Gene Therapy (2008) 15, 1067–1071 & 2008 Nature Publishing Group All rights reserved 0969-7128/08 $30.00 www.nature.com/gt SHORT COMMUNICATION Potential genotoxicity from integration sites in CLAD dogs treated successfully with gammaretroviral vector-mediated gene therapy

M Hai1,3, RL Adler1,3, TR Bauer Jr1,3, LM Tuschong1, Y-C Gu1,XWu2 and DD Hickstein1 1Experimental Transplantation and Immunology Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA and 2Laboratory of Molecular Technology, Scientific Applications International Corporation-Frederick, National Cancer Institute-Frederick, Frederick, Maryland, USA

Integration site analysis was performed on six dogs with in hematopoietic stem cells. Integrations clustered around canine leukocyte adhesion deficiency (CLAD) that survived common insertion sites more frequently than random. greater than 1 year after infusion of autologous CD34+ bone Despite potential genotoxicity from RIS, to date there has marrow cells transduced with a gammaretroviral vector been no progression to oligoclonal hematopoiesis and no expressing canine CD18. A total of 387 retroviral insertion evidence that vector integration sites influenced cell survival sites (RIS) were identified in the peripheral blood leukocytes or proliferation. Continued follow-up in disease-specific from the six dogs at 1 year postinfusion. A total of 129 RIS animal models such as CLAD will be required to provide an were identified in CD3+ T-lymphocytes and 102 RIS in accurate estimate of the genotoxicity using gammaretroviral neutrophils from two dogs at 3 years postinfusion. RIS vectors for hematopoietic stem cell gene therapy. occurred preferentially within 30 kb of transcription start Gene Therapy (2008) 15, 1067–1071; doi:10.1038/gt.2008.52; sites, including 40 near oncogenes and 52 near genes active published online 27 March 2008

Keywords: hematopoietic stem cells; transplantation; retrovirus; LAM–PCR; integration site

Although the theoretical risk of retroviral-mediated (SCID-X1) and adenosine deaminase (ADA) in SCID- oncogene activation through insertional mutagenesis ADA.4–6 Initial excitement in the early reports from these had been considered to be low, this risk has been shown trials was subsequently tempered by serious adverse to be significant in recent human gene therapy clinical events which occurred in four of the eight children with trials using gammaretroviral vectors.1–3 The importance therapeutic levels of in vivo transgene expression in the of the individual variables responsible for genotoxicity French SCID-X1 trial2 and in one of 10 children in the UK in hematopoietic stem cell gene therapy trials using SCID-X1 trial.7 The selective advantage conferred by the gammaretroviral vectors remain uncertain. Human transgene was thought to not only contribute to the clinical trials, as well as small and large animal studies, observed clinical success but also to genotoxic clonal using gammaretroviral vectors have implicated the outgrowth.6 vector, transgene and retroviral insertion sites (RIS) in In the recent non-myeloablative gene therapy trial the development of serious adverse events. In each case, for chronic granulomatous disease (CGD), in which the gammaretroviral vector integration itself has been no selective advantage was provided by the CYBB thought to contribute to the genotoxicity observed, but (gp91phox) transgene, two patients developed clonal may not be the only factor responsible for clonal dominance of retrovirally marked cells, accompanied outgrowth. by multiple RV insertions in EVI1-MDS1, PRDM16 and Successful hematopoietic stem cell gene therapy in SETBP1.3 Thus, factors other than the transgene, such as humans using gammaretroviral vectors have generally RIS, appear to play a major role in clonal dominance. relied upon an in vivo selective growth advantage for the Previous animal studies, including mice, dogs and transduced cells provided by a therapeutic transgene, non-human primates, also point to genotoxic risk from such as the interleukin-2 receptor g chain (IL2Rg)in gammaretroviral vectors. Insertional activation of Evi-1 X-linked severe combined immunodeficiency disease by gammaretroviral vectors was first described in mice.8 A study of 14 non-human primates had integrations in Correspondence: Dr DD Hickstein, Experimental Transplantation EVI-1 after gammaretroviral gene transfer, with one and Immunology Branch, National Cancer Institute, National animal developing a granulocytic sarcoma.9 A recent Institutes of Health, 10 Center Drive, MSC1203, Bldg 10-CRC, study examining RIS in dogs marked with a gamma- Rm 3-3142, Bethesda, MD 20892-1203, USA. retroviral vector also revealed a bias of RIS occurring E-mail: [email protected] 10 3These authors contributed equally to this work. in and near proto-oncogenes. Received 21 December 2007; revised 18 February 2008; accepted 18 In the current study, we conducted integration site February 2008; published online 27 March 2008 analysis in six dogs with canine leukocyte adhesion Integration sites in RV vector-treated CLAD dogs M Hai et al 1068 deficiency (CLAD) that had been successfully treated M 6 12 18 24 -C M 6 12 18 -C M 6 12 -C with gammaretroviral vector-mediated ex vivo gene bp bp bp 11 500 500 500 therapy. CLAD is analogous to leukocyte adhesion 400 400 400 deficiency in children, and thus represents a preclinical 300 300 300 large animal model for leukocyte adhesion deficiency. 200 The characteristics of the six treated dogs with CLAD, 200 200 including age at infusion, conditioning regimen and transduced cell dose infused, are shown (Supplementary 100 100 100 Table 1). In this study, the therapeutic CD18 transgene A1 A2 A3 did not confer a strong selective growth advantage to the M 6 12 1824 30 36 -C M 6 12 1824 -C M 6 12 1824 30 -C transduced cells, and a clinically applicable non-myelo- bp bp bp 500 500 ablative conditioning regimen was used to facilitate 500 400 400 400 engraftment of the transduced cells. The percentages of 300 300 300 CD18+ leukocytes in the peripheral blood in the six dogs 200 ranged from 0.7 to 11.2% at 12–36 months follow-up and 200 200 represent therapeutic levels (Supplementary Table 1). To evaluate the diversity of clones contributing to 100 100 100 long-term hematopoiesis after autologous transplanta- B1 B2 C1 tion of CD34+ BM transduced with a gammaretroviral MNL-C MNL-C vector carrying canine CD18, we identified integration bp bp sites by linear amplification-mediated (LAM) PCR from 500 500 all six treated animals (Figure 1). LAM–PCR, performed 400 400 at 6 month intervals using DNA from peripheral blood 300 300 leukocytes demonstrated multiple banding patterns 200 200 indicative of polyclonal hematopoiesis in all six dogs (Figure 1a). Similar polyclonal hematopoiesis was evi- 100 100 dent for both myeloid and lymphoid lineages in dogs B1 B1 C1 and C1 examined at 36 and 31 months postinfusion, Figure 1 Evaluation of polyclonality by linear amplification- respectively (Figure 1b). mediated PCR (LAM-PCR). Genomic DNA was collected from six To confirm the polyclonal nature of the samples, CLAD dogs (A1, A2, A3, B1, B2 and C1) treated by gammaretroviral linker-mediated (LM)–PCR was used to search for vector PG13/Mscv-cCD1811 at designated months postinfusion of potentially over-represented clones or RIS within genes autologous transduced CD34+ cells. DNA was isolated from (a) peripheral blood leukocytes and (b) flow sorted CD3+ lymphocytes and gene regions. When LM–PCR amplicons were and neutrophils using either the Blood and Cell Culture DNA Kit sequenced, the integration sites in all six dogs were (Qiagen Inc., Valencia, CA, USA) or the Wizard Genomic DNA polyclonally derived (complete list of integration sites in purification Kit (Promega Corp., Madison, WI, USA). Cells were Supplementary Table 2) with no predominant clone(s) sorted by immunostaining for T-cells using an antibody to canine observed. We obtained a total of 618 RIS from the treated CD3 (CA17.2A12; AbD Serotec, Raleigh, NC, USA) conjugated to CLAD dogs by sequence analysis of LM–PCR amplicons: AlexaFluor 647 by using a Zenon labelling kit, and for neutrophils using an antibody to canine neutrophils (CADO48A; VMRD Inc., 387 RIS from peripheral blood leukocytes from all six Pullman, WA, USA) conjugated to phycoerythrin. Labelled cells dogs at 5–21 months postinfusion (36 to 88 RIS per dog), were sorted to purities of 91.6 and 87.2% for CD3+ cells in dogs B1 129 RIS from sorted CD3+ lymphocytes and 102 RIS from and C1, respectively, with o0.5% neutrophil contamination. sorted neutrophils from dogs B1 and C1 at 31–36 months Neutrophils were sorted to purities of 97.8 and 93.5% for B1 and postinfusion. We identified 535 unique RIS after adjust- C1, respectively, with 0.1% CD3+ cell contamination. Cells were ment of 83 RIS that overlapped between the cell types sorted using a FACSVantage SE with DiVa software (BD Bios- ciences, San Jose, CA, USA) equipped with 488 and 633 nm and time points examined (see Supplementary Table 2). excitation lasers. LAM–PCR was performed on extracted DNA, as RIS isolated separately from lymphocytes and neutro- previously described.11 LAM–PCR samples were electrophoresed phils confirmed polyclonal insertion sites in each lineage. on a high-resolution Spreadex gel (Elchrom Scientific AG, Cham, In addition, 28 RIS were shared between the sorted Switzerland) and photographed using a gel documentation system lymphocytes and neutrophils RIS, indicating that a (Biochemi system, UVP, Upland, CA, USA). Bands represent common hematopoietic precursor was transduced. The insertion sites. M indicates size markers; bp, base pairs; ÀC, negative water control; N, neutrophils; L, lymphocytes. positions corresponding to human sequences listed in the Reference Sequence database (RefSeq) that mapped to the dog genome were examined for RIS that occurred within 30 kilobases (kb) upstream or downstream of the from a random dataset: 39 CIS with two insertions within RefSeq genes. Comparison of the RIS to a computer- 30 kb (expected: 3.7 CIS), 13 CIS with three insertions generated dataset of 1200 random positions in the dog within 50 kb (expected: 0.035 CIS) and six CIS with four genome revealed a modest propensity of RIS to be insertions within 100 kb (expected: 0.001 CIS) (Supple- located within the RefSeq genes (43.4%), although this mentary Table 3). Several genes were found near CIS, was not significant compared to the random dataset including PTP4A2, ABHD1, TCF23, NID1, NACA and (39.3%). HSD17B6. Overall, 71 RIS (13.3%) of the 535 unique RIS Further examination of the RIS showed a tendency to were found within CIS, similar to the percentage occur at common insertion sites (CIS).12 CIS may observed by other investigators with gammaretroviral represent either insertional hotspots or loci that confer vectors.1,7 some selective advantage. A large number of CIS were RIS located near oncogenes were determined by present among the RIS, when compared to that expected comparison to a list of human cancer genes compiled

Gene Therapy Integration sites in RV vector-treated CLAD dogs M Hai et al 1069 from the University of New South Wales (UNSW) To examine the propensity of the retrovirus to Embryology DNA-Tumor Suppressor and Oncogene integrate near genes active in early hematopoiesis, we Database (http://embryology.med.unsw.edu.au/DNA/ compared the genes near RIS to a database of genes DNA10.htm) and from the Cancer Gene Consensus expressed in HSCs.14 We found 52 RIS (9.7%; expected: database at the Wellcome Trust Sanger Institute (http:// 4.7%; Po0.001) located near 47 HSC genes, indicating www.sanger.ac.uk/genetics/CGP/Census/). Forty RIS preferential retroviral integration (Figure 2). This in- (7.5%, expected: 3.2%, Po0.001) were located in or creased propensity of RIS near HSC genes was not within 30 kb of 32 oncogenes (Figure 2). None of the unexpected given that canine CD34+ cells were used and RIS were near the genes implicated in clonal proliferation several HSC genes were likely activated by the G-CSF, in the gene therapy trial for CGD (EVI1-MDS1, PRDM16 SCF and Flt-3 Ligand growth factors used. and SETBP1). Although no RIS in our study were present The non-myeloablative conditioning regimen used in near the major gene implicated in leukemia in several both the CGD trial and our study, in conjunction with the SCID-X1 children treated by gene therapy, LMO2,2 fact that both diseases involve a transgene without a several of the RIS were near other prominent cancer- strong selective advantage, prompted us to compare our related genes such as ABL1, BCL2, ETV6, FLI1 and NF1, RIS profile with that of the CGD trial. Genes within 30 kb including some that were also associated with a CIS of RIS in treated CLAD dogs were compared to the genes (Supplementary Table 3). within 30 kb of RIS found in the two CGD gene therapy

Cancer-related genes Genes expressed in hematopoeitic stem cells RIS - Exp: 7.5%, Rnd: 3.2%, P < 0.001 RIS - Exp: 9.7%, Rnd: 4.7%, P < 0.001

ABL1 FLI1 NCOA2 ABCF2 EGLN3 HNRPAB RFFL ATF1 FLT1 NF1 ABL1 EPS8 IDH2 RPS25 BCL2 GATA1 NF2 ACAD10 ERCC1 IMPA2 SNRPB2 CBFB GLI2 RHOH ARL3 ETV6 LIG3 SUFU CCND2 GNAS SEPT9 ATP5G2 FLNB LRRC33 SYN2 CDH1 HCK SH3GL1 BCL2 FOSB MEIS1 SYT11 CDK6 ITK SPECC1 CASP8 GBP3 MMP12 TCF7 CIITA LYN SUFU CBFB GCAT NR4A1 TEK CTTN MAF TEC CCL27 GNAZ PAK1 TMSB10 EPHA1 MLLT6 VAV1 CCND2 GPR56 PRKAG2 WEE1 ETV6 NACA CD34 GSTA3 PRKCE ZBTB20 CHN2 GSTK1 RBX1

Cancer- HSC related 6

32 0 47 77

59 CGD

AMICA1 CIITA FLI1 LOC196264 MLL4 PHF15 SH2D3A ZBTB32 ARHGAP25 CMIP GPR108 LOC283551 MLLT6 PPM1L SH2D6 ZCD2 C19orf55 COMMD7 GPR56 LOC493856 NF1 PRKCE TMEM149 ZNF143 C3 EGLN3 GRM5 LRP2 OMG PSENEN TRIP10 ZNF382 CALN1 EPS8 HSPB6 LYK5 OVGP1 RAB37 U2AF1L4 ZNF710 CD300C EVI2A IDH2 LYN PHACTR1 RFFL UBAC2 CD300LB EVI2B JARID2 MAF PHF15 RFTN1 UBE2D3 CD300LF F25965 LIN37 MAML3 PPM1L SEPT9 ZBTB20 Genes found near RIS in CGD patients RIS -Exp: 9.5%, Rnd: 3.9%, P < 0.001 Figure 2 Genes identified near retroviral insertion sites (RIS). LM–PCR was performed as previously described,13 with slight modifications (see Supplementary methods) and amplicons were sequenced to identify the precise viral integration sites within the dog genome. Genes within 30 kb of each RIS were determined (Supplementary Table 2) and further characterized as either cancer-related, expressed in HSCs, or related to genes found near RIS in the CGD clinical trial and are listed as three different sets. The numbers in the outer circle of the VENN diagram indicated the number of genes identified for each data set. The numbers of genes that overlap between the three gene sets are represented in the common area of the VENN diagram. The percentage of integrations expected within genes is based on random RIS analysis (Rnd) generated in silico (see Supplementary methods). The experimental percentage of RIS close to genes in each of these sets is shown (Exp) with the corresponding statistical significance (P-value).

Gene Therapy Integration sites in RV vector-treated CLAD dogs M Hai et al 1070 35% RIS_All_dogs 30% Random 25% 20% 15% 10% % of total genes 5% 0%

GTP binding GTPase activity Protein binding Immune response Response to stress

Guanyl nucleotide binding Response to biotic stimulus

35% RIS_Early 30% RIS_Late 25% Random * 20% 15% 10% % of total genes * * 5% * 0% e II

Signal transduction Response to stress RNA polymeras transcription factor activity Response to biotic stimulus

35% RIS_Lymphs 30% RIS_Neuts 25% Random 20% 15% 10% * *

% of total genes ** 5% * * * *

0%

Response to Response to Response to Response to biotic stimulus other organismspest, pathogen, stress Ubiquitin cycle Defense responseImmune response or parasite RNA polymerase II transcription factor activity

Figure 3 Ontological classes of genes within 30 kb of retroviral insertion sites (RIS). RIS from (a) pooled, unique RIS, (b) RIS from early (p21 months posttransplantation) and late (>21 months posttransplantation), or (c) sorted neutrophils and CD3+ lymphocytes were compared to an averaged random dataset (see Supplementary methods). Percentages represent the number of genes for each particular gene class divided by the total number of genes found in or near insertion sites. The asterisk (*) indicates those classes over-represented using a modified one- tailed Fisher’s Exact test.

patients (Figure 2).3 There was a significant overlap between the two studies were also oncogenes (CIITA, between the RIS profiles in both studies: 51 RIS in our FLI1, LYN, MAF, MLLT6, NF1 and SEPT9). study (9.5%) were close to 61 genes associated with RIS To determine whether RIS occurred close to genes in the CGD trial (expected: 3.9%, Po0.001). It is with particular functions that may influence in vivo noteworthy that seven of these 51 (13.7%) genes shared selection, we used gene ontology classification to identify

Gene Therapy Integration sites in RV vector-treated CLAD dogs M Hai et al 1071 gene families among genes within 30 kb of the RIS. We Orlando, FL, December 9, 2006, and at the 9th and 10th observed over-representation (Po0.05) of seven cate- annual meetings of the American Society of Gene gories when compared to the expected distribution, Therapy, Baltimore, MD, May 31, 2006 and Seattle, WA, however, none of these categories involved cell prolif- May 30, 2007. eration or differentiation (Figure 3a). Similarly, a com- parison of gene families at either early (p21 months) or late (421 months) time points (Figure 3b), or in either References lymphocytes or neutrophils (Figure 3c) in dogs B1 and C1, revealed over-represented families, but none 1 Deichmann A, Hacein-Bey-Abina S, Schmidt M, Garrigue A, appeared to be related to cell proliferation or survival. Brugman MH, Hu J et al. Vector integration is nonrandom and Despite the increased frequency of retroviral inser- clustered and influences the fate of lymphopoiesis in SCID-X1 tions near oncogenes in both the CGD and CLAD gene therapy. J Clin Invest 2007; 117: 2225–2232. studies, to date the animals in this CLAD study show 2 Hacein-Bey-Abina S, Von Kalle C, Schmidt M, McCormack MP, no evidence of clonal expansion up to 3 years post- Wulffraat N, Leboulch P et al. LMO2-associated clonal T cell treatment, a time period representing approximately 25 proliferation in two patients after gene therapy for SCID-X1. percent of the canine lifespan. In comparison, clonal Science 2003; 302: 415–419. expansion in two CGD patients occurred relatively 3 Ott MG, Schmidt M, Schwarzwaelder K, Stein S, Siler U, Koehl U rapidly, with clonal dominance occurring within 5 et al. Correction of X-linked chronic granulomatous disease by months after gene therapy. In the SCID-X1 clinical trials, gene therapy, augmented by insertional activation of MDS1- clonal expansion led to leukemia in five patients at 24–68 EVI1, PRDM16 or SETBP1. Nat Med 2006; 12: 401–409. months postinfusion. In the CLAD studies, it would be 4 Aiuti A, Slavin S, Aker M, Ficara F, Deola S, Mortellaro A et al. expected that leukemia or lymphoproliferation would Correction of ADA-SCID by stem cell gene therapy combined with nonmyeloablative conditioning. Science 2002; 296: occur much earlier because dogs have a faster HSC 2410–2413. replication rate (assuming greater similarity to cats) than 15 5 Gaspar HB, Bjorkegren E, Parsley K, Gilmour KC, King D, humans. Sinclair J et al. Successful reconstitution of immunity in ADA- The difference in outcomes between the CGD trial and SCID by stem cell gene therapy following cessation of PEG-ADA our study may have been influenced by several factors, and use of mild preconditioning. Mol Ther 2006; 14: 505–513. including vector design (the highly active spleen focus 6 Hacein-Bey-Abina S, Le Deist F, Carlier F, Bouneaud C, Hue C, forming virus LTR used in the CGD trial versus the De Villartay JP et al. Sustained correction of X-linked severe murine stem cell virus LTR used in this study), the combined immunodeficiency by ex vivo gene therapy. N Engl J source of target cells (G-CSF mobilized peripheral blood Med 2002; 346: 1185–1193. CD34+ cells used in the CGD trial versus bone marrow 7 Schwarzwaelder K, Howe SJ, Schmidt M, Brugman MH, CD34+ cells in this study), and length of time in culture Deichmann A, Glimm H et al. Gammaretrovirus-mediated (5 days in the CGD trial versus 3 days in our trial). There correction of SCID-X1 is associated with skewed vector integra- was also a higher transduction efficiency with subse- tion site distribution in vivo. J Clin Invest 2007; 117: 2241–2249. quent increased vector MOI levels in the cells infused 8 Li Z, Dullmann J, Schiedlmeier B, Schmidt M, von Kalle C, into the CGD patients. However, the multiple insertions Meyer J et al. Murine leukemia induced by retroviral gene near or within oncogenes that we identified raises marking. Science 2002; 296: 497. concerns that clonal dominance remains a distinct 9 Seggewiss R, Pittaluga S, Adler RL, Guenaga FJ, Ferguson C, Pilz possibility, and that long-term follow-up of these animals IH et al. Acute myeloid leukemia is associated with retroviral will be necessary to accurately estimate the risk from the gene transfer to hematopoietic progenitor cells in a rhesus 107 gammaretroviral vector used in this study. macaque. Blood 2006; : 3865–3867. 10 Beard BC, Keyser KA, Trobridge GD, Peterson LJ, Miller DG, These results also emphasize the value of disease- Jacobs M et al. Unique integration profiles in a canine model of specific, pre-clinical animal models in assessing risks long-term repopulating cells transduced with gammaretrovirus, from the complex interaction of vector, transgene, lentivirus, or foamy virus. Hum Gene Ther 2007; 18: 423–434. insertion site and disease in gene therapy studies. In 11 Bauer Jr TR, Hai M, Tuschong LM, Burkholder TH, Gu Y-C, particular, the CLAD model enables one to identify Sokolic RA et al. Correction of the disease phenotype in canine potential interactions between the transgene and the leukocyte adhesion deficiency using ex-vivo hematopoietic stem integration site, such as the one that occurred in the cell gene therapy. Blood 2006; 108: 1767–1769. SCID-X1 trial.2 12 Suzuki T, Shen H, Akagi K, Morse HC, Malley JD, Naiman DQ et al. New genes involved in cancer identified by retroviral tagging. Nat Genet 2002; 32: 166–174. Acknowledgements 13 Wu X, Li Y, Crise B, Burgess SM. Transcription start regions in the human genome are favored targets for MLV integration. We thank Dr William Telford and Veena Kapoor for Science 2003; 300: 1749–1751. assistance with flow cytometry. This research was 14 Ivanova NB, Dimos JT, Schaniel C, Hackney JA, Moore KA, supported by the Intramural Research Program of the Lemischka IR. A stem cell molecular signature. Science 2002; 298: NIH, National Cancer Institute, Center for Cancer 601–604. Research. Presented in part in abstract form at the 47th 15 Shepherd BE, Guttorp P, Lansdorp PM, Abkowitz JL. Estimating and 48th annual meetings of the American Society of human hematopoietic stem cell kinetics using granulocyte Hematology, Atlanta, GA, December 12, 2005 and telomere lengths. Exp Hematol 2004; 32: 1040–1050.

Supplementary Information accompanies the paper on Gene Therapy website (http://www.nature.com/gt)

Gene Therapy