bio-science

jhh19970724

the_journal_of_gene_medicine.pdf

Home >Biology homework help >bio-science

R E V I E W A R T I C L E

Gene therapy for hemophilia

Marinee K. L. Chuah

Desire Collen

Thierry VandenDriessche*

Center for Transgene Technology and Gene Therapy, Flanders Interuniversity Institute for Biotechnology, University of Leuven, 49 Herestraat, B-3000 Leuven, Belgium

*Correspondence to: T. VandenDriessche, Center for Transgene Technology and Gene Therapy, Flanders Interuniversity Institute for Biotechnology, University of Leuven, 49 Herestraat, B-3000 Leuven, Belgium. E-mail: thierry.vandendriessche @med.kuleuven.ac.be or marineekhim.chuah@med. kuleuven.ac.be

Received: 3 October 2000

Revised: 13 November 2000

Accepted: 14 November 2000

Summary

Hemophilia A and B are X-chromosome linked recessive bleeding disorders that result from a de®ciency in factor VIII (FVIII) and factor IX (FIX) respectively. Though factor substitution therapy has greatly improved the lives of hemophiliac patients, there are still limitations to the current treatment that have triggered interest in alternative treatments by gene therapy. Signi®cant progress has recently been made in the development of gene therapy for the treatment of hemophilia A and B. These advances parallel the technical improvements of existing vector systems including MoMLV-based retroviral, adenoviral and AAV vectors, and the development of new delivery methods such as lentiviral vectors, helper-dependent adenoviral vectors and improved non-viral gene delivery methods. Thera- peutic and physiologic levels of FVIII and FIX could be achieved in FVIII- and FIX-de®cient mice and hemophilia dogs by different gene therapy approaches. Long-term correction of the bleeding disorders and in some cases a permanent cure has been realized in these preclinical studies. However, the induction of neutralizing antibodies often precludes stable phenotypic correction. Another complication is that certain promoters are prone to transcriptional inactivation in vivo, precluding long-term FVIII or FIX expression. Several gene therapy phase I clinical trials are currently ongoing in patients suffering from severe hemophilia A or B. No signi®cant adverse side-effects were reported, and semen samples were negative for vector sequences by sensitive PCR assays. Most importantly, some subjects report fewer bleeding episodes and occasionally have very low levels of clotting factor activity detected. The results from the extensive preclinical studies in normal and hemophilic animal models and encouraging preliminary clinical data indicate that the simultaneous development of different strategies is likely to bring a permanent cure for hemophilia one step closer to reality. Copyright # 2001 John Wiley & Sons, Ltd.

Keywords hemophilia; factor VIII; factor IX; coagulation; gene therapy

Rationale and general principles of hemophilia gene therapy

Hemophilia is characterized by spontaneous and prolonged bleeding in the joints, muscle and internal organs [1]. It is potentially life-threatening and is often associated with disabling arthropathy resulting from the recurring joint bleeding episodes. Hemophilia A and B are congenital X-chromosome linked coagulation disorders, due to a de®ciency in coagulation factor VIII (FVIII) or factor IX (FIX), respectively, which are normally expressed in the liver. Hemophilia A occurs in 1 in 10 000 whereas hemophilia B affects 1 in 30 000 males. FVIII has no intrinsic enzymatic activity and functions as a co-factor to accelerate the activation of factor X by activated FIX in the presence of calcium and phospholipids. Ultimately, the coagulation cascade leads to the localized generation of thrombin and conversion of ®brinogen to insoluble

THE JOURNAL OF GENE MEDICINE J Gene Med 2001; 3: 3±20.

®brin polymers, which in conjunction with platelet aggregation maintains hemostasis.

Current treatment for hemophilia consists of protein replacement therapy in response to bleeding crises with infusion of plasma-derived or, more recently, recombi- nant FVIII or FIX. Although this treatment markedly improves both the life expectancy and the quality-of-life of the hemophilia patients, it has several drawbacks. The treated patients are still at risk for life-threatening bleeding episodes and chronic joint damage, especially since prophylactic treatment is restricted by the limited availability and high cost of puri®ed FVIII and FIX. In some industrialized countries, patients with hemophilia have problems getting adequate health insurance cover- age and therapy is virtually unavailable for most patients in developing countries.

Prior to the development of viral screening and inactivation techniques, plasma-derived FVIII and FIX has resulted in transmission of HIV and hepatitis in the hemophiliac population. Hepatitis B virus (HBV) and especially chronic hepatitis C (HCV) infections currently affect nearly all adult hemophilic patients. Though HCV often persists asymptomatically, it can lead to progressive liver disease and chronic hepatocyte proliferation and hepatoma and is now the second leading cause of death in the adult hemophilia population. Though current viral inactivation procedures such as solvent-detergent treat- ment effectively kills all lipid-enveloped viruses this type of treatment is less effective for non-enveloped viruses such as hepatitis A virus. Other techniques include heat- inactivation (pasteurisation) but this has lead to several outbreaks of neutralizing antibodies to FVIII (clinically referred to as inhibitors; see below), possibly due to conformational changes of the FVIII protein which may have resulted in the exposure of neo-epitopes [2]. Despite the improvements in viral inactivation and puri®cation, transmission of other blood-borne viruses or infectious agents such as prions, the causative agent of Creutzfeldt- Jakob disease, cannot be excluded when plasma-derived products are used. This is a concern even for recombinant FVIII products since they are still stabilized with human plasma-derived albumin.

An important side-effect of clotting factor substitution therapy is that some patients develop neutralizing antibodies against FVIII or FIX and render further substitution ineffective. Inhibitors occur in 10±40% of the hemophilia A patients and in about 5% of the hemophilia B patients that are treated by protein replacement therapy. The exact reason why certain patients develop inhibitors remains elusive but appears to depend on several confounding variables including treatment regimen or the products used and genetic factors, such as the nature of the underlying mutation in the FVIII or FIX gene. The majority of reported inhibitor cases have nonsense mutations or deletions in their FVIII gene. In contrast, patients with missense mutations rarely develop inhibitory antibodies. The inhibitory response to FIX is signi®cantly less simply because the smaller size of the FIX gene minimizes the number of deletions. It has

also been suggested that major trauma, surgery or in¯ammation may be a co-factor contributing to the development of inhibitors. Within hemophilic brother pairs, it is not unusual for only one to develop an inhibitor implicating other acquired in¯uences in inhibitor devel- opment. The development of strategies that prevent inhibitor formation represents one of the most important challenges associated with hemophilia treatment.

Hemophilia is well suited for gene therapy since it is due to a single gene defect and since the therapeutic window is relatively broad (reviewed also in [1,3±14]). Whereas levels as high as 150% of normal levels are not associated with any thrombotic side-effects, a slight increase in plasma FVIII or FIX levels can potentially convert severe to mild hemophilia. The concentration of FVIII in normal plasma is low (100±200 ng/ml), but it is relatively dif®cult to express high levels of FVIII. Instead, FIX is easier to express, but the normal plasma level is high (5 mg/ml). Gene therapy for hemophilia can potentially provide constant, sustained FVIII or FIX synthesis within the patient, thereby obviating the risk of spontaneous bleeding, the need for repeated FVIII or FIX infusions and the risk of viral infections associated with plasma-derived FVIII or FIX. Gene therapy for hemophilia could provide a cure for this disease.

Ex vivo and in vivo approaches are being pursued for hemophilia gene therapy. Ex vivo gene therapy involves the isolation of cells from the patient followed by expansion and genetic modi®cation in culture with vectors expressing either FVIII or FIX and subsequent re-administration of the engineered cells to the patient. Successful ex vivo gene therapy requires ef®cient engraft- ment of the engineered cells leading to sustained FVIII or FIX production in the circulation. Different cell types including skin ®broblasts [15±17], keratinocytes and skin cells [18±20], endothelial cells [21±23], epithelial cells [21], myoblasts [24±28], hematopoietic stem/progenitor cells [29,30] and bone marrow (BM) stromal cells [31] have been considered as potential targets for hemophilia gene therapy and express FVIII and/or FIX following gene transfer. Most ex vivo gene therapy strategies have relied on the use of retroviral vectors and to a more limited extent on non-viral gene transfer systems. Alternatively, in vivo gene therapy involves the administration of a gene transfer vector encoding FVIII or FIX directly to the patient leading to in situ genetic modi®cation of the target cells. Intravenous administration of a FVIII or FIX vector would provide a more cost-effective treatment than ex vivo protocols that involve ex vivo expansion of the target cells and sometimes require surgical procedures, which could be undesirable in hemophilia patients. However, the main disadvantage of in vivo gene therapy based on viral vectors is that a host immune response towards the viral vector would preclude vector re-administration, if more than one injection would be required to achieve therapeutic FVIII or FIX levels. Repeated administration of ex vivo engineered cells would be possible instead, provided the engineered cells do not express foreign antigens. Most importantly, germline gene transfer

4 M. K. L. Chuah et al.

cannot be excluded following systemic administration of gene therapy vectors, whereas germline gene transfer would not occur following ex vivo gene therapy (provided no replication-competent viruses were generated by homologous recombination), which is an important safety and ethical issue. Finally, unlike most ex vivo gene therapy approaches, direct in vivo gene transfer can lead to the inadvertent transduction of antigen-presenting cells, depending partly on the type of vector used, which could in turn trigger an immune response directed against the transduced cells and/or the transgene products.

Since FVIII and FIX are naturally produced in the hepatocytes (and also in hepatic sinusoidal endothelial cells in the case of FVIII) [32], the liver appears to be an attractive target for hemophilia gene therapy. FVIII or FIX produced in the transduced hepatocyte should have readily access to the circulation. In addition, access of FVIII to von Willebrand factor (vWF) in the circulation or to vWF produced in sinusoidal endothelial cells promotes stable secretion of FVIII. In contrast, coagulation factors produced in non-hepatic tissues might not be properly processed due to differences in post-translational mod- i®cation that might in turn affect the speci®c activity and/or the potential immunogenicity. However, several independent studies have shown that even certain cell types of non-hepatic origin have the capacity for proper post-translational modi®cation of the FVIII or FIX protein. Moreover, the recombinant clotting factors that are currently used for substitution therapy have also been produced in vitro in non-hepatic cells [such as Chinese hamster ovary (CHO) cells] [1] with no evidence of increased immunogenicity or decreased speci®c activity compared to plasma-derived FVIII or FIX. However, the liver may not necessarily be the ideal target organ for hemophilia gene therapy, particularly in hemophiliacs that are infected with HBV and/or HCV viruses. The prevalence of hepatitis in the adult hemophilia population represents an additional level of complexity that can potentially interfere with the outcome of the gene therapy strategy, especially when the liver is targeted. In hemophiliacs with chronic HCV infection liver-directed gene transfer may in¯uence HCV progression and vice versa. Hence the outcome of such a complex interaction between the resident virus and the gene delivery approach cannot be predicted and will depend on several confounding variables including baseline liver disease, the type of vector used, the transduced cell types and the immune response.

Hemophilia gene therapy requires the use of a gene delivery system that is ef®cient, safe, non-immunogenic and allows for long-term gene expression. Most impor- tantly, gene therapy for hemophilia A and B must compare favourably with existing protein replacement therapies. The availability of animal models including FVIII and FIX-knockout mice [33±36] and hemophilia A and B dogs [37,38], which mimic the clinical symptoms of hemophilia, signi®cantly facilitated preclinical ef®cacy and safety studies of gene therapy strategies. Both viral vectors as well as non-viral vectors have been considered

for the development of hemophilia gene therapy. In general, viral vector-mediated gene transfer is far more ef®cient than non-viral gene transfer and has therefore been the method of choice. These vectors include retroviral, lentiviral, adenoviral and adeno-associated viral (AAV) vectors, each with their own advantages and limitations. However, there are several reasons why a non-viral treatment would still be desirable. Non-viral vectors can be assembled in cell-free systems from well- de®ned components and have the potential to be less immunogenic than viral vectors. Hybrid non-viral vector systems that include some viral components to increase gene transfer ef®ciency [39,40] may combine the advantages of both systems and constitute an attractive alternative for hemophilia gene therapy. Details of each of these gene therapy strategies for hemophilia will be discussed below (Tables 1 and 2).

Gene therapy for hemophilia A

The FVIII gene has been cloned, is 186 kb in length and is spread over 26 exons. This is too large to be packaged ef®ciently into any commonly used gene therapy vector [41±43]. The gene encodes a 9029 nucleotide FVIII mRNA which is translated in a single chain polypeptide of 2351 amino acids that is post-translationally glycosylated and sulphated at various positions. The FVIII protein consists of three types of domains: the triplicated A-domain of 330±380 amino acids, a unique B-domain of 908 amino acids and a duplicated C-domain of about 160 amino acids, and are organized in the order A1-A2-B-A3-C1-C2. The A-domain is important for Ca2+ binding and the C-domains are essential for phospholipid and vWF binding. In the plasma, the FVIII protein predominately exists as an inactive two chain derivative consisting of a 200 kDa heavy chain (A1-A2-B) and a 80 kDa light chain (A3-C1-C2) that are stabilized by Ca2+ and by vWF. The FVIII protein is subsequently activated by the progressive proteolytic cleavage of the heavy chain by factor Xa or thrombin. The B-domain is not required for FVIII function and is post-translationally cleaved from the FVIII heavy chain during proteolytic activation. A fully functional 4.6 kb version, which is derived from the FVIII cDNA by deleting the central B-domain without affecting its activity or immunogenicity [44], can be easily packaged into most gene therapy vectors. This B-domain deleted FVIII cDNA had been shown to express a higher level of FVIII protein as compared to the full-length FVIII cDNA. B-Domain deleted FVIII has already been used to treat hemophilia with similar pharmacokinetics as plasma- derived FVIII [45] and no adverse side-effects were noted.

FVIII production using the human FVIII cDNA has been dif®cult. First, the FVIII cDNA has been found to contain sequences that repress its expression, resulting in low levels of FVIII mRNA [17,21,46±50]. Some sequences (i.e. the 1.2 kb INS element) [21,47,51] inhibit transcrip- tional elongation; other elements seem to have an inhibitory effect on transcription initiation, presumably

Hemophilia Gene Therapy 5

Table 1. Gene therapy in hemophilia A animal models

Vector Promoter Transgene Features Animal Dose Site

Max. level (%) Correction Stability

Duration (> 1%)

aFVIII Ab Comment Ref.

MoMLV LTR hFVIIIDB Intron FVIII-KO 1±2r108 TU iv 1250 + (Survival) + > 14 m t Neonates [77] MoMLV LTR hFVIIIDB Intron FVIII-KO

/SCID Ex vivo is 15 NT ± 1 w ± BM stroma NP

MoMLV LTR hFVIIIDB Intron Hemophilia A dog

14r108

TU/kg iv + (WBCT) ± 1-2 w + [85]

HIV CMV cFVIIIDB WPRE /cPPT

FVIII-KO /SCID

Ex vivo iv 25 NT ± 1 w ± BM stroma NP

Ad Albumin hFVIIIDB Intron FVIII-KO 4±6r1010 vp iv 210 + (Survival) ± > 9 m ± DE1DE3 [109] 255 + (Survival) DE1DE2aDE3

Ad Albumin cFVIIIDB Intron FVIII-KO 6r1010 vp iv 700 + (aPTT) ± > 12 m NT DE1DE2aDE3 [111] Ad Albumin hFVIIIFL Intron FVIII-KO 1011 vp iv < 1 + (Survival) ± NT NT DE1DE2aDE3 [110] Ad Albumin cFVIIIDB Intron Hemophilia

A dog 3r1010

vp/kg iv 800 + (ACT/aPTT) ± 1±2 w + DE1DE3 [100]

Ad CMV mFVIIIDB Intron FVIII-KO 3r107 pfu ip 600 NT ± 3 w NT In utero [114] Ad CMV mFVIIIDB Intron FVIII-KO 1011 vp iv 100 + (Survival) ± 4 w + DE1DE3 [113] Ad Albumin hFVIIIFL Intron

/stuffer FVIII-KO 2r1011 vp iv 800 + (Survival) ± > 12 m + Gutless [124]

AAV, Adeno-associated virus; ACT, activated clotting time; Ad, adenoviral vector; a-FVIII Ab, FVIII-speci®c neutralizing antibodies (inhibitors); a-FIX Ab, FIX- speci®c neutralizing antibodies (inhibitors); aPTT, (activated) partial thromboplastin time; BM, bone marrow; BT, bleeding time; cFVIIIDB, canine B-domain deleted factor VIII cDNA; CMV, human cytomegalovirus promoter; cPPT, central polypurine tract; duration, duration of expression of therapeutic levels (> 1%); D, deleted; EF-1a, elongation factor 1-a; ex vivo, ex vivo gene therapy approach; FVIII-KO, factor VIII knock-out hemophilic mouse; FVIII-KO/SCID, severe combined immune de®cient factor VIII knock-out hemophilic mouse; hFVIIIDB, human B-domain deleted factor VIII cDNA; hFVIIIFL, human full-length factor VIII cDNA; HIV, human immunode®ciency virus; ip, intraperitoneal; ipv, intraportal vein; iv, intravenous; is, intrasplenic; in utero, injection in utero; LSP, liver-speci®c promoter : thyroid hormone-binding globulin promoter combined with a1-microglobuline/bikunin enhancer and intron-containing leader sequence; LTR, long terminal repeat; m, months; max. level, maximum level of expression as a percentage relative to normal physiologic levels; mFVIIIDB, mouse B-domain deleted factor VIII cDNA; MoMLV, Moloney murine leukemia virus; neonates, injection in 2±3-day-old animals; NT, not tested; NP, not published (VandenDriessche et al.); PHx, partial (2/3) hepatectomy; pfu, plaque forming units; stuffer, genomic albumin fragment; survival, survival following injury (tail-clipping); Tn, transposon; ts125, conditional temperature sensitive mutation (in adenoviral E2A protein); RSV, Rous sarcoma virus long terminal repeat; TU, transducing units; UTR, untranslated region; vp, vector particles; w, weeks; WBCT, whole blood clotting time; WPRE, woodchuck hepatitis virus post-regulatory element; Ref., reference number; t, heterogenous antibody response (not all recipients develop Abs to FVIII or FIX).

Table 2. Gene therapy in hemophilia B animal models

Vector Promoter Transgene Features Animal Dose Site

Max. level (%) Correction Stability

Duration (> 1%)

aFIX Ab Comment Ref.

MoMLV LTR cFIX Hemophilia B dog

iv 0.1 + (aPTT, WBCT) + > 9 m ± PHx [73]

Ad CMV mFIX FIX-KO 109 pfu iv 155 + (aPTT) ± 1±2 w NT DE1 [33] Ad CMV mFIX 3kUTR FIX-KO 1011 vp iv 1±2 + (aPTT, BT) ± > 12 w t DE1 [154] Ad RSV cFIX Hemophilia

B dog 2±3r1012

pfu ipv 300 + (WBCT, PTT) ± 1±2 m NT DE1 [152]

Ad RSV cFIX Hemophilia B dog

1011 pfu/kg iv 500 No bleeding

NT ± 2±3 m ± DE1

E2A ts125 [161] Ad RSV cFIX Hemophilia

B dog 2r1011

pfu/kg iv 500 + (WBCT) ± 7±8 m NT DE1

Cyclosporin [159] AAV LTR hFIX Intron FIX-KO 6r1010 vp ipv 36 + (BT) + > 4 m ± [163] AAV LSP cFIX Intron FIX-KO 6r1011 vp ipv 400 + Survival, aPTT + > 5 m ± [165] AAV LTR cFIX Intron Hemophilia

B dog 2r1012 vp ipv 1 + (WBCT, aPTT) + > 8 m ± [163]

AAV LSP cFIX Intron /WPRE

Hemophilia B dog

5r1012

vp/kg ipv 5 + (WBCT, aPTT) + > 7 m ± No bleeding [166]

AAV CMV cFIX Hemophilia B dog

2r1012

vp/kg im 1.5 + (WBCT, aPTT) + > 12 m t Transient Ab [169]

AAV CMV hFIX Hemophilia B dog

2.5r1012

vp im < 0.1 + (WBCT) ± 1±2 w + [168]

AAV CMV cFIX Hemophilia B dog

1012±1013

vp im 1±2 + (WBCT, aPTT) + > 12 m ± [170]

Tn EF-1a hFIX + Transposase

FIX-KO 25 mg DNA iv 1±2 + (BT) + > 5 m ± Hydrodynamic Transfection

[172]

For abbreviations see notes to Table 1.

6 M. K. L. Chuah et al.

because of the presence of A/T rich nuclear matrix attachment regions (MAR) within the A2 region of the FVIII cDNA [48,49]. Second, FVIII is transported inef®- ciently from the endoplasmic reticulum to the Golgi system [52±54]. Third, in the blood, FVIII is most susceptible to proteolytic degradation and the stability of the protein is dependent upon the formation of a complex with vWF [52±54]. In addition, the protein undergoes extensive post-translational modi®cation and needs to be proteolytically cleaved for its functional activation [52±54]. For a gene therapy strategy to be successful and achieve FVIII accumulation in the plasma, it is important to express high levels of FVIII in tissues that permit direct secretion of FVIII into the bloodstream, given its short half-life (12 h) and since it is a relatively large protein (280 kDa) that remains essentially con®ned to the intravascular space [22].

MoMLV-based retroviral vectors

Moloney murine leukemia virus (MoMLV)-based retro- viral vectors are the ®rst and most commonly used vectors for clinical trials and offer the potential for long-term gene expression by virtue of their stable chromosomal integration and lack of viral gene expression [55,56]. Their integration allows passage of the transgene to all progeny cells. However, a disadvantage of MoMLV-based retroviral vectors is that cell division is required for transduction and integration, thereby limiting retroviral- mediated gene therapy to actively dividing target cells.

Gene therapy for hemophilia with MoMLV-based vectors would require either direct in vivo transduction of cells that are naturally proliferating or induced to proliferate, or ex vivo expansion and transduction of target cells followed by their re-administration. Stable packaging cell lines are available that facilitate the large- scale production and characterization of recombinant virus for potential use in clinical trials. No acute toxicity or adverse effects have been reported in more than 100 clinical trials with retroviral vectors [57]. Furthermore, retroviral vector preparations devoid of replication- competent retroviruses (RCR) have not been shown to cause malignant transformation in animals or in patients [58,59]. Rare clonal transformation events have only been observed in severely immunocompromised primates receiving a high dose of contaminating RCR [60]. Retroviral vector-mediated gene transfer is therefore relatively safe for gene therapy of hemophilia. Based on some of the above considerations, retroviral vectors [46,47] were developed containing the B-domain deleted FVIII gene [44] (designated as MoMLV-FVIII). Early studies demonstrated the feasibility of using retroviral vectors for transfer and expression of FVIII and suggested that cells of non-hepatic origin have the capacity for proper post-translational modi®cation of the FVIII protein and can be used as targets for hemophilia A gene therapy [17,46].

These ®rst generation retroviral vectors were charac- terized by low titers and low FVIII expression levels that

hampered their clinical applications for gene therapy. The presence of an intact FVIII cDNA into MoMLV retroviral vectors resulted in a 100±1000-fold decrease in vector titer in comparison with the parental vector or vectors carrying other similarly sized cDNAs [17,21,46±48]. The inhibition of FVIII expression and the low viral titer are due to the presence of sequences within the FVIII cDNA that inhibited RNA accumulation by interfering with transcriptional initiation or elongation (see above) [47±49,51]. Conservative mutagenesis of the entire 1.2 kb INS element failed to increase FVIII expression or vector titer [21]. However, we and others showed that this impediment could be circumvented by including an intron upstream of the FVIII cDNA which led to a signi®cant increase in FVIII expression and restored retroviral titer to normal levels (105 cfu/ml) [21,22]. This design was based on the observation that splicing can improve mRNA accumulation by increasing transcrip- tional ef®ciency, stabilizing RNA and/or increasing transport from the nucleus to the cytoplasm [61,62] but other post-transcriptional mechanisms may also be involved [63].

Attempts at achieving long-term human FVIII expres- sion by ex vivo gene therapy using a variety of retrovirally- transduced primary cells have shown that the implanta- tion site, the target cell type and the vector design (particularly the strength of the expression cassette) are critically important in obtaining detectable FVIII levels in the circulation. The reason for the lack of circulating FVIII following subcutaneous or intramuscular injection [22,64] is not fully understood. Neo-organs containing human ®broblasts transduced with an intron-based FVIII- retroviral vector (MFG) and embedded into collagen- coated arti®cial ®bers, were implanted into the peritoneal cavities of immunode®cient mice. This approach yielded high level expression of human FVIII with peak levels of 100 ng/ml but expression declined gradually to basal levels over a period of 2 weeks. Limited survival of the transduced cells within the collagen implants and transcriptional inactivation of the FVIII expression cassette may have contributed to the cessation of FVIII expression. Though cells belonging to the lympho- hematopoietic lineage, including T-cells and BM progeni- tor/stem cells, are obvious candidates to develop ex vivo gene therapy for hemophilia, expression of FVIII is extremely low [30,65]. Nevertheless, BM-derived cells can express FVIII undetectable by standard assays but suf®ciently high to establish immunologic tolerance to FVIII following transplantation of engineered BM cells into myeloablated FVIII-de®cient mice [65]. These observations suggest that it might be possible to express therapeutic levels of FVIII in the lympho-hematopoietic compartment by using other promoters or increasing transduction ef®ciencies. Moreover, this model system will prove useful for the evaluation of genetic therapies for FVIII immunomodulation and tolerance induction.

Since cells of the lympho-hematopoietic lineage expressed only very low levels of the FVIII protein, these observations prompted us to identify new target

Hemophilia Gene Therapy 7

cells for hemophilia A gene therapy. BM stroma is a potentially attractive target for gene therapy of hemo- philia A since it is relatively easy to obtain by needle aspiration from the iliac crest or the sternum, can be transduced by retroviral vectors [66,67] and has been used safely in phase I/II clinical trials [68]. Moreover, BM stroma contains mesenchymal stromal stem/precursor cells that can stably engraft and give rise to multiple lineages [31,69,70]. It has been shown that transiently transfected canine BM stromal cells expressed human growth hormone or FIX in vitro and persisted for several months after autologous re-infusion in recipient dogs [31]. Based on these observations, human BM stromal cells were transduced with a FVIII-retroviral vector using optimized transduction methods. High in vitro FVIII expression levels were achieved without selective enrich- ment of the transduced BM stromal cells [71]. Long-term engraftment of the engineered human BM stromal cells could be achieved in non-myeloablated immunode®cient mice, and therapeutic human FVIII levels could be attained (peak levels of up to 20 ng/ml) but eventually declined to basal levels due to MoMLV LTR promoter inactivation [72]. These ®ndings indicate that retroviral vector mediated gene therapy using engineered BM stromal cells may lead to therapeutic levels of FVIII in vivo and that long-term engraftment of human BM stromal cells was achieved without having to rely on neo-organs or conditioning regimens. This suggest a more practical ex vivo gene therapy approach for hemophilia A, provided prolonged expression can be achieved by using alternative promoters.

In vivo gene therapy for hemophilia A with MoMLV- based vectors requires direct transduction of cells that are either naturally proliferating in vivo or that were induced to proliferate. Most efforts have been aimed at improving hepatic transductions [73,74] and indicate that, in addition to the rate of hepatocyte proliferation, viral titer is also a limiting factor in hepatic transduction. This underscores the importance of using high-titer MoMLV vectors for in vivo liver transduction. In addition, recent studies have shown that keratinocyte growth factor (KGF) or hepatocyte growth factor (HGF) administration can lead to hepatocyte proliferating and ef®cient liver transduction with MoMLV-based vectors [75,76].

We have generated FVIII retroviral vectors with titers equivalent to 109±1010 transducing unit (TU)/ml by combining improved vector design with improved meth- ods to concentrate and purify MoMLV vectors [77]. This constitutes a 106±107 fold improvement in viral titer compared to ®rst-generation, non-concentrated MoMLV- FVIII retroviral vectors. The high-titer retroviral stocks were concentrated 100±1000-fold by ultracentrifugation after pseudotyping MoMLV-derived FVIII retroviral vectors with the G glycoprotein of vesicular stomatitis virus (VSV) [78] using conditional human packaging cell lines that express VSV-G in a tetracycline-dependent manner. These VSV-G pseudotyped MoMLV vectors are composed of the MoMLV genome encapsidated by the VSV-G envelope protein and were shown to ef®ciently

transduce primary hepatocytes in vitro. Human packaging cell lines for generating high-titer vectors for in vivo gene therapy of hemophilia have at least two advantages compared to murine packaging cell lines. First, no endogenous murine retroviral sequences are present in the human cell lines, reducing the possibility for RCR generation and eliminating the risk of co-packaging these sequences into retroviral vector particles. Second, human cells do not express a1-4 galactosyl-transferase, resulting in increased resistance of viral vector particles to human complement [79,80] though other mechanisms may also be involved [78]. These features may increase the safety and ef®cacy of MoMLV-based retroviral vectors for in vivo gene therapy of hemophilia. Furthermore, patients do not have pre-existing antibodies speci®c for MoMLV or VSV-G that would preclude or interfere with in vivo transduction.

High-titer VSV-G pseudotyped MoMLV-FVIII vectors were injected intravenously into newborn, FVIII-de®cient mice. High-levels (i20% of normal human physiologic levels) of functional human FVIII production could be detected in about 50% of the recipients, some of which expressed stable physiologic or supranormal levels (up to 1250%). Most high-expressers survived an otherwise lethal tail-clipping, unequivocally demonstrating pheno- typic correction of the bleeding disorder. FVIII expression was sustained for more than 14 months. This is the ®rst demonstration that hemophilia A can be cured by gene therapy in a clinically relevant animal model. These ®ndings also indicate that a genetic disease can be corrected by in vivo gene therapy using retroviral vectors that may be applicable to the treatment of other diseases. Ef®cient gene transfer occurred into liver, spleen and lungs but not in other organs including testes, with predominant FVIII mRNA expression in the liver. The ef®cient hepatic gene transfer in neonatal mice could be due primarily to the higher hepatocyte turnover rate in newborn versus adult animals. Alternatively, hepatic stem/progenitor cells [81,82] may have been transduced by virtue of their relative abundance in neonates and their relative proximity to the vasculature [83]. In addition, the use of high-titer vectors may have been essential to achieve these high levels of gene transfer and FVIII expression.

Long-term correction of hemophilia has been achieved in about 50% of the FVIII-de®cient animals, whereas the other half was not corrected, due to the induction of human FVIII-speci®c neutralizing antibodies. The cause of the heterogeneous antibody response and FVIII levels among the different recipients is not clear. Since the FVIII-de®cient mouse is not an inbred strain, genetic differences may have contributed to this variability. Alternatively, high levels of FVIII may have been required to induce immune tolerance to FVIII in the FVIII-de®cient mice and prevent induction of inhibitory antibodies consistent with the induction of tolerance of FVIII in neonatal mice receiving high doses of human FVIII protein [84]. Since there is a lower level of FVIII gene transfer in the transient and non-expresser recipient mice compared to the long-term expressers, a mechanism

8 M. K. L. Chuah et al.

whereby transduced cells were eliminated by a cellular immune response cannot be ruled out.

In a separate study conducted by Jolly and co-workers, intravenous administration of a high dose of amphotropic retroviral vector (5r109 cfu eq) encoding a B-domain deleted human FVIII cDNA to immunocompetent juvenile rabbits resulted in long-term expression of therapeutic levels of FVIII (75% of the normal human levels) which was sustained for >6 months [85]. This could again be due primarily because of the relatively high hepatocyte turnover rate in juveniles allowing ef®cient retroviral transduction. Re-administration of the vector could lead to increased FVIII despite the presence of vector-speci®c antibodies. Therapeutic FVIII levels could also be obtained in adult recipient rabbits, which varied between 10% and 75% of the normal human levels. The exact reason for this unexpected ®nding is not fully understood but is at least partly due to the use of high-titer vector preparations. Vector sequences were detectable in spleen and liver with small amounts in bone marrow and lymph nodes but no vector in the semen. Direct detection of FVIII activity was not possible due to the endogenous rabbit FVIII. The same vector construct was injected intravascu- larly in hemophilia A dogs which led to transient correction of whole-blood clotting time during the ®rst 2 weeks. However, antibody formation to human FVIII confounded extensive analysis of FVIII activity.

Lentiviral vectors

In order to overcome the need for active cell division to achieve ef®cient transduction, retroviral vectors derived from lentiviruses have been developed, with HIV as prototype [86±89]. The lentivirus life cycle differs from that of onco-retroviruses in that the pre-integration complex is readily transported into the nucleus, thus enabling ef®cient transduction of non-dividing cells, including neurons. Gene transfer with lentiviral vectors in dividing cells is more ef®cient than with MoMLV-based retroviral vectors [90,91]. While present efforts are aimed at the generation of stable packaging cell lines [92] and safe [89], clinically acceptable vectors, the attributes of the lentivirus vector system may be well suited for the treatment of hemophilia. The latest generation lentiviral vector packaging system relies on a self-inactivating HIV-derived vector backbone which does not contain any HIV-1 genes but only the cis-acting elements necessary for integration, reverse transcription and expression [89]. Trans-complementation occurs with only a minimum of HIV genes (gag and pol) which are expressed condition- ally by the HIV Rev protein [93]. The envelope typically is VSV-G or the retroviral amphotropic 4070A envelope. Other non-HIV-based lentiviral vector systems have also been developed.

Studies involving direct in situ injection of liver and muscle demonstrated ef®cient transduction localized to the injection site as well as sustained transgene expres- sion by hepatocytes or muscle cells [94]. However, in vitro studies using wild-type HIV or HIV-based vectors

have shown that, in some cases, cell cycle activation is required for infection or transduction, even though cellular mitosis is not an absolute requirement for integration [90]. In contrast to previous reports [94], more recent studies have shown that ef®cient lentiviral transduction of hepatocytes may require cell division. At high lentiviral vector dose, hepatotoxic effects were apparent possibly due to virion proteins or to concen- trated contaminants in the viral preparations [91]. The liver injury associated with vector administration has been shown to induce hepatocellular cycling and may be involved in enhancing the ef®ciency of lentiviral vector transduction. For certain cell types gene transfer is in¯uenced by the presence of HIV-1 accessory proteins in the vector particles [95] but whether this in¯uences hepatic gene transfer remains controversial [90,91,94].

Therapeutic levels of FVIII corresponding to 15% of the normal human plasma levels could be achieved following systemic administration of lentiviral vectors containing an EF1a enhancer/promoter driven human FVIII gene in normal mice but only after induction of hepatocyte proliferation by partial hepatectomy [91]. These levels were stable in immunode®cient mice but were transient in normal mice due to the induction of inhibitory antibodies to human FVIII. In contrast, other studies indicate that arti®cial induction of hepatocyte prolifera- tion is not necessary to achieve therapeutic FVIII levels following lentiviral transduction in mice (Verma, perso- nal communication). The exact reason for this difference is not fully understood but it may be attributable, at least partly, to differences in promoter strength and/or differences in vector production or puri®cation. The incorporation of cis-acting sequences from the HIV-1 pol gene (central polypurine tract or cPPT) that facilitate nuclear translocation of the vector genome may further improve hepatic gene transfer [96]. We have recently observed that such a cPPT+ lentiviral vector can transduce the liver of adult mice without partial hepatectomy and that the target cells include hepatocytes and non-hepatocytes, presumably liver sinusoidal endothelial cells (VandenDriessche et al., unpublished observations). In addition, ef®cient transduction of putative splenic ICAM-1+CD11b+ antigen presenting cells, possibly macrophages was observed. This has important implications for gene therapy of hemophilia since the possible inadvertent expression of clotting factors in antigen-presenting cell (APC) may theoretically increase the risk of inhibitor formation. The use of liver-speci®c instead of ubiquitous promoters may circumvent this potential concern [97].

To further develop a lentiviral vector-based strategy for hemophilia A, human BM stromal cells were transduced with a cPPT+ lentiviral vector that contained the B-domain deleted canine factor VIII cDNA (Vanden- Driessche et al., unpublished observations). Transduction was more ef®cient than when MoMLV-based retroviral vectors were used and following xenografting of engi- neered BM stromal cells into the spleen of FVIII-de®cient SCID mice, canine plasma FVIII levels rose to levels

Hemophilia Gene Therapy 9

corresponding to 25% of the normal human plasma levels. However, FVIII expression declined over time possibly due to transcriptional inactivation of the CMV promoter. These ®ndings indicate that lentiviral vector mediated gene therapy using engineered BM stromal cells may lead to therapeutic levels of FVIII in vivo and that alternative promoters will be necessary to prolong FVIII expression.

Adenoviral vectors

Adenoviral vectors can transduce a broad range of dividing as well as non-dividing cells, unlike MoMLV- based retroviral vectors. The adenoviral genome remains episomal in the transduced cells implying that there is virtually no risk for neoplastic transformation due to insertional mutagenesis. However, dividing cells will gradually lose the adenoviral vector along with its potentially therapeutic gene. Most adenoviral vectors are derived from human adenovirus serotypes 2 and 5 and rendered replication-de®cient by removal of essential viral regulatory genes (e.g. E1 gene), which are complemented in trans using appropriate packaging cells that express the missing gene product [98]. Since most humans have been naturally infected with adeno- virus, it is possible that the presence of adenovirus- speci®c antibodies would interfere with adenoviral transduction in vivo. To circumvent this potential problem and to allow for repeated administration, other adeno- viral serotypes or non-human adenoviral vectors could be used.

The main disadvantage of adenoviral vectors is that the host immune response, in general, appears to limit the duration of transgene expression and the ability to re-administer the vector. The immune response can be directed towards the transgene product itself and/or towards the adenoviral vector particles or the adenoviral vector backbone gene products expressed in the trans- duced cells. First-generation adenoviral vectors still contain all other viral genes that are expressed at low levels even in the absence of E1, presumably as a result of activation by cellular E1-like proteins. In addition, expression of transgene products can break tolerance to homologous self-proteins when adenoviral vectors are employed [99], which would be particularly disconcert- ing in the case of hemophilia. Furthermore, acute and chronic toxic and lethal effects are frequently observed in non-human primates and other animals including mice, rabbits and dogs, receiving high doses of replication- de®cient adenoviral vectors [100,101]. Finally, several clinical trials based on adenoviral vectors had to be discontinued because of acute in¯ammatory responses and severe morbidity and mortality in some patients.

As the cells transduced with E1-deleted recombinant adenoviral vectors appeared to be eliminated by cytotoxic T lymphocytes (CTLs) directed towards late viral gene products, it was thought that this cellular immune response could be attenuated if not completely abolished by further reducing viral gene expression. This was

initially accomplished by alteration or deletion of additional early viral genes such as E2 or E4 [102±105]. Despite evidence that adenoviral vectors containing additional deletions in E2 or E4 are less toxic and less immunogenic, there is no conclusive evidence that such vectors are capable of persistent or even signi®cantly prolonged expression.

Intravenous administration of an adenoviral vector in which the human or canine B-domain deleted FVIII cDNA was driven by an albumin promoter to normal or hemophilic mice resulted in expression of therapeutic and even physiologic levels of biologically active FVIII which corrected the bleeding de®ciency in the hemophilic mice [100,106±111] However, FVIII levels gradually declined at least partly due to dose-dependent vector toxicity but no antibodies to human FVIII were detected. Intravenous injection of human or canine FVIII adenoviral vectors into hemophilia A dogs resulted in short-term phenotypic correction and transient expression of ther- apeutic levels of human or canine FVIII [100,112] (Connelly, personal communication) due to the develop- ment of human or canine FVIII neutralizing antibodies, possibly in combination with acute and chronic liver toxicity. When an autologous murine FVIII gene was administered using E1/E3-deleted adenoviral vectors to FVIII-de®cient mice that expressed non-functional FVIII heavy chains, expression was short-term due to a humoral and cellular immune response [113]. Hence despite the presence of endogenous FVIII protein, the immune system still recognized a species-speci®c transgene protein as a neo-antigen, eliciting an immune response. Short-term correction of factor VIII de®ciency in a murine hemophilia A was also observed after delivery of adenovirus murine factor VIII in utero [114] but in this case the gradual loss in expression is mainly due to the postnatal hepatocyte proliferation since adenoviral vector do not integrate.

The ultimate adenoviral vector modi®cation comprises a vector containing only the cis elements necessary for replication and packaging, but lacking all adenoviral genes. These so-called `gutted' or `gutless' vectors depend on the use of helper viruses that provide all missing functions in trans but cannot be packaged ef®ciently because of mutations or CRE-lox mediated excision of the packaging signal [115±120]. `Gutless' vectors that are of approximately the same size as wild-type adenovirus (32±34 kb) and contained the human genomic a1-antitrypsin gene are stable and could give rise to prolonged high-level transgene expression with signi®- cantly reduced acute and chronic hepatotoxicity in mice and baboons [121]. However, expression is not constant and decreases very slowly over time perhaps due to slow hepatocyte turnover and/or to CTL-mediated elimination of transduced cells triggered by the proteins present in the capsid of the virus. The latter possibility would be consistent with a recent report showing that psoralen- treated UV-crosslinked ®rst-generation vectors can stimulate the production of CTLs, suggesting that cells infected with inactive vector can present antigens for

10 M. K. L. Chuah et al.

recognition by MHC class I molecules in the absence of viral replication or de novo protein synthesis [122]. Another concern with current gutless adenoviral vectors is that helper contamination may still occur in the vector preparations which might perhaps help to stabilize the gutless vectors [123] but which would need to be eliminated or reduced further to minimize possible adverse effects in future clinical studies.

Recently, a gutless adenoviral vector that carries the full-length human FVIII cDNA under the control of the human 12.5 kb albumin promoter was injected at a relatively high dose (2r1011 viral particles per mouse) into FVIII-de®cient mice and resulted in ef®cient hepatic gene transfer and long-term therapeutic FVIII expression (100±800 ng/ml) leading to phenotypic correction [124]. However, expression eventually declined possibly due to induction of inhibitory antibodies to human FVIII. No signi®cant histopathologic ®ndings or toxicities were observed to be associated with the vector. At a ten-fold lower vector dose, no FVIII could be detected, suggesting a non-linear threshold effect. The exact reason for this `threshold' effect is not clear but has also been observed using ®rst-generation adenoviral vectors [101,111]. When the vector was injected into hemophilia A dogs at a dose of 3r1012 vp/dog, no human FVIII could be detected, suggesting that a threshold dose levels may be required.

Adeno-associated viral (AAV) vectors

Another vector that holds promise for achieving long- term gene expression in the absence of expression of viral genes is the AAV [125]. AAV is a naturally occurring non- pathogenic virus that is inherently replication defective since adenovirus co-infection is required to initiate a productive infection. Unlike retroviral vectors, AAV vectors are able to transduce non-dividing cells in vivo including liver, muscle and brain [126±128]. AAV vectors can integrate into the target cell genome or alternatively persist as an episome in the transduced cells prior to integration. These features justify the use of AAV for developing gene therapy for hemophilia. However, since most humans are seropositive for AAV-2, it is possible that the presence of these AAV-speci®c antibodies would interfere with AAV transduction in vivo. To circumvent this potential problem, other AAV serotypes could be used. The large-scale production of high-titer prepara- tions of AAV is still dif®cult, but this is being resolved by developing stable packaging cell lines and by further improving vector puri®cation strategies using chromato- graphy. Ef®cient hepatic transduction with AAV requires direct intraportal vein injection instead of peripheral intravenous infusion. While virtually all hepatocytes take up vector, only about approximately 5% of hepatocytes contain high molecular weight concatemers and conse- quently express the transgene [129]. Unknown cellular factors are required for stable transduction and dimer formation is a critical event in the transduction pathway.

One major challenge in developing AAV-FVIII vectors is

to design a FVIII expression cassette that is small enough to be accommodated within the AAV vector without compromising viral titer or FVIII expression levels. Because of rAAV packaging constraints, an AAV vector encoding a B-domain deleted FVIII cDNA was generated using a small nuclear RNA promoter which provided the ®rst step in the utilization of AAV to treat hemophilia [130]. In another study, small regulatory elements designed for liver-speci®c transgene expression were linked to B-domain deleted human FVIII cDNA making the recombinant vector slightly larger than wt AAV (109%) [131]. Concentrated vector particles (1010±1011) were administered via portal vein injection to C57BL/6 and NOD-SCID mice. NOD-SCID mice expressed hFVIII reaching 27% of normal human plasma levels but all C57BL/6 mice developed anti-hFVIII antibodies that preclude long-term high FVIII expression. About 5% of the hepatocytes were transduced and FVIII mRNA expression was detected primarily in the liver. Histologic analysis of the liver revealed no pathologic abnormalities.

Because of the size constraints of recombinant AAV, an alternative strategy was developed whereby the heavy and light chains of the human FVIII cDNA were delivered independently by using two separate vectors. The FVIII heavy or light chain was expressed using the human EF1a promoter [132]. In animals injected with both the heavy and light chain vectors (3r1011 vp), greater than physiological levels (200±400 ng/ml) of biologically active hFVIII were produced. Although the amount of biologically active protein was only a fraction of the total amount of light chain antigen produced (about 5%, which is consistent with the requirement of co-transduction), it represents greater than normal physiological levels of FVIII and would be expected to completely correct the bleeding disorder in humans. This suggests that co-expression of the heavy and light chains of hFVIII may be a feasible approach for treatment of hemophilia A. However, it is not clear what the rami®cations of production and secretion of inactive FVIII chains are in vivo. There are some concerns that this may have an untoward effect on inhibitor formation in hemophilia patients.

Gene therapy for hemophilia B

Expression of FIX

The FIX gene has been cloned, is 34 kb in length and is spread over 8 exons, too large to be packaged ef®ciently into the commonly used gene therapy vectors, but the FIX gene is only about 1.4 kb and can therefore easily be packaged [8,133]. Unlike FVIII, FIX is relatively easy to express and has a wide volume of distribution, since FIX readily equilibrates between extravascular and intravas- cular spaces given its smaller size [134]. This implies that, in contrast to FVIII, direct secretion into the circulation is no prerequisite for FIX accumulation into the plasma, which creates new possibilities for the delivery of FIX

Hemophilia Gene Therapy 11

gene therapy vectors or FIX-engineered cells, including intramuscular or subcutaneous administration. However, ectopic expression of FIX can result in binding to collagen type IV and hereby reduce the ef®ciency by which it can reach the circulation [135]. Activated FIX functions as a serine protease and it is secreted as an inactive precursor protein that is activated by proteolytic cleavage. The expression of functional FIX does require post-transla tional modi®cation, particularly vitamin K-dependent c-carboxylation of 12 glutamic-acid residues, which is essential for Ca2+ binding and FIX function. This occurs ef®ciently in hepatocytes but also in many other tissues that normally do not express FIX but can be considered targets for hemophilia B gene therapy. However, very high levels of FIX may saturate the cellular ability to process biologically active FIX, hence affecting the overall ability to secrete FIX.

MoMLV-based retroviral vectors

Transduction of ®broblasts with retroviral vectors con- taining the FIX gene led to signi®cant amounts of functional FIX in vitro. [15,16,136,137]. However, following transplantation of the engineered ®broblast cells into mice, FIX production in the plasma was transient and lower than what would be expected based on FIX secretion in vitro. In some of the recipients the disappearance of functional FIX could be attributed to the development of FIX neutralizing antibodies [15]. In addition, the retroviral LTR-promoter that drives expres- sion of the gene of interest was found to be inactivated in ®broblasts in vivo, possibly as a consequence of hyper- methylation [138]. Prolonged expression could be achieved in a rabbit model [139,140] and was the basis for a phase I clinical trial in China [141,142] (see below).

In parallel, many other cell types have been shown to express functional FIX following ex vivo transduction with FIX retroviral vectors. These cell types include myoblasts [25,26,143], endothelial cells [23], hemato- poietic cells [29,144] hepatocytes [145], keratinocytes [18,19,146,147] and gut epithelial cells [148]. Trans- plantation of these engineered cells in vivo often resulted in detectable but transient levels of FIX, reminiscent of the ®broblast transplantation experiments. Again the use of alternative promoters and modi®cations in vector design may provide a solution to this problem. Indeed, sustained systemic expression of FIX could be achieved by ex vivo gene therapy following transduction of myoblasts with MoMLV-FIX retroviral vectors expressing FIX using muscle creatinin enhancer/promoter elements [26,27].

An alternative ex vivo implantation strategy involves the use of immuno-isolation chambers. When surgically implanted, the outer surface of these membrane devices becomes vascularized and the chambers can be loaded with genetically modi®ed cells [149,150]. When devices containing FIX-secreting human ®broblasts were main- tained in athymic nude rodents, sustained plasma levels of human FIX were achieved. One potential safety feature of this approach is that the device could be removed in the

event of an adverse reaction. However, necrosis of encapsulated cells or ®brosis can sometimes be proble- matic. As an alternative, implanted cells can be enclosed in alginate-polylysine-alginate microcapsules that are permeable to FIX diffusion, but impermeable to the hosts' immune mediators. These capsules do not cause necrosis or ®brosis and can even be used to implant non- autologous cells. This strategy can give rise to persistent human FIX delivery in vivo in immunode®cient mice but expression is short-term in immunocompetent animals due to anti-FIX antibodies [24,151].

The potential of in vivo gene therapy for hemophilia B with MoMLV-based vectors was evaluated by direct infusion of MoMLV retroviral vectors containing the canine FIX cDNA into the portal vasculature of hemophi- lia B dogs [73]. The animals were subjected to partial hepatectomy prior to the in vivo gene transfer to induce hepatocyte proliferation and constitutively expressed low levels of canine FIX (0.1% of normal levels) for over 9 months. Persistent low level expression of the clotting factor resulted in 60% reductions of whole-blood clotting and partial thromboplastin times of the treated animals.

Lentiviral vectors

Stable therapeutic levels of FIX slightly above 1% (50±60 ng/ml) of the normal human plasma levels could be achieved following systemic administration of lentiviral vectors containing an EF1a enhancer/promoter driven human FIX gene in normal mice. Stable FIX levels increased to 7% (350 ng/ml) after induction of hepato- cyte proliferation by partial hepatectomy, independent of the presence or absence of HIV-1 accessory proteins in the vector particles. However, incorporation of cis-acting cPPT sequences from the HIV-1 pol gene (see above) improved hepatic gene transfer and in combination with stronger promoters and the Woodchuck hepatitis virus post-regulatory element, signi®cantly higher FIX expres- sion levels (2.5% of normal levels or 120 ng/ml) could be achieved requiring a 50-fold lower vector dose [96] and without relying on partial hepatectomy. Studies in larger animal models and continued efforts to improve the vector design are warranted to explore the full potential of lentiviral vectors for hemophilia gene therapy.

Adenoviral vectors

Partial improvement in hemostatic parameters has been achieved using retroviral or lentiviral vectors expressing FIX, but the relatively low levels of FIX expression in the plasma were insuf®cient to fully correct the bleeding phenotype. This could be overcome by using improved expression cassettes and more ef®cient gene delivery vectors such as adenoviral vectors. Adenoviral vectors expressing the canine or human FIX cDNAs were injected intravenously into hemophilia B dogs [152] or in normal [153] or FIX-de®cient mice [33,154]. This resulted in ef®cient liver transduction and therapeutic levels of FIX which transiently corrected the bleeding diathesis in the

12 M. K. L. Chuah et al.

hemophilia B mice and dog models. However, levels slowly declined to baseline due to cell-mediated and humoral immune responses as con®rmed in independent studies [155] and were not re-established by a second vector injection due to high amounts of circulating antibodies that neutralized the vector upon re-challenge. Intramuscular administration of an adenoviral vector ef®ciently activated FIX-speci®c CTLs and T helper cells of both Th1 and Th2 subsets, leading to in¯ammation and destruction of transduced muscle tissue and activation of B cells also [156].

Adenoviral vector-mediated factor IX gene transfer in rhesus macaques gave rise to transient near-physiologic levels of human FIX but was associated with dose-limiting liver toxicity and coagulopathy due to persistent hypo®- brinogenemia [101]. Despite the fact that human FIX protein injections did not induce antibodies given the high degree of homology between human and macaque FIX [157] antibodies to FIX were induced (suggesting an immune adjuvant effect of the ®rst-generation E1/E3-deleted adenoviral vector [101]). In utero gene transfer of FIX using E1/E3-deleted adenoviral vectors resulted in transient FIX levels (up to 600% of normal human physiologic levels) and may pave the way towards prenatal gene therapy in order to prevent haemorrhagic complications such as intracranial bleeding during delivery [158], provided improved vectors yielding long-term expression can be employed.

To prolong FIX expression, the administration of the adenoviral vector was combined with immunosuppressive drugs, which allowed expression to persist for 6 months [155,159]. However, the induction of neutralizing anti- bodies again rendered re-administration of the vector ineffective. By combining different immunosuppressive drugs, repeated vector administration became possible [160]. However, for clinical applications it would be preferred to modify the vector components instead of using immunosuppressive treatments. A recombinant adeno- virus expressing canine FIX was therefore modi®ed so that they also contained a temperature-sensitive mutation (ts125) in the DNA binding protein encoded within the viral E2A region, leading to reduced late adenoviral gene expression [161]. The effects of the inclusion of the ts125 mutation on transgene expression in vivo were evaluated in Balb/c mice and hemophilia B dogs by comparison with adenoviral vectors containing the same transgene but lacking the ts125 mutation. No signi®cant differences in the duration of transgene expression were observed in either animal model. The inability of the ts125 mutation in the prolongation of transgene expression in these two animal models suggests that further modi®cation of the vector backbone and ultimately the generation of `gutless' adenoviral vector containing the FIX cDNA may be required to achieve long-term gene expression.

AAV vectors

Intravenous or intrahepatic injection of mice with AAV vectors encoding human FIX from the CMV promoter

resulted in synthesis of low levels of human FIX for at least 5 months [162] and up to 3% of all hepatocytes could be transduced after a single vector injection. c-Irradiation and wild-type AAV was necessary to achieve these transduction ef®ciencies in vivo. However, a single intraportal administration to normal or FIX-de®cient mice of AAV vectors containing MoMLV LTR-driven human FIX cDNA resulted in 3±5% of the hepatocytes expressing FIX yielding stable, therapeutic concentrations of func- tional human FIX (20±60% of normal). These levels could be achieved without the need for c-irradiation [128] or wild-type AAV and adenovirus and were suf®cient to partially correct the bleeding de®ciency in the hemophilic mice [163]. Similar results were obtained in normal mice with the human EF1a promoter whereas the CMV promoter was inactivated in vivo [164]. In hemophilic dogs, hepatic AAV-FIX gene transfer resulted in 0.5±1% of normal canine factor IX levels, the absence of inhibitors, and a sustained partial correction of the coagulation defect for at least 8 months using a MoMLV LTR-driven transgene [163].

The highest FIX levels in mice were achieved using a AAV-canine FIX vector using a chimeric liver-speci®c promoter (LSP) that contains the thyroid hormone-binding globulin promoter sequences and a1-microglobulin/biku- nin enhancer sequences and an intron-containing leader sequence, as described previously [50]. Stable, supra- normal levels of canine FIX of up to 15±20 mg/ml were obtained by a single intraportal injection which success- fully cured the bleeding disorder in hemophilia B mice [165]. A similar AAV-LSP-cFIX construct that also included the woodchuck hepatitis virus post- transcriptional element led to stable expression of the therapeutic level of cFIX (5% of normal level) for 7 months following intraportal vector administration of 4.6r1012 particles/kg. Importantly, no bleeding has been observed in the dog that expresses these therapeutic levels of canine FIX. Moreover, no persistent signi®cant hepatic enzyme abnormalities nor anti-canine factor IX antibodies were detected in the treated dogs [166].

The muscle is an attractive target for delivery of FIX to the circulation and is readily transducable with AAV vectors [126] Intramuscular injection of high-titer AAV vectors encoding CMV-driven human FIX into immuno- de®cient Rag-1 knock-out mice led to ef®cient transduc- tion of muscle tissue in vivo resulting in therapeutic plasma levels of FIX (200±350 ng/ml, i.e. 5±7%) [135]. Improved muscle-derived expression of FIX could be achieved with a skeletal actin/CMV hybrid enhancer/ promoter [167]. Intramuscular injection of an AAV vector expressing FIX fails to activate FIX-speci®c CTLs in hemostatically normal immunocompetent mice or in hemophilia B mice. However, transgene-derived FIX can cause B cell responses characterized by production of T helper cell-dependent antibodies (predominantly IgG1, but also IgG2 subclasses) resulting from activation of CD4+ T helper cells primarily of the Th2 subset [156]. Consequently, no FIX could be detected in the plasma. It therefore appears that FIX is more immunogenic when

Hemophilia Gene Therapy 13

produced from muscle tissue than when expressed in the liver. The exact reason for this difference is not fully understood but may re¯ect a difference in capacity of antigen presenting cells in different tissues. In hemophilic dogs, intramuscular delivery of rAAV encoding human FIX from a CMV promoter resulted in transient reduction in the whole-blood clotting time with the anticipated development of an antihuman FIX inhibitor antibody which correlated with the loss of phenotypic correction [168]. However, when the autologous canine instead of human FIX was used sustained FIX levels (>1 year) could be corresponding to 1±2% of normal human physiologic levels, obtained with a vector dose of 8.5r1012 vp/kg. This lead to a partial stable correction of the coagulation de®ciency [169,170]. A transient inhibitor was observed in one of the dogs.

Other gene delivery methods

The development of non-viral gene transfer technologies that can support stable chromosomal integration and persistent gene expression in vivo is desirable for hemophilia gene therapy. However, the use of non-viral vectors for gene therapy has been hampered by low ef®ciency and/or transient expression of potentially therapeutic transgene. To overcome some of the inherent limitations of non-viral-based strategies, incorporation of viral elements has been pursued either to enhance stable integration [171] or endosome-disrupting viral peptides to augment gene transfer ef®ciency [39]. Recently, transposon technology has been used for the non- homologous insertion of foreign genes into the genomes of adult mammals using naked DNA [172]. Using the Sleeping Beauty transposase ef®cient insertion of trans- poson DNA has been achieved into the mouse genome in approximately 5±6% of transfected mouse liver cells which is similar to that obtained with safe doses of AAV or lentiviral vectors. Chromosomal transposition resulted in long-term expression (>5 months) of human blood coagulation factor IX at levels (3% of normal) that led to a partial correction of the bleeding diathesis in hemophilia B mice. Re-administration resulted in increased FIX levels up to 10%. These integrating vectors are widely amenable to industrial-scale manufacture. However, the potential immunological and cytological consequences of expres- sing a bacterial transposase in hepatocytes remains to be addressed. In this model, ef®cient hepatic transfection (up to 40% transiently) was achieved by a hydrodynamic transfection method which requires infusion of plasmids in large volumes of aqueous solutions [173]. Though this transfection method may not be clinically applicable, the incorporation of transposon technology should be useful to improve existing non-viral and viral vectors for hemophilia gene therapy applications. Surprisingly, even without transposon-mediated gene transfer, long- term expression of therapeutic levels of FIX (10±40% of normal) could be achieved for several months by mere hydrodynamic transfection using plasmid DNA containing potent FIX expression cassettes that contain the hepatic

locus control and the FIX intron and FIX untranslated region, which could not be predicted from in vitro transfection [174]. These data underscore the importance of cis sequences for enhancing in vivo hepatic gene expression and re-emphasized the lack of correlation of gene expression in tissue culture versus in vivo studies.

Most gene therapy approaches rely on the introduction of a functional gene copy rather than correcting the mutation in the defective gene like FVIII or FIX. However, by using chimeric RNA/DNA oligonucleotides it is now possible to actually repair the defective genes resulting in stable phenotypic correction as has recently been shown in the Gunn rat model for Criggler-Najjar syndrome [175] suggesting a potentially powerful strategy for gene repair without the use of viral vectors. As a proof-of-concept, a sequence mutation was induced in the rat factor IX gene using chimeric RNA/DNA oligonucleotides that were delivered by receptor-mediated non-viral delivery to the hepatocytes [176]. Nucleotide conversion was both site- speci®c and dose-dependent and the mutated gene was associated in vivo with signi®cantly reduced factor IX coagulant activity and a marked prolongation of the activated partial thromboplastin time. The results demon- strate that single base-pair alterations can be introduced in hepatocytes in situ by RNA/DNA oligonucleotides. However, there are multiple different point mutations in hemophilic families that would require `tailor-made' RNA/DNA oligonucleotides. In addition, this strategy may not correct a common, recurrent FVIII gene inversion or other more extensive deletions. Nevertheless, it could prove useful for correction of, for instance, frame-shift mutations leading to severe hemophilia.

Clinical trials

Four phase I clinical trials are currently ongoing in patients suffering from severe hemophilia A or B. In the ®rst gene therapy trial for hemophilia A, a non-viral ex vivo transfection approach is being explored in severe hemophilia A patients based on yet to be disclosed preclinical animal data and a phase I trial of human growth hormone delivery. Dermal skin ®broblasts obtained by biopsy from patients were electroporated ex vivo with an expression plasmid containing a B-domain deleted FVIII and a neomycin resistance gene. Trans- fected cells were selected and a single clone was expanded prior to laparoscopical implantation in the patient's peritoneum. Preliminary reports suggest that some subjects reported fewer bleeding episodes and occasionally have very low levels of clotting factor activity detected, without signi®cant adverse effects (R. Selden, personal communication). Based on preclinical studies in rabbits, an ex vivo gene therapy clinical trial had also been initiated in China [141] whereby autologous skin ®broblasts were transduced with a FIX retroviral vector that were transplanted into two brothers suffering from hemophilia B. It appeared that one of the two patients responded and exhibited increased FIX clotting activity

14 M. K. L. Chuah et al.

that persisted over a year [142]. These results demon- strate the potential of gene therapy to correct hemophilia, albeit partially. However, it remains to be seen how reproducible these results are in a larger patient group.

A clinical trial of intramuscular injection of an AAV vector expressing human FIX has been initiated in adults with severe haemophilia B [177] based on preclinical studies demonstrating ef®cacy and absence of vector- related toxicity in hemophilic mice and dog models. The study has a dose-escalation design, and all patients have now been enrolled in the initial dose cohort (2r1011 vp/kg). Assessment in the ®rst three patients of safety and gene transfer and expression show no evidence of gene transfer in semen or formation of inhibitory antibodies against FIX, nor systemic or local toxicity. Muscle biopsies reveal the presence of vector sequence and FIX expression. Modest changes in clinical endpoints were observed including circulating levels of FIX and decreased frequency of FIX protein infusion. The evidence of gene expression at low doses of vector suggests that dose calculations based on animal data may have overestimated the amount of vector required to achieve therapeutic levels in humans, and that the approach offers the possibility of converting severe haemophilia B to a milder form of the disease.

A phase I clinical trial in severe hemophilia A patients started recently based on preclinical ef®cacy studies in rabbits and hemophilic dogs and other toxicological animal studies that involves peripheral intravenous infusion of high-titer FVIII retroviral vectors. Preliminary data support safety of vector administration of doses studied, with no related serious adverse events. No acute exacerbation of HCV or HIV disease has occurred, and all tests for vector in semen and for RCR have been negative, consistent with the preclinical studies. Gene transfer and FVIII expression data have not yet been reported.

Conclusions

The recent advances in viral and non-viral gene transfer technology has expedited the development of gene therapy for the treatment of hemophilia A and B. Several studies indicate that long-term therapeutic levels of FVIII or FIX can be achieved either in normal or hemophilic mice and even in hemophilic dogs. Some of these encouraging preclinical studies culminated in several phase I clinical trails for hemophilia A and B that were initiated in 1999. More recently, Inder Verma's group [165] and our own [77] unequivocally demonstrated stable physiologic levels of FVIII or FIX, effectively curing hemophilia in hemophilic mice but this has not yet been achieved in hemophilic dogs (Tables 1 and 2).

The ex vivo gene therapy strategies for hemophilia based on MoMLV-vectors have evolved from short-term or no expression of FVIII or FIX in vivo to more prolonged and increased levels of production by altering vector design and promoters and by improving re-implantation strategies. For the ®rst time, potentially therapeutic levels

of FVIII could be achieved both by ex vivo and in vivo gene therapy with MoMLV vectors containing the FVIII gene. Until recently, the titers of retroviral vectors were modest or extremely poor, particularly in the case of FVIII retroviral vectors, but it has now become possible to generate high-titer, concentrated retroviral vectors that permit direct in vivo gene therapy in animal models. In particular, hemophilia A mice have recently been cured by peripheral intravenous administration of FVIII- retroviral vectors [77] and partial correction of hemo- philia A has also been achieved in canine hemophilia but long-term expression was thwarted due to anti-human FVIII antibodies [85]. Though retroviral vectors require cell division for stable gene transfer, stable therapeutic FVIII levels could also be achieved in adult rabbits, paving the way towards a phase I clinical trial in severe hemophilia A patients [85].

Since lentiviral vectors are more ef®cient than onco- retroviral vectors in their ability to transduce both dividing and non-dividing cells, they are a logical choice for hemophilia gene therapy. Recent studies indicate that therapeutic levels of coagulation factors can be achieved in the absence of arti®cial induction of hepatocyte proliferation when improved vectors are employed [96]. However, other studies suggest that this may sometimes be dif®cult to achieve and that cell cycling would still be required to facilitate lentiviral gene transfer in the liver [90,91]. The continued evaluation of ef®cacy, safety and toxicity of lentiviral vectors will be essential before they can be considered for the treatment of hemophilia.

High-titer adenoviral vectors are still the most ef®cient at achieving high ef®ciency transduction of FVIII or FIX genes in vivo leading to phenotypic correction in hemophilia animal models [100,109,152±154]. However, the ongoing development of improved high-titer retro- viral, lentiviral and AAV vectors gradually diminishes this difference in transduction ef®ciency. The hepatotoxic effects of FVIII-containing adenoviral vectors could be circumvented but not eliminated by using stronger expression cassettes, allowing the use of lower vector doses. However, the immune response towards adeno- viral gene products in the transduced cells contributes to the decline in FVIII or FIX expression. The development of helper-dependent `gutless' adenoviral vectors, that do not contain adenoviral genes and which have a signi®cantly reduced acute and chronic toxicity, may provide a solution to this problem. Gutless FVIII-adenoviral vectors have led to long-term therapeutic levels of FVIII in hemophilic mice, but the induction of antibodies to FVIII made it dif®cult to prove that the mice could actually be cured with this approach [124]. The use of gutless adenoviral vectors has not yet proven to be effective in hemophilia dogs [178]. One concern that remains to be addressed further is that even in the absence of de novo expression of adenoviral gene products, adenoviral particles as such may still mount a CTL response that could curtail transgene expression.

Progress in the use of AAV for hemophilia gene therapy

Hemophilia Gene Therapy 15

has been very signi®cant and therapeutic levels of FVIII and FIX could be demonstrated in mice [128,135]. Given the size constraints of AAV, functional FVIII expression could be achieved by incorporating small promoter or by using two different vectors that expressed the heavy and light chains independently [131,132]. Again, antibodies to human FVIII hampered long-term expression in mice. A complete cure could be achieved by intrahepatic admin- istration of AAV-FIX containing improved expression cassettes [165]. Partial correction of the bleeding phenotype and stable expression of canine FIX has been demonstrated in hemophilia B dogs by intraportal or intramuscular injection of recombinant AAV-FIX [163,166,169,170]. Levels as high as 5% of normal physiologic levels could be achieved with no subsequent spontaneous bleeding in hemophilia B dogs that received an improved AAV-FIX vector intraportally. Some of these preclinical data constituted the basis of a phase I clinical trail whereby severe hemophilia B patients received high- titer AAV-FIX intramuscularly. No signi®cant adverse effects were recorded and encouraging clinical endpoints revealed detectable FIX in the plasma and reduced FIX usage [177].

In general, up-scaling of viral vector production for preclinical evaluation in canine models and for eventual clinical use remains an issue but is currently being improved through the development of ef®cient stable packaging cell lines and improved vector puri®cation strategies. In contrast, non-viral vectors are widely amenable to industrial-scale manufacture. Improved transfection techniques using naked plasmid DNA [174] or transposon-mediated integration [172] has recently been shown to lead to therapeutic FIX levels in normal or hemophilic mice paving the way towards better vector systems. In addition, the ability to speci®cally correct genes using RNA-DNA chimeric oligonucleotides [176] opens new perspectives for hemophilia gene therapy.

Another important development is the generation of mouse models for hemophilia A and B by targeted disruption of the FVIII and FIX genes, respectively [33,34]. The availability of these mouse models facilitated the preclinical evaluation of various gene therapy strategies prior to assessing ef®cacy, safety and immuno- genicity in the canine hemophilia models. The use of the canine models may require the development of vectors containing the autologous canine FVIII or FIX gene instead of the corresponding human genes, to avoid xenogeneic immune responses, which would more closely mimic the actual clinical gene therapy situation [179]. However, even when autologous proteins are used, inhibitory antibodies developed on several occasions which may be related to several confounding variables including the type of vector used, the purity of the vector preparation, the promoter used to drive FVIII or FIX expression, the site of administration, the transduced cell types (antigen-presenting cells) or the underlying genetic defect in the hemophilic animal.

The same variables may also in¯uence the outcome of an inhibitor response in hemophilia patients. Whether

gene therapy would increase or decrease the likelihood of inhibitor formation compared to protein replacement therapy is one of the important questions that still needs to be addressed. Continuous production of high levels FVIII or FIX in situ following gene therapy may actually induce immune tolerance reminiscent of current immune tolerization strategies by repeated high-dose clotting factor administration. Alternatively, gene transfer to naive patients could evoke the same or even a more potent immune response to FVIII or FIX. A major concern is that the use of viral vectors expressing coagulation factors or that impurities in the vector preparations may provide immunological `danger signals' that may facilitate inhibitor formation [180]. In addition, gene transfer may result in the presentation of endogenously synthesized FVIII- or FIX-derived peptides in the context of MHC class I molecules potentially resulting in cytotoxic T-cell responses that could eliminate the FVIII- or FIX-engineered target cells. The secreted FVIII protein that is produced in vivo may also be presented in the context of MHC class II as in the case of infused clotting factors. It is not clear whether gene therapy could break tolerance in patients that are tolerant to FVIII and whether inhibitors can be suppressed once they occur following gene therapy. Since many adult hemophilia patients have an infectious or in¯ammatory disease, complex interactions can in¯uence the therapeutic ef®cacy of the gene therapy procedure or the propensity for inhibitor formation and caution is warranted not to exacerbate these underlying conditions.

Despite the tremendous progress in the ®eld over the past few years many questions remain largely unexplored. Extensive gene therapy studies in preclinical hemophilia models are needed to anticipate the possible outcome in patients and to increase the overall ef®ciency of the various gene therapy strategies while further improving their safety. The development of hemophilia gene therapy will undoubtedly continue to contribute to a better understanding of vector±host interactions that will bene®t the entire ®eld of gene therapy.

References

1. Kaufman RJ. Advances toward gene therapy for hemophilia at the millennium. Hum Gene Ther 1999; 10: 2091±2107.

2. Peerlinck K, Arnout J, Di Giambattista M, et al. Factor VIII inhibitors in previously treated haemophilia A patients with a double virus-inactivated plasma derived factor VIII concentrate. Thromb Haemost 1997; 77: 80±86.

3. Lozier JN, Brinkhous KM. Gene therapy and the hemophilias [see Comments]. JAMA 1994; 271: 47±51.

4. Morgan RA, Anderson WF. Human gene therapy. Annu Rev Biochem 1993; 62: 191±217.

5. Hoeben RC. Gene therapy for the haemophilias: current status. Biologicals 1995; 23: 27±29.

6. Eisensmith RC, Woo SL. Viral vector-mediated gene therapy for hemophilia B. Thromb Haemost 1997; 78: 24±30.

7. Connelly S, Kaleko M. Gene therapy for hemophilia A. Thromb Haemost 1997; 78: 31±36.

8. Kurachi K, Yao SN, Furukawa M, Kurachi S. De®ciencies in factors IX and VIII: what is now known. Hosp Pract (Off Ed) 1992; 27: 41±51.

16 M. K. L. Chuah et al.

9. Fallaux FJ, Hoeben RC. Gene therapy for the hemophilias. Curr Opin Hematol 1996; 3: 385±389.

10. Thompson AR. Status of gene transfer for hemophilia A and B. Thromb Haemost 1991; 66: 119±122.

11. Chuah MK, Collen D, VandenDriessche T. Gene therapy for hemophilia: hopes and hurdles. Crit Rev Oncol Hematol 1998; 28: 153±171.

12. White IIGC, Roberts HR. Gene therapy for hemophilia: a step closer to reality. Mol Ther 2000; 1: 207±208.

13. Kay MA, High K. Gene therapy for the hemophilias [comment]. Proc Natl Acad Sci U S A 1999; 96: 9973±9975.

14. Walter J, High KA. Gene therapy for the hemophilias. Adv Vet Med 1997; 40: 119±134.

15. St Louis D, Verma IM. An alternative approach to somatic cell gene therapy. Proc Natl Acad Sci U S A 1988; 85: 3150±3154.

16. Axelrod JH, Read MS, Brinkhous KM, Verma IM. Phenotypic correction of factor IX de®ciency in skin ®broblasts of hemophilic dogs. Proc Natl Acad Sci U S A 1990; 87: 5173±5177.

17. Hoeben RC, van der Jagt RC, Schoute F, et al. Expression of functional factor VIII in primary human skin ®broblasts after retrovirus-mediated gene transfer. J Biol Chem 1990; 265: 7318±7323.

18. Page SM, Brownlee GG. An ex vivo keratinocyte model for gene therapy of hemophilia B. J Invest Dermatol 1997; 109: 139±145.

19. Fenjves ES, Yao SN, Kurachi K, Taichman LB. Loss of expression of a retrovirus-transduced gene in human keratinocytes. J Invest Dermatol 1996; 106: 576±578.

20. Fakharzadeh SS, Zhang Y, Sarkar R, Kazazian Jr HH. Correction of the coagulation defect in hemophilia A mice through factor VIII expression in skin. Blood 2000; 95: 2799±2805.

21. Chuah MK, VandenDriessche T, Morgan RA. Development and analysis of retroviral vectors expressing human factor VIII as a potential gene therapy for hemophilia A. Hum Gene Ther 1995; 6: 1363±1377.

22. Dwarki VJ, Belloni P, Nijjar T, et al. Gene therapy for hemophilia A: production of therapeutic levels of human factor VIII in vivo in mice. Proc Natl Acad Sci U S A 1995; 92: 1023±1027.

23. Yao SN, Wilson JM, Nabel EG, et al. Expression of human factor IX in rat capillary endothelial cells: toward somatic gene therapy for hemophilia B. Proc Natl Acad Sci U S A 1991; 88: 8101±8105.

24. Hortelano G, Al Hendy A, Ofosu FA, Chang PL. Delivery of human factor IX in mice by encapsulated recombinant myoblasts: a novel approach towards allogeneic gene therapy of hemophilia B. Blood 1996; 87: 5095±5103.

25. Yao SN, Kurachi K. Expression of human factor IX in mice after injection of genetically modi®ed myoblasts. Proc Natl Acad Sci U S A 1992; 89: 3357±3361.

26. Yao SN, Smith KJ, Kurachi K. Primary myoblast-mediated gene transfer: persistent expression of human factor IX in mice. Gene Ther 1994; 1: 99±107.

27. Dai Y, Roman M, Naviaux RK, Verma IM. Gene therapy via primary myoblasts: long-term expression of factor IX protein following transplantation in vivo. Proc Natl Acad Sci U S A 1992; 89: 10892±10895.

28. Baru M, Sha'anani J, Nur I. Retroviral-mediated in vivo gene transfer into muscle cells and synthesis of human factor IX in mice. Intervirology 1995; 38: 356±360.

29. Hao QL, Malik P, Salazar R, et al. Expression of biologically active human factor IX in human hematopoietic cells after retroviral vector-mediated gene transduction. Hum Gene Ther 1995; 6: 873±880.

30. Hoeben RC, Einerhand MP, Briet E, et al. Toward gene therapy in haemophilia A: retrovirus-mediated transfer of a factor VIII gene into murine haematopoietic progenitor cells. Thromb Haemost 1992; 67: 341±345.

31. Hurwitz DR, Kirchgesser M, Merrill W, et al. Systemic delivery of human growth hormone or human factor IX in dogs by reintroduced genetically modi®ed autologous bone marrow stromal cells. Hum Gene Ther 1997; 8: 137±156.

32. Do H, Healey JF, Waller EK, Lollar P. Expression of factor VIII by murine liver sinusoidal endothelial cells. J Biol Chem 1999; 274: 19587±19592.

33. Wang L, Zoppe M, Hackeng TM, et al. factor IX-de®cient mouse

model for hemophilia B gene therapy. Proc Natl Acad Sci U S A 1997; 94: 11563±11566.

34. Bi L, Lawler AM, Antonarakis SE, et al. Targeted disruption of the mouse factor VIII gene produces a model of haemophilia A [Letter]. Nat Genet 1995; 10: 119±121.

35. Bi L, Sarkar R, Naas T, et al. Further characterization of factor VIII-de®cient mice created by gene targeting: RNA and protein studies. Blood 1996; 88: 3446±3450.

36. Kundu RK, Sangiorgi F, Wu LY, et al. Targeted inactivation of the coagulation factor IX gene causes hemophilia B in mice. Blood 1998; 92: 168±174.

37. Giles AR, Tinlin S, Hoogendoorn H, et al. Development of factor VIII : C antibodies in dogs with hemophilia A (factor VIII : C de®ciency). Blood 1984; 63: 451±456.

38. Evans JP, Brinkhous KM, Brayer GD, et al. Canine hemophilia B resulting from a point mutation with unusual consequences. Proc Natl Acad Sci U S A 1989; 86: 10095±10099.

39. Zatloukal K, Cotten M, Berger M, et al. In vivo production of human factor VII in mice after intrasplenic implantation of primary ®broblasts transfected by receptor-mediated, adeno- virus-augmented gene delivery. Proc Natl Acad Sci U S A 1994; 91: 5148±5152.

40. Lozier JN, Thompson AR, Hu PC, et al. Ef®cient transfection of primary cells in a canine hemophilia B model using adenovirus- polylysine-DNA complexes. Hum Gene Ther 1994; 5: 313±322.

41. Gitschier J, Wood WI, Goralka TM, et al. Characterization of the human factor VIII gene. Nature 1984; 312: 326±330.

42. Wood WI, Capon DJ, Simonsen CC, et al. Expression of active human factor VIII from recombinant DNA clones. Nature 1984; 312: 330±337.

43. Toole JJ, Knopf JL, Wozney JM, et al. Molecular cloning of a cDNA encoding human antihaemophilic factor. Nature 1984; 312: 342±347.

44. Toole JJ, Pittman DD, Orr EC, et al. large region (approxi- mately equal to 95 kDa) of human factor VIII is dispensable for in vitro procoagulant activity. Proc Natl Acad Sci U S A 1986; 83: 5939±5942.

45. Berntorp E. Second generation, B-domain deleted recombinant factor VIII. Thromb Haemost 1997; 78: 256±260.

46. Israel DI, Kaufman RJ. Retroviral-mediated transfer and ampli®cation of a functional human factor VIII gene. Blood 1990; 75: 1074±1080.

47. Lynch CM, Israel DI, Kaufman RJ, Miller AD. Sequences in the coding region of clotting factor VIII act as dominant inhibitors of RNA accumulation and protein production. Hum Gene Ther 1993; 4: 259±272.

48. Hoeben RC, Fallaux FJ, Cramer SJ, et al. Expression of the blood-clotting factor-VIII cDNA is repressed by a transcriptional silencer located in its coding region. Blood 1995; 85: 2447±2454.

49. Fallaux FJ, Hoeben RC, Cramer SJ, et al. The human clotting factor VIII cDNA contains an autonomously replicating sequence consensus- and matrix attachment region-like sequence that binds a nuclear factor, represses heterologous gene expression, and mediates the transcriptional effects of sodium butyrate. Mol Cell Biol 1996; 16: 4264±4272.

50. Ill CR, Yang CQ, Bidlingmaier SM, et al. Optimization of the human factor VIII complementary DNA expression plasmid for gene therapy of hemophilia A. Blood Coagul Fibrinolysis 1997; 8 (Suppl. 2): S23±S30.

51. Koeberl DD, Halbert CL, Krumm A, Miller AD. Sequences within the coding regions of clotting factor VIII and CFTR block transcriptional elongation. Hum Gene Ther 1995; 6: 469±479.

52. Pittman DD, Kaufman RJ. Structure±function relationships of factor VIII elucidated through recombinant DNA technology. Thromb Haemost 1989; 61: 161±165.

53. Dorner AJ, Kaufman RJ. The levels of endoplasmic reticulum proteins and ATP affect folding and secretion of selective proteins. Biologicals 1994; 22: 103±112.

54. Kaufman RJ, Wasley LC, Dorner AJ. Synthesis, processing, and secretion of recombinant human factor VIII expressed in mammalian cells. J Biol Chem 1988; 263: 6352±6362.

55. Miller AD. Retroviral vectors. Curr Top Microbiol Immunol 1992; 158: 1±24.

56. Miller AD, Miller DG, Garcia JV, Lynch CM. Use of retroviral vectors for gene transfer and expression. Methods Enzymol 1993; 217: 581±599.

57. Anderson WF. Human gene therapy. Nature 1998; 392: 25±30.

Hemophilia Gene Therapy 17

58. Cornetta K, Morgan RA, Anderson WF. Safety issues related to retroviral-mediated gene transfer in humans. Hum Gene Ther 1991; 2: 5±14.

59. Anderson WF, McGarrity GJ, Moen RC. Report to the NIH Recombinant DNA Advisory Committee on murine replication- competent retrovirus (RCR) assays (February 17, 1993). Hum Gene Ther 1993; 4: 311±321.

60. Donahue RE, Kessler SW, Bodine D, et al. Helper virus induced T cell lymphoma in nonhuman primates after retroviral mediated gene transfer. J Exp Med 1992; 176: 1125±1135.

61. Hamer DH, Leder P. Splicing and the formation of stable RNA. Cell 1979; 18: 1299±1302.

62. Palmiter RD, Sandgren EP, Avarbock MR, et al. Heterologous introns can enhance expression of transgenes in mice. Proc Natl Acad Sci U S A 1991; 88: 478±482.

63. Krall W, Kohn DB. Expression levels by retroviral vectors based upon the N2 and the MFG backbones [Letter]. Gene Ther 1996; 3.

64. Hoeben RC, Fallaux FJ, Van Tilburg NH, et al. Toward gene therapy for hemophilia A: long-term persistence of factor VIII- secreting ®broblasts after transplantation into immunode®cient mice. Hum Gene Ther 1993; 4: 179±186.

65. Evans GL, Morgan RA. Genetic induction of immune tolerance to human clotting factor VIII in a mouse model for hemophilia A. Proc Natl Acad Sci U S A 1998; 95: 5734±5739.

66. Drize NJ, Surin VL, Gan OI, et al. Gene therapy model for stromal precursor cells of hematopoietic microenvironment. Leukemia 1992; 6 Suppl 3: 174S±175S.

67. Nolta JA, Hanley MB, Kohn DB. Sustained human hematopoi- esis in immunode®cient mice by cotransplantation of marrow stroma expressing human interleukin-3: analysis of gene transduction of long-lived progenitors. Blood 1994; 83: 3041±3051.

68. Prockop DJ. Marrow stromal cells as stem cells for nonhema- topoietic tissues. Science 1997; 276: 71±74.

69. Pittenger MF, Mackay AM, Beck SC, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999; 284: 143±147.

70. Li KJ, Dilber MS, Abedi MR, et al. Retroviral-mediated gene transfer into human bone marrow stromal cells: studies of ef®ciency and in vivo survival in SCID mice. Eur J Haematol 1995; 55: 302±306.

71. Chuah MK, Brems H, Vanslembrouck V, et al. Bone marrow stromal cells as targets for gene therapy of hemophilia A. Hum Gene Ther 1998; 9: 353±365.

72. Chuah MK, Van Damme A, Zwinnen H, et al. Long-term persistence of human bone marrow stromal cells transduced with factor VIII-retroviral vectors and transient production of therapeutic levels of human factor VIII in nonmyeloablated immunode®cient mice. Hum Gene Ther 2000; 11: 729±738.

73. Kay MA, Rothenberg S, Landen CN, et al. In vivo gene therapy of hemophilia B: sustained partial correction in factor IX- de®cient dogs [see Comments]. Science 1993; 262: 117±119.

74. Rettinger SD, Kennedy SC, Wu X, et al. Liver-directed gene therapy: quantitative evaluation of promoter elements by using in vivo retroviral transduction. Proc Natl Acad Sci U S A 1994; 91: 1460±1464.

75. Bosch A, McCray JrPB, Chang SM, et al. Proliferation induced by keratinocyte growth factor enhances in vivo retroviral- mediated gene transfer to mouse hepatocytes. J Clin Invest 1996; 98: 2683±2687.

76. Patijn GA, Lieber A, Schowalter DB, et al. Hepatocyte growth factor induces hepatocyte proliferation in vivo and allows for ef®cient retroviral-mediated gene transfer in mice. Hepatology 1998; 28: 707±716.

77. VandenDriessche T, Vanslembrouck V, Goovaerts I, et al. Long- term expression of human coagulation factor VIII and correction of hemophilia A after in vivo retroviral gene transfer in factor VIII-de®cient mice [see Comments]. Proc Natl Acad Sci U S A 1999; 96: 10379±10384.

78. Ory DS, Neugeboren BA, Mulligan RC. A stable human-derived packaging cell line for production of high titer retrovirus/ vesicular stomatitis virus G pseudotypes. Proc Natl Acad Sci U S A 1996; 93: 11400±11406.

79. Takeuchi Y, Porter CD, Strahan KM, et al. Sensitization of cells and retroviruses to human serum by (alpha 1-3) galactosyl- transferase. Nature 1996; 379: 85±88.

80. Cosset FL, Takeuchi Y, Battini JL, et al. High-titer packaging

cells producing recombinant retroviruses resistant to human serum. J Virol 1995; 69: 7430±7436.

81. Alison M. Liver stem cells: a two compartment system. Curr Opin Cell Biol 1998; 10: 710±715.

82. Kay MA, Fausto N. Liver regeneration: prospects for therapy based on new technologies. Mol Med Today 1997; 3: 108±115.

83. Tian YW, Smith PG, Yeoh GC. The oval-shaped cell as a candidate for a liver stem cell in embryonic, neonatal and precancerous liver: identi®cation based on morphology and immunohistochemical staining for albumin and pyruvate kinase isoenzyme expression. Histochem Cell Biol 1997; 107: 243±250.

84. Pittman DD, Alderman EM, Tomkinson KN, et al. Biochemical, immunological, and in vivo functional characterization of B- domain-deleted factor VIII. Blood 1993; 81: 2925±2935.

85. Greengard JS, Jolly DJ. Animal testing of retroviral-mediated gene therapy for factor VIII de®ciency. Thromb Haemost 1999; 82: 555±561.

86. Naldini L, Blomer U, Gage FH, et al. Ef®cient transfer, integration, and sustained long-term expression of the transgene in adult rat brains injected with a lentiviral vector. Proc Natl Acad Sci U S A 1996; 93: 11382±11388.

87. Naldini L, Blomer U, Gallay P, et al. In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector [see Comments]. Science 1996; 272: 263±267.

88. Blomer U, Naldini L, Kafri T, et al. Highly ef®cient and sustained gene transfer in adult neurons with a lentivirus vector. J Virol 1997; 71: 6641±6649.

89. Zufferey R, Nagy D, Mandel RJ, et al. Multiply attenuated lentiviral vector achieves ef®cient gene delivery in vivo. Nat Biotechnol 1997; 15: 871±875.

90. Park F, Ohashi K, Chiu W, et al. Ef®cient lentiviral transduction of liver requires cell cycling in vivo. Nat Genet 2000; 24: 49±52.

91. Park F, Ohashi K, Kay MA. Therapeutic levels of human factor VIII and IX using HIV-1-based lentiviral vectors in mouse liver. Blood 2000; 96: 1173±1176.

92. Kafri T, van Praag H, Ouyang L, et al. packaging cell line for lentivirus vectors. J Virol 1999; 73: 576±584.

93. Dull T, Zufferey R, Kelly M, et al. third-generation lentivirus vector with a conditional packaging system. J Virol 1998; 72: 8463±8471.

94. Kafri T, Blomer U, Peterson DA, et al. Sustained expression of genes delivered directly into liver and muscle by lentiviral vectors. Nat Genet 1997; 17: 314±317.

95. Chinnasamy D, Chinnasamy N, Enriquez MJ, et al. Lentiviral- mediated gene transfer into human lymphocytes: role of HIV-1 accessory proteins. Blood 2000; 96: 1309±1316.

96. Follenzi A, Ailles LE, Bakovic S, et al. Gene transfer by lentiviral vectors is limited by nuclear translocation and rescued by HIV-1 pol sequences. Nat Genet 2000; 25: 217±222.

97. Pastore L, Morral N, Zhou H, et al. Use of a liver-speci®c promoter reduces immune response to the transgene in adenoviral vectors. Hum Gene Ther 1999; 10: 1773±1781.

98. Berkner KL. Development of adenovirus vectors for the expression of heterologous genes. Biotechniques 1988; 6: 616±629.

99. Tripathy SK, Black HB, Goldwasser E, Leiden JM. Immune responses to transgene-encoded proteins limit the stability of gene expression after injection of replication-defective adeno- virus vectors. Nat Med 1996; 2: 545±550.

100. Connelly S, Mount J, Mauser A, et al. Complete short-term correction of canine hemophilia A by in vivo gene therapy. Blood 1996; 88: 3846±3853.

101. Lozier JN, Metzger ME, Donahue RE, Morgan RA. Adenovirus- mediated expression of human coagulation factor IX in the rhesus macaque is associated with dose-limiting toxicity [see Comments]. Blood 1999; 94: 3968±3975.

102. Yeh P, Dedieu JF, Orsini C, et al. Ef®cient dual transcomple- mentation of adenovirus E1 and E4 regions from a 293-derived cell line expressing a minimal E4 functional unit. J Virol 1996; 70: 559±565.

103. Armentano D, Sookdeo CC, Hehir KM, et al. Characterization of an adenovirus gene transfer vector containing an E4 deletion. Hum Gene Ther 1995; 6: 1343±1353.

104. Wang Q, Jia XC, Finer MH. A packaging cell line for propagation of recombinant adenovirus vectors containing two lethal gene-region deletions. Gene Ther 1995; 2: 775±783.

105. Gao GP, Yang Y, Wilson JM. Biology of adenovirus vectors with

18 M. K. L. Chuah et al.

E1 and E4 deletions for liver-directed gene therapy. J Virol 1996; 70: 8934±8943.

106. Connelly S, Smith TA, Dhir G, et al. In vivo gene delivery and expression of physiological levels of functional human factor VIII in mice. Hum Gene Ther 1995; 6: 185±193.

107. Connelly S, Gardner JM, McClelland A, Kaleko M. High-level tissue-speci®c expression of functional human factor VIII in mice. Hum Gene Ther 1996; 7: 183±195.

108. Connelly S, Gardner JM, Lyons RM, et al. Sustained expression of therapeutic levels of human factor VIII in mice. Blood 1996; 87: 4671±4677.

109. Connelly S, Andrews JL, Gallo AM, et al. Sustained phenotypic correction of murine hemophilia A by in vivo gene therapy. Blood 1998; 91: 3273±3281.

110. Connelly S, Andrews JL, Gallo-Penn AM, et al. Evaluation of an adenoviral vector encoding full-length human factor VIII in hemophiliac mice. Thromb Haemost 1999; 81: 234±239.

111. Gallo-Penn AM, Shirley PS, Andrews JL, et al. In vivo evaluation of an adenoviral vector encoding canine factor VIII: high-level, sustained expression in hemophiliac mice. Hum Gene Ther 1999; 10: 1791±1802.

112. Tinlin S, Webster S, Giles AR. The development of homologous (canine/anti-canine) antibodies in dogs with haemophilia A (factor VIII de®ciency): a ten-year longitudinal study. Thromb Haemost 1993; 69: 21±24.

113. Sarkar R, Gao GP, Chirmule N, et al. Partial correction of murine hemophilia A with neo-antigenic murine factor VIII. Hum Gene Ther 2000; 11: 881±894.

114. Lipshutz GS, Sarkar R, Flebbe-Rehwaldt L, et al. Short-term correction of factor VIII de®ciency in a murine model of hemophilia A after delivery of adenovirus murine factor VIII in utero. Proc Natl Acad Sci U S A 1999; 96: 13324±13329.

115. Mitani K, Graham FL, Caskey CT, Kochanek S. Rescue, propagation, and partial puri®cation of a helper virus- dependent adenovirus vector. Proc Natl Acad Sci U S A 1995; 92: 3854±3858.

116. Clemens PR, Kochanek S, Sunada Y, et al. In vivo muscle gene transfer of full-length dystrophin with an adenoviral vector that lacks all viral genes. Gene Ther 1996; 3: 965±972.

117. Kochanek S, Clemens PR, Mitani K, et al. new adenoviral vector: replacement of all viral coding sequences with 28 kb of DNA independently expressing both full-length dystrophin and beta-galactosidase. Proc Natl Acad Sci U S A 1996; 93: 5731±5736.

118. Chen HH, Mack LM, Kelly R, et al. Persistence in muscle of an adenoviral vector that lacks all viral genes. Proc Natl Acad Sci U S A 1997; 94: 1645±1650.

119. Parks RJ, Chen L, Anton M, et al. helper-dependent adenovirus vector system: removal of helper virus by Cre-mediated excision of the viral packaging signal. Proc Natl Acad Sci U S A 1996; 93: 13565±13570.

120. Parks RJ, Graham FL. A helper-dependent system for adenovirus vector production helps de®ne a lower limit for ef®cient DNA packaging. J Virol 1997; 71: 3293±3298.

121. Morral N, O'Neal W, Rice K, et al. Administration of helper- dependent adenoviral vectors and sequential delivery of different vector serotype for long-term liver-directed gene transfer in baboons. Proc Natl Acad Sci U S A 1999; 96: 12816±12821.

122. Kafri T, Morgan D, Krahl T, et al. Cellular immune response to adenoviral vector infected cells does not require de novo viral gene expression: implications for gene therapy. Proc Natl Acad Sci U S A 1998; 95: 11377±11382.

123. Lieber A, He CY, Kirillova I, Kay MA. Recombinant adeno- viruses with large deletions generated by Cre-mediated excision exhibit different biological properties compared with ®rst-generation vectors in vitro and in vivo. J Virol 1996; 70: 8944±8960.

124. Balague C, Zhou J, Dai Y, et al. Sustained high-level expression of full-length human factor VIII and restoration of clotting activity in hemophilic mice using a minimal adenovirus vector. Blood 2000; 95: 820±828.

125. Kotin RM. Prospects for the use of adeno-associated virus as a vector for human gene therapy. Hum Gene Ther 1994; 5: 793±801.

126. Xiao X, Li J, Samulski RJ. Ef®cient long-term gene transfer into muscle tissue of immunocompetent mice by adeno-associated virus vector. J Virol 1996; 70: 8098±8108.

127. Kaplitt MG, Leone P, Samulski RJ, et al. Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat Genet 1994; 8: 148±154.

128. Snyder RO, Miao CH, Patijn GA, et al. Persistent and therapeutic concentrations of human factor IX in mice after hepatic gene transfer of recombinant AAV vectors. Nat Genet 1997; 16: 270±276.

129. Miao CH, Nakai H, Thompson AR, et al. Nonrandom transduction of recombinant adeno-associated virus vectors in mouse hepatocytes in vivo: cell cycling does not in¯uence hepatocyte transduction. J Virol 2000; 74: 3793±3803.

130. Gnatenko DV, Saenko EL, Jesty J, et al. Human factor VIII can be packaged and functionally expressed in an adeno-associated virus background: applicability to haemophilia A gene therapy. Br J Haematol 1999; 104: 27±36.

131. Chao H, Mao L, Bruce AT, Walsh CE. Sustained expression of human factor VIII in mice using a parvovirus-based vector. Blood 2000; 95: 1594±1599.

132. Burton M, Nakai H, Colosi P, et al. Coexpression of factor VIII heavy and light chain adeno-associated viral vectors produces biologically active protein. Proc Natl Acad Sci U S A 1999; 96: 12725±12730.

133. Kurachi K, Furukawa M, Yao SN, Kurachi S. Biology of factor IX. Hematol Oncol Clin North Am 1992; 6: 991±997.

134. Liles D, Landen CN, Monroe DM, et al. Extravascular administration of factor IX: potential for replacement therapy of canine and human hemophilia B. Thromb Haemost 1997; 77: 944±948.

135. Herzog RW, Hagstrom JN, Kung SH, et al. Stable gene transfer and expression of human blood coagulation factor IX after intramuscular injection of recombinant adeno-associated virus. Proc Natl Acad Sci U S A 1997; 94: 5804±5809.

136. Anson DS, Hock RA, Austen D, et al. Towards gene therapy for hemophilia B. Mol Biol Med 1987; 4: 11±20.

137. Scharfmann R, Axelrod JH, Verma IM. Long-term in vivo expression of retrovirus-mediated gene transfer in mouse ®broblast implants. Proc Natl Acad Sci U S A 1991; 88: 4626±4630.

138. Hoeben RC, Migchielsen AA, van der Jagt RC, et al. Inactivation of the Moloney murine leukemia virus long terminal repeat in murine ®broblast cell lines is associated with methylation and dependent on its chromosomal position. J Virol 1991; 65: 904±912.

139. Zhou JM, Qiu XF, Lu DR, et al. Long-term expression of human factor IX cDNA in rabbits. Sci China B 1993; 36: 1333±1341.

140. Chen L, Nelson DM, Zheng Z, Morgan RA. Ex vivo ®broblast transduction in rabbits results in long-term (>600 days) factor IX expression in a small percentage of animals. Hum Gene Ther 1998; 9: 2341±2351.

141. Lu DR, Zhou JM, Zheng B, et al. Stage I clinical trial of gene therapy for hemophilia B. Sci China B 1993; 36: 1342±1351.

142. Qiu X, Lu D, Zhou J, et al. Implantation of autologous skin ®broblast genetically modi®ed to secrete clotting factor IX partially corrects the hemorrhagic tendencies in two hemo- philia B patients. Chin Med J Engl 1996; 109: 832±839.

143. Roman M, Axelrod JH, Dai Y, et al. Circulating human or canine factor IX from retrovirally transduced primary myo- blasts and established myoblast cell lines grafted into murine skeletal muscle. Somat Cell Mol Genet 1992; 18: 247±258.

144. Nelson DM, Metzger ME, Donahue RE, Morgan RA. In vivo retrovirus-mediated gene transfer into multiple hematopoietic lineages in rabbits without preconditioning. Hum Gene Ther 1997; 8: 747±754.

145. Armentano D, Thompson AR, Darlington G, Woo SL. Expres- sion of human factor IX in rabbit hepatocytes by retrovirus- mediated gene transfer: potential for gene therapy of hemophilia B. Proc Natl Acad Sci U S A 1990; 87: 6141±6145.

146. Gerrard AJ, Hudson DL, Brownlee GG, Watt FM. Towards gene therapy for haemophilia B using primary human keratinocytes. Nat Genet 1993; 3: 180±183.

147. Gerrard AJ, Austen DE, Brownlee GGRecombinantfactor. IX secreted by transduced human keratinocytes is biologically active. Br J Haematol 1996; 95: 561±563.

148. Lozier JN, Yankaskas JR, Ramsey WJ, et al. Gut epithelial cells as targets for gene therapy of hemophilia. Hum Gene Ther 1997; 8: 1481±1490.

149. Brauker JH, Carr Brendel VE, Martinson LA, et al. Neovascu-

Hemophilia Gene Therapy 19

larization of synthetic membranes directed by membrane microarchitecture. J Biomed Mater Res 1995; 29: 1517±1524.

150. Carr Brendel VE, Geller RL, Thomas TJ, et al. Transplantation of cells in an immunoisolation device for gene therapy. Methods Mol Biol 1997; 63: 373±387.

151. Hortelano G, Xu N, Vandenberg A, et al. Persistent delivery of factor IX in mice: gene therapy for hemophilia using implantable microcapsules. Hum Gene Ther 1999; 10: 1281±1288.

152. Kay MA, Landen CN, Rothenberg SR, et al. In vivo hepatic gene therapy: complete albeit transient correction of factor IX de®ciency in hemophilia B dogs. Proc Natl Acad Sci U S A 1994; 91: 2353±2357.

153. Smith TA, Mehaffey MG, Kayda DB, et al. Adenovirus mediated expression of therapeutic plasma levels of human factor IX in mice. Nat Genet 1993; 5: 397±402.

154. Kung SH, Hagstrom JN, Cass D, et al. Human factor IX corrects the bleeding diathesis of mice with hemophilia B. Blood 1998; 91: 784±790.

155. Dai Y, Schwarz EM, Gu D, et al. Cellular and humoral immune responses to adenoviral vectors containing factor IX gene: tolerization of factor IX and vector antigens allows for long- term expression. Proc Natl Acad Sci U S A 1995; 92: 1401±1405.

156. Fields PA, Kowalczyk DW, Arruda VR, et al. Role of vector in activation of T cell subsets in immune responses against the secreted transgene product factor IX. Mol Ther 2000; 1: 225±235.

157. Lozier JN, Metzger ME, Donahue RE, Morgan RA. The rhesus macaque as an animal model for hemophilia B gene therapy. Blood 1999; 93: 1875±1881.

158. Schneider H, Adebakin S, Themis M, et al. Therapeutic plasma concentrations of human factor IX in mice after gene delivery into the amniotic cavity: a model for the prenatal treatment of haemophilia B. J Gene Med 1999; 1: 424±432.

159. Fang B, Eisensmith RC, Wang H, et al. Gene therapy for hemophilia B: host immunosuppression prolongs the thera- peutic effect of adenovirus-mediated factor IX expression. Hum Gene Ther 1995; 6: 1039±1044.

160. Smith TA, White BD, Gardner JM, et al. Transient immuno- suppression permits successful repetitive intravenous admin- istration of an adenovirus vector. Gene Ther 1996; 3: 496±502.

161. Fang B, Wang H, Gordon G, et al. Lack of persistence of E1-recombinant adenoviral vectors containing a temperature- sensitive E2A mutation in immunocompetent mice and hemophilia B dogs. Gene Ther 1996; 3: 217±222.

162. Koeberl DD, Alexander IE, Halbert CL, et al. Persistent expression of human clotting factor IX from mouse liver after intravenous injection of adeno-associated virus vectors. Proc Natl Acad Sci U S A 1997; 94: 1426±1431.

163. Snyder RO, Miao C, Meuse L, et al. Correction of hemophilia B in canine and murine models using recombinant adeno- associated viral vectors [see Comments]. Nat Med 1999; 5: 64±70.

164. Nakai H, Iwaki Y, Kay MA, Couto LB. Isolation of recombinant adeno-associated virus vector-cellular DNA junctions from mouse liver. J Virol 1999; 73: 5438±5447.

165. Wang L, Takabe K, Bidlingmaier SM, et al. Sustained correction of bleeding disorder in hemophilia B mice by gene therapy. Proc Natl Acad Sci U S A 1999; 96: 3906±3910.

166. Wang L, Nichols TC, Read MS, et al. Sustained expression of therapeutic level of factor IX in hemophilia B dogs by AAV- mediated gene therapy in liver. Mol Ther 2000; 1: 154±158.

167. Hagstrom JN, Couto LB, Scallan C, et al. Improved muscle- derived expression of human coagulation factor IX from a skeletal actin/CMV hybrid enhancer/promoter. Blood 2000; 95: 2536±2542.

168. Monahan PE, Samulski RJ, Tazelaar J, et al. Direct intramus- cular injection with recombinant AAV vectors results in sustained expression in a dog model of hemophilia. Gene Ther 1998; 5: 40±49.

169. Herzog RW, Yang EY, Couto LB, et al. Long-term correction of canine hemophilia B by gene transfer of blood coagulation factor IX mediated by adeno-associated viral vector [see Comments]. Nat Med 1999; 5: 56±63.

170. Chao H, Samulski R, Bellinger D, et al. Persistent expression of canine factor IX in hemophilia B canines. Gene Ther 1999; 6: 1695±1704.

171. Young SM Jr, Xiao W, Samulski RJ. Site-speci®c targeting of DNA plasmids to chromosome 19 using AAV cis and trans sequences. Methods Mol Biol 2000; 133: 111±126.

172. Yant SR, Meuse L, Chiu W, et al. Somatic integration and long- term transgene expression in normal and haemophilic mice using a DNA transposon system. Nat Genet 2000; 25: 35±41.

173. Zhang G, Budker V, Wolff JA. High levels of foreign gene expression in hepatocytes after tail vein injections of naked plasmid DNA. Hum Gene Ther 1999; 10: 1735±1737.

174. Miao CH, Ohashi K, Patijn GA, et al. Inclusion of the hepatic locus control region, an intron, and untranslated region increases and stabilizes hepatic factor IX gene expression in vivo but not in vitro [in process citation]. Mol Ther 2000; 1: 522±532.

175. Kren BT, Parashar B, Bandyopadhyay P, et al. Correction of the UDP-glucuronosyltransferase gene defect in the Gunn rat model of Crigler-Najjar syndrome type I with a chimeric oligonucleotide. Proc Natl Acad Sci U S A 1999; 96: 10349±10354.

176. Kren BT, Bandyopadhyay P, Steer CJ. In vivo site-directed mutagenesis of the factor IX gene by chimeric RNA/DNA oligonucleotides [see Comments]. Nat Med 1998; 4: 285±290.

177. Kay MA, Manno CS, Ragni MV, et al. Evidence for gene transfer and expression of factor IX in haemophilia B patients treated with an AAV vector [see Comments]. Nat Genet 2000; 24: 257±261.

178. Zhang WW, Josephs SF, Zhou J, et al. Development and application of a minimal-adenoviral vector system for gene therapy of hemophilia A. Thromb Haemost 1999; 82: 562±571.

179. Cameron C, Notley C, Hoyle S, et al. The canine factor VIII cDNA and 5k ¯anking sequence. Thromb Haemost 1998; 79: 317±322.

180. Matzinger P. An innate sense of danger. Semin Immunol 1998; 10: 399±415.

20 M. K. L. Chuah et al.