664 Identification of TLR2 as a Mediator of the Innate Immune Response to Adenoviral Vectors 662 Study of Adenovirus Vector Induced Activation of Muscle Dendritic Cells and Analysis of Mouse Dendritic[.]
Trang 1662 Study of Adenovirus Vector-Induced
Activation of Muscle Dendritic Cells and Analysis
of Mouse Dendritic Cell Migration from the
Tibialis Anterior Muscle
Lei Wang, PaulaR Clemens
'Neurology University ofPittsburgh, Pittsburgh , I'll;
Weurol-ogy Service, Department ofVeterans Affairs Medical Center;
Pittsburgh, PA.
Dendritic cells (DCs),known to be increased in skeletal muscle
undergoing dystrophic change, are the most potent professional
APCs that initiate the adaptive immune response to Ad v
ector-mediated gene delivery Therefore, these cells may be an excellent
target to mediate immune suppression to facilitate successful gene
transfer Toward that end,we have studied the localization and
traf-ficking ofDCs in the setting ofAd vector-mediated gene transferto
skeletal muscle First, we explored theAd vector-induced activation
of muscle DCs The tibialis anterior muscle (TA) of C57BLllO
mice was injected with PBS or a first-generation AdEOFPvector
Immunohistochemistry revealed that the majorityofactivated DCs
(CDI l c" CD86+)were observed in vector-infected muscle at 7
days after gene transfer,when EOFP expression reached its highest
level CDI lc" CD86+ DCs were also found at the margin of white
and red pulp in spleen,and in the paracortical area of ILNs on the
side of injection.PBS-injected mice did not show activated DCs
in TA muscle, LNs or spleen Although migration ofDCs through
lymphatics to draining LNs is an important step in the initiation of
an adaptive immune response, very little is known about how the
muscle DCs migrate from muscle treated with gene transfer To study
DC trafficking we adoptively transferred BM-derived mature DCs
labeled with CFSE ex vivo to syngeneic recipient mice, We
quanti-fied CFSWDCs in PLNs,ILNs and spleen by flow cytometryalter
footpad (FP) orTA injection On day 2 after adoptive transfer to Fp,
most recoveredCFSWDCs were found in PLNs, with fewer found
in ILNs There were very few recovered DCs found in spleen on
day I or 2 This study confirmed the results of others showing that
CFSE-Iabeled DCs migrated to draining LNs after FP injection,and
a vel)'small fraction ofsuch cells further migrated to spleen through
the thoracic duct, In contrast, after adoptive transfer ofCFSWDCs to
TA,a few DCs appeared in the spleen within 12 h.Over the next 36
hours, the number decreased and became undetectable On day I,we
recovered the largest number ofCFSE+DCsfrom PLNs, with fewer
from ILNs.The numbers were reduced in PLNs and ILNs after 2
days,and were undetectable thereafter.Interestingly, our study found
that the peak number of CFSWDCs recovered in spleen occurred
at 12h,not 24h afterTA injection when the largest number of these
cells appeared in the PLNs This finding suggests that a subfraction
of the transferred DCs from the muscle injection site could travel
through the bloodstream directly to spleen perhaps gaining access
through the well-developed capillary network of skeletal muscle
Subsequently,the largest fraction oftransferred DCs trafficked to the
PLNs and ILNs, likely through lymphatic vessels.Previous studies
have shown that DCs migrating to the spleen and draining LNs is
required to induce humoral immunity and primary immune response
respectively Therefore, our study yielded important evidence
impli-cating DCs in the deleterious immunological consequences induced
by Ad-mediated gene transfer to skeletal muscle
S254
663 Methotrexate Selection of Gene Modified
T Cells with an Engineered Human Dihydrofolate Reductase (DHFR)
Christine E Brown,' Wen-Chung Chang,' Christine Wright,1
Araceli Naranjo,' Jamie Wagner,IHunsarB.Meechoovet,' Julie
R Ostberg,'Michael C Jensen.'
'Division a/Cancer lmmunotherapeutics & Tumor
Immunol-ogy, Beckman Research Institute, City ofHope National Medical Center, Duarte, CA
Endowing T-cells with resistance to Iymphotoxic drugs by their genetic modification affords the opportunity forin vitroandin vivo
selection The clinical usc of bacterial drug resistance genes such
as neomycin and hygromycin phophotransferases is problematic due to the immunogcnicity of these transgcnes,as well as, the incompatibility of 0418 and hygromycin tor direct administration
to humans forin vivoselection Here,we report on the utility of a human dihydrofolate reductase double mutant (DHFRdm; Leu22 -7 Phe22, Phe31 -7Ser31) transgene for conferring resistance of hu-man 'l-cclls to the drug methotrexate (MTX) Studies using Jurkat
T cells,peripheral blood OKT3-stimulated 'l-cells and established human T-celilincs demonstrated their sensitivity to the lympholytic effects of MTX at concentrations of 0.05 JlM or more.The stable expression in Jurkat T cells with DHFR constructs containing either both mutations or each individual mutation alone revealed that the DHFRdm was the most efficient at conferring MTX resistance Using a OFP-DHFRdm fusion construct introduced into PBMC derived T cells by c1ectrotransfer, we were able to derive OFP' stable integrants in the presence of cyctocidal concentrations of MTX (0.1 JlM) These MTX-resistant T cells retained dependency
on antigen receptor/gamma-c cytokines for survivaland proliferation and displayed T cell phenotypic markers similar to their DHFRdm' counterparts We next evaluated the utilityofthe selection transgene
to enforce the expression ofa secondtherapeutic transgene, in this case a zetakine chimeric antigen receptor,by fusion using a 2A cleavable linker Electroporation of a DHFRdm-2A-IL 13zetakine construct into T cells resulted in MTX resistance and coordinated zetakine expression; these CTLs killed glioma target cells (express-ing the IL 13zetakine ligand, IL 13R(2), as determined us(express-ingin vitro
chromium release assays These data demonstrate the utility of this human enzyme transgene selection system for human T cell genetic modification Our studies are now focusing on the potential to select for transgene expressing MTX-resistant cells in vivo.
664 Identification of TLR2 as a Mediator of the Innate Immune Response to Adenoviral Vectors Daniel M.Applcdorn,IJeannine M Scott,'Andrea Amalfitano.'
'Microbiology and Molecular Genetics , Mlchlgan State Univer-sity , East Lansing, MI
The interaction between adenovirus (Ad) and pathogen recogni-tion receptors (PRR) of the toll-like receptor family (TLR) has recently received increased attention TLR9,found predominantly
in endosomes, recognizes viral CpO dsDNA that,in the presence
of an agonist, leads to increased NF-KB activation,and increased plasma levels of Type I interferons,IL-6 and IL- I2 Recent reports have identified TLR9 as a PRR activated following high titer Ad injection,and it has been shown that the production of type I IFNs was mediated by TLR9 in a MyD88 dependent manner in plasma-cytoid dendritic cells (pDCs), but not conventional DCs (cDCs) However,the induction of numerous additional pro-inflammatory cytokines and chcmokincs,i.e IL-Ia,RANTES and type II IFN-y among others,were not addressed in those studies Furthermore, those studies observed a TLR9 independent induction ofmonocyte chemoattractant protein-I (MCP-I) suggesting the specific anti-Ad innate immune response is multi-faceted Because of this,we
hy-Molecul ar Therapy Yofume 15 S upplement I, \b y 2007
Co pyright © '111C Am erican Soc iety o f G ene TI ICr.lpr
Trang 2pothesized an additional PRR was required to initiate inflammatory
signals in conjunction with TLR9 in the context ofAd Studies
utiliz-ing herpes-simplex viruses,Trypanosoma cruzi ,andMycobacterium
tuberculosis,have reported cooperation between TLR2 and TLR9 in
controlling cytokine and chemokine responses to these pathogens
In conjunction with TLRs I and 6,known agonists ofTLR2 include,
but arc not limited to lipoprotein,modulin, porin and peptidoglycan
To determine ifTLR2 was partly responsible for these anti-Ad
in-nate immune responses, we investigated the cytokine/chemokine
response,and thrombocytopenia, following high titer adenovirus
injection into TLR2 deficient mice We found that like TLR9, TLR2
is also involved in the production ofIL-12 We also found that Ad
induction ofIL-Ia, IFN-y,MCP-I and RANTES were in part, TLR2
dependent Although we saw reduced levels of IL-6 in Ad treated
TLR2 deficient mice,these results were not statistically significant
It is well established that Ad injection induces thrombocytopenia in
a complement dependent fashion,and recently we observed that this
may be a response that is both mediated by the TLR adaptor TRI F
and MyD88 at relatively lower viral particle dosages However, Ad
injection into TLR2-KO mice still elicited thrombocytopenia Based
upon these findings,and those of others,we propose a model where
TLR2 and TLR9 synergistically cooperate to detect both viral capsid
and viral DNA, and subsequently induce the immune responses
noted shortly after intravenous Ad delivery,
665 Host Response to Adenovirus,
Helper-Dependent Adenovirus and Adeno-Associated
Virus in Mouse Liver
Paul Fawcett,' Hiroyuki Nakai,'Anja Ehrhardt,'Mark A.Kay ,~
Anton McCamey.s
'Internal Medicine, Virginia Commonwealth University,
Rich-mond , I~; lMoleclllar Genetics and Biochemistry; University of
Pennsylvania, Pittsburgh, PA; JVirology
Ludwig-Maximilians-University Munich, Munich, Bavaria, Germany; "Pediatrics and
Genetics, Stanford University, Stanford, CA; Sinternal Medicine,
University ofIowa , Iowa City , IA
Understanding host responses to viral gene therapy vectors is
necessary for the development of safe and efficaciousin vivogene
transfer agents We describe the usc of high-density spotted cDNA
mieroarrays to monitor thein vivohost transcriptional responses in
mouse liver upon administration ofeither a "first-generation"
adeno-virus vector (Ad),a helper-dependent "gutless" adenovirus vector
(HD),or an adeno-associated virus vector (AAV) containing
identi-cal human factor IX expression cassettes (hFlX) These
replication-defective vectors allowed us to assess the contribution of the viral
capsid and leaky viral gene expression to the host response in whole
animals,since HD and MV vectors additionally contain no viral
genes While all three vectors induced characteristic
temporally-se-quenced programs ofgene expression, the gene expression programs
induced by the Ad and HD adenovirus vectors were largely similar,
including the induction ofa prominent interferon-dependent cluster
within 6 hours of administration Differentially expressed genes
were validated by quantitative PCR Our results show that the viral
capsid is the main driver of the host transcriptional response Upon
infection with Ad and HD,we observe modulation ofmembers ofthe
AP-I signaling pathway, innate immune pathways, chemokines and
cytokines,immune-dampening factors as well as proteins involved
in antigen presentation We observe paraerine signaling events
that only occur after cross talk between different cell types.This
highlights the importance of virus-host interaction studies in whole
mammals In contrast to infection with Ad and HD, theAAV-based
vector caused far fewer alterations ofhost-gene expression, Notably,
AAV infection induced significantly fewer genes that arc involved
in immunity and defense, signal transduction as well as protein
metabolism and modification.AAV infection caused a decrease in a
Molecular Therapy Volume15.Supple ment I , M ay 2007
C opyright © T he American Soc ie ty o f Gen e Therap y
number ofAP-I components,again in contrast to Ad and HD AAV also altered the expression ofa number of'immunornodulatory genes Taken as a whole, the host response to Ad and HD was remarkably similar suggesting that the initial innate response is driven largely
by the viral capsid The host response to AAV, in comparison, is blunted Though they share some common features,the response
to AAV differs significantly from that to the Ad and HD We have identified a number of novel genes involved in the host response
to Ad,HD and AAV These virus-host interaction studies in mice should complement previous studies in cultured cells
666 Effect of Anti-Adenovirallmmunity on Long-Term Vector and Islet Graft Function Jun Cheng,' Jianmin Sun,' Randall S Sung.'
'Transplant Division, General Surgery Section, Department of Surgery University ofMichigan Health System, Ann Arbon MI.
Purpose: Adenoviral gene transfer is a potential ex vivogene therapy for islet transplantation However, the inflammatory and immune responses induced by adenovirus may be detrimental to islet survival and transgene expression in transduced islets We investigated the effect of immune response to adenovirus on long-term vector and islet graft function Methods: C57BLl6 (n=17 per group) mice were used for syngeneic islet transplantation Islets were isolated and transduced with AdCMVntp-galaetosidase (Ad, type 5 adenovirus, EI and E3-deleted) at a multiplicity of infec-tion of I x \Os pfu/islet Islets transduced with Ad or untransduced were implanted under the kidney capsule ofstreptozotocin-induced diabetic syngeneic recipients Blood glucose and glucose tolerance were utilized to assess to graft function All grafts were collected at intervals after transplantation Graft inflammation was estimated by H&E and immunohistochemical staining Intragraft transgene and cytokine/chemokine expressions were assayed by X-gal staining and quantitative real time PCR,respectively ELISPOT and ELISA were used to measure systemic cytokines and anti-adenoviral antibodies (IgG I and IgG2a), respectively Results: Diabetes was reversed in one day in recipients ofa marginal mass ofuntransduced islets, while Ad-transduced islets failed to reverse diabetes until 10 days post-transplant (pS:O.048).A profound and progressive infiltration with CD4+,CD8+lymphocytes and CD II b"maerophages was developed
in adenovirus transduced grafts from 10 to 25 days after transplanta-tion and paralleled a decline in p-galactosidase expression However, transgene expression still persisted in transduced islet grafts over 3 months Glucose tolerance was impaired in transduced grafts at 25 days (p=O.008) (Fig A), but restored at 3 months (p'=O.275)(Fig B) post-transplant Levels of intragraft intcrfcron-y(IFN-y) and CC chemokine ligand 5 (CCLS) mRNA expression were not different between transduced and untransduced islet grath at 5 and 25 days aftertransplantation, despite the presence ofincreased expression of
CC chemokine receptor 5 (CCR5) at 5 days (pS:O.029) No significant splenocyte IFN-yand interleukin-4 (IL-4) production or serum anti-adenoviral antibody was detected in recipients of transduced islet grafts Conclusions: Although transduction of islets with adenoviral vectors results in overt local inflammationof islet grafts and impairs early graft function, long-term(3 months) graft function is not sig-nificantly impaired, and transgene expression persists in syngeneic islet transplant Systemic anti-adenoviral immunity is minimal Thus,
8255