Modified HIV-1 Based Lentiviral Vectors Have an Effect on ...med.stanford.edu/content/dam/sm/kaylab/documents/MT.9.2001.PARK.KAY.pdf · cppt (41% versus 81%, respectively; Table 1)

ARTICLE doi:10.1006/mthe.2001.0450, available online at http://www.idealibrary.com on IDEAL

Modified HIV-1 Based Lentiviral Vectors Have an Effect on Viral Transduction Efficiency and Gene Expression

in Vitro and in VivoFrank Park and Mark A. Kay*

Program in Human Gene Therapy, Departments of Pediatrics and Genetics, Stanford University, Stanford, California 94305, USA

*To whom correspondence and reprint requests should be addressed. Fax: (650) 498-6540. E-mail: [email protected].

Gene transfer using lentiviral vectors has been recently shown to be enhanced with cis-actingelements in a cell-type-dependent manner in vivo. For this reason, the study reported here wasdesigned to modify lentiviral vectors that express lacZ, human factor IX (FIX), or human �1-anti-trypsin (AAT) to study the effect of different cis DNA elements on transduction efficiencies. Wefound that incorporation of the central polypurine tract sequence (cppt) increased transductionefficiency in vitro while increasing the transduction of non-cell-cycling hepatocytes in vivo. C57Bl/6scid mice that were administered lentiviral vectors devoid of the cppt (2 � 108 transducing units(T.U.)/mouse) had 81% of their lacZ-transduced hepatocytes colabeled with the cell cycle marker5‘-bromo-2‘-deoxyuridine (BrdU). In contrast, inclusion of the cppt reduced the colabeling inmouse hepatocytes by 50%. Further modifications in the lentiviral vectors were performed toenhance viral titer and gene expression. We found that the inclusion of a matrix attachmentregion (MAR) from immunoglobulin-� (Ig�) significantly increased the transduction efficiency,as measured by transgene protein expression and proviral DNA copy number, compared withvectors without Ig� MAR. In vitro studies using human hepatoma cells demonstrated a signifi-cant increase (two- to fourfold) in human AAT and human FIX production when the Ig� MARwas incorporated. In vivo transduction of partially hepatectomized C57Bl/6 mice given an opti-mized lentiviral vector containing the cppt and Ig� MAR (2 � 108 T.U./mouse) resulted in sus-tained therapeutic levels of serum FIX (~ 65 ng/ml). Our study demonstrates the importance ofcis-acting elements to enhancing the transduction ability of lentiviral vectors and the expressionof vector transgenes.

Key Words: lentivirus, hemophilia, liver, matrix attachment region, central polypurine tract,gene expression, gene therapy

INTRODUCTION

Retroviruses have been an attractive vector system for use ingene therapy, but their use has been hampered by poor invivo transduction efficiencies [1], in part due to the need fornuclear membrane dissolution for efficient integration tooccur [2]. A variety of terminally differentiated cells in vitroand in vivo have been shown to be transduced by lentiviralvectors based on human immunodeficiency virus (HIV)-1[3–5]. However, we found that the status of the cell cycle mayhave an important role in promoting efficient transductionof first- and second-generation lentiviral vectors into hepa-tocytes in vivo [4]. In quiescent mouse liver, 83% of the lacZ-positive hepatocytes progressed through the S-phase of thecell cycle. Moreover, proliferative stimulation of liver regen-eration by surgical hepatectomy or drug therapy enhancedhepatocyte transduction efficiencies by ~ 30-fold [4,6].

164

Because the replication-defective lentiviral vectors areessentially “gutted” for their native viral genes, whichpotentially affects their ability to infect nondividing cellsin vivo and in vitro, several investigators have recently stud-ied the role of cis-acting elements in HIV and how theyaffect lentiviral transduction and transgene expression[7–10]. The central polypurine tract sequence (cppt), whichis found within the gene pol of HIV, has been one of themore extensively studied cis-acting elements. Presence ofthe cppt can enhance viral titer in vitro [7–10], and signif-icantly increase transduction of neuronal cells in vivo [9],and human hematopoietic cells ex vivo [7]. The inclusionof the cppt has continued the debate as to whether qui-escent cells in vivo, such as the hepatocytes in the liver,require cell cycling for transduction by lentiviral vectors.Pfeifer, et al. [11], attempted to demonstrate lentiviral vec-

MOLECULAR THERAPY Vol. 4, No. 3, September 2001Copyright © The American Society of Gene Therapy

1525-0016/01 $35.00

ARTICLEdoi:10.1006/mthe.2001.0450, available online at http://www.idealibrary.com on IDEAL

tor transduction of noncycling hepatocytes in vivo usinga modified third-generation lentiviral vector, which con-tained cis-acting elements that were not found in previousstudies using second-generation vectors [3–6]. Becausethere were several modifications made in the third-gener-ation lentiviral vector, it remains to be determinedwhether the inclusion of the cppt alone would allow forthe enhancement of lentiviral transduction in vivo.

Additional cis-acting elements are the scaffold or matrixattachment regions (MARs) [12,13], which have beenfound to increase gene expression within retroviral vectorsand may have an important role in lentiviral vectors.MARs are believed to bind to the nuclear matrix ofgenomic DNA and establish local access of transcriptionfactors to the enhancer/promoter sequences within thedomain. Agarwal, et al. [12], demonstrated that the MARfrom the human �-interferon gene incorporated into aMoloney murine leukemia virus-based vector increasedgene expression in primary human lymphocytes in vitro,and that long-term expression was related to the ability ofthe MAR to inhibit methylation of the long terminal
repeats (LTR) [13]. However, thereremains a plethora of MARs that havebeen shown to possess position- andcell-type-independent enhancingeffects on transgene expression [14–18],yet their mechanism of action is notfully understood.
For this reason, our study wasdesigned to better understand the roleof cis-acting elements, such as the cpptand MARs from the chicken lysozymeand immunoglobulin-� gene, in termsof their effects on transduction effi-ciency and transgene expression withinthe context of lentiviral vectors in vitroand in vivo.

RESULTS

Effect of the cppt on ViralTransduction Efficiency in Vitroand in VivoTo study the effect of the cppt on trans-duction efficiencies, we constructed avariety of different lentiviral vectors(Fig. 1A) and then titered them on HeLacells in vitro. Insertion of the cppt intosecond-generation integrating transferplasmids 5� of the expression cassetteincreased the transduction efficiency ofthe vector in HeLa cells by two- tothreefold (Fig. 1B), consistent with aprevious study [7]. No significant dif-ference was observed in the transduc-tion efficiency when the cppt was


cloned into the middle of the integrating vector (betweenthe EF1� promoter and the start site of lacZ). Two copiesof the cppt (Fig. 1B) resulted in a slightly higher trans-duction efficiency compared with the cppt-deficient vec-tor, but it was not significantly different (P < 0.06).

As previous in vitro studies found that lentiviral vectorsare not as efficient in nuclear translocation without thiscis-acting cppt element, we determined the role of the cpptin liver transduction of naive C57Bl/6 mice in vivo.Lentiviral vectors (2 � 108 transducing units (T.U.)/mouse)were administered through the portal vein into mouseliver and osmotic minipumps were inserted subcuta-neously to deliver a steady-state concentration of BrdU for7 days. Consistent with our previous study, we observedvery mild liver injury at day 1 following lentiviral admin-istration as measured by serum alanine aminotransferase(ALT; 105 ± 20 international units (I.U.)) compared withvehicle (control) mice (35 ± 5 I.U.). The serum ALT levelsreturned to normal levels when measured at day 3 [4,19].In lentiviral-treated mice with or without the cppt, theproportion of BrdU-positive hepatocytes increased about

FIG. 1. Role of the cppt in transduc-tion efficiency of lentiviral vectors invitro. (A) Schematic of the second-generation lentiviral transfer con-structs. EEFIA, EFI� promoter. Thecppt (black box) was cloned into dif-ferent regions of the integrating vec-tor. (B) Graph demonstrating thetransduction efficiency as titered onHeLa cells (5 � 105 cells) by X-galstaining and p24 ELISA for the differ-ent lentiviral vectors containing theEF1�cytoLacZ expression cassettewith and without the cppt (n = 3–4independent vector preparationstitered for each group). An asteriskdenotes a significant differencebetween EF1�cytoLacZ versus theother groups (P < 0.01).

A

B

165


FIG. 2. Role of the cppt on lentiviral vector transduction in mouse hepatocytes in vivo. C57Bl/6 scid mice (8 weeks of age) were injected into the portal veinwith lentiviral vectors (2 � 108 T.U.) expressing cytoplasmic lacZ driven by the EF1� promoter with (n = 5) and without (n = 4) the cppt. BrdU was adminis-tered with an osmotic minipump inserted into the subcutaneous space for a period of 7 d. Liver sections were labeled with BrdU and X-gal from a mouse with(A) and without (B) the cppt. The brown nuclei indicate BrdU incorporation.

A B

fourfold (7–8%) compared with that of vehicle (control)-treated mice (Table 1). The mild liver injury, which couldbe attributed to factors in the viral preparation, may havecaused the increased number of BrdU-positive hepatocytescompared with the control mice (Fig. 2 and Table 1). Theduodenum was used as a positive control for BrdU incor-poration (data not shown).

In terms of the cppt and its role in mediating trans-duction into non-cell-cycling hepatocytes, it was foundthat both groups with or without the cppt had between0.1% and 0.2% of their hepatocytes transduced (no sig-nificant difference was observed between the groups).However, mice receiving the lentiviral vector (n = 5), whichcontained the cppt, had a 50% lower proportion of lacZ-positive cells that were colabeled with BrdU comparedwith mice (n = 4) given the lentiviral vector devoid of thecppt (41% versus 81%, respectively; Table 1). For this rea-son, subsequent integrating transfer plasmids used in thisstudy included the insertion of the cppt.

Because the cppt had a role in enhancing viral trans-duction efficiency in vitro and reducing the need for cellcycle progression of lentiviral vectors to transduce hepa-

TABLE 1: Role of the cppt in lentiviral transduction in mouse hepatocytes in vivo

% of LacZ hepatocytes

% of BrdU positive that are BrdUProtocol hepatocytes positive

PBS control (n = 5) 1.8 ± 0.3 N/A

pHR2EF1�cytoLacZ (n = 4) 7.7 ± 0.4* 81

pHR2cEF1�cytoLacZ (n = 5) 7.2 ± 0.6* 41

166

tocytes in vivo, several modifications were made in thetransfer vector to potentially enhance the efficiency oftransduction and also reduce biosafety concerns.

Initial modifications were carried out to develop a self-inactivating (SIN) vector, whereby the 3‘ LTR was deletedas described [20] to remove 400 bp of the U3 region, whichultimately results in a 5‘ LTR deletion in the vector dur-ing reverse transcription. The woodchuck post-regulatoryelement (WPRE) was included in the integrating vectorbecause previous studies showed the ability of this cis act-ing element to increase mRNA stability of the transgenein a retroviral construct [21].

Also, an additional 509 bp of the envelope sequence,which contain the rev responsive element (RRE), wereremoved from the vector. The RRE (253 bp) was re-insertedinto the 3‘ end of the transfer plasmid (5‘ to the WPRE),which effectively removed 256 bp of native HIV envsequence and potentially helped to minimize homologybetween wild-type HIV. The lentiviral transfer plasmidscontaining the cytomegalovirus (CMV) promoter drivingthe expression of cytoplasmic lacZ are shown schemati-cally in Fig. 3. The vectors were transduced into HeLa cellsand gene transfer was measured by quantifying the num-ber of lacZ-positive cells, the amount of enzyme, and/orvector DNA copy number. There was no significant dif-ference in transduction efficiency regardless of the locationof the RRE (Fig. 4A). The presence of two RREs in the trans-fer vector increased the transduction efficiency by a fac-tor of ~ 2.5 from 4.9 to 12.3 T.U./pg p24.

As we made several alterations in the lentiviral trans-fer vector, we wanted to demonstrate that the cppt couldenhance viral transduction efficiency as in the previoussecond-generation vectors. As before, the cppt was cloned5‘ to the expression cassette and the inclusion of this 118-



FIG. 3. Schematic of modified second-generation SIN transfer plasmids. WPRE, woodchuck post-regulatory ele-ment; RRE, rev-responsive element; SIN, 3� LTR with 400-bp deletion; CMV, cytomegalovirus promoter; solLacZ,cytoplasmic lacZ transgene. The cppt is shown in gray.

bp DNA flap sequence significantly enhanced the trans-duction efficiency to 17.3 T.U./pg p24, which was three-fold higher than the lentiviral vector without the cppt(Fig. 4B).

To determine the role of MARs, lentiviral vectors weregenerated containing either the chicken lysozyme 5‘ MAR orthe immunoglobulin-� (Ig�) MAR (Fig. 4B). The individualMAR sequences were cloned 5‘ to the cppt (Fig. 3) in bothsense and antisense directions. The transduction efficiencydecreased to 4 T.U./pg p24 when the chicken lysozyme 5‘MAR, regardless of its orientation, was inserted into the vec-tor. However, the Ig� MAR significantly increased the trans-duction efficiency to 37.7 and 43.8 T.U./pg p24 in the senseor antisense directions, respectively.

To quantify the transgene product, protein lysates fromHeLa cells were prepared 48 hours following infection withdifferent lentiviral preparations (Fig. 4C). The steady-statelevel of �-galactosidase (�-gal) protein was 77 ± 5 pg/mgprotein following infection with the CMVsolLacZ lentivi-ral vector. Consistent with the increase in transduction efficiency, the insertion of the cppt resulted in a twofold


increase in �-gal protein to 158 ±2 pg/mg protein. The chickenlysozyme MAR inserted 5‘ into theexpression cassette resulted in 108± 10 and 122 ± 14 pg/mg protein(in the sense and antisense orien-tations, respectively). For the Ig�MAR, the steady state level of �-gal was four- to fivefold higher(836 ± 154 and 668 ± 161 pg/mgprotein in the sense and antisenseorientations, respectively) com-pared with the pHR(-)cCMVsolLacZR(+)W(+) lentiviralvector.

To ascertain whether theincreased protein expression wasattributable to an increased copynumber of lentiviral vectors or cis-acting enhancing effects by theMARs, we carried out quantitativereal-time PCR analysis using iso-lated DNA from infected HeLa cellsto demonstrate a twofold increaseof lentiviral vector copies (from0.11 ± 0.05 to 0.24 ± 0.02copies/cell; P < 0.05) following theinsertion of the cppt. Incorporationof the chicken lysozyme 5� MARdid not significantly alter thecopies of vector genomes (Fig. 4D),consistent with the viral transgeneexpression results. However, thepresence of the Ig� MAR, regardlessof the orientation, significantly

increased the number of vector genomes by ~ 2.5-fold to0.57 ± 0.04 and 0.65 ± 0.06 from 0.24 ± 0.02 copies/cell (P< 0.001). No lentiviral DNA copies were detected in the PCRreaction using genomic DNA from un-infected HeLa cells orthe control (no template) reaction. Thus, the increased �-gal transgene expression observed was due in part to anincrease in the proviral genome copy number.

Subsequent experiments were carried out using lentivi-ral vectors containing similar cis DNA elements and poten-tially therapeutic cDNAs, specifically human factor IX(FIX; Fig. 5A) and human �1-antitrypsin (AAT; Fig. 5B). Inaddition to developing lentiviral vectors with liver-spe-cific promoters, the AAT promoter including the cis-actingDNA elements from ApoE lipoprotein (A) and the hepa-tocyte control locus (H) [22], was cloned into the transfervectors to drive the expression of AAT and FIX cDNA.Different lentiviral preparations were generated andinfected into Huh7 cells (a hepatoma cell line). Human FIXlevels were measured by ELISA as described [19] after 24hours. The levels of FIX (calculated as ng/106 cells/24 h)were significantly higher using the lentiviral vector with

167


the cppt. The FIX levels were 47 ± 6 ng/106 cells/24 hoursand significantly increased to 127 ± 25 ng/106 cells/24 hours(P < 0.001) when the cppt was inserted into the integratingconstruct. The insertion of the MAR from Ig� regardless ofthe orientation significantly elevated FIX levels to 275 ± 15(sense) and 270 ± 6 ng/106 cells/24 hours (antisense), respec-

A

B

C

D

168

tively (P < 0.001). The 5� MAR from the chicken lysozymegene was inserted into the integrating plasmid, whichresulted in a statistically insignificant increase in FIX levelsto 80 ± 3 (sense) and 67 ± 7 ng/106 cells/24 h (antisense).Expression from HeLa cells infected with the lentiviral vec-tor containing the AAT(A/H) promoter with or without theIg� MAR did not produce any detectable FIX (data notshown), which was not unexpected because of the hepato-cyte specificity of this promoter.

We found that the AAT level in the medium followinginfection with the lentiviral vector without the cppt was 142± 14 ng/106 cells/24 hours (Fig. 5B). Insertion of the cpptincreased AAT levels approximately threefold to 415 ± 47ng/106 cells/24 hours (P < 0.05). Moreover, the presence ofthe Ig� MAR with the cppt resulted in a significant increasein steady-state levels of AAT in the medium to 1710 ± 178ng/106 cells/24 hours (P < 0.005). These in vitro expressionstudies demonstrated the importance of cis-acting DNA ele-ments that had a role in enhancing lentiviral titer, transgeneexpression, and viral DNA copy numbers.

In Vivo Transduction of Vectors Expressing FIX cDNA Using the AAT PromoterC57Bl/6 mice (8 weeks of age) were partially hepatectomized48 hours before the administration of lentiviral vectors (50�g p24/mouse) containing the AAT(A/H)hFIX expressioncassette with the cppt into the portal vein (Fig. 6). The pHR(-)chAAT(A/H)hFIXR(+)W(+) lentiviral vector demonstratedlong-term but sub-therapeutic levels of FIX (~ 20 ng/ml; n= 3 mice) over a 10-week period. Insertion of the chickenlysozyme 5� MAR decreased the steady-state level of FIX by~ 50% to ~ 10 ng/ml (n = 3 mice), which was consistent withthe in vitro results. On the other hand, therapeutic levels ofFIX expression were obtained from mouse serum (65 ± 10ng/ ml; n = 3), which had the lentiviral vector with the Ig�MAR administered through the portal vein. Serum FIX wasundetectable (< 1.6 ng/ml) in control (vehicle) C57Bl/6 mice(n = 6) and non-hepatectomized mice, which were givensimilar doses of lentiviral vector (50 �g p24/mouse; n = 3–6mice/group; data not shown).

FIG. 4. Effect of cis-acting elements on lentiviral vectors in vitro. Lentiviral vec-tors using transfer plasmids as diagrammed in Fig. 3 were generated andtested in HeLa cells in vitro. (A) The effect of altering the RRE from the 5� end(white bar; pHR’CMVLacZR(-)W(+)) to the 3� end (black bar;pHR(+)CMVsolLacZR(-)W(+)) of the transfer vector or increasing the numberof RREs (hatched bar; pHR’CMVsolLacZ(R+)(W+)) on transduction efficiency.HeLa cells were infected with different lentiviral vectors to demonstrate the roleof the cppt and MARs on the transduction efficiency (B), protein expression(C), and viral DNA copy number (D). The groups are denoted by the nomen-clature of the transfer plasmid (Fig. 3). Black, pHR(-)CMVsolLacZR(+)W(+);gray, pHR(-)cCMVsolLacZR(+)W(+); horizontal lines, pHR(-)cCMVsolLacZR(+)W(+)ChMAR(s); vertical lines, pHR(-)cCMVsolLacZR(+)W(+)ChMAR(as); dia-monds, pHR(-)cCMVsolLacZR(+)W(+)Ig�(s); squares, pHR(-)cCMVsolLacZR(+)W(+)Ig�(as); S, sense direction; AS, antisense direction of the MAR; n = 3–6different lentiviral preparations for each group. *Significant difference withgroups with and without the cppt (P < 0.01). Open squares, significant dif-ference with groups with and without the Ig� MAR (P < 0.0001).



DISCUSSION

Role of the cppt in Lentiviral Transduction in Mouse Hepatocytes in VivoPrevious studies in our lab have found that first- and sec-ond-generation lentiviral vector transduction into mouseor rat hepatocytes in vivo was inefficient and predomi-nantly required hepatocytes to progress into the cell cycle[4]. Investigators have found that insertion of the cpptenhances transduction of lentiviral vectors in HeLa and293T cells in vitro. Previous studies [7] demonstrated invivo that inclusion of the cppt resulted in a twofold ele-vated steady-state level of human FIX in mouse serumcompared with lentiviral vectors devoid of the cppt.However, this study did not directly determine the role ofthe cppt in vivo within the mouse liver, specificallywhether this cis element could enhance transduction intoquiescent hepatocytes. Moreover, it is possible that a sub-stantial portion of FIX gene expression comes from non-parenchymal liver, splenic cells or other non-hepatic tis-sue after intravascular administration of the vector[4,11,20].

For this reason, the role of the cppt was investigated todetermine its ability to circumvent the need for hepato-cytes to enter the cell cycle for lentiviral vector transduc-tion to occur efficiently. We found that the insertion ofthe cppt resulted in a 50% reduction in the requirement

FIG. 5. Role of cis-acting elements on FIX and AAT expression in Huh7 cells in vitro. The basic transfer plasmid backbone used to make the lentiviral vectors wassimilar to that of the CMVsolLacZ vector (Fig. 3), except the AAT promoter and FIX or AAT cDNA transgene. Huh7 cells were infected with different lentiviralvectors with the AAT promoter containing the ApoE and HCR cis-acting elements expressing FIX (A) and AAT (B) cDNA. FIX or AAT levels were analyzed by ELISA24 h following medium replacement. The groups for (A) are denoted by the nomenclature of the transfer plasmid as follows: black, pHR(-)hAAT(A/H)hFIXR(+)W(+);gray, pHR(-)chAAT(A/H)hFIXR(+)W(+); horizontal lines, pHR(-)chAAT(A/H)hFIXR(+)W(+)ChMAR(s); vertical lines, pHR(-)chAAT(A/H)hFIXR(+)W(+)ChMAR(as); dia-monds, pHR(-)chAAT(A/H)hFIXR(+)W(+)Ig�(s); squares, pHR(-)chAAT(A/H)hFIXR(+)W(+)Ig�(as); S, sense direction; AS, antisense direction of the MAR; n = 3 dif-ferent lentiviral preparations for each group. *Significant difference with group with and without the cppt (P < 0.01). Open squares, significant difference withgroups with the Ig� MAR and the group with the cppt alone (gray; P < 0.0001). The groups in (B) demonstrate the expression of AAT in the cell culture mediumof Huh7 cells. The groups are denoted by the following nomenclature of the transfer plasmid: no cppt, white, pHR(+)hAAT(A/H)hAATR(-)W(+); with cppt, black,pHR(+)chAAT(A/H)hAAT(-)W(+); and with the Igk MAR, gray, pHR(+)chAAT(A/H)hFIXR(-)W(+)Ig�(as). *Significant difference between groups with and withoutthe cppt (P < 0.001). Open squares, significant difference between groups with and without the Ig� MAR (P < 0.0001).

A B


for cell cycling of hepatocytes in vivo. Recent studies [11]have suggested that the cell cycle was not required fol-lowing intravenous administration of lentiviral vectors.Although the authors suggested that vectorologicalchanges were a possible reason for their observed differ-ences compared with second-generation vectors, severalother possibilities exist that may have contributed to thedifferent results between these two studies, which includedthe age and strain differences of the mice, but, more likely,the method by which the BrdU was administered. Pfeifer,et al. [11], administered the BrdU into the intraperitonealspace of the mice, whereas we administered the BrdU usinga subcutaneously placed osmotic minipump. It has beenpreviously shown that intraperitoneal injections or slow-releasing pellets of BrdU substantially underestimate thenumber of slowly cycling cells, such as hepatocytes [23,24],compared with osmotic minipumps. This may explain thecurious result by Pfeifer, et al. [11], who found very fewBrdU labeled hepatocytes to begin with. It is important tonote that, consistent with our previous study using highdoses of lentivirus [4], Pfeifer, et al. [11], found that thisvector system can also transduce non-parenchymal liverand splenic cells.

Moreover, because no comparisons with previous ver-sions of the vector were made, our study confirms that thecppt can, in part, alleviate the need for cell cycle progres-sion in order for lentiviral vectors to integrate into mouse

169


hepatocytes in vivo. Further biological studies are neededto elucidate the exact mechanism by which the DNA flapsequence can help in circumventing the need for cellcycling to allow for enhanced nuclear transport of lentivi-ral vectors within hepatocytes in vivo.

Effect of Cis Elements on Viral Titer andTransduction Efficiencies in VitroBecause the cppt had a role in enhancing viral titer in vitroand minimizing the requirement for hepatocytes to enterthe cell cycle to transduce lentiviral vectors in vivo, wemade further modifications to the lentiviral integratingvectors to optimize the viral titer, transgene expression,and, potentially, biosafety of the vector system. Initialalterations in the second-generation transfer plasmids weremade to delete 400 bp of the U3 region in the 3� LTR,which results in the production of a SIN vector. This wasbelieved to help in reducing the likelihood for generatingreplication-competent retroviruses from producing celllines. In addition, the integrated provirus would be unableto initiate transcription from the LTR, avoiding the possi-bility of activating a “dormant” gene within the cell andaffecting transcription from the internal promoter withinthe vector. The env sequence from the HIV genome sur-rounding the RRE was removed and replaced with a PCR-amplified RRE (253 bp) without surrounding env sequencein the 3� region of the transfer plasmid 5� to the WPRE.No significant effect on transduction efficiency was deter-mined, which suggests that the RRE can function regard-less of the position in the lentiviral transfer vector.

We found that insertion of the cppt regardless ofwhether it is near the 5� end or in the center of the trans-fer plasmid significantly increased viral transduction effi-ciency (as measured by transducing units/pg p24). We

FIG. 6. Role of the MARs on FIX levels in mouse serum in vivo. C57Bl/6 mice (8weeks) were injected into the portal vein with lentiviral vectors (50 �g/mouse)containing the chAAT(A/H)hFIX expression cassette with either the Ig� or chickenlysozyme 5� MAR. Open squares, control (vehicle) group (n = 6); filled squares,pHR(-)chAAT(A/H)hFIX(R+)W(+)Ig�(as); filled circles, pHR(-)chAAT(A/H)hFIX(R+)W(+); �, pHR(-)chAAT(A/H)hFIX(R+)W(+)ChMAR(as). *Significant differencebetween groups with and without Ig� MAR (P < 0.05).

170

found that the transduction efficiency and/or gene expres-sion increased in the presence of the cppt in a promoter-and cell-type-independent manner. The increased viraltiter and transduction efficiency was consistent with pre-vious studies examining the role of the cppt in vitro [7–10].

MARs are stretches of DNA that are thought to deter-mine chromatin structure and may have a role in open-ing specific domains for transcription to potentially aug-ment gene expression. For this reason, lentiviral constructswere developed that contained either the MAR from the5� chicken lysozyme or the Ig� gene. These MARs werepreviously found to have cis-acting enhancing effects ontransgene expression in a cell-type- and position-inde-pendent manner in vitro [14–18], but the mechanism bywhich these MARs function biologically is still unknown.Previous studies using human �-interferon MARs inmurine Moloney-based retroviruses resulted in long-termincreases in gene expression in primary T cells [12,13].Here, we found that the chicken lysozyme 5� MARdecreased the viral transduction efficiency, but did havesome cis-acting enhancer effect on transgene expression.The cis-acting enhancer effect is consistent with previousin vitro and transgenic mouse studies using the chickenlysozyme 5� MAR.

In contrast, the inclusion of the MAR from the Ig� genesignificantly enhanced the transduction efficiency andtransgene expression using either the CMV or the AATpromoter in the transfer vector. The cis-enhancing effectby the Ig� MAR was consistent with previous studies,which found that the deletion of the MAR significantlyblunted the transgene expression of the Ig� gene [18].Although the mechanism by which the Ig� MAR enhancedviral transduction efficiency and gene expression was notstudied, it may act through lentiviral integration byenhanced lentiviral RNA encapsidation or perhaps bybinding to the DNA matrix and stabilizing the vector,allowing it more favorable access to integration sites,which may aid in facilitating transcription factors to acti-vate the internal promoter of the lentivirus. Further inves-tigation into the mechanism of the Ig� MAR is needed.

Role of Cis-Acting Elements on In VivoLentiviral TransductionAlthough we have previously achieved therapeutic levelsof FIX in the serum of mice given lentiviral vectors usingthe EF1 �promoter, the dose of the viral vector was veryhigh (more than 100 �g p24) or lentiviral administrationwas performed following proliferative induction by surgi-cal partial hepatectomy [19].The results in the presentstudy also demonstrated that a liver-specific promoter(human AAT) driving the expression of FIX cDNA in apotentially more biosafe lentiviral vector produced long-term therapeutic levels (> 50 ng/ml) of FIX in partiallyhepatectomized mice (n = 3/group). Significantly lowerlevels of FIX, which were transiently expressed, have beenreported [22] using naked DNA containing an expression



cassette with FIX cDNA driven by the AAT promoter withthe ApoE and HCR cis-acting elements. From our previouswork in the context of a lentiviral vector [19], the AAT pro-moter seems to be ~ 700% weaker compared with the EF1�enhancer/promoter as determined by measuring FIX inlentiviral-treated mice. Because the EF1� promoter, butnot the AAT promoter, is ubiquitous and the lentiviralvector will transduce many cell types in the liver andspleen [4], it is likely a proportion of the serum FIX wasderived from nonhepatocytes. It is important to note thatcis-acting elements, specifically MARs from Ig�, elevatedthe serum levels of FIX to therapeutic levels. The levels ofFIX were substantially lower than that found in adeno-associated viral (AAV) vectors [25,26], which used a humanFIX “minigene” construct that contains intronic sequencesfound to enhance gene expression in vivo [22]. Future stud-ies may be performed which will include the “minigene”in lentiviral vectors along with other cis-acting elementsthat may optimize transgene expression in vivo.

This study demonstrated that the cppt could reducethe need for cell cycle progression for lentiviral vectors totransduce quiescent hepatocytes in vivo. Additionallentivector manipulation will potentially allow for the pro-duction of viral vectors, which will further reduce the needfor cell cycle progression in order for efficient transductionto occur in quiescent hepatocytes. In all, our study demon-strated that modifications in the lentiviral transfer vectorthrough the addition of cis-acting elements enhanced theviral titer and transgene expression in vitro and in vivo.

METHODS

SIN vector (3� LTR deletion). The KpnI/NheI fragment from pHR�CMVlacZ[5] was cloned into pBSII-SK(-) digested with KpnI and SpeI to makepBSLTR(KpnI/NheI), which was subsequently digested with PvuII. The DNAfragment was isolated and self-ligated to make pBSLTR(KpnI/NheI-PvuII).Then pBSLTR(KpnI/NheI) was digested with PvuII and NdeI and ligatedwith a DNA fragment from pBSLTR(KpnI/NheI-PvuII), which was digestedwith EcoRV and NdeI. This plasmid, designated pBSLTR(KpnI/NheI)SIN,was digested with StuI and an annealed oligo linker with a novel NheI sitewas inserted generating pBSLTR(KpnI/NheI)SINlink. PCR amplification ofthe 3� LTR was carried out using the following oligos: 5�-TAGCTAGCT-GCTAGAGATTTTCCACACTGA-3� and 5�-CCTGGGAGCTCTCTGGCTAA-3�. The gel-isolated PCR product was digested with SacI and NheI andinserted into pBSLTR(KpnI/NheI-PvuII)SINlink resulting in the final clone,pBSLTRSIN.

To create the SIN vector backbone, pHR2mPGKNLSLacZ was alteredsuch that the PGKNLSLacZ expression cassette was removed and replacedby a multicloning site. This plasmid, pHR2multiclone, was subsequentlydigested with KpnI, and NheI allowed for the ligation with the KpnI/NheIfragment from pBSLTRSIN, resulting in the making ofpHR2PacI/PmeIlinkerSIN. This construct has removed the remaining 1456bp of human sequence from the original cloning of the HXB2 proviralgenome downstream of the 3� SIN LTR.

SIN vector with WPRE insertion. Initial steps were carried out to removethe RRE. The XhoI/NotI fragment from pHR2mPGKNLSLacZ was isolatedand cloned into pBS.SKII(-), making pBS(XhoI/NotI). This plasmid was sub-sequently digested with EcoNI and HindIII, which removed 509 bp of theenvelope sequence including the 253 bp of the RRE, and then bluntedwith T4 DNA polymerase. The blunted linear fragment was isolated and


either self-ligated resulting in pBS(XhoI/NotI-EcoNI/HindIII) or ligated withblunted BamHI/BglII RRE fragment to generate pBS(XhoI/NotI-EcoNI/HindIII)RRE253. The I/NotI fragment was inserted into thepHR2PacI/PmeIlinkerSIN resulting in a plasmid, which was devoid of theRRE and known as pHR2PacI/PmeIlinkerSINRRE(-). Subsequently, theWPRE sequence was cloned from WPRE-BSKII (provided by Thomas J.Hope, Salk Institute, La Jolla, CA) as a BamHI/KpnI fragment intopHR2PacI/PmeIlinkerSINRRE(-).

Central polypurine tract. PCR amplification of the 118 bp frompCMV{D}R8.74, which is found in pol, was performed using primers asdescribed [7] with a slight modification at the ends of each primer to incor-porate unique restriction sites to facilitate cloning into different transferplasmids. An example of a primer set used with the addition of restrictionsite is as follows: 5�-GATGGATCCTTTTAAAAGAAAAGGGGGGAT-3� (sense)and 5�-CGAAGATCTAAAATTTTGAATTTTTGTAAT-3� (antisense). PCRamplification was performed as follows: 94°C for 3 min, then 30 cycles of94°C for 30 s, 60°C for 30 s and 72°C for 7 s, followed by an extension at72°C for 7 min.

Human AAT promoter with ApoE/hepatocyte control locus (HCR) enhancer.pBS-ApoE-HCR-hAAT-FIXmg-bpA as described [22] was digested with SpeI andSalI (1.1 kb) and inserted into the pHR2EF� linker [19] creatingpHRhAAT(A/H). pHR2hAAT(A/H)hFIX was made by inserting the FIX cDNA(BamHI/BamHI fragment) from pHR2EF1� hFIX [19]. The cppt was PCR ampli-fied and inserted into the XhoI and SpeI site of the pHR2hAAT(A/H)hFIX tomake pHR2chAAT(A/H)hFIX. The XhoI/BglII fragment frompHR2hAAT(A/H)hFIX and pHR2chAAT(A/H)hFIX was inserted intopHR2PacI/PmeIlinkerSINRRE(-) to create pHR(-)hAAT(A/H)hFIXR(-)W(+) andpHR(-)hAAT(A/H)hFIXR(-)W(+). The RRE (253 bp) was PCR amplified andinserted into a novel SmaI site 5� to the WPRE sequence to generate pHR(-)hAAT(A/H)hFIXR(+)W(+) and pHR(-)chAAT(A/H)hFIXR(+)W(+).

AAT cDNA expression cassette construction. pHR(-)hAAT(A/H)hFIXR(+)W(+) was digested with XhoI and NotI allowing for ligation with a XhoI/NotI frag-ment from pBS(XhoI/NotI-EcoNI/HindIII)RRE253. This construct was thendigested with EcoRI and the AAT cDNA (EcoRI/EcoRI fragment was inserted tomake pHR(+)hAAT(A/H)hAATR(-)W(+)). The Ig� MAR was inserted into theXhoI site to create pHR(+)hAAT(A/H)hAATR(-)W(+)Ig�(as).

Cytoplasmic LacZ expression cassette construction. The CMV promoter(blunted ClaI/blunted BamHI) was isolated from pHR�CMVLacZ and insertedinto the blunted SpeI site of pHR2chAATNLSLacZ to makepHR2cCMVNLSLacZ. The NdeI/NdeI fragment from pHR�CMVLacZ wasinserted into the NdeI/NdeI site of pHR2cCMVNLSLacZ to createpHR2cCMVsolLacZ.

PHR�CMVlacZ was digested with KpnI and NheI and the KpnI/NheI frag-ment from pBSLTRSIN was ligated to make pHR�CMVlacZSIN-18F. This plas-mid was subsequently digested with NotI and XhoI and the NotI/XhoI frag-ment from pBS(XhoI/NotI-EcoNI/HindIII) was inserted to makepHR(-)CMVlacZSIN-18F. The WPRE fragment (EcoRI/KpnI) was inserted tomake pHR(-)CMVlacZSIN-18W(+)F. The final step was the inclusion of theRRE (253 bp; EcoRI/EcoRI) to make pHR(-)CMVsolLacZR(+)W(+) andpHR�CMVlacZR(+)W(+). The XhoI/NdIeI fragment from pHR2cCMVsolLacZwas inserted into pHR(-)CMVsollacZR(+)W(+) to create pHR(-)cCMVsollacZR(+)W(+).

Ig� MAR. The Ig� MAR (415 bp) was PCR amplified from mouse genomicDNA. This region of DNA, spanning 3361–3774 bp of the Ig� gene asdescribed [27], was cloned into the SrfI site of pPCR using the PCR-ScriptAmp Cloning kit, and was subsequently cloned into pBS-SKII(-). Ig� MAR(XhoI/XhoI fragment) was inserted 5� to the cppt in pHR(-)chAAT(A/H)hFIXR(+)W(+) to create pHR(-)chAAT(A/H)hFIXR(+)W(+)Ig�(s) and pHR(-)chAAT(A/H)hFIXR(+)W(+)Ig�(as).

Chicken lysozyme gene MAR. The 5� MAR from the chicken lysozymegene (2.77 kb; XbaI/XbaI fragment) was cloned from pBS-2x(B-1-X1),which was provided by Wolf H. Stratling (Hamburg, Germany), into anovel XbaI site flanked by XhoI sites in pBS-SKII(-). pBS.ChickenMAR wasdigested with XhoI and the MAR fragment was cloned into the XhoI siteof pHR(-)]hAAT(A/H)hFIXR(+)W(+), and both orientations of the chickenMAR were isolated (pHR(-)hAAT(A/H)hFIXR(+)W(+)ChMAR).

171


XhoI fragments for chicken lysozyme 5� MAR were inserted into the novel XhoI site of pHR(-)cCMVsolLacZR(+)W(+) to make pHR(-)]CMVsollacZR(+)W(+)ChMAR(s) and pHR(-)CMVsollacZR(+)W(+)ChMAR(as).

Envelope and packaging construct. pMD.G contains the expressionsequence for the pseudotyped VSV-G envelope as described [22]. ThepCMVR8.74 was used as the packaging construct [22].

Authenticity of PCR products. The accuracy of the PCR fragments usedfor cloning was assessed by PCR sequencing of the plasmids (Biotech Core,Inc., Mountain View, CA).

Viral production and assays. The 293T cells (provided by Garry P. Nolan,Stanford University) were transiently transfected to produce replication-defective HIV-1 lentivirus as described [4,19]. For in vitro and in vivo stud-ies, lentiviral preparations were generated in 10-cm and 15-cm plates,respectively. The medium (10 ml and 15 ml for in vitro and in vivo prepa-rations, respectively) was replaced 12 h following plasmid transfection andthe conditioned medium was collected 30–36 h later. The viral vectorswere filtered and concentrated as described [20]. The final pellet was resus-pended in PBS in the presence of Polybrene (8 �g/ml).

To determine the titer of the HIV stocks for in vivo delivery of thelentivirus, serial dilutions of the vector stocks were made and infected into5 � 105 HeLa cells in a 6-well plate in the presence of Polybrene (8 �g/ml).In addition, HIV p24 Gag antigen was measured by ELISA (Alliance;Dupont-NEN) following serial dilution of the concentrated virus. The trans-duction efficiency for each individual lentiviral preparation was calculatedas T.U./pg p24. Vector batches were tested for the absence of replication-competent lentivirus by monitoring p24 antigen expression in the culturemedium of transduced SupT1 lymphocytes for 3 weeks. In all cases tested,p24 antigen was undetectable (detection limit, 3 pg/ml).

�-Galactosidase protein and DNA analysis. HeLa cells (2 � 106 cells) wereinfected with 100 ng p24 lentiviral vectors (three individual dishes pervector), the medium was changed 24 h later, and the HeLa cells were iso-lated 24 h later and separated into two tubes, which allowed for the isola-tion of protein lysates and genomic DNA from the same cell preparations.Protein lysates were analyzed by ELISA as described by the manufacturer(Boerhinger Mannheim, Germany). Quantitative PCR was performed usingthe GeneAmp 5700 Sequence Detection System (PE Biosystems, Foster City,CA). The primers used to detect lentiviral vector genomes were as follows:5�-CCGCGAGGTGCGGATTGAAAAT-3� (sense) and 5�-AAGTTGTTCT-GCTTCATCAGCAGG-3� (antisense). The amplification procedure was 95°Cfor 5 min, then 95°C for 30 s and 60°C for 30 s for 40 cycles.

ELISA for human FIX and human AAT. Mouse serum was obtained viaretro-orbital bleeding or cell culture conditioned medium was obtained formeasurement by ELISA as described [19,28].

Animal preparation. Female C57Bl/6 and C57Bl/6 scid mice (8 weeks ofage) were purchased from Jackson Laboratories. All animal protocols wereperformed according to the Stanford University and NIH guidelines. Micewere anesthetized with a mixture of Ketamine (6 mg/kg) and acepromazine(2 mg/kg) before surgical manipulations. Replication-defective lentiviralparticles (50 �g p24, which is an index of viral titer) expressing FIX wereinfused into the portal vein at a volume of 0.5 ml for mice over a periodof 120 s using a 30-gauge needle. In the human FIX experiments, a 2/3 par-tial hepatectomy was performed 48 h before lentiviral injection in somemice, which represents the maximal period of liver regeneration for mice.

BrdU and X-gal double-labeling experiment. pHR2EF1�cytoLacZ [4] wasdigested with XhoI and SpeI to insert the cppt (118 bp) as described [7].Lentiviral vectors expressing cytoplasmic LacZ driven by the EF1� enhancerpromoter in the presence or absence of the cppt (2 � 108 T.U./mouse) wereinfused into the portal vein of C57Bl/6 scid mice. Osmotic minipumps(model 2001; Alzet) were placed subcutaneously to administer BrdU (1mg/day) for 7 d. The animals were sacrificed and the liver and duodenum(positive control for BrdU incorporation) were harvested, fixed in OCTbuffer on dry ice, and stored frozen at –80°C.

Sections (7 �m) were stained for �-gal expression using 5-bromo-4-chloro-3-indolyl-�-D-galactoside (X-gal; Fisher Scientific) overnight. Thetissue was then processed for BrdU incorporation as follows. First, endoge-

172

nous peroxidase activity was removed by incubation with 0.3% hydrogenperoxide and the tissue was blocked in 10% normal goat serum (VectorLabs) overnight at 4°C. The tissue sections were incubated with a primaryantibody against BrdU (1:200; mouse anti-BrdU; Becton Dickinson) for 1h and the secondary antibody (1:500) was used for 30 min. 3�-3�-Diaminobenzidine (DAB; Sigma) was the brown substrate used for the pre-cipitation of the BrdU sample. The tissue sections were then counterstainedwith hematoxylin, dehydrated, and coverslipped for analysis. To minimizedifferences in BrdU labeling in liver lobes, 2–4 separate lobes of the mouseliver were counted for BrdU-labeled nuclei (1500 nuclei in all).

ALT assay. Mouse serum was obtained at days 1 and 3 after vector infu-sion to measure serum concentrations of ALT, which was determined usinga colorimetric diagnostic kit (Sigma, St. Louis, MO) as described [4].

Statistical analysis. The significance of differences between groups wastested by a one-way ANOVA with the use of StatView 5.0 software (SASInstitute Inc., Cary, NC). If a probability value of P < 0.05 was obtained,the Tukey test was then used for comparison for each individual group withthe appropriate control.

ACKNOWLEDGMENTSThis work was supported by NIH-R01 HL64274. F.P. is a recipient of the JudithGraham Pool Fellowship through the National Hemophilia Foundation.

RECEIVED FOR PUBLICATION JUNE 11; ACCEPTED JULY 5, 2001.

REFERENCES1. Kay, M. A., et al. (1993). In vivo gene therapy of hemophilia B: sustained partial cor-

rection in factor IX-deficient dogs. Science 262: 117–119.2. Miller, D., Adam, M., and Miller, A. (1990). Gene transfer by retrovirus vectors occurs

only in cells that are actively replicating at the time of infection. Mol. Cell. Biol. 10:4239–4242.

3. Miyoshi, H., Takahashi, M., Gage, F. H., and Verma, I. M. (1997). Stable and efficientgene transfer into the retina using an HIV-based lentiviral vector. Proc. Natl. Acad. Sci.USA 94: 10319–10323.

4. Park, F., Ohashi, K., Chiu, W., Naldini, L., and Kay, M. A. (2000). Efficient lentiviral trans-duction of liver requires cell cycling in vivo. Nat. Genet. 24: 49–52.

5. Naldini, L., et al. (1996). In vivo gene delivery and stable transduction of nondividingcells by a lentiviral vector. Science 272: 263–267.

6. Ohashi, K., Park, F., and Kay, M. A. (2001). Development of experimental models forliver regeneration in mice: the effects on lentiviral mediated gene transduction in vivo.Mol. Ther. 3: S128.

7. Follenzi, A., Ailles, L. E., Bakovic, S., Geuna, M., and Naldini, L. (2000). Gene transferby lentiviral vectors is limited by nuclear translocation and rescued by HIV-1 polsequences. Nat. Genet. 25: 217–222.

8. Zennou, V., et al. (2000). HIV-1 genome nuclear import is mediated by a central DNAflap. Cell 101: 185–193.

9. Zennou, V., et al. (2001). The HIV-1 DNA flap stimulates HIV vector-mediated cell trans-duction in the brain. Nat. Biotechnol. 19: 446–450.

10. Mautino, M. R., Ramsey, W. J., Reiser, J., and Morgan, R. A. (2000). Modified humanimmunodeficiency virus-based lentiviral vectors display decreased sensitivity to trans-dominant Rev. Hum. Gene Ther. 11: 895–908.

11. Pfeifer, A., et al. (2001). Transduction of liver cells by lentiviral vectors: analysis in livinganimals by fluorescence imaging. Mol. Ther. 3: 319–322.

12. Agarwal, M., et al. (1998). Scaffold attachment region-mediated enhancement of retro-viral vector expression in primary T cells. J. Virol. 72: 3720–3728.

13. Dang, Q., Auten, J., and Plavec, I. (2000). Human � interferon scaffold attachmentregion inhibits de novo methylation and confers long-term, copy number-dependentexpression to a retroviral vector. J. Virol. 74: 267–2678.

14. Phi-Van, L., and Stratling, W. H. (1996). Dissection of the ability of the chicken lysozymegene 5� matrix attachment region to stimulate transgene expression and to dampen posi-tion effects. Biochemistry 20: 10735–10742.

15. McKnight, R. A., Shamay, A., Sankaran, L., Wall, R. J., and Hennighausen, L. (1992).Matrix-attachment regions can impart position-independent regulation of a tissue-spe-cific gene in transgenic mice. Proc. Natl. Acad. Sci. USA 89: 6943–6947.

16. Phi-Van, L., von Kries, J. P., Ostertag, W., and Stratling, W. H. (1990). The chickenlysozyme 5� matrix attachment region increases transcription from a heterologous pro-moter in heterologous cells and dampens position effects on the expression of trans-fected genes. Mol. Cell. Biol. 10: 2302–2307.

17. Okada, S., Tsutsui, K., Tsutsui, K., Seki, S., and Shohmori, T. (1996). Subdomain struc-ture of the matrix attachment region located within the mouse immunoglobulin k geneintron. Biochem. Biophys. Res. Commun. 222: 472–477.

18. Goyenechea, B., et al. (1997). Cell strongly expressing Ig� transgenes shown clonalrecruitment of hypermutation: a role for both MAR and the enhancers. EMBO J. 16:



3987–3994.19. Park, F., Ohashi, K., and Kay, M. A. (2000). Therapeutic levels of human factor VIII and

IX using HIV-1 based lentiviral vectors in mouse liver. Blood 96: 1173–1176.20. Dull, T., et al. (1998). A third-generation lentivirus vector with a conditional packaging

system. J. Virol. 72: 8463–8471.21. Zufferey, R., Donello, J. E., Trono, D., and Hope, T. J. (1999). Woodchuck hepatitis virus

posttranscriptional regulatory element enhances expression of transgenes delivered byretroviral vectors. J. Virol. 73: 2886–2892.

22. Miao, C. H., et al. (2000). Inclusion of the hepatic locus control region, an intron, anduntranslated region increases and stabilizes hepatic factor IX gene expression in vivo butnot in vitro. Mol. Ther. 1: 522–532.

23. Weghorst, C. M., Henneman, J. R., and Ward, J. M. (1991). Dose response of hepaticand renal DNA synthetic rates to continuous exposure of bromodeoxyuridine (BrdU) viaslow-release pellets or osmotic minipumps in male B6C3F1 mice. J. Histochem. Cytochem.39: 177–184.


24. Eldridge, S. R., Tilbury, L. F., Goldsworthy, T. L., and Butterworth, B. E. (1990).Measurement of chemically induced cell proliferation in rodent liver and kidney: a com-parison of 5-bromo-2�-deoxyuridine and [3H]thymidine administered by injection orosmotic pump. Carcinogenesis 11: 2245–2251.

25. Nakai, H., et al. (1998). AAV-mediated gene transfer of human blood coagulation fac-tor IX into mouse liver. Blood 91: 4600–4607.

26. Snyder R. O., et al. (1997). Persistence and therapeutic concentrations of human factorIX in mice after hepatic gene transfer of recombinant AAV vectors. Nat. Genet. 16:270–276.

27. Max, E. E., Maizel, Jr., J. V., and Leder, P. (1981). The nucleotide sequence of a 5.5-kilo-base DNA segment containing the mouse � immunoglobulin J and C region genes. J.Biol. Chem. 256: 5116–5120.

28. Kay, M. A., Graham, F., Leland, F., and Woo, S. L. C. (1995). Therapeutic serum con-centration of human �1-antitrypsin after adenoviral-mediated gene transfer into mousehepatocytes. Hepatology 21: 815–819.

173

Documents

Modified HIV-1 Based Lentiviral Vectors Have an Effect on ...med.stanford.edu/content/dam/sm/kaylab/documents/MT.9.2001.PARK.KAY.pdf · cppt (41% versus 81%, respectively; Table 1)