481 analysis of host cell and viral transcriptomes during infection by wild type species c ad6 and species d ad26

481 Analysis of Host Cell and Viral Transcriptomes during Infection by Wild Type Species C Ad6 and Species D Ad26 Molecular Therapy Volume 21, Supplement 1, May 2013 Copyright © The American Society o[.]

Molecular Therapy Volume 21, Supplement 1, May 2013

Copyright © The American Society of Gene & Cell Therapy S185

ADENOVIRUS AND OTHER DNA VIRUS VECTORS

IIIc, n = 10; IVM1a, n = 16; IVM1b, n = 4; IVM1c, n = 20) received a

median of six injection sets; 74% of patients had received one or more

nonsurgical prior therapies for active disease, including dacarbazine/

temozolomide or interleukin-2 (IL-2) The overall initial response

rate by RECIST was 26% (complete response [CR], n = 8; partial

response [PR], n = 5) Remarkably, despite technology being limited

to local-regional IT injection, both local and distal, non-injected,

metastatic sites (including visceral) demonstrated response Overall

survival was previously reported as 58% at 1 year We now report ³

5 year follow up of a subset of all 15 patients managed at MCCRC

Four of these patients were previously identifi ed as having achieved

complete response involving both local regional and metastatic

disease sites including lung and liver Three of these patients remain

alive 2528, 2031, and 1833 days, and one survived 868 days before

disease related mortality No long term adverse events were identifi ed

in all 15 patients including the 3 surviving ³ 5 years Five year Kaplan

Meier survival from time of fi rst treatment of the 15 patients treated at

MCCRC is shown below These results support long term durability

of response without evidence of adverse toxicity

Adenovirus Suppress Tumor Growth in Hamsters

Brittany A Young,1 Karoly Toth,1 Jacqueline F Spencer,1 William

S M Wold.1

1 Molecular Microbiology & Immunology, Saint Louis University

School of Medicine, St Louis, MO.

The Syrian hamster model has been described and used by our

laboratory and others for evaluating anti-tumor effi cacy, toxicity

and biodistribution of oncolytic adenovirus serotype 5 (Ad5)-based

vectors Hamster tumors and tissues are semi-permissive for Ad5

replication, which allows us to use permissive immunocompetent

animals for our studies We have previously reported that the

Ad5-based vector (VRX-007) replicates and suppresses tumor growth

when it is injected directly into subcutaneous tumors formed by

injection of hamster HaK renal cancer cells We also reported that

VRX-007 replication is increased and tumor growth is suppressed

when hamsters are treated with using cyclophosphamide (CP)

Effects that can be attributed to CP (an alkylating agent) include

immunosuppression of the hamsters that results in increased

VRX-007 replication and oncolysis in tumors as well as a chemotherapeutic

effect of CP on the tumor cells We now report that short-term CP

treatment, given for only one week before VRX-007 injection into

tumors, had a similar effect on enhancing the ability of VRX-007

to retard tumor growth as did long-term continuous CP treatment

Further, hamsters treated with short-term CP therapy did not have

detectable VRX-007 in their tumors, in contrast to modest levels

of VRX-007 in tumors of animals treated with CP continuously These results imply that (1) long term immunosuppression by

CP is not required for the CP-mediated enhancement of tumor growth control by VRX-007, and (2) that the CP may be acting as a chemotherapeutic agent in combination with VRX-007 but without increasing VRX-007 replication in tumors To further address these issues, we used a luciferase-expressing VRX-007 vector to view virus location and replication over time in HaK tumors via the IVIS

in vivo imaging system We found that short-term CP treatment, dosed either one week before or one week after tumor infection, had

an additive effect on VRX-007 tumor growth control even though virus replication in tumors was absent after one week Therefore, in this immunocompetent situation, the anti-tumor effects of VRX-007 and CP, alone or in combination, are exerted early after treatment and

do not involve long-term virus replication We have added radiation therapy as a third treatment modality to vector and CP therapies We found that tumor-specifi c irradiation has an additive effect with

VRX-007 therapy and further augments the inhibition of tumor growth, in both CP-treated and non-CP-treated animals, even though radiation does not lead to increased viral replication in tumors when compared

to those treated with virus alone These results further suggest that long term viral replication is not the cause for tumor growth inhibition

in our model This project aims to better understand the mechanism

of action of VRX-007 and possibly enhancing its anti-tumor effi cacy

481 Analysis of Host Cell and Viral Transcriptomes during Infection by Wild-Type Species C Ad6 and Species D Ad26

Mallory A Turner,1,2 Sean E Hofherr,3 Michael A Barry.2

1 Virology and Gene Therapy Program, Mayo Graduate School, Rochester, MN; 2 Department of Medicine, Mayo Clinic, Rochester, MN; 3 Laboratory Medicine, Children’s National Medical Center, Washington, DC.

There are more than 57 human Ad serotypes, which fall into seven species with diverse cell and tissue tropisms: A, B, C, D,

E, F, and G To date, species C Ad5 has dominated the fi eld of Ad based therapeutics However, in recent years limitations to Ad5 have emerged, including pre-existing immunity, blood factor affi nity, and off-target effects To combat these issues, many research groups are exploring the utility of vectors based on other Ad species and serotypes To better understand the differences in Ad serotypes, we compared the transcription and replication programs of of species D Ad26 and species C Ad6 in permissive cells qPCR for viral DNA showed that Ad6 and 26 have similar kinetics of viral DNA synthesis beginning between 6 and 12 hours after cell binding TruSeq mRNA libraries were generated for samples at 6 and 12 hours post infection and 101bp paired-end mRNA-seq was conducted 99.8% of 150 million reads were used to analyze both viral and host mRNAs Viral transcript sequences at 6 hours showed induction of Ad ‘early’ gene transcripts (E1, E2, E3, E4) for both Ad6 and Ad26 However, viral transcriptional output varied between Ad6 and Ad26 at some sites Notably, Ad6 E1A transcripts at 6 hours were about 4-fold higher than E1A transcripts of Ad26 Likewise, E3 total mRNAs varied between the two viruses in regions encoding homologous E3 proteins; fold changes varied between two and four E1B mRNAs were not signifi cantly different between the two viruses Adenoviral transcripts

at 12 hours post infection showed substantial increase in late gene transcripts (L1, L2, L3, L4, and L5) compared to the 6 hour time point

as expected In some cases Ad6 mRNAs exceeded those for Ad26 (L1, L3, L4) Interestingly, several presumptive early genes were not activated during the early phase, but were instead induced with the late genes mRNAs of the E2b region coding for DNA polymerase (pol) and the preterminal protein (pTP) were increased over 300-fold at 12 hours compared to 6 hours for both viruses The increase

Molecular Therapy Volume 21, Supplement 1, May 2013 Copyright © The American Society of Gene & Cell Therapy

S186

ADENOVIRUS AND OTHER DNA VIRUS VECTORS

of E2b products from 6 to 12 hours coincided with induction of

replication for both viruses Within 12 hours of infection, both viruses

commandeered host cell transcription with viral mRNAs constituting

14.66 to 19.6% of total transcripts in the cell Analysis of the effects

of virus infection on cellular mRNAs showed that 819 host genes

were differentially expressed at a level of 1.5 fold or greater (p < 0.05)

in Ad6 infected cells and 782 genes were differentially expressed in

Ad26 infected cells 333 genes were differentially expressed between

Ad6 and Ad26 infected samples These data suggest that species C

and D adenoviruses induce different viral and cellular programs in

permissive cells Understanding differences in the activation of these

viruses in permissive and non-permissive cells will enable better

engineering of vector platforms for translation, but also contributes

to the understanding of basic adenoviral biology

Adenovirus Genes Following Transduction with

a Replication-Incompetent Adenovirus Vector by

Incorporating microRNA-Targeted Sequences into

the 3’-Untranslated Region of the E2A, E4, or pIX

Gene

Kahori Shimizu,1,2 Fuminori Sakurai,1 Shin-ichiro Nakamura,3

Yasuhito Nagamoto,1 Kyoko Tomita,1 Masashi Tachibana,1

Hiroyuki Mizuguchi.1,4,5

1 Graduate School of Pharmaceutical Sciences, Osaka University,

Suita, Japan; 2 Japan Society for the Promotion of Science

Research Fellow, Suita, Japan; 3 Shiga University of Medical

Sciences, Otsu, Japan; 4 National Institute of Biomedical

Innovation, Ibaraki, Japan; 5 The Center for Advanced Medical

Engineering and Informatics, Osaka University, Suita, Japan.

A conventional replication-incompetent adenovirus (Ad) vector

exhibits the leaky expression of Ad genes following transduction,

leading to an induction of cellular immunity against Ad proteins and

Ad protein-induced toxicity In order to suppress the leaky expression

of Ad genes, we developed novel Ad vectors by incorporating four

tandem copies of sequences with perfect complementarity to miR-122a

or miR-142-3p, which exhibits liver-, or spleen-specifi c expression,

respectively, into the 3’-untranslated region (UTR) of the E2A,

E4, or pIX gene (Shimizu K et al., ASGCT annual meeting 2012)

These Ad vectors easily grew to high titers comparable to those of a

conventional Ad vector in normal 293 cells The leaky expression of

these Ad vectors in mouse organs (liver and spleen) was signifi cantly

suppressed by 2- to 100-fold, compared with a conventional Ad vector,

by insertion of the miRNA-targeted sequences Notably, the Ad vector

carrying the miR-122a-targeted sequences into the 3’-UTR of the E4

gene (Ad-E4-122aT) expressed about 100-fold lower levels of all the

Ad early and late genes examined in the liver than a conventional

Ad vector Furthermore, Ad-E4-122aT mediated approximately 2- to

4-fold lower levels of serum alanine aminotransferase (ALT) levels,

compared with a conventional Ad vector However, the numbers

of CTL against the hexon, which is the dominant epitope of Ad

vector, in the splenocytes were not signifi cantly different between

the Ad vectors examined Numbers of T-cells infi ltrated in the liver

following intravenous administration were comparable between the

mice treated with a conventional Ad vector and those treated with

Ad-E4-122aT These results indicate that insertion of

miR-122a-targeted sequences into the E4 gene 3’-UTR did not reduce the

immune responses to Ad proteins In order to examine whether Ad

vector-induced hepatotoxicity via immune-independent pathway

was suppressed by insertion of miR-122a-targeted sequences into

the E4 gene 3’-UTR, Ad vectors were administered to the

immune-incompetent mice (Rag2/IL2Rc knockout mice) The conventional

Ad vector signifi cantly elevated the serum ALT levels following

administration, on the other hand, no increases in the serum ALT

levels were found in Ad-E4-122aT-treated immune-incompetent mice

These results indicate that miR-122a-mediated suppression of the E4 gene expression in the liver signifi cantly reduced the hepatotoxicity

which an Ad vector causes via non-immunological pathway, and that

Ad-E4-122aT would be a valuable vector for safe and effective gene therapies and basic researches

Vectors in DSG2-Transgenic Mice

Maximilian Richter,1 Roma Yumul,1 Hongjie Wang,1 Dirk Nettelbeck,2 Andre Lieber.1

1 Medicine, University of Washington, Seattle, WA; 2 DKFZ, Heidelberg, Germany.

One of the fi rst fi ber-chimeric adenovirus (Ad) vectors generated, contained the tail and long fi ber shaft of Ad5 and the fi ber knob domain of Ad3 (Ad5/3L) This tropism-modifi ed vector has been shown to efficiently transduce cancer cells Oncolytic vectors possessing Ad5/3L capsids have been used in clinical trials by several groups We recently identifi ed human desmoglein 2 (DSG2)

as a new Ad receptor and showed that Ad5/3L uses DSG2 in in vitro transduction studies Because the mouse orthologue of DSG2 cannot function as a receptor for Ad5/3L, biodistribution studies in mice were

so far of limited value We have therefore generated transgenic mice that express human DSG2 in a pattern and at a level comparable to humans Furthermore, we modifi ed a syngeneic epithelial tumor cell line to overexpress human DSG2 (TC1-DSG2) For biodistribution studies we produced vectors with different capsids that expressed

fi refl y luciferase under the control of the CMV promoter Two vectors were based on Ad5 and either contained the long Ad5 fi ber shaft and the Ad3 fi ber knob (Ad5/3L) or the short Ad3 fi ber shaft and knob (Ad5/3S) In vitro, both Ad5/3L and Ad5/3S vectors transduced cells with similar effi cacy in a DSG2-dependent manner For in vivo studies, we used DSG2-transgenic mice with TC1-DSG2 tumors and non-transgenic littermates with TC1 tumors We injected the luciferase expressing Ad5/3 vectors intravenously and studied transgene expression three days later Non-invasive in vivo imaging of luciferase expression revealed strong signals for Ad5/3L in the liver for both, DSG2-transgenic and non-transgenic animals DSG2-independent liver transduction is most likely mediated through the Ad5 hexon-factor X pathway Importantly, transgenic mice, but not DSG2-negative mice showed signals in subcutaneous tumors, indicating DSG2-dependent transduction No specifi c signals were detected

in mice injected with Ad5/3S by in vivo imaging Measurements

of luciferase activity in tissue homogenates corroborated DSG2-independent liver transduction and DSG2-dependent transduction

of tumor In addition we found luciferase expression form Ad5/3L

in the gastro-intestinal tract of DSG2 transgenic mice Luciferase activity in the liver was 2-3 logs lower with Ad5/3S indicating that the Ad5 hexon-factor X pathway is not effi cient for short-shafted

Ad vectors such as Ad5/3S Luciferase immunoreactivity was found

in the small intestine and colon in lamina propria macrophages as well as in intestinal epithelial cells We also attempted to generate a luciferase vector that was entirely based on Ad serotype 3 During virus propagation in HeLa cells, we noticed however undesired rearrangements in the Ad3-luciferase viral genome We are therefore currently re-designing the vector Biodistribution data with the Ad3-luciferase vector will be reported Our fi ndings have implications for assessing the safety profi le of Ad5/3 vectors and reveal insights into determinants of in vivo tropism of these vectors