction of the number of immunostimulatory sequences in the genome of adenoviruses,26 we and others have demonstrated that adenoviruses are able to elicit antitumor immune responses,5,9 and mechanistic studies have revealed molecular interactions with different arms of the innate immune system.27,28 We hypothesized that the combination of oncolytic adenovirus with specific oligonucleotides to stimulate TLR9 (e.g., ODN2395) would result in a more robust antitumor immune response and ultimately in improved efficacy. To this end, nude mice were implanted with subcutaneous A549 lung cancer xenografts and treated with media only, CpG oligos only, oncolytic adenovirus Ad5D24 (which bears a 24 bp deletion in E1A gene to allow selective replication in p16/Rb pathway deficient cancer cells)29 and a combination of Ad5D24 and CpG (Figure 1a). Interestingly, we observed that Ad5D24 + CpG significantly reduced tumor growth not only compared with mock (P = 0.005) but also when compared with Ad5D24 alone (P = 0.016) and CpG alone (P = 0.027) suggesting that TLR9 stimulation has a synergistic effect with oncolytic virotherapy. The activation of NK cells is one of the most prominent mechanisms that have been proposed to explain the activity of TLR9 agonists.16 Activated DCs secrete IL12, which activates NK cells to secrete IFN-γ. To study this, mouse splenocytes were harvested and cultured for 72 hours, media was collected and IL-12 and IFN-γ concentration were analyzed. We found that the levels of IL-12 (Figure 1b) and IFN-γ (Figure 1c) were significantly higher in the group featuring the combination of Ad5D24 and CpG. This suggested that the efficacy observed in this group might have been mediated by the innate arm of the immune system, in particular by NK cells.

A genetically modified CpG-rich oncolytic adenovirus (Ad5D24-CpG) mediates increased NFκB activation but retains oncolytic potency As the combination of CpG and oncolytic adenovirus (Ad5D24) resulted in an increased antitumor effect (Figure 1) and given that adenoviruses genome has evolved to reduce TLR9 stimulatory Molecular Therapy vol. 20 no. 11 nov. 2012

2,000

Mock Ad5D24 CpG Ad5D24 + CpG

Tumor volume (mm3)

1,500

**

1,000

*

*

500

0 0

b

1,000 800

IL12 (pg/ml)

Results Combination of stimulatory CpG oligonucleotides (CpG) and oncolytic adenovirus results in an enhanced antitumor efficacy in a xenograft model of lung cancer

a

2

4

6 Time (days)

*

Mock CpG Ad5D24 Ad5D24 + CpG

8

10

12

*

800 400 200 0 Mock

CpG

Ad5D24 Ad5D24 + CpG

Treatment

c

80

IFN-γ (pg/ml)

(constituted by the two epitopes Ly6G and Ly6C) and the α1βM integrin CD11b.11 MDSC represent a heterogeneous cell population, further subdivided by the differential expression of the Ly6G and Ly6C epitopes and the differential expression of Ly6G has been used to distinguish between MDSC subpopulations.24,25 One possible way to increase DC activation is stimulation of TLR9; however, nearly all DNA viruses, including Ad5, have evolved to avoid TLR9 recognition by reducing their genomic content of CpG dinucleotides by 50–94% from what would be expected.26 For this reason, we generated a CpG-rich oncolytic adenovirus to combine the effect of oncolysis (which releases tumor-associated antigens) with targeted delivery of a CpG stimulating double-stranded DNA molecule to the endosome where TLR9 is located.

CpG-enriched Oncolytic Adenovirus

60

*

Mock CpG Ad5D24 Ad5D24 + CpG

*

40 20 0 Mock

CpG

Ad5D24 Ad5D24 + CpG

Treatment

Figure 1 Stimulatory oligonucleotides (CpG) enhance antitumor efficacy of oncolytic adenoviruses. Nude mice (N = 5 per group) bearing established A549 lung cancer xenografts were injected intratumorally with Ad5D24 (closed squares), CpG stimulatory oligos (CpG) (open squares), and combination of CpG and Ad5D24 (open circles); saline solution was used as control (closed triangles). Viruses were administered at 1 × 108 virus particle/tumor on day 1 and 3. CpG (ODN2395) was administered intratumorally at 5 g on day 1, 3, 6 and 9. (a) Tumor growth was measured and plotted as total tumor volume. Mice were euthanized on day 12 and splenocytes were harvested and cultured. Media was collected and (b) IL12 and (c) IFN-γ was measured by ­facsarray technology as indicators of activation of dendritic cells and NK cells, respectively. P values; *P < 0.05, **P < 0.01. IFN-γ, interferon-γ; IL, interleukin; NK cells, natural killer cells.

sequences,26 we hypothesized that inserting CpG islands back into the genome would result in a synergistic antitumor efficacy with no additional need for administration of oligonucleotides. To this end, Ad5D24-CpG was generated by direct cloning of synthetic CpG island (18 islands) into the virus genome with no alterations of any of the adenovirus open reading frames (Figure 2a). Because TLR9 stimulation culminates in nuclear factor κB (NFκB) activation, we assessed the functionality of the new virus in 293 cells expressing TLR9 but no other TLRs and featuring luciferase driven by an NFκB-inducible promoter. At different doses and time points, we constantly observed a significant 2077

© The American Society of Gene & Cell Therapy

CpG-enriched Oncolytic Adenovirus

a ∆24E1

CpG AATT CpG AATT

06

6 x 10

06

***

AdCpG Ad5D24

***

***

4 x 1006 2 x 1006

8 x 10

06

6 x 10

06

***

*

2 x 1006

120 Time (minutes)

720

A549 cell line

120 Time (minutes)

06

8 x 10

***

6 x 1006 4 x 1006 2 x 1006

720

30

120 Time (minutes)

720

SKOV3 cell line

AdWT Ad5Luc1 Ad5D24 Ad5D24-CpG

150

AdCpG Ad5D24

07

1 x 10

0 30

AdWT Ad5Luc1 Ad5D24 Ad5D24-CpG

150

% of cell viability

100

50

100

50

0

0 0

0.1

1

10

100

0

Virus (concentration in VP/cell)

0.1

1

10

100

Virus (concentration in VP/cell) AdWT Ad5Luc1 Ad5D24 Ad5D24-CpG

PC3MM3 cell line

200

100

50

AdWT Ad5Luc1 Ad5D24 Ad5D24-CpG

Hela cell line 150

% of cell viability

150 % of cell viability

AdCpG Ad5D24

0 30

% of cell viability

***

4 x 1006

0

c

NFκB activation (10 VP/cell)

NFκB activation (1 VP/cell)

NFκB activation (0.1 VP/cell)

8 x 10

RLU/2 x 1005 cells

RLU/2 x 1005 cells

b

CpG AATT

RLU/2 x 1005 cells

6X

E3

100

50

0

0 0

0.1

1

10

100

Virus (concentration in VP/cell)

0

0.1

1

10

100

Virus (concentration in VP/cell)

Figure 2  Ad5D24-CpG adenovirus displays increased NFκB activation compared with Ad5D24 while maintaining oncolytic efficacy in vitro. (a) Schematic representation of Ad5D24-CpG. This virus bears a 24 bp deletion in E1A gene for selective replication in cancer cells and 18 CpG-rich islands to increase Toll-like receptor 9 (TLR9) stimulation. (b) NFκB activation upon infection with Ad5D24 and Ad5D24-CpG at different MOI of viruses and different time points. P values; *P < 0.05, **P < 0.01, ***P < 0.001 (c) Oncolytic efficacy of Ad5D24-CpG by MTS cell-killing assay in ­different human tumor cell lines. MOI, multiplicity of infection; MTS, 3-(4,5,dimethylthiazol-2-yl)-5-(3-carboxymethoxy-phenyl)-2-(4-sulfophenyl)2H-tetrazolium; NF-κB, nuclear factor-κB; RLU, relative luciferase unit; VP, virus particle.

increase in NFκB activation when the cells were infected with the CpG-rich adenovirus (Figure 2b). We next sought to investigate whether or not the capability of killing cancer cells in vitro was maintained. To this end, 3-(4,5,dimethylthiazol-2-yl)-5-(3-carboxymethoxy-phenyl)-2(4-sulfophenyl)-2H-tetrazolium (MTS) assay was performed on different human cancer cell lines. We observed that Ad5D24-CpG was slightly less effective than Ad5D24 only in SKOV3 at 1 and 10 2078

virus particle (VP)/cell and in PC3MM2 10 VP/cell (Figure 2c). These small differences could not be due to the VP/PFU (plaqueforming unit) ratio that was in slight favor of Ad5D24-CpG (10 versus 50 VP/PFU), but we can speculate that the more immunogenic nature of the CpG virus could lead the cell in a more stringent antiviral state that is able to slower its replication. However, Ad5D24-CpG was able to kill most cells and therefore we concluded that the virus retained its lytic capacity. www.moleculartherapy.org vol. 20 no. 11 nov. 2012

© The American Society of Gene & Cell Therapy

CpG-enriched Oncolytic Adenovirus

Ad5D24-CpG retains the safety profile of the maternal Ad5D24 but it shows enhanced immunogenicity Used at high doses, adenoviruses can constitute a real concern in terms of toxicity.30 Although the safety profile of cancer treatment with oncolytic adenoviruses has been excellent,31–36 it is conceivable that increasing the CpG content in the adenovirus genome could result in increased toxicity due to higher stimulation of TLR9 and consequent cytokine storm. To address this, we studied immediate and late cytokine levels, liver tropism and toxicity of the viruses, and ultimately the kinetics of viral clearance. To this end, immunocompetent C57BL/6 mice were intravenously injected with a high dose of virus. Mice were bled at early time point (6 hours) and at later time points (72 hours) and proinflammatory cytokines were analyzed. The first time point (6 hours) is indicative of the immediate toxicity, mainly due to the interaction of the viral capsid with different cell types of the immune system. The second time point (72 hours) is indicative of intermediate toxicity due to the replication of the genome and the production of viral proteins.30,37–39 Although both viruses (Ad5D24 and Ad5D24-CpG) induced a significant increase of cytokines at 6 and 72 hours compared with phosphate-buffered saline (PBS)-treated mice, we did not observe any significant difference between them (Figure 3a). The histological examination of the liver of mice infected with either virus, identified similar changes. At 6 hours postinfection, pathological changes were restricted to marked periportal hepatocellular glycogen loss, indicating cellular energy deprivation and evidence of some degree of T-cell recruitment, indicated by the presence of some T cells in central veins (Supplementary Figure IL6

250

Mock Ad5D24 Ad5D24-CpG

200

100

50

IL12p40

200

Mock Ad5D24 Ad5D24-CpG TNF alpha (pg/ml)

150

IL-12p40 (pg/ml)

IL-6 (pg/ml)

a

S1a,b). At 72 hours postinfection, both groups of mice exhibit marked pathological changes, represented by extensive random hepatocyte apoptosis and marked T-cell recruitment and infiltration in portal areas and sinuses (Supplementary Figure S1c–f). At 6 days postadministration, there is still evidence of hepatocytes’ apoptosis and T-cell infiltration, but to a far lower degree than at 72 hours postinfection. The T-cell infiltrate is seen predominantly in portal areas and lesser in sinusoids, without evidence of further cell recruitment. The degree of cell death and T-cell infiltration is lower in animals infected with Ad5D24-CpG (Supplementary Figure S2a–d). At 9 days postinfection, the inflammatory (T cell) infiltration has further subsided significantly and there is no evidence of cell death. Again, the degree of leukocyte infiltration is lower in animals infected with Ad5D24-CpG (Supplementary Figure S2e,f). Weight loss and survival were monitored and no significant differences were observed between the three different groups (Supplementary Figure S3a,b). To test the immunogenicity of the CpG-rich backbone, splenocytes of immunocompetent mice intramuscularly vaccinated with the viruses (Ad5D24, Ad5D24-CpG) or PBS as control, were harvested pulsed with Ad5-specific peptides and γ-ELISPOT was performed. Interestingly, we found a significant increase of splenocytes-secreting IFN-γ in the group of mice immunized with Ad5D24-CpG, which was not observed when splenocytes were pulsed with an irrelevant peptide (Figure 3b). To assess whether the new viral backbone would have an impact on replication of the virus and consequently on its liver toxicity, the kinetics of the viral genomes was studied by quantitative PCR (Figure 4). As expected, we did not observe significant

150 100 50

Time (hours)

6 hours

RANTES

150

Mock Ad5D24 Ad5D24-CpG

400

200

0

0

72 hours

IFN-γ (pg/ml)

RANTES (pg/ml)

600

6 hours

6 hours

Time (hours)

72 hours

IFN γ Mock Ad5D24 Ad5D24-CpG

100

50

0

b

80

72 hours Time (hours)

50

6 hours

72 hours Time (hours)

Mock Ad5D24 Ad5D24-CpG

60

** *

40 *

20

0 6 hours

100

0

72 hours Time (hours)

SFU/106 splenocytes

0

150

TNF-α Mock Ad5D24 Ad5D24-CpG

Mock

PSA peptide

Ad5 peptide

Figure 3  Ad5D24-CpG retains the safety profile of the maternal Ad5D24 but shows enhanced immunogenicity. C57BL/6 mice (N = 8 per group) were intravenously injected with 5 × 1012 VP/kg (~1 × 1010 VP/mouse) of Ad5D24 (white), Ad5D24-CpG (gray), and PBS (black). (a) Proinflammatory cytokines were analyzed at 6 and at 72 hours. (b) C57BL/6 mice (N = 4 per group) were intramuscularly injected with ~1 × 108 VP/mouse of Ad5D24 (white), Ad5D24-CpG (gray), and PBS (black). Three weeks later, splenocytes were harvested from the mice pulsed with adenovirus 5-derived peptides and IFN-γ ELISPOT was performed. P values; *P < 0.05, **P < 0.01. IFN-γ, interferon-γ; IL, interleukin; PBS, phosphate-buffered saline; PSA, prostate-specific antigen; RANTES, regulated upon activation, normal T-cell expressed, and secreted; SFU, standard fluorescence unit; TNF-α, tumor necrosis factor-α; VP, virus particle.

Molecular Therapy vol. 20 no. 11 nov. 2012

2079

© The American Society of Gene & Cell Therapy

CpG-enriched Oncolytic Adenovirus

Viral particles in the clot (VP/100 ml)

a

100,000

10,000

Mock Ad5D24 Ad5D24-CpG

*

1,000

100

10

1 6 hours

72 hours

6 days

9 days

6 days

9 days

Time

b

100,000

VP/ng of genomic DNA

10,000

Mock Ad5D24 Ad5D24-CpG

*

1,000

100

10

1

6 hours

72 hours Time

Figure 4 The insertion CpG islands into the adenoviral genome did not impact biodistribution but altered DNA replication. C57BL/6 mice (N = 12 per group) were intravenously injected with 5 × 1012 VP/ kg (~1 × 1010 VP/mouse) of Ad5D24 (white), Ad5D24-CpG (gray), and PBS (black). (a) qPCR was performed at different time points to assess the amount of viral genomes in the blood clots. (b) Total liver DNA (N = 3 mice) was extracted to assess amount of viral genomes at different time points. P values; *P < 0.05. PBS, phosphate-buffered saline; qPCR, quantitative PCR; VP, virus particle.

differences in the liver (P = 0.16) or in the clot (P = 0.13) at 6 hours indicating that the transduction of the virus has remained unaltered. Also, identical amounts of virus DNA detected in the liver confirm that dosing and the titers of the viruses were equivalent. At 72 hours, we observed a significant increase of Ad5D24 genomes in both the clot (P = 0.01) and liver (P = 0.007) suggesting a more active replication of this virus. These findings well correlate with the in vitro experiment results described in Figure 2c. No differences were observed at the later time points (6 and 9 days after virus administration).

Ad5D24-CpG displays a TLR9-mediated NK-dependent enhanced antitumor effect in a xenograft model of lung cancer In a mouse model of human lung cancer, Ad5D24 was more effective than mock but adding recombinant CpG oligonucleotide did not significantly enhance efficacy (Figure  5a). In contrast, Ad5D24-CpG was superior to the combination group, highlighting the utility of targeted delivery of the CpG-rich adenoviral genome to the endosome where TLR9 is located. On day 12, mice were euthanized and spleens were collected for further analysis. The cytokine profile (Figure  5b) suggested that NK activation might be involved in the enhanced antitumor 2080

response, especially in absence of T cells in athymic nude mice. To study this further, we analyzed the proportion of NK cells in spleens of treated mice (Supplementary Figure S4). Interestingly, we found that in mice treated with Ad5D24-CpG, the percentage of NK cells was higher than in controls. Stimulated splenocytes were then cocultured with tumor cells at different ratios and after 72 hours the viability of target cells was measured (Supplementary Figure S5). Intriguingly, splenocytes harvested from mice treated with Ad5D24-CpG showed a significantly increased cell-killing activity which is in line with the hypothesis that the treatment might have upregulated NK cells (Supplementary Figure S4). When splenocytes from this experiment were cocultured with different target cell line (SKOV3, BT474, and 293), we also observed some degree of cell killing (Supplementary Figure S6). These results are in line with an NK cells-mediated cell-killing effect. To confirm the role of the NK cells in this tumor model, the experiment was repeated with prior depletion of NK cells. Interestingly, we observed that in the absence of NK cells, Ad5D24CpG lost its advantage on the regular Ad5D24 virus (Figure 5c). To assess the role of TLR9-binding on the activity of the CpGrich oncolytic adenovirus, an experiment was performed with simultaneous TLR9-blockage.28 This also resulted in Ad5D24CpG losing its advantage on Ad5D24 (Figure 5d). Taken together, these experiments suggest that the CpG-enriched adenovirus genome is able to stimulate TLR9 resulting in an NK-mediated improvement in antitumor response.

Ad5D24-CpG displays increased efficacy and enhanced antitumor immune response in a syngeneic model of melanoma Xenograft models in nude mice are limited by the lack of an intact immune system. Although we found NK cells to be important players in this model, we could not exclude that other cells might also be relevant in the presence of a full immune system. It has been previously demonstrated that adenovirus-based therapy can stimulate a tumor-specific immune response.5,9 However, the impact of stimulation of TLR9 with a CpG-rich oncolytic adenovirus on the tumor-specific immune response has not been previously studied. To answer these questions, we performed an additional set of experiments using a syngeneic murine model of melanoma stably transfected with chicken ovalbumin (B16-OVA), that allows accurate assessment of tumor specific T-cell responses. The syngeneic model has the advantage of the intact immune system but it has the limitation that serotype 5-based oncolytic adenoviruses do not kill murine cells due to species-specific incompatibility.40 Hence the effect observed in the following experiments can only rely on the activation of the immune system. We observed that Ad5D24-CpG significantly enhanced tumor control compared with Ad5D24 or the combination of Ad5D24 with recombinant CpG oligonucleotides (Figure  6a). Because stimulation of TLR9 can lead to T-cell activation, we analyzed the anti-OVA–specific T cell as a surrogate of antitumor immunity in the spleen and directly in the tumors (Figure 6b,c). Interestingly, we found that the number of CD8+ and the number of the antiOVA specific T cells was significantly higher in mice treated with Ad5D24-CpG in the spleen and in the tumor (Figure 6b,c). www.moleculartherapy.org vol. 20 no. 11 nov. 2012

© The American Society of Gene & Cell Therapy

a

CpG-enriched Oncolytic Adenovirus

1,500

1,250

Tumor volume (mm3)

§

Mock Ad5D24 Ad5D24+CpG Ad5D24CpG

1,000

***

750

*** 500

*

250

0 0

b

Mock Ad5D24 Ad5D24-CpG Ad5D24CpG

40

**

2

4

6

8 Time (days)

100

**

12

14

Mock Ad5D24 Ad5D24-CpG Ad5D24CpG

**

IFN-γ (pg/ml)

20 10

60 40

Mock

Ad5D24

0

Ad5D24-CpG Ad5D24+ODN

Ad5D24

Mock

Ad5D24-CpG Ad5D24+ODN Treatment

Treatment Mock Ad5D24 Ad5D24-CpG Ad5D24CpG

40

* TNF-α (pg/ml)

RANTES (pg/ml)

600

400

200

0

Ad5D24

Mock

Mock Ad5D24 Ad5D24-CpG Ad5D24CpG

*

30 20 10 0

Ad5D24-CpG Ad5D24+ODN

Mock

Ad5D24 Ad5D24-CpG Ad5D24+ODN Treatment

Treatment 2,500

Mock Ad5D24 Ad5D24-CpG

2,000

d

NK depleted

1,500

** **

1,000 500 0

3 Tumor volume (mm )

3 Tumor volume (mm )

*

20

0

c

**

80

30

IL12 (pg/ml)

10

TLR9 KO-Oligos 2,000

Mock Ad5D24 Ad5D24-CpG

1,500

1,000

** **

500

0 0

2

4

6

8 10 12 Time (days)

14

16

18

20

0

2

4

6

8 10 12 Time (days)

14

16

18

20

Figure 5  Ad5D24-CpG displays a TLR9-mediated NK-dependent enhanced antitumor effect in a xenograft model of lung cancer. Nude mice (N = 5 per group) bearing A549 lung cancer xenografts were injected intratumorally with 1 × 108 VP/tumor on day 0. Ad5D24-CpG (open circles), Ad5D24 (closed squares), and a combination of Ad5D24 and CpG stimulatory oligos ODN 2395 (open squares) were used in the experiment. (a) Tumor growth was measured over time. (b) Mice were euthanized on day 12 and splenocytes were harvested and cultured. Media was ­collected and the expression of proinflammatory cytokines IL12, IFN-γ, RANTES, and TNF-α were measured. (c) Nude mice (N = 5 per group) bearing A549 lung cancer xenografts were depleted of NK cells and injected intratumorally with 1 × 108 VP/tumor on day 0 of Ad5D24-CpG (open circles), Ad5D24 (closed squares), and PBS (closed triangles). Tumor growth was measured over time. NK cells depletion from mice was maintained for all the duration of the experiment. (d) Nude mice (N = 5 per group) bearing A549 lung cancer xenografts were treated with TLR9-blockers and injected intratumorally with 1 × 108 VP/tumor on day 0 of Ad5D24-CpG (open circles), Ad5D24 (closed squares), and PBS (closed triangles). Tumor growth was measured over time. TLR9-blockers treatment was performed for all the duration all of the experiment. P values; *P < 0.05, **P < 0.01, ***P < 0.001. § Mice euthanized. IFN-γ, interferon-γ; IL, interleukin; NK cells, natural killer cells; PBS, phosphate-buffered saline; RANTES, regulated upon activation, normal T-cell expressed, and secreted; TLR, Toll-like receptor; TNF-α, tumor necrosis factor-α; VP, virus particle.

Molecular Therapy vol. 20 no. 11 nov. 2012

2081

© The American Society of Gene & Cell Therapy

CpG-enriched Oncolytic Adenovirus

a

1,000

Mock Ad5D24 Ad5D24+CpG Ad5D24CpG

Tumor volume (mm3)

800 600

***

400 200 0

2

0

8

6

4

10

Time (days)

3

% Of CD8+ cells (spleen)

Analysis of the spleen

*** ***

2

***

***

1

0

Mock

Ad5D24

Ad5D24+CpG Ad5D24 CpG

8

***

***

6 4 2 0

Mock

Ad5D24 Ad5D24+CpG Ad5D24 CpG

10

*

**

8 6 4 2 0

Mock

Ad5D24

Ad5D24+CpG Ad5D24 CpG

% of OVA-specific T cell (tumor)

Analysis of the tumor

c % of CD8+ cells (tumor)

% Of OVA-specific T cell (spleen)

b

**

1.5

**

1.0

0.5

0.0 Mock

Ad5D24

Ad5D24+CpG Ad5D24 CpG

Analysis of the spleen

*** 8

80

6

60 % Ly-6Ghigh

% Cr+ CD11b+

d

4

***

40 20

2 0

***

Mock

0

Ad5D24 Ad5D24+CpG Ad5D24 CpG

Mock

Ad5D24 Ad5D24+CpG Ad5D24 CpG

Analysis of the tumor

e

30

*

*

*

*

6

% Ly-6Ghigh

% Cr+ CD11b+

8

4 2 0 Mock

Ad5D24

Ad5D24+CpG Ad5D24 CpG

20

10

0

Mock

Ad5D24

Ad5D24+CpG Ad5D24 CpG

Figure 6  Ad5D24-CpG displays enhanced antitumor efficacy compared with Ad5D24 and coadministration of Ad5D24 + CpG oligonucleotides in immunocompetent mice bearing B16-OVA tumors. C57BL/6 mice (N = 4 per group) bearing established tumors were injected intratumorally with 1 × 108 VP/tumor on day 0, 2, and 4 of the following viruses: Ad5D24-CpG (open circles), Ad5D24 (closed squares), and a combination of Ad5D24 and CpG oligonucleotide (open squares). (a) Tumor growth was measured over time. Mice were euthanized on day 8 and (b,d) splenocytes and (c,e) tumors were harvested and analyzed for percentage of total CD8+ and anti-OVA specific T cell defined as CD8+CD19-pentamerSIINFEKL+. Percentage of myeloid-derived suppressor cells, defined as positive for granzyme (Gr+) and CD11b. Percentage of Ly6G epitope positive cells was analyzed as a functional marker of the immunosuppressivity of the CD11b+Gr+ population. P values where not indicated; *P < 0.05, **P < 0.01, ***P < 0.001. VP, virus particle.

2082

www.moleculartherapy.org vol. 20 no. 11 nov. 2012

© The American Society of Gene & Cell Therapy

Because it has been recently demonstrated that CpG oligonucleotide therapy blocks immune suppression by MDSCs in tumorbearing mice,19,41 we also analyzed the percentage of MDSCs. Interestingly, in spleen although we did not observe quantitative differences between the different groups (Figure  6d), when a marker of their suppressivity (Ly-6G) was taken into account, a significant reduction was seen with Ad5D24-CpG (Figure 6d). When the analysis was repeated in the tumors, we observed a significant reduction of total number and activation of MDSCs (Figure 6e).

Discussion

This report outlines a novel immunotherapeutic approach to genetically modify oncolytic adenoviruses by enriching their genome with CpG islands. The combination of selective and selfpropagating tumor cell killing with immunostimulation is important, as it may facilitate changes in the tumor microenvironment. Specifically, this approach could result in switching from limited antigen presentation and activation to an effective priming of tumor-specific CD8+ T cells with potent effector function. We found that a genetically modified CpG-rich oncolytic adenovirus (Ad5D24-CpG) shows enhanced antitumor activity in conjunction with a capacity for activation of the innate and adaptive immune systems toward the tumor. First, we showed that the combination of CpG containing oligonucleotide CpG and oncolytic adenovirus therapy is synergistic (Figure  1). These results are well in line with what was previously demonstrated by Van Oosten and colleagues that a TRAILexpressing adenovirus in combination with CpG oligonucleotide treatment resulted in enhanced efficacy dependent on CD8+ T cells.10 In our model, the lack of T cells is compensated by the activation of NK cells as demonstrated in similar animal models where only CpG oligonucleotide therapy was given.16 The limitations of combining two separate agents are several, from the cost of the oligonucleotides to their inactivation by nuclease. Also, two agents would disseminate to different cell compartments resulting in suboptimal concentrations at the target. The amount of oligonucleotides reaching the endosome after intravenous delivery is not known but it is likely to constitute a small fraction of the initial administered dose. In contrast, oncolytic replication amplifies locally at tumors, and adenovirus is an optimal vehicle for delivery of DNA to endosomes, because infection proceeds through them. In a xenograft model of lung cancer, we showed that the CpGrich adenovirus displayed increased antitumor activity compared not only with an oncolytic adenovirus alone but also with the combination of the oncolytic adenovirus and CpG oligonucleotides. One of the reasons for this phenomenon could be that while the pharmacokinetics of the oligonucleotide results in metabolic elimination from the host, the CpG-rich virus can constantly and sustainably replicate in tumor cells, transduce nearby antigenpresenting cells and stimulate the TLR9. Along this line, Raykov and colleagues inserted CpG motifs in a parvovirus to enhance its antitumor activity.42 However, because the animal model used was a rat, only limited immunological analyses were performed and thus the mechanism for additive effects was not fully analyzed. Our experiments in nude mice highlight the involvement of the innate immune system (in absence of T cells) and in particular Molecular Therapy vol. 20 no. 11 nov. 2012

CpG-enriched Oncolytic Adenovirus

the cytokine profile of the splenocytes harvested from the treated mice indicated an activation of NK cells.16,17 This was confirmed in a coculture experiment where splenocytes from the treated mice were cocultured with different cell lines. We observed that splenocytes from mice treated with the CpG-rich adenovirus displayed a more pronounced ability to kill cancer cells within 24 hours, which is typical of fast-acting NK cells. Also, increased numbers of NK cells were seen in mice treated with Ad5D24-CpG. Interestingly, when we performed the same experiment in NK cell-depleted mice, the CpG-virus lost its advantage suggesting a direct involvement of NK cells in tumor clearance. These results are well in line with what has been previously observed with CpG therapy alone and what has been proposed as the mechanism of action of CpG in nude mice.16,17 The proposed mechanism of action of TLR9 agonists is the stimulation of the innate and adaptive antitumor immune responses via secretion of IFN-α from immature plasmacytoid DCs, which activates NK cells to lyse tumor cells and release tumor antigens. DCs activated directly or indirectly by the TLR9 agonist therapy can capture and process tumor-associated antigens, increase surface expression of major histocompatibility complex and costimulatory receptors, and migrate to the lymph nodes where antigens can be presented to lymphocytes, inducing expansion of lymphocytes that recognize the tumor antigen. Because the response occurs in a Th1-like cytokine environment, the lymphocyte response tends to be Th1-like, including cytotoxic T lymphocytes.16 To assess whether the CpG-rich adenovirus would use the same mechanisms, an immunecompetent mouse model was utilized. In this model, we observed that Ad5D24-CpG was able to control tumor growth more efficiently than the control viruses. Thus, the effect is probably related to the ability of the viruses to stimulate an immune response towards the tumor. Our results are well in line with Tuve and colleagues, who observed that even unarmed replication deficient adenovirus, if used at high enough doses, can induce a specific antitumor immune response.9 These data are not surprising given that adenovirus has the capability to stimulate several receptors of the innate immune system including TLR 2 and 9,27,28 NOD-like receptors,43 and complement.44 However, several clinical studies have indicated that adenovirus alone is not enough to cure advanced tumors.5,7–8,45 Thus, the approach of genetically modifying the adenovirus to improve its capacity for stimulation of receptors involved in tumor rejection (e.g., TLR9) could be an exciting opportunity for improving oncolytic adenovirus efficacy. Interestingly, we found that Ad5D24-CpG showed increased efficacy compared with the combination of Ad5D24 and recombinant CpG oligonucleotides. We can speculate that this difference is due to two mechanisms: (i) oncolytic adenoviruses have the ability to replicate in tumors resulting in amplification of the amount of DNA with CpG islands; (ii) the adenovirus genome is directly and efficiently delivered to the endosome in contrast to oligonucleotides which have to overcome several obstacles to reach the target receptor following intratumoral or intraperitonel administration. We believe that most of the effect shown by Ad5D24-CpG is due its capacity to trigger NK and T cell antitumor response; however, 2083

CpG-enriched Oncolytic Adenovirus

given the recent finding that TLR9 stimulation by CpG oligos can trigger a suppression of myeloid-derived suppressor cells,18 we sought to investigate whether a similar effect could be observed also in the context of adenovirus-mediated TLR9 stimulation. Interestingly, we observed that Ad5D24-CpG therapy significantly reduced the suppression mediated by MDSCs in B16-OVA model. Although these are preliminary findings that would deserve a deeper experimentation and understanding of the all phenomenon, we do believe they are worth to be reported as overall contribution to Ad5D24-CpG activity and to eventually open new frontiers in the study of the mechanisms of action of oncolytic adenovirus. In summary, our results indicate that Ad5D24-CpG shows an appealing efficacy and immunogenicity improvement compared with Ad5D24. The insertion of CpG islands into the genome of the adenovirus has conferred the ability to overstimulate TLR9 and triggers a more robust immune response. This approach addresses the weaknesses of both oncolytic adenovirus-mediated immunotherapy (low immunogenicity of an unmodified backbone) and CpG therapy (poor delivery to TLR9, systemic side effects). Further, the small genome size and the intact multi-cloning sites of this novel oncolytic adenovirus represent an optimal backbone to insert transgenes by rationale design that can synergize with TLR9 stimulation and innate immune system activation to increase tumor clearance even further.

Materials and Methods

Construction of adenoviruses. Ad5D24 was generated and propagated with standard protocol as previously described.5 Ad5D24-CpG was generated recombining a CpG-rich shuttle plasmid (pTHSN-CpG1) with a plasmid containing the 24 adenovirus backbone. pTHSN-CpG1 was constructed by inserting NdeI-digested CpG1 (derived by digestion from custom designed pU57-CpG1; GenScript, Piscataway, NJ) and SpeI-digested CpG2 sequences (derived by digestion from custom designed pU57-CpG1; GenScript) respectively into NdeI- and SpeIlinearized plasmid pTHSN previously generated. CpG insertion was confirmed by enzyme digestion and sequencing using the following primers: CpG-NdeI400 forward: 5′-TTGTGTGGAATTGTGAGCGG-3′; CpGNdeI400 reverse: 5′-CTCCTGGCTGCAAACTTTCT-3′; CpG-SpeI400 forward: 5′-AACGACGGCCAGTGAATTGT-3′; CpG-SpeI400 reverse: 5′-GTAGGGCGTGGGAATTTCCT-3′. pTHSN-CpG1 was then recombined with pAd5D24 to obtain pAd5D24-CpG. The CpG-rich oncolytic adenovirus Ad5D24-CpG was then obtained with standard procedure. After obtaining the virus, its structure and insertion of the CpG islands were analyzed with the following primers: E3 transg forward: 5′-CAACCCAGACGGAGTGAGT-3′; E3 transg reverse: 5′-CCTGACTTCAAGGTTGTAGC-3′; CpG-MscI forward: 5′-CGATTGGCCATGAATTCCATATGTCGT-3′; CpG-MscI reverse: 5′-TTACTGGCCACTTCATATGATCAACGT-3′. Cell lines. Human transformed embryonic kidney cell line 293 stably

expressing human TLR9 (293XL-TLR9) was purchased from InvivoGen Europe (Toulouse, France). Human lung adenocarcinoma cell line A549, epithelian cervical carcinoma Hela, and 293 cell line were purchased from the American Type Culture Collection (Manassas, VA). Ovarian carcinoma cells expressing firefly luciferase, SKOV3-luc were provided by Dr Robert Negrin (Stanford Medical School, Stanford, CA). PC3MM2 are highly metastatic prostate cancer cells that were gently obtained from Isaiah J Fidler, MD Anderson Cancer Center, Houston, TX. B16-OVA cell line was kindly provided by Professor Richard Vile at Mayo Clinic (Rochester, MN). All cell lines were cultured in the recommended conditions.

2084

© The American Society of Gene & Cell Therapy

NFκB activation assay. To assay TLR9 stimulation via CpG island,

293XL-hTLR9 were plated at a density of 2 × 105 cells/well on 2 × 24 well/ plates and cultured overnight. Monolayers in triplicates were transfected using SuperFect reagent (Qiagen, Valencia, CA) with 1 µg of reporter plasmid pNiFty-Luc (InvivoGen) which carries the luciferase reporter gene driven by an NFκB-inducible ELAM-1 composite promoter able to respond to TLR9 stimulation. Green fluorescent protein-expressing plasmid was used as control of transfection. After 24 hours, cells were stimulated for 2 hours with Ad5D24-CpG or Ad5D24 and a cell lysate was obtained by adding a report lysis buffer in each well. After a freeze/thaw cycle, luciferase activity in 20 µl of cell lysates was assayed, using 100 µl of a commercially available buffer system (Promega, Mannheim, Germany), according to the manufacturer’s instructions.

Tumor cell-killing assay. Cell viability was assessed by MTS assay as described elsewhere. Briefly, cells were seeded at the concentration of 0.5 × 105 cells/well on 96-well plate and maintained under appropriate condition (RPMI 1640 or DMEM, completed with 10% FCS, 2 mmol L-glutamine, and 100 units/ml of penicillin (all from Sigma-Aldrich, St Louis, MO)). At the indicated time point, cell viability was assessed using CellTiter 96 Aqueous One Solution Cell Proliferation Assay (Promega), i.e., the inner salt (MTS) assay and compared with their control. Mice and animal experiments. All animal protocols were reviewed and approved by the Experimental Animal Committee of the University of Helsinki and the Provincial Government of Southern Finland. Mice were obtained from Taconic (Ejby, Denmark) at 4–5 weeks of age and quarantined at least for 2 weeks before the study. Health status of the mice was frequently monitored and as soon as sign of pain or distress was evident they were killed. The toxicity analysis was performed in C57BL/6 mice. Immuno­ competent C57BL/6 mice (N = 8 per group) were intravenously injected with high dose (5 × 1012 VP/kg and ~1 × 1010 VP/mouse) of virus (Ad5D24 and Ad5D24-CpG) while PBS was used as negative control. Histo­ pathological analysis of the livers was assessed as described previously.46–48 The immunogenicity of the CpG-rich backbone was tested in immunocompetent C57BL/6 mice (N = 4 per group) intramuscularly injected with 1 × 108 VP of virus (Ad5D24, Ad5D24-CpG, and PBS as control), 21 days after the immunization mice were euthanized, splenocytes were harvested, pulsed with Ad5-specific peptides and IFN-γ ELISPOT was performed (Mabtech, Stockholm, Sweden). Viral replication in clot and liver was assessed in immunocompetent C57BL/6 mice. Three groups of mice (N = 12 per group) were intravenously injected with Ad5D24, Ad5D24-CpG, and PBS at the dose of 1 × 1012 VP/ kg, three mice per group were euthanized at 6 hours, 72 hours, 6 days, and 9 days and quantitative PCR was performed on liver and clots. For the immune-deficient models, 106 A549 cells were injected subcutaneously into flanks of nude mice (n = 5–7 mice/group). When tumors reached the size of ~5 × 5 mm, virus was injected intratumorally at 108 VP/tumor at indicated time points. For the immunocompetent model, 5 × 105 B16-OVA cells were injected subcutaneously on shaved flanks of C57 black mice (n = 4 mice/group). Virus was injected intratumorally at 1 × 109 VP/tumor on days 0, 2, and 4, starting when tumors reached the size of ~5 × 5 mm. CpG oligonucleotides (TLR9 stimulator, ODN 2395 5′-tcgtcgttttcggcgc:gcgccg-3′; InvivoGen) was resuspended in PBS at a concentration of 10 μg/ml and premixed to the vector before intratumor injection. When injected alone ODN 2395 was simply resuspended in PBS. All the injections were performed with a total volume of 500 μl. TLR9-blockers (ODN2088, 5′-tcctggcggggaagt-3′; InvivoGen) were intraperitoneally administered every second day. Oligos were administrated at the dose of 100 μg for mouse per injection. NK depletion was obtained by intraperitoneal administration of 50 μl of anti-asialo GM1 (rabbit) antibody every 4 days (Wako Chemicals, www.moleculartherapy.org vol. 20 no. 11 nov. 2012

© The American Society of Gene & Cell Therapy

Neuss, Germany) according to manufacturer’s instruction. Each package of antibody comes with a titration for in vivo and in vitro use expressed in microliter. Quantitative PCR for detection of adenovirus was performed as described by us elsewhere.49 Tumor growth was followed and organs/ tumors were collected at the end of the experiments. Spleens were minced and splenocytes were cultured in 10% Dulbecco’s modified Eagle’s medium supplemented with 1% L-glutamine and penicillin/streptomycin for flow cytometry and cytokine analysis. Flow cytometry analysis. Single cell suspensions were stained with

fluorochrome-conjugated monoclonal antibodies and analyzed using a FACSARIA flow cytometer (BD Biosciences, Palo Alto, CA) and FlowJo software (TreeStar, Ashland, OR). For MDSCs, the following antibodies were used: mouse CD11b (PE-Cy7 labeled, RAT IgG2b,k isotype), mouse Gr-1 (Ly6G and Ly6C) (FITC labeled, RAT IgG2b,k isotype), and mouse Ly-6G (PE öabeled, RAT IgG2a,k isotype) all antibodies were purchased from (BD Biosciences). MDSCs-like population was defined as Gr-1+CD11b+ while reduction in the proportion of Ly6G was used as a marker of suppressive activity of the cell.11,18 For anti-OVA–specific pentamer analysis, we used mouse CD8 (FITC labeled, IgG2a isotype), mouse CD19 (PE-Cy7 labeled, RAT IgG2a isotype), and mouse SIINFEKL MHC-I pentamer (PE-labeled). Antibodies were purchased from PROIMMUNE (Oxford, UK). NK cells and lysis was performed using the anti mouse NK1.1 (FITC labeled, Mouse IgG2a,k isotype) from eBioscience (San Diego, CA). Tumor cell suspensions were obtained by digesting dissected tumor tissues with collagenase D (1 mg/ml) and DNase I (0.05 mg/ml, both from Sigma-Aldrich, Steinheim, Germany) for 40 minutes at 37 ºC and subsequent passage through a 70 μm pore cell mesh. Analysis was performed using a FACSCanto II flow cytometer (BD Biosciences) and FlowJo software (TreeStar). Cytokine analysis. Mouse IFN-γ, IL-2, TNF-α, RANTES (regulated upon

activation, normal T-cell expressed, and secreted) were assayed using the BD cytokine multiplex bead array system (BD Biosciences), and analyzed using BD FacsArray (BD Biosciences) according to the manufacturer’s instructions.

Statistical analysis. Statistical significance was determined by unpaired Student’s t-test using GraphPad Prism 5 (GraphPad Software, San Diego, CA). Comparison of multiple groups was performed using one-way analysis of variance with pair-wise Bonferroni post-tests. Tumor sizes were analyzed using the Mann–Whitney U-test. All data are expressed as mean ± SEM. A two-tailed P value of <0.05 was considered significant.

SUPPLEMENTARY MATERIAL Figure S1.  Morphological changes in livers of C57BL/6J mice intravenously infected with (a,c,e) Ad5D24 and (b,c,f) Ad5D24-CpG at 6 and 72 hours postinfection. Figure S2.Morphological changes in livers of C57BL/6J mice intravenously infected with (a,c,e) Ad5D24 and (b,c,f) Ad5D24-CpG at 6 and 9 days postinfection. Figure S3.Ad5D24-CpG retains the safety profile of the maternal Ad5D24. Figure S4.Ad5D24-CpG increases the number of NK cells in spleen of treated mice. Figure S5.Splenocytes derived from mice implanted with A459 tumors and treated with Ad5D24-CpG showed an increased capability to kill A549 cells ex vivo. Figure S6.Splenocytes derived from mice implanted with A549 tumors and treated with Ad5D24-CpG showed an increased capability to kill different cell lines in vitro.

ACKNOWLEDGMENTS This work was supported by the Finnish Institute for Molecular Medicine, University of Helsinki, Marie Curie FP7-IRG-2008, K. Albin Molecular Therapy vol. 20 no. 11 nov. 2012

CpG-enriched Oncolytic Adenovirus

Johansson Foundation, European Research Council, ASCO Foundation, HUCH Research Funds (EVO), Sigrid Juselius Foundation, Academy of Finland, Biocentrum Helsinki, Biocenter Finland. A.H. is K. Albin Johansson Research Professor of the Foundation for the Finnish Cancer Institute. A.H. is shareholder and paid consultant in Oncos Therapeutics Ltd. The other authors declared no conflict of interest.

REFERENCES

1. Real, LM, Jimenez, P, Kirkin, A, Serrano, A, García, A, Cantón, J et al. (2001). Multiple mechanisms of immune evasion can coexist in melanoma tumor cell lines derived from the same patient. Cancer Immunol Immunother 49: 621–628. 2. Immonen, A, Vapalahti, M, Tyynelä, K, Hurskainen, H, Sandmair, A, Vanninen, R et al. (2004). AdvHSV-tk gene therapy with intravenous ganciclovir improves survival in human malignant glioma: a randomised, controlled study. Mol Ther 10: 967–972. 3. Pearson, S, Jia, H and Kandachi, K (2004). China approves first gene therapy. Nat Biotechnol 22: 3–4. 4. Yu, W and Fang, H (2007). Clinical trials with oncolytic adenovirus in China. Curr Cancer Drug Targets 7: 141–148. 5. Cerullo, V, Pesonen, S, Diaconu, I, Escutenaire, S, Arstila, PT, Ugolini, M et al. (2010). Oncolytic adenovirus coding for granulocyte macrophage colony-stimulating factor induces antitumoral immunity in cancer patients. Cancer Res 70: 4297–4309. 6. Nokisalmi, P, Pesonen, S, Escutenaire, S, Särkioja, M, Raki, M, Cerullo, V et al. (2010). Oncolytic adenovirus ICOVIR-7 in patients with advanced and refractory solid tumors. Clin Cancer Res 16: 3035–3043. 7. Pesonen, S, Nokisalmi, P, Escutenaire, S, Särkioja, M, Raki, M, Cerullo, V et al. (2010). Prolonged systemic circulation of chimeric oncolytic adenovirus Ad5/3-Cox2L-D24 in patients with metastatic and refractory solid tumors. Gene Ther 17: 892–904. 8. Koski, A, Kangasniemi, L, Escutenaire, S, Pesonen, S, Cerullo, V, Diaconu, I et al. (2010). Treatment of cancer patients with a serotype 5/3 chimeric oncolytic adenovirus expressing GMCSF. Mol Ther 18: 1874–1884. 9. Tuve, S, Liu, Y, Tragoolpua, K, Jacobs, JD, Yumul, RC, Li, ZY et al. (2009). In situ adenovirus vaccination engages T effector cells against cancer. Vaccine 27: 4225–4239. 10. VanOosten, RL and Griffith, TS (2007). Activation of tumor-specific CD8+ T Cells after intratumoral Ad5-TRAIL/CpG oligodeoxynucleotide combination therapy. Cancer Res 67: 11980–11990. 11. Gabrilovich, DI and Nagaraj, S (2009). Myeloid-derived suppressor cells as regulators of the immune system. Nat Rev Immunol 9: 162–174. 12. Della Chiesa, M, Romagnani, C, Thiel, A, Moretta, L and Moretta, A (2006). Multidirectional interactions are bridging human NK cells with plasmacytoid and monocyte-derived dendritic cells during innate immune responses. Blood 108: 3851–3858. 13. Tabeta, K, Georgel, P, Janssen, E, Du, X, Hoebe, K, Crozat, K et al. (2004). Toll-like receptors 9 and 3 as essential components of innate immune defense against mouse cytomegalovirus infection. Proc Natl Acad Sci USA 101: 3516–3521. 14. Hovav, AH, Cayabyab, MJ, Panas, MW, Santra, S, Greenland, J, Geiben, R et al. (2007). Rapid memory CD8+ T-lymphocyte induction through priming with recombinant Mycobacterium smegmatis. J Virol 81: 74–83. 15. Cerullo, V, Seiler, MP, Mane, V, Brunetti-Pierri, N, Clarke, C, Bertin, TK et al. (2007). Toll-like receptor 9 triggers an innate immune response to helper-dependent adenoviral vectors. Mol Ther 15: 378–385. 16. Krieg, AM (2008). Toll-like receptor 9 (TLR9) agonists in the treatment of cancer. Oncogene 27: 161–167. 17. Krieg, AM (2007). Development of TLR9 agonists for cancer therapy. J Clin Invest 117: 1184–1194. 18. Zoglmeier, C, Bauer, H, Nörenberg, D, Wedekind, G, Bittner, P, Sandholzer, N et al. (2011). CpG blocks immunosuppression by myeloid-derived suppressor cells in tumor-bearing mice. Clin Cancer Res 17: 1765–1775. 19. Shirota, Y, Shirota, H and Klinman, DM (2012). Intratumoral injection of CpG oligonucleotides induces the differentiation and reduces the immunosuppressive activity of myeloid-derived suppressor cells. J Immunol 188: 1592–1599. 20. Serafini, P, Borrello, I and Bronte, V (2006). Myeloid suppressor cells in cancer: recruitment, phenotype, properties, and mechanisms of immune suppression. Semin Cancer Biol 16: 53–65. 21. Kusmartsev, S and Gabrilovich, DI (2006). Role of immature myeloid cells in mechanisms of immune evasion in cancer. Cancer Immunol Immunother 55: 237–245. 22. Bronte, V, Apolloni, E, Cabrelle, A, Ronca, R, Serafini, P, Zamboni, P et al. (2000). Identification of a CD11b(+)/Gr-1(+)/CD31(+) myeloid progenitor capable of activating or suppressing CD8(+) T cells. Blood 96: 3838–3846. 23. Liu, C, Yu, S, Kappes, J, Wang, J, Grizzle, WE, Zinn, KR et al. (2007). Expansion of spleen myeloid suppressor cells represses NK cell cytotoxicity in tumor-bearing host. Blood 109: 4336–4342. 24. Movahedi, K, Guilliams, M, Van den Bossche, J, Van den Bergh, R, Gysemans, C, Beschin, A et al. (2008). Identification of discrete tumor-induced myeloid-derived suppressor cell subpopulations with distinct T cell-suppressive activity. Blood 111: 4233–4244. 25. Peranzoni, E, Zilio, S, Marigo, I, Dolcetti, L, Zanovello, P, Mandruzzato, S et al. (2010). Myeloid-derived suppressor cell heterogeneity and subset definition. Curr Opin Immunol 22: 238–244. 26. Krieg, AM, Wu, T, Weeratna, R, Efler, SM, Love-Homan, L, Yang, L et al. (1998). Sequence motifs in adenoviral DNA block immune activation by stimulatory CpG motifs. Proc Natl Acad Sci USA 95: 12631–12636. 27. Suzuki, M, Cerullo, V, Bertin, TK, Cela, R, Clarke, C, Guenther, M et al. (2010). MyD88-dependent silencing of transgene expression during the innate and adaptive immune response to helper-dependent adenovirus. Hum Gene Ther 21: 325–336.

2085

CpG-enriched Oncolytic Adenovirus

28. Cerullo, V, Seiler, MP, Mane, V, Brunetti-Pierri, N, Clarke, C, Bertin, TK et al. (2007). Toll-like receptor 9 triggers an innate immune response to helper-dependent adenoviral vectors. Mol Ther 15: 378–385. 29. Heise, C, Hermiston, T, Johnson, L, Brooks, G, Sampson-Johannes, A, Williams, A et al. (2000). An adenovirus E1A mutant that demonstrates potent and selective systemic anti-tumoral efficacy. Nat Med 6: 1134–1139. 30. Seiler, MP, Cerullo, V and Lee, B (2007). Immune response to helper dependent adenoviral mediated liver gene therapy: challenges and prospects. Curr Gene Ther 7: 297–305. 31. Rajecki, M, Raki, M, Escutenaire, S, Pesonen, S, Cerullo, V, Helminen, A et al. (2011). Safety of glucocorticoids in cancer patients treated with oncolytic adenoviruses. Mol Pharm 8: 93–103. 32. Pesonen, S, Diaconu, I, Cerullo, V, Ranki, T, Kangasniemi, L, Escutenaire, S et al. (2011). Chimeric oncolytic adenovirus Ad5/3-hTERT-E1A-CD40L for the treatment of advanced solid tumors: preclinical and clinical evaluation. Human Gene Ther 22: A120–A120. 33. Pesonen, S, Diaconu, I, Cerullo, V, Escutenaire, S, Raki, M, Kangasniemi, L et al. (2012). Integrin targeted oncolytic adenoviruses Ad5-D24-RGD and Ad5-RGD-D24GMCSF for treatment of patients with advanced chemotherapy refractory solid tumors. Int J Cancer 130: 1937–1947. 34. Cerullo, V, Diaconu, I, Kangasniemi, L, Rajecki, M, Escutenaire, S, Koski, A et al. (2011). Immunological effects of low-dose cyclophosphamide in cancer patients treated with oncolytic adenovirus. Mol Ther 19: 1737–1746. 35. Nokisalmi, P, Pesonen, S, Escutenaire, S, Sarkioja, M, Raki, M, Cerullo, V et al. (2010). Oncolytic adenovirus ICOVIR-7 in patients with advanced and refractory solid tumors. Clin Cancer Res 16: 3035–3043. 36. Cerullo, V, Pesonen, S, Diaconu, I, Escutenaire, S, Arstila, PT, Ugolini, M et al. (2010). Oncolytic adenovirus coding for granulocyte macrophage colonystimulating factor induces antitumoral immunity in cancer patients. Cancer Res 70: 4297–4309. 37. Vetrini, F, Brunetti-Pierri, N, Palmer, DJ, Bertin, T, Grove, NC, Finegold, MJ et al. (2010). Vasoactive intestinal peptide increases hepatic transduction and reduces innate immune response following administration of helper-dependent Ad. Mol Ther 18: 1339–1345. 38. Brunetti-Pierri, N and Ng, P (2009). Progress towards liver and lung-directed gene therapy with helper-dependent adenoviral vectors. Curr Gene Ther 9: 329–340.

2086

© The American Society of Gene & Cell Therapy

39. Brunetti-Pierri, N, Palmer, DJ, Mane, V, Finegold, M, Beaudet, AL and Ng, P (2005). Increased hepatic transduction with reduced systemic dissemination and proinflammatory cytokines following hydrodynamic injection of helper-dependent adenoviral vectors. Mol Ther 12: 99–106. 40. Edukulla, R, Ramakrishna, E, Woller, N, Mundt, B, Knocke, S, Gürlevik, E et al. (2009). Antitumoral immune response by recruitment and expansion of dendritic cells in tumors infected with telomerase-dependent oncolytic viruses. Cancer Res 69: 1448–1458. 41. Zoglmeier, C, Bauer, H, Norenberg, D, Wedekind, G, Bittner, P, Sandholzer, N et al. (2011). CpG blocks immunosuppression by myeloid-derived suppressor cells in tumor-bearing mice. Clin Cancer Res 17: 1765–1775. 42. Raykov, Z, Grekova, S, Leuchs, B, Aprahamian, M and Rommelaere, J (2008). Arming parvoviruses with CpG motifs to improve their oncosuppressive capacity. Int J Cancer 122: 2880–2884. 43. Suzuki, M, Cela, R, Bertin, TK, Sule, G, Cerullo, V, Rodgers, JR et al. (2011). NOD2 signaling contributes to the innate immune response against helper-dependent adenovirus vectors independently of MyD88 in vivo. Hum Gene Ther 22: 1071–1082. 44. Hartman, ZC, Appledorn, DM, Serra, D, Glass, O, Mendelson, TB, Clay, TM et al. (2008). Replication-attenuated Human Adenoviral Type 4 vectors elicit capsid dependent enhanced innate immune responses that are partially dependent upon interactions with the complement system. Virology 374: 453–467. 45. Pesonen, S, Helin, H, Nokisalmi, P, Escutenaire, S, Ribacka, C, Sarkioja, M et al. (2010). Oncolytic adenovirus treatment of a patient with refractory neuroblastoma. Acta Oncol 49: 117–119. 46. Vaquero, J, Bélanger, M, James, L, Herrero, R, Desjardins, P, Côté, J et al. (2007). Mild hypothermia attenuates liver injury and improves survival in mice with acetaminophen toxicity. Gastroenterology 132: 372–383. 47. Hughes, DJ, Kipar, A, Sample, JT and Stewart, JP (2010). Pathogenesis of a model gammaherpesvirus in a natural host. J Virol 84: 3949–3961. 48. Antoine, DJ, Williams, DP, Kipar, A, Jenkins, RE, Regan, SL, Sathish, JG et al. (2009). High-mobility group box-1 protein and keratin-18, circulating serum proteins informative of acetaminophen-induced necrosis and apoptosis in vivo. Toxicol Sci 112: 521–531. 49. Escutenaire, S, Cerullo, V, Diaconu, I, Ahtiainen, L, Hannuksela, P, Oksanen, M et al. (2011). In vivo and in vitro distribution of type 5 and fiber-modified oncolytic adenoviruses in human blood compartments. Ann Med 43: 151–163.

www.moleculartherapy.org vol. 20 no. 11 nov. 2012