From www.bloodjournal.org by guest on February 6, 2015. For personal use only. Retroviral Vector Design for Long-Term Expression in Murine Hematopoietic Cells In Vivo By Pamela H. Correll, Susan Colilla, and Stefan Karlsson A series of retroviral vectors containing the human glucocerebrosidase (GC) cDNA driven by various promoters have been constructed in an attempt t o discover which vector design can most efficiently transduce murine hematopoietic stem cells (HSCs) and drive expression of the transferred gene in hematopoietic cells of mice reconstituted with the transduced stem cells. The simplest vector, LG, in which the GC gene is driven by the viral LTR, was the most efficient vector at infecting HSCs, with an average viral copy number in hematopoietic tissues of 3 copies/cell in recipient mice. In general, the viral vectors that contained any additional promoters or enhancers t o drive expression of either the GC gene or a selectable marker gene (NeoR)had lower titers and/or transduced HSCs at a lower efficiency. This was seen most markedly when the human phosphoglycerate (PGK) promoter was used t o drive the human GC cDNA. Despite repeated attempts t o obtain a high titerproducer clone, this virus consistently produced low titersand subsequently resulted in the lowest proviral copy numbers in long-term reconstituted mice. Only the viral LTR and the PGK promoter were capable of drivingsignificant levels of human GC RNA in hematopoietic cells of long-term reconstituted mice, with a much lower level of RNA generated by an internal herpes TK or SV40 immediate early promoter. Insertionof the internal transcription unit in the opposite orientation relative t o the viral LTRs had a detrimental effect on gene expression. The levels of RNA generated by a hybrid LTR containing the myeloproliferative sarcoma virus enhancer were higher in bone marrow-derived macrophages than in nonadherent cells of the bone marrow when compared with the LG vector. The presence of an internal promotert o drive expression of the human GC cDNA did notseem t o have a detrimental effect on expression levels from theviral LTR. In fact, in the presence of an internal TK or PGK promoter expression from the LTR was increased despite the presence of lower proviral copy numbers. Insertion of a second gene (NeoR)into the vector had a negative impact on long-term expression in hematopoietic cells in vivo; however, this seems t o be due solely t o the lower transduction efficiency of this vector. Overall, the highest levels of GC activity in macrophages of long-term reconstituted mice were generated by the LG vector; however, these levels were variable. Vectors containing an internal SV40,TK, or PGK promoter produced consistent levels of GC activity in these cells, but because of the lowertransduction efficiency obtained in the presence of these promoters, simple vectors containing a single gene driven bythe viral LTR currently remainthe mostpromising viruses for gene therapy of human hematopoieticdisorders. This is a US government work. There are no restrictions on its use. R platelets, all mature hematopoietic cells derived from these pluripotent stem cells, are replaced continuously throughout life. Transfer of the appropriate gene into HSCs could be used for treatment of any disorder that affects one or more of these cell lineages. Retroviral vectors have been used successfully to transfer a number of genes into murine HSCs; however, persistent, long-term expression in the progeny of these cells in vivo has been historically difficult to obtain.' Vectors that produce high levels of expression in immortalized fibroblasts or hematopoietic cell lines do not necessarily result in expression in differentiated progeny of infected HSCs in v ~ v o .It~has .~ beenunclearwhether the low levels of expression in the majority of these studies were due to insufficient transcription or low transduction of HSCs. In addition, it is impossible to directly compare the results from these studies because of differences in methods of infection, transduction efficiency, and vector design (genes being transferred, regulatory elements present, direction of transcription, and vector backbone used). Inan attempt to determine what vector design would be mostusefulto drive expression of a transferred gene in hematopoietic cells, a series of retroviral vectors containing the human glucocerebrosidase (GC) cDNA under the transcriptional regulation of a variety of promoters and enhancers has been constructed. GC deficiency, commonly known as Gaucher's d i ~ e a s eis, ~a leading candidate for HSC gene therapy. It is an autosomal recessive lysosomal storage disorder characterized by an accumulation of GC in macrophages of the BM, spleen, liver, lung, and brain.6 We previously transferred the human GC gene into murine hemato- ETROVIRAL VECTORS have been used for the last decade to transfer genes into mammalian cells for the purpose of developing gene therapy for inherited disorders, cancer, and acquired immunodeficiency syndrome (AIDS). The majority of research to date has focused on bone marrow (BM) as a target tissue for retroviral gene therapy.' Hematopoietic stem cells (HSCs) are able to persist throughout life by undergoing proliferation to produce daughter stem cells (self-renewal) and are also able to differentiate to form all cells of the hematopoietic system. Erythrocytes, lymphocytes (B and T), granulocytes, monocytes/macrophages, and From the Molecular and Medical Genetics Section, Developmental and Metabolic Neurology Branch, National lnstitute of Neurological Disease and Stroke, National Institutes of Health, Bethesda, MD. Submitted October 25, 1993: accepted May 17, 1994. Supported in part bythe National Gaucher Foundation by a grant to S.K. Performed in partial fuljillment of the requirements for the PhD degree in genetics at the George Washington University, Washington, DC, by P.H.C. Address reprint requests to Stefan Karlsson, MD, PhD, Molecular and Medical Genetics Section, Developmental and Metabolic Neurology Branch, National lnstitute of Neurological Disease and Stroke, National Institutes of Health, Bldg 10, Room 3004, Bethesda, MD 20892. The publication costsof this article were defrayedin part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. section I734 solely to indicate this fact. This is a US govemment work There are no restrictions on its use. OOO6-4971/94/8406-0033$0.00/0 1812 Blood, Vol84, No 6 (September 15). 1994: pp 1812-1822 From www.bloodjournal.org by guest on February 6, 2015. For personal use only. 1813 VECTOR DESIGN FOR LONG-TERMEXPRESSION poietic progenitor cells’ and stem cells,8 and expression of potentially therapeutic levels of human GC in the majority of macrophages from long-term reconstituted mice has been shown?.“ In this report, we directly compare the transduction efficiency of several retroviral vectors in murine HSCs and the ability of these viruses to stably express human GC in macrophages derived from the transduced stem cells. In comparing these vectors, we are asking several fundamental questions about retroviral vector design and its effect on transduction efficiency and expression. How does the inclusion of an internal promoter to the viral construct effect its ability to infect HSCs and drive expression of the transferred gene, and how do different promoters compare with respect to these issues? What happens when the internal transcription unit is placed in the opposite orientation with respect to the viral LTRs? Does the addition of a selectable marker gene affect the ability of the virus to infect HSCs or the ability of the viral LTR to direct expression of the transferred gene in hematopoietic cells? Finally, does the choice of retroviral enhancer have an impact on the ability of the virus to transduce HSCs and/or direct expression of the transferred gene in the differentiated progeny of these cells? The answers to these questions have universal importance when designing retroviral vectors for gene therapy of any disorder involving the hematopoietic system. MATERIALS AND METHODS Retroviral Vectors All vectors described in this study are based on the LN series of retroviral vectors.” Construction of the LG9and LGSN” vectors has been described elsewhere. All other vectors were constructed from a plasmid (Gl) in which the NeoRgene from LN was removed and replaced with a multiple cloning site (Genetic Therapy Inc, Gaithersburg, MD). A 2.3-kb fragment of the human GC cDNA” was subcloned into the EcoRI site of the polylinker in bluescript sk(-) (Stratagene, LaJolla, CA), with the 5’ end of the gene towards the Sac1 site and the 3’ end of the gene towards the Kpn I site (SKGCS). A polyadenylation signal was added to the 3’ end of the GC cDNA by ligation of the Nae IIHindIII fragment of PMCINeoPolA into the EcoRV site in SKGCS (SKGCNPolA). All intenal promoters used were subcloned into the polylinker of bluescript ks(-) with the 3’ end of the promoter oriented towards the Sac I site. The 520-bp EcoRYBamHI fragment of PGK-pucl9 (provided by Dr S.H. Orkin, Children’s Hospital, Boston, MA) containing the human phosphoglycerate (PGK) promoter was cloned into the EcoRI to BamHI sites of bluescript (KSPGK). The 334-bp Pvu IIIHindIII fragment of pCHllO (Stratagene) containing the SV40 promoter was blunted and cloned into the EcoRV site of bluescript and oriented toward the Sac I site of the polylinker (KSSV). The 280-bp Xho UPsr I fragment of PMCINeo containing the herpes thymidine kinase promoter and mutant polyoma enhancerI4was cloned into the Xho I to Pst I sites of bluescript (KSTK). The myeloproliferative sarcoma virus (MPSV) enhancer was cloned into the 3‘ Moloney’s murine leukemia virus (MoMLV) LTR in G1 by replacing the Sac I-Cla I fragment of the MoMLV LTR with the 600-bp Sac I-Cla I fragment from the MPSV LTR (provided by Dr D. Bodine, NHLBI, NIH, Bethesda, MD) (GIMP). Expression cassettes containing the internal promoters in front of the human GC cDNA were constructed in bluescript by digesting the plasmids KSSV, KSTK, and KSPGK with EcoRI, S m I, and Xba I, respectively, all of which cut at the 3’ end of the respective promoters. A second digest was performed with Xmn I, which cuts in the ampicillin-resistance gene of the bluescript plasmid. ‘Ihe 1.35-, 1.3-, and 1.5-kb fragments containing the respective promoters were isolated. SKGCS was cut with a partial EcoRI digest, S m I, or Xba I, which all cut at the 5’ end of the GC cDNA, and Xmn I in the ampRgene. The 4.2-bp fragments containing the human GC cDNA were isolated and ligated to the respective promoter fragments. The MG vector was constructed by ligating the Nor UXho I fragment of SKGCS into the Not VXho I fragment of GlMP. The SG, TG, and PG vectors were created by ligation of the Xho I fragments of SVGC, TKGC, and PGKGC into the Xho I site of G1. Both orientations of the cassettes were isolated relative to the viral LTR. A polyadenylation site was added to the vectors in which the GC gene was in the backwards orientation by removal ofthe Sal I fragment containing the last 828-bp of the GC cDNA and replacing this with the Sal I fragment of SKGCNPolA that contains the last 828 bp of the GC cDNA plus the inserted polyadenylation signal. Virus-Producing Cells The GP + E86 virus-producing cells usedin this study were grown in Dulbecco’s modified Eagle’s medium (DMEM), 10% heatinactivated fetal bovine serum, and glutamine. The virus producing clones were created by transfection (LGSN) or cotransfection with pSV2Neo by calcium phosphate precipitation of the recombinant viral constructs into GP + E86 cells. Two days later, the cells were split by serial dilution and replated in media containing 1 g/L geneticin sulfate ((3418). Ten days to 2 weeks after plating, individual colonies were picked and expanded. Titration of LGSN was performed by plating 1 X lo6 TK- 3T3 cells onto a 10-cm tissue culture dish on day 1. The following day, the 3T3 cells were infected by serial dilutions of an overnight supernatant from the virus producing cells in the presence of 8 pg/ mL of polybrene. The following day, the infected 3T3 cells were split 1:lO in media containing 1 g/L G418. Eight to 10 days later, G418-resistant colonies were counted. Viral titers are expressed as colony-forming units ( C m ) per milliliter of viral supernatant. Titration of viruses with no selectable marker was performed by plating 1 X lo6 TK- 3T3 cells on 10-cm tissues culture dishes and infecting them on day 2 with 5 mL of an overnight supernatant from the virus-producing cells in the presence of 8 pg/mL of polybrene. On day 3, viral supernatant was removed and replaced with normal media. On day 4,the 3T3 cells were harvested and DNA was extracted for Southern blot analysis. The blots were probed with the full-length 2.2-kb human cDNA, and the signal of DNA from infected 3T3 cells was compared with the signal of equal amounts of DNA from uninfected cells containing the equivalent of 1 copykell and 0.1 copieskell of vector-containing plasmid DNA. Viral titers are expressed in copies per cell. Detection of helper virus by the SfL- assay was performed by plating D-56 cells (R.H. Bassin, National Institutes of Health, Bethesda, MD) at a density of 2 X [email protected] cells per 6-cm tissue culture dish. The following day, the cells were infected with serial dilutions of viral supernatant in the presence of 8 pg/mL of polybrene. Two hours later, viral Supernatants were aspirated and replaced with normal media. Foci (plaques) were scored on days 5 through 10. Detection of helper virus by marker rescue assay was performed by plating 3T3 cells transduced with a replication defective retrovirus containing the NeoR selectable marker gene at a density of 1 X lo6 cells per 10-cm dish. The following day, the cells were infected with serial dilutions of viral supernatant in the presence of 8 pg/mL of polybrene. After the infected cells were passaged for 1 to 2 weeks From www.bloodjournal.org by guest on February 6, 2015. For personal use only. 1814 toallowtimeforpotentialvirustospread,thesupernatantwas cells. These cells werethen harvestedandtestedonvirgin3T3 selected in 1 g& G418 and any resistant colonies scored 8 to 10 days later. CORRELL,COLILLA, AND KARLSSON RESULTS Retroviral Vectors A series of retroviral vectors containing the human GC cDNA was constructed (Fig 1). In these vectors, expression of the humanGC gene is directed by either theMoMLV LTR BM Infection and Transplantation (LGSNandLG),theMoMLVLTRcontainingenhancer Donor C57BY6 mice were injected with150 mgikg 5-fluorouracil sequences from the MPSV LTR (MG), or an internal simian (5-W;Fluka BioChemika,Ronkonkoma, N Y ) andBMwasharvirus 40 immediate early promoter (SG), thymidine kinase vested from these mice 3 days postinjection. Thecells were plated at a densityof 1 to 2X IO7 celldl0-cm dish and prestimulatedfor 2 promoter and mutant polyoma enhancer (TG), or phosphodays in DMEM containing 10% heat-inactivated fetal bovine serum, glycerate kinase promoter(PG) in either orientation relative glutamine, PedStrep, and 200 U/mL interleukin-3(L-3). L-6, and to the viral LTR sequences. The selectable marker neomycin100 ng/mL stem cell factor (Immunex, Seattle, WA). On day 3, l resistance gene has been deleted from all constructs but one to 2 X lo7 cells wereplatedona10-cmdishcontainingvirus (LGSN). All of the vectors were packaged in the GP + E86 producersplatedthepreviousdayat1to2 X lo6 cellddish and packaging cell line. The expected transcripts generated by cocultured for 2 days under the same growth factor conditions in each of the vectors are also shown in Fig 1. the presence of 8 pg/mL of polybrene. Lethally irradiated (850 rad) Several individual clonesof each vector packaged in GP C57BV6 recipient mice were injected with 5 X 104 celldmouse for + E86 cells were used to infect 3T3 cells, and DNA from individual CFU-S foci and 1 to 2 X IO6 celldmouse for long-term these infected cells was analyzedby Southern blot analysis. reconstitutedmice. CW-S foci were examined on12to 14 days posttransplantationand BM, spleen,andthymusfromlong-term A Southern blot of the highest titer clone found for each reconstituted mice were analyzed 6 to 12 months posttnlnsplantation. vector is shown in Fig 2. The titers of these viruses range from 0.3 to 2.6 copiedcell. Although there is no guarantee that the highest possible titer will be obtained for each vector, DNA and RNA Analysis for every vector but one, a clone withof aattiter least1 copy/ DNA and Southern blot analysis were performed using standard techniques. Proviral copy numbers were determined by scanning the cell was isolated. However, despite screening more than 50 clones containing the PG vector, the highest titer obtained intensity of each band and comparing it with copy numbercontrols, was only 0.3 copiedcell on 3T3 cells. The virus-producing with the endogenous GC bandused as an internal control.All scanning and quantitation was performed with the clones shown in here were used for all subsequent experiNIH Image l .49 software (DrWayne Rasband, NIMH,NIH, Bethesda, MD). Total celluments. lar RNA was extracted by guanidine thiocyanate'J and separated on To estimatehow these titers relate to other published titers, a formaldehyddagarose gel. The RNA was transferred toa nitroceleight LGSN clones were analysed by colony-forming assay lulose filter, prehybridized. hybridized, and washedas described.16 and Southern analysis for comparison (Fig 3). Sixofthe eight clones produced titers of close to 1.0 copylcell(O.6 to Isolation of Macrophages 1.2 copieskell) on the Southern blot. The colony-forming titers of these clones ranged from 2 X IO6 to 1 X 10' CFU/ Macrophages from BM and spleen were isolated by plating 2X mL. In most cases, the intensity of the band on the Southern io7cells from a single-cell suspension derived from two these tissues in 10 mL of RPM medium (Biofluids, Rockville, MD) with 10% blot corresponded well to the G418 titer. Two of the eight heat-inactivated fetal bovineserum, 2 mmol/L glutamine, 10 mmoY LGSNcloneswithtitersof 3 X lo3 and 4 X 104 CFU/ L HEPES, pH 7.3, 5 X IO-' m o w B-mercaptoethanol, 100 U/mL mL did not produce a detectable band on the Southern blot penicillin, and 100 pg/mL streptomycin. The plates were incubated analysis. This indicates that titration by Southern blot analyat37°C for 3hours.Thenonadherent cells weregentlyremoved sis will not detect a titer of less than approximately5 X 10" from the dish and the remaining adherent cells were grown in the CFU/mL. Furthermore, this indicates that all of the clones above medium plus 10% L929 (American Type Culture Collection, used in this study most likely have a titer of at least 1X lo6 Rockville. MD) cell-conditionedmedium.Themacrophageswere CFU/mL or greater, with the exception of the PG virus, for cultured for 8 days and were refed with fresh medium every other which the titer is considerably lower. Clone no. 6 with a day. titer of 1 x 10' CFU/mL is the one shown in Fig2 and used in the experiments presented here. The difference in copy GC Enzymatic Assay number between the two experiments reflects the variation Cellpelletswereextractedina 50 mmoK potassiumcitrate/ sometimes seen from one infection to another. potassiumphosphatebuffer(pH5.9)containingTritonX-100 (2 Viral supernatants from eachof the producer clones used mg/mL) and freeze-thawed for three cycles. Cell extracts were spun in this study were tested for the presence of wildtype virus for30minutes at 12,000 rpmand clearedcellularlysateswere by the S+L- and/or the marker rescue assay. Helper virus assayed for GC activity. GC activity was assayed by cleavage ofthe was not detected by these methods in any of the supernatants synthetic substrate Cmethylumbelliferyl glucopyranoside (Sigma, St tested. Louis,MO)at 4.8 mmol/L ina0.1 mol& potassiumphosphate buffer(pH5.9)with 1.5 m g / d TritonX-100and1.25 mg/mL The Effect of Vector Design on Transduction of sodium taurocholate at 37OC. The reaction was terminated with 0.4 Hematopoietic Cells moVL NaOW0.4 mom glycine, and the fluoresenceof the cleaved 4-methylumbelliferone was measured using a fluorimeter (excitation: Because certain as yet unidentified characteristics of particular viral producer clonesmay be important for transduc360 nm, emission 430 nm). From www.bloodjournal.org by guest on February 6, 2015. For personal use only. VECTOR EXPRESSION FOR LONG-TERM 1815 "_ LG "_ """""_ .0 -I './ b " " " " " " _ ~ MG MPSV EHANCER "m "-( _"\ .0 "_ ~ , ,"""-b"""" b \ H Fig l. A diagram of the vectors used in this study. All vectors are based on the LNseries of retroviral vectors. Open boxes denoteMoMLV LTR; light hatched boxes, human GCcDNA; dark hatched boxes, polyadenylation signal fromSV40; SV, SV40 immediate early promoter; NEO, neomycin-resistance gene;TK, herpes thymidine kinase promoter and mutant polyoma enhancer; PGK, human phosphoglycerate kinase promoter; MPSV, meyloproliferative sarcoma virus; Kb, kilobase. Arrows indicate origin and direction of transcription. The expected RNA transcripts are shown below each vector. The spliced transcripts are shown as dotted lines because they are not always detected. - "_ D .H 4 - 4 TG -" , .,"""""""_"" I * U 1 Kb 7 . . 4.4- 2.32.0- 1 H 4 (D 9.6 6.6 - I TK 03 w W , ,""""""" ~~ Fig 2. Southern blot analysis of DNA from NlH3T3 cells infected by each of the viralsupernatants used for this study. The DNA was digested with Nhe 1, which cutsin the viralLTRs, and 10 p g was loaded per lane. The blot was probedwith the human GC cDNA. 1 and 0.1 copieslcell, genomic DNA fromuninfected 3T3 cells the equivalent of 1 + 3T3 cells infected with supernatantfrom untransfected GP E86 cells. Molecular weight standards are shown on the left in kilobases. The proviral copy number, which is shown below eachlane, was determined by scanning with the genomic GC + From www.bloodjournal.org by guest on February 6, 2015. For personal use only. COLILLA, 1816 CORRELL, AND KARLSSON B A LGSN I 9.6- I ~ LGSN Clone ## Titer CFU/ml 1 2 3 4 2x106 4x106 4x104 3x103 6 1 x108 8 6 x 10' 4.4 - Proviral copy no. 1.2 0.9 0 0 0.8 1.1 0.6 0.9 + Fig 3. (A) Southern blot analysis of DNA from 3T3 calls infected by supernatants from eight separate c l o n e of GP E86 cells transfected with LGSN. The DNA was digested with Nhe 1, which cuts in the viral LTRs, and 10 p g was loaded per lane. The blot was probed with the human GC cDNA. 1 and 0.1 copieslcell, genomic DNA from uninfected 3T3 cells + the equivalent of 1 and 0.1 copieslcell of LGSN plasmid DNA; 3T3, DNA from uninfected 3T3 cells. Molecular weight standards are shown on theleft in kilobases. Proviral copy numbers, determined by scanning, are shown below each lane. (B)Titers of the same eight LGSN clones shown in (A)as determined by Neomycin-resistance.CFUl mL, colony-forming units per milliliter of viral supernatant. tion of hematopoietic progenitor cells and it would be impossible to analyze all of the clones on HSCs, a second screen of the virus producer lines that gave high titers on 3T3 cells was performed on CFU-S multipotential hematopoietic progenitor cells. Each of the producer clones was used to infect primary murine hematopoietic cells in the presence of the growth factors L-3, IL-6, and stem cell factor. DNA from individual day 12 to 14 CFU-S foci was screened by Southern blot analysis for the presence of the provirus. The results from each of the viruses are summarized in Table 1. With each of the viruses used for this study, with the exception ofSG, the initial experiment resulted in a 100% infection efficiency of CFU-S progenitor cells. When the experiment Table l.Infection Efficiency of Viral Producer Cells in CFU-S 100 100 Exp. 100 100 Vector No. Positiveflotal Efficiency Infection LGSN LG MG SG no. 1 Exp. no. 2 TG PG 9/9 1%) 100 10110 loll0 3i7 100 43 10/10 m - 10110 TG 1111 1 PG 10/10 100 100 was repeated with SG, this virus also infected 100% of the CFU-S. Even though the PG virus has a much lower titer than the others by Southern blot analysis of infected 3T3 cells, this virus was also capable of infecting CFU-S with a 100% infection efficiency. It was not possible to compare gene transfer efficiency of clonogenic progenitor cells by G418 resistance using these vectors, because all the vectors, except one, were single gene vectors lacking a selectable marker gene. Because there is no direct way to determine the infection efficiency of the viruses in repopulating HSCs, wehave based the transduction efficiency in these cells on the proviral copy number obtained in BM, spleen, and thymus harvested from lethally irradiated long-term reconstituted mice at 1ea.t 6 months posttransplantation. Southern blot analysis of DNA from the nonadherent cells of the BM and spleen and total thymus of several individual mice transplanted with vectorinfected BM was performed. The viral copy number for each sample was determined by comparison of the signal intensity to copy number controls. Results from each of themice analyzed in this fashion and the average viral copy number obtained with each virus are summarized in Table 2. The transduction efficiencies of these vectors in HSCs correspond well with their titers on 3T3 cells. The LG vector has the highest infection efficiency of the vectors tested, with an average copy number of 3 copied cell. This vector has consistently transduced HSCs at a high From www.bloodjournal.org by guest on February 6, 2015. For personal use only. VECTORDESIGN 1817 FOR LONG-TERMEXPRESSION Table 2. Proviral Copy Number in LongTermReconstituted Mice Vector LGSN Mouse No. 1 2 3 4 '5 '6 '7 '8 Avg LG 1-5 6-10 11 12 13 14 15 '16 '17 '18 '19 '20 SG S T 2.0 3.1 1.a 0.2 0.3 0.5 0.9 0.8 1.2 1.a 0.4 0.8 1.2 0.6 0.6 1.o 1.3 1.o 0.6 3.2 2.9 2.2 2.6 3.7 4.4 3.5 2.5 2.0 2.9 3.2 3.8 4.4 2.6 2.1 2.0 3.3 2.4 1 2 3 4 '5 '6 '7 '8 '9 0.4 1.3 2.8 0.7 1.3 0.2 0.9 0.6 Avg 1.3 0.8 1 2 '3 '4 '5 '6 '7 0.7 3.2 3.3 2.5 3.6 0.4 0.8 2.6 3.2 2.7 3.9 2.7 2.3 Avg MG BM Ave Vector TG 0.7 0.4 0.1 0.7 PG 0.7 1.5 2.0 2.1 3.1 4.0 1.4 2.1 1 2 3 4 5 6 7 8 9 10 11 Avg BM 5 T 0.4 0.8 0.5 1.o 4.5 2.2 1.2 1.5 0.4 1.5 0.4 0.2 1.6 0.9 0.5 0.9 0.7 0.3 1.1 0.2 0.9 0.4 0.9 0 0 0.3 0.7 0.3 0 0.5 0 0 0 0.2 0.2 0.1 0 0.9 1.5 0.2 0 0.9 0.3 0 0 0.1 0.1 0.2 0 0 0.4 0.2 0.3 1.3 0.2 0.8 0.6 0.3 0.2 0.2 0.3 0.3 0.2 0.2 0.3 0.4 0.1 1.2 0.4 0.6 0 2.5 0.3 0.6 1.o 0.3 0.7 0.1 1.o 0.2 0.7 0.2 1.7 0.3 1.o 1.? 0.4 0.7 0.9 0.1 2.1 0 0 0 Avg 0.8 0.4 0.4 0.6 0.6 L 3.0 TG 2.9 0.2 0.9 1.o 1.2 1.2 2.8 1.3 1.4 1 2 3 4 5 6 7 Avg 1.o 5.0 2.7 4.7 3.6 3.2 3.2 3.1 2.6 2.0 3.3 Mouse No. '1 '2 '3 Avg PG 1.1 1 2 3 4 5 6 7 8 '9 '10 Avg 0 0.2 0.4 1.8 0.8 0.5 2.4 Abbreviations: AVG, average; S, spleen; T, thymus. Animals used for Northern blot analysis (Fig 4) and enzyme analysis (Fig 5). efficiency, with 2 to 5 copiedcell detected in every mouse analyzed thus far. Consistent with the viral titer on 3T3 cells, the PG virus had the lowest transduction efficiency in HSCs, with an average copy number of 0.3 copiedcell (approximately 1 log lower than the efficiency of the LG vector) in long-term reconstituted mice. The remainder of the viruses tested had intermediate infection efficiencies, with average copy numbers ranging from 0.4 to 2.4 copieskell. Vectors containing additional promoters, enhancers, or selectable markers, in general, generated lower transduction efficiencies than the LG virus. The vector containing an internal SV40 promoter (SG) had a greater transduction efficiency than those containing the TK (TG) or PGK (PG) promoters. It is also interesting to note that when the PGK promoter was used in the opposite orientation relative to the viral LTR, a higher transduction efficiency was obtained, correlating with the higher titer of this virus on 3T3 cells. The virus in which the MPSV enhancer was substituted for the MoMLV enhancer (MG) also had a lower transduction efficiency of both 3T3 cells and HSCs. Finally, the virus containing an internal neomycin-resistance gene (LGSN) had a lower transduction efficiency in HSCs than the LG vector or the same vector without the selectable marker (SG). DNA from BM, spleen, and thymus of several of the longterm reconstituted mice was also analyzed by Southern blot analysis for the presence of the Y chromosome using a Y- From www.bloodjournal.org by guest on February 6, 2015. For personal use only. 1818 9.5 - 7.5 4.4- ]LTR Transcripts 2.4 - Internal ]Transcripts -Mouse GC 1.3- -GAPDH 2 3 3 5 5 5 5 4 2 3 3 5 5 5 5 4 0.5 0.2ND 1.8 3.1 ND 3.1 0.6 Number of Mice Average proviralno. Fig 4. Northern blot analysis of RNA from (A) nonadherent cellsand (B) macrophages from the BM of long-term reconstituted mice transplanted with BM infected by each of the viruses as indicated. RNA was extracted from pooled cells from the number of mice indicated beneath the blot, corresponding to the mice marked by an asterisk in Table 2. Ten micrograms of RNA was loaded per lane. The blot was probed with thehuman GC cDNA, stripped, and reprobed with GAPDH. NC,RNAfrom nonadherentcells and macrophages an of untransplanted mouse. Molecular weight standards are shown to the left in kilobases. The expected size of each transcript is shown in Fig 1. The average proviral copy numbershown for the nonadherentcells was determined by scanning densitometry for the same cells as used forthe Northern. specific probe.I7 The disappearance of the Y-specific band in male mice transplanted withBM from female donors indicates that, in all animals tested, the recipient mice were fully reconstituted with the donor marrow (data not shown). These results support the idea that the variability in copy number obtained with the different viruses in fact reflects a variability in transduction efficiency rather than differences in the extent of repopulation from one experiment to another. The Effect of Vector Design on Gene Expression in Hematopoietic Cells To determine how efficiently the various promoters can direct expression of the GC cDNA in hematopoietic cells in vivo, expression of these vectors was examined in the progeny of the transduced HSCs. Nonadherent cells (Fig 4A) from the BM of several long-term reconstituted mice transplanted with BM transduced by each of the vectors were analyzed for expression of human GC RNA. The RNA in Fig 4A was isolated from a pool of the same nonadherent BM cells analyzed for copy number in the mice indicated by asterisks in Table 2. The number of mice and the average proviral copynumber for those cells are indicated below each lane. For Gaucher's disease, we are ultimately interested in obtaining expression in differentiated macrophages derived from the transduced HSCs. Therefore, macrophages isolated from BM of these same mice were also analyzed for human GC RNA (Fig 4B). Due to a limited number of cells, proviral copy number was not assessed directly in this cell population; however, the results from the macrophages correlates well with the expression pattern seen in the nonadherent cell population for which proviral copy number data is available. The Northerns shown in Fig 4 were quantitated and the results from this analysis are shown in Table 3. The values obtained for both the nonadherent cells and macrophages are a sum of all transcripts expressing the human GC cDNA. The values for the GAPDH control are also shown as well as the corrected values for human GC transcripts. For the nonadherent cells, the values are also corrected for proviral copy number. The values shown in parentheses are those for which the corresponding copy number data are not available. For these vectors, the overall average copy numbers obtained with these vectors are used. Therefore, there is likely some error in these numbers. Comparison of promoters/enhancers for expression in hematopoietic cells. The results from these experiments indicate that, overall, the viral LTR is the most efficient of the promoters tested in all of the hematopoietic cells analyzed, although the levels of LTR-driven transcript vary among the vectors tested. Of the internal promoters tested, the PGK promoter in the forward orientation directs the highest levels of human GC message in these experiments. There are much From www.bloodjournal.org by guest on February 6, 2015. For personal use only. VECTOR EXPRESSION DESIGN FOR LONG-TERM 1819 Tablo 3. Exprodon kv.h of Human GC RNA In Long-Twm Roc0nrtitut.d Mica - - LGSN LG MG SG TG PG TG PG GC RNA nonadherent GAPDH nonadherent GC RNNGAPDH nonadherent 465 2,908 0.16 1,761 1,995 0.90 494 3,061 0.16 2.947 2.01 4 1.46 4,938 1,839 2.69 2,674 2,056 1.3 0 1,710 0 0 618 0 GC RNA macrophages GAPDH macrophages GC RNNGAPDH macrophages 0 2,469 0 1,396 2,314 0.60 1,545 2,812 0.67 2,039 1,842 1.1 1 4,373 2,101 2.08 4,156 2,388 1.74 456 1,963 0.23 0 1,027 0 Proviral copy number nonadherent GC RNNcopy 3, no. of nonadherent 0.6 3.1 (1.1) .8 (0.3) 0.2 0.5 0.3 0.3 (4.3) 0 0 1.5 (0.15) 1 3.1 0.5 The numbers represent arbitrary densitometry units. the TG virus is present at a lower copy number than the LG lower, but detectable, levels of GC RNA generated by the internal TK and SV40 promoters. These results correspond virus in the BM of these long-term reconstituted mice (1.8 to those obtained when expression of the vectors was tested v 3.1 copiedcell, respectively), thelevelsofLTR-derived message generated by this vector are higher than those generin three separate CFU-S experiments (data not shown). ated by the LG virus. In addition, levels of LTR-generated Interestingly,levels of RNA fromtheviralLTRcontaining theMPSV enhancer (MG)are lower than those from transcript from the PG vector are also high in these mice. Although no proviral copy number data is available for the LG in the nonadherent cells; however,in the macrophages, these levels are roughly equivalent. Because these respective BM of these mice because of experimental error, the overall average copy number obtained in BM with the PG vector is samples are from the same mouse BM, it appears as though 0.3 copiedcell, and,of the 11 animals tested, the highest transcription from the hybrid LTRmay be more efficient in copy number in an individual mouse was 0.9 copiedcell. thedifferentiatedmacrophagesthaninothernonadherent These data indicate that the presence of an internal transcrip hematopoietic cells. tion unit driven by the TK or PGK promoters does not deEffect of a second transcription unit on expression from the viral LTR. We were also interested in determining what crease, but may even increase expression from the viral LTR in hematopoietic cells. effect, if any, the inclusion of an internal transcription unit has onthe levelsof expression from the viral LTR. For these Effectof gene orientation on gene expression. Those vectors (TG and PG) that contain internal transcripts in the comparisons, it is most informative to focus on expression in the nonadherent fraction of the BM in long-term reconsti- opposite orientation relative to the viral LTRs consistently showverylowlevelsofboththeLTRmessageandthe tuted mice for which proviral copy number data has been internally generated message in all hematopoietic cells tested determined. It is difficult to assess the effect of integration site of the provirus on viral expression; however, due to the in this study. This is most likely caused by the production fact that the RNA is pooled from several transplanted mice of antisense RNA from the internal promoter. rather than individual mice, some general observations can Protein Production in Macrophages of Long-Term be made. Reconstituted Mice Thelevels ofhuman GC transcriptsgenerated by the LGSNvector are lowerthanthosegeneratedbytheLG Macrophages isolated from BM of the same set of mice vector; however, the lower viral copy number obtained with for which the RNA is shown were analyzed for expression the LGSN virus (0.6copykell) when compared with the LG of GC activity. The activity for each mouse is shown in Fig virus(3.1 copiedcell) almostcertainlycontributestothe 5 . In this experiment, the PG virusproducedthehighest lower levelsof expression from the LTR seen with this virus. overall levels of GC activity in BM macrophages, with an When corrected for the copy numbers, the levels of these average activity 1.77-fold higher than the uninfected contranscripts are roughly equivalent. On the other hand, the trols. This activity was generated by a combination of tranlevels of LTR-generated transcript in long-term reconstituted scripts from the LTR and PGK promoters because both tranmice infected with the SG virus are not reduced when com- scripts can be translated to produce GC protein. This was pared with those generated by the LG virus. Because the LG followed by the TG, LG, and SG viruses, which increased and SG mice in this experiment contained similar proviral the enzyme activity in BM macrophages by an average of copy numbers,it appears as though the presenceof the inter1.46-, 1.45-, and 1.44-fold, respectively. The majority of the nal SV40 promoter alone or in combination with the NeoR enzyme activity generated by the TG and SG viruses is most gene in the viral construct has a minimal effect on the expreslikely caused by transcription from the viral LTR, because sion levels from the viral LTR. levels of transcript from the SV40 and TK promoters in The levels of message from the viral LTR generated by these cells are low. The LGSN virus produced lower levels both the TG and PG viruses also remain consistently high of activity on average than the LG virus, and the TG and in all the hematopoietic cells tested. Interestingly, although PG viruses in which the transcriptional unit was placed in From www.bloodjournal.org by guest on February 6, 2015. For personal use only. CORRELL, COLILLA, ANDKARLSSON 1820 the opposite orientation relative to the viral LTRs also produced very low levels of GC activity. These results follow veryclosely the levelsof RNAproduced by thevarious vectors in these cells. The MC virus did not showanydetectable increasein GC protein in these cells. This is most likely caused by a mutation in the GC cDNA that may have occurred during subcloning, because the packaging cells also do not express any human GC cDNA (data not shown). When a different clone of this plasmid was transfected into amphotropic packaging cells, it was capable of expressing human GC protein. Althoughthe highest averagelevels of GC activityin this experiment were produced by the PG virus, the highest expressing single mouse in this experiment was an LC-infected mouse in which the LC virusproduced a 2.1-fold increase in GC activity in BMmacrophages.Inaddition, much higher levels of GC activity in macrophages of longterm reconstituted mice have been produced by the L C virus." In the 20 mice tested thus far, theL C virus has produced average increases in activity in BM and spleen macrophages ranging from zero to fivefold over those from normal uninfected mice despite consistently high proviral copy numbers, with an overallaverage increase of 2.23-fold (data not shown). DISCUSSION In the studies presented here, we have compared the utility of several retroviral vectors for use in the gene therapy of hematopoietic disorders. We have found that the elements present in the viral vector can affect the retroviral titer, its efficiency at infecting murine HSCs, andits ability to express human GC in hematopoietic cells of long-term reconstituted mice. With the data presented here. we can begin to address the questions posed at the outset of this study. In general, the simplest viral construct (LC) is the most efficientat infectingHSCs.This virusresults in ahigher vectorcopy number in thehematopoietictissues and, on average, the highest overall enzyme production level of the vectors tested. However, the level of enzyme generated by theLGvirus is quite variable anddoes notseem to be dependenton viral copy number,because all of the LGinfected mice in these studies had multiple copies of the viral genome in all hematopoietic tissuestested. These results compare withthose obtained by Kalekoet who reported similar resultswithavectorcontainingthe hADA gene driven by the MoMLV LTR. They concluded that, although the viral LTR was capableof generating high levels ofhADA in long-termreconstituted mice,thepresence of multiple copies was not sufficient to guarantee long-term expression. Replacement of the MoMLV LTR enhancer with the enin the viral construct (MC) hancerfromtheMPSVLTR resulted in a lower viral titer and a subsequent decrease in proviral copy number in long-term reconstituted mice. Results from Beck-Engeser et a l l g indicate that the MPSV enhancer caused a reduction in the retroviral infection of primitive hematopoietic cells. In further support of these results, the transduction efficiency of the M C virus, packaged in two different amphotropiccell lines, in both 3T3 and murine myeloid (MI) cells was lower than that for the LC virus in these same cells.'" Expression of human GC RNA from the MG vector was lower than expression from the LC vector in the nonadherent cells from the BM of long-term reconstituted mice, possibly reflecting a lower copy number in these cells. However, in the BM-derived macrophages, the expression levels of human GC RNA generated by these two viruses were roughly equivalent.Therefore,fordiseases in which macrophages are the target cell, suchas in Gaucher's disease, the increased levels of transcription in these cells may compensate for the lower infection efficiency. This conclusion is based on the assumption that the macrophage compartment was reconstituted toanequivalentextent in thesemice.Although an increase in GC enzyme couldnot be detectedin these macrophages because of a defect in the viral construct, we believe that, in this experiment, the level of enzyme production would be similar to that obtained with the LC virus because the levels of RNA are comparable. Addition of an internal promoter to the viral construct to drive expression of thehuman CC cDNA also seemed to haveanegativeeffectonthe viral titer and its ability to infect HSCs. In addition,some internalpromoters had a more deleterious effectonthe viral titer than others. The presence of an internal PGK promoter(PG) dramatically reduced the viral titer and its infection efficiency in HSCs. Again, when the PC vector was packaged in amphotropic cell lines and used to infectM1 cells in culture,asimilar decrease in the infection efficiency when compared with the LC and M C vectorswasobserved."' However, when the PGK promoter was placed in the opposite orientation relative to the viral LTR, the viral titer increased somewhat. Similar I T 50 40 - VECTOR LGSN LG MG SG TG PG TG 4 PG Fig 5. GC activity in macrophages isolated from B M of the same long-term reconstituted mice as shown in Fig 4 and marked with an asterisk in Table2. Mouse numbers from Table 2 are increasing from left to right, ie, LGSN mice are 5,6,7, and 8 from leftto right. Enzyme activities of individual mice are shown. Specific activity is presented as nanomoles of 4MU cleaved per minute per milligram of protein. From www.bloodjournal.org by guest on February 6, 2015. For personal use only. 1821 VECTOR DESIGN FOR LONG-TERMEXPRESSION results have been shown with the c-fos promoter that did not allow transduction of the target cells in the forward orientation, although the reverse orientation did.” The presence of the SV40 promoter in the viral construct (SG) seemed to have the least overall effect on viral titer. The TK promoter (TG) had an intermediate effect on the ability of the virus to infect HSCs, and placing the transcriptional unit in the reverse orientation relative to the viral LTRs did not enhance the titer of the virus. Expression of the human GC cDNA from an internal promoter in hematopoietic cells is most efficient when the PGK promoter is employed. It has been suggested that the viral LTR can sometimes cause transcriptional interference with a downstream internal promoter. To decrease this interference, some groups have chosen to delete the 3’ LTR enhan~er.”~’~ However, when the PGK promoter was used todrive expression of the ADA gene, interference from the LTR was absent or minimal and a deletion of the 3’ LTR was not necessary to improve expression in CFU-S colonies.24When the SV40 (SG) and TK (TG) promoters were used to drive expression of the human GC cDNA, they generated low, but detectable levels of human GC RNA in the hematopoietic cells. The presence of an internal promoter did not appear to have a detrimental effect on expression levels from the viral LTR in the hematopoietic cells. Levels of LTR-generated transcript in nonadherent BM cells from long-term reconstituted mice transduced with the SG, TG, and PG vectors were equal to or greater than the levels generated by the LG vector. This increase was observed despite the lower proviral copy number present in the mice transplanted with the TG- and PG-infected BM. Our double gene vector, LGSN, resulted in a lower infection efficiency in HSCs than the LG and SG viruses, indicating that the presence of the NeoRgene may have a deleterious effect on retroviral titer. It has been generally agreed that the size of the retrovirus is inversely proportional to the viral titer that it can produce. The LGSN vector also resulted in decreased expression of human GC in long-term reconstituted mice when compared with the LG and SG vectors. However, this decrease in expression seems to be due solely to the lower proviral copy number present in these cells. Others have reported detrimental effects on expression levels when a second gene is In support of this, all reports of expression of ADA and GC at levels comparable to endogenous mouse levels in hematopoietic cells have used vectors that contain only the ADA or GC gene and no selectable markers. We conclude that the reduction in expression seen with double gene vectors may be due more to a lower transduction efficiency of HSCs than to the presence of a second transcriptional unit. In this study, the LG virus was capable of transducing HSCs at the highest efficiency and generated the mouse expressing the highest levels of human GC. However, expression from this vector in long-term reconstituted mice is not consistent. Overall, the TG and PG vectors in this study generated the highest levels of human GC expression per copy of the provirus. However, particularly in the case of the PGK promoter, this high level of expression comes at the expense of viral titer. Unfortunately, at the present time, transduction of human hematopoietic progenitors is not as efficient as the efficiencies in the mouse. Therefore, when designing retroviral vectors for human gene therapy, having a high viral titer is important. The TG vector generates higher titers than the PG vector and may be useful, although very little of the expression from this virus is actually coming from the TK promoter itself. The SV40 promoter is expressed at low levels in hematopoietic cells in vivo and its presence in the viral construct seems to be quite neutral in terms of viral titer and expression from the viral LTR, therefore no advantage is gained by its use. Addition of a selectable marker gene causes a decrease in long-term expression from the viral LTR, mostlikely due to the lower proviral copy numbers it generates, and should be avoided if possible. Given these results, it would be beneficial to continue focusing efforts to design an “ideal” vector that can transduce HSCs with high efficiency and consistently generate high expression levels of the transferred gene in the progeny of those cells. Of the vectors tested here, the LG vector remains the most promising for transduction of primitive human hematopoietic cells due to its high titer.*’ ACKNOWLEDGMENT We thank DrR.O. Brady for generous support and encouragement, M. Amiri and S. Stahl for technical assistance, and D.L. Freas-Lutz and Dr M. Brennan for valuable discussions. REFERENCES 1. Karlsson S: Treatment of genetic defects in hematopoietic cell function by gene transfer. Blood 78:2481, 1991 2. Williams DA: Expression of introduced sequences in hematopoietic cells following retroviral-mediated gene transfer. Hum Gene Therapy 1:229,1990 3. Williams DA, Orkin SH, MulliganRC:Retrovirus-mediated transfer of human adenosine deaminase gene sequences into cells in culture and into murine hematopoietic cells in vivo. Proc Natl Acad Sci USA 83:2566, 1986 4. MagliMC, Dick JE,Huszar D, Bernstein A, Phillipes RA: Modulation of gene expression in multiple hematopoietic cell lineages following retroviral vector gene transfer. Proc Natl Acad Sci USA 84:789, 1987 5. Brady RO, Kanfer JN, ShapiroD: Metabolism of glucocerebrosides 11. Evidence of an enzymatic deficiency in Gaucher’s disease. Biochem Biophys Res Commun 18:221, 1965 6. Barranger JA, Ginns EI: Glucosylceramide lipidoses: Gaucher disease, in ScriverCR,Beaudet AL, Sly WS, Valle D (eds): The Metabolic Basis of Inherited Disease, v01 2. New York, NY, McGraw-Hill, 1989, p 1677 7. Correll PH, Fink JK, Brady RO, Perry LK, KarlssonS: Production of human glucocerebrosidase in mice after retroviralgene transferintomultipotentialhematopoieticprogenitor cells. ProcNatl Acad Sci USA 863912, 1989 8. Correll PH, Kew Y, Perry LK, Brady RO, Fink JK, Karlsson S: Expression of human glucocerebrosidase in long-term reconstituted mice following retroviral-mediated gene transfer into hematopoietic stem cells. Hum Gene Ther 1:277, 1990 9. Correll PH, Colilla S, Dave HPG,Karlsson S: High levels of human glucocerebrosidase activity in macrophages of long-term reconstituted mice afterretroviral infection of hematopoietic stem cells. Blood 80:331, 1992 From www.bloodjournal.org by guest on February 6, 2015. For personal use only. 1822 10. Ohashi T, Boggs S, Robbins P, Bahnson A, Patrene K, Wei F-S, Wei J-F,Li J, LuchtL,FeiY,Clark S, KimakM,HeH, Mowery-Rushton P, Barranger JA: Efficient transfer and sustained high expression of the human glucocerebrosidase gene in mice and their functional macrophages following transplantation of bone marrowtransducedby aretroviralvector.ProcNatlAcadSci USA 89: 11332, 1992 l 1. Miller AD, Rosman GJ: Improved retroviral vectors for gene transfer and expression. Biotechniques 7:980, 1989 12.FreasDL,CorrellPH,DoughertySF,Karlsson S, Pluznik DH: Evaluation of expression of transferred genes in differentiating myeloid cells: Expression of human glucocerebrosidasein murine macrophages. Hum Gene Ther 4:283, 1993 13. Sorge J, West C, Westwood B, BeutlerE: Molecular cloning and nucleotide sequence of human glucocerebrosidase cDNA. Proc Natl Acad Sci USA 82:7289, 1985 14. Capecchi MR Altering the genome by homologous recombination. Science 244:1288, 1989 15. Chomczynski P, Sacchi W: Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162156, 1987 16. Reiner 0, Wilder S, Givol D, Horowitz M Efficient in vitro and in vivo expression of human glucocerebrosidase cDNA. DNA 6:101,1987 17. Mardon G, Page D: The sex-determining region of the mouse Y chromosome encodes a protein with a highly acidic domain and 13 zinc fingers. Cell 56:765, 1989 18. Kaleko M, Garcia JV, Osbom WRA, Miller AD: Expression of human adenosine deaminase in mice after transplantationof genetically-modified bone marrow. Blood 751733, 1990 19. Beck-Engeser G, Stocking C, JustU, Albritton L, Dexter M, Spooncer E, Ostertag W Retroviral vectors relatedto the myeloproliferative sarcoma virus allow efficient expression in hematopoietic CORRELL,COLILLA, AND KARLSSON stem and precursor cell lines, but retroviral infection is reduced in more primitive cells. Hum Gene Ther 2:61, 1991 20. Freas DL, Correll PH, Dougherty SF, Pluznik DH, Karlsson S : Evaluation of retroviral vector expression in differentiated myeloid lineages: Expression of human glucocerebrosidase in murine macrophages. Clin Res 41:163A, 1993 21.BelmontJW,MacGregorGR,Wager-SmithK,FletcherF, Mom KA, HawkinsD,VillalonD,ChandSM-W,Caskey CT Expression of human adenosine deaminase in murine hematopoietic cells. Mol Cell Biol 85116, 1988 22. Yu S-F, vonRuden T, KantoffPW,Garber C, Seilber M, Ruther W, Anderson W F , Wagner EF, Gilboa E Self-inactivating retroviral vectors designed for transfer of whole genes into mammalian cells. Proc Natl Acad Sci USA 83:3194, 1986 23. Guild BC, Finer MH, Houseman DE, Mulligan RC: Development of retrovirus vectors useful for expressing genes in cultured murineembryonalcellsandhematopoieticcellsinvivo.JVirol 62:3795,1988 24.Lim B, ApperleyJF,OrkinSH,WilliamsDA:Long-term expression of human adenosine deaminase in mice transplanted with retrovirus-infectedhematopoieticstemcells. h o c NatlAcadSci USA8623892,1989 25.ApperleyFJ,LuskeyBD,WilliamsDA:Retroviralgene transferofhumanadenosinedeaminaseinmurinehematopoietic on long-term expression. cells: Effectof selectable marker sequences Blood 78:310, 1991 26. Bowtell DDL, Cory S,Johnson GR, Gonda TJ: Comparison of expression in hematopoietic cells by retroviral vectors carrying two genes. J Virol62:2464, 1988 27. Xu L, Stahl SK, Dave HPG, SchiffmannR, Correll PH, Kessler S , Karlsson S: Correction of the enzyme deficiency in hematopoietic cellsof Gaucher patients usinga clinically acceptable retroviral supematant transduction protocol. Exp Hematol 22:223, 1994 From www.bloodjournal.org by guest on February 6, 2015. For personal use only. 1994 84: 1812-1822 Retroviral vector design for long-term expression in murine hematopoietic cells in vivo PH Correll, S Colilla and S Karlsson Updated information and services can be found at: http://www.bloodjournal.org/content/84/6/1812.full.html Articles on similar topics can be found in the following Blood collections Information about reproducing this article in parts or in its entirety may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://www.bloodjournal.org/site/subscriptions/index.xhtml Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.
© Copyright 2021