Correction in Female PKU Mice by Repeated

2 downloads 0 Views 469KB Size Report
another set of CD1 nude mice were repeatedly injected with an integration vector ..... Thomason, LC, Calendar, R and Ow, DW (2001). Gene insertion and ...
original article

© The American Society of Gene Therapy

Correction in Female PKU Mice by Repeated Administration of mPAH cDNA Using phiBT1 Integration System Li Chen1 and Savio LC Woo1 Department of Gene and Cell Medicine, Mount Sinai School of Medicine, New York, New York, USA

1

Phenylketonuria (PKU) is a metabolic disorder secondary to a hepatic deficiency of phenylalanine hydroxylase (PAH) that predisposes affected children to develop severe and irreversible mental retardation. We have previously reported the complete and permanent correction of the hyperphenylalaninemic and hypopigmentation phenotypes in male, but not female, PKU mice after genometargeted delivery of murine PAH (mPAH) complementary DNA (cDNA) in a phiBT1 bacteriophage integration system. Here we show that sequential administration of green fluorescent protein (GFP)– and red fluorescent protein (RFP)–expressing cassettes in the phiBT1 integration system led to distinct and non-overlapping populations of green and red fluorescent hepatocytes in vivo. The hyperphenylalaninemic and hypopigmentation phenotypes of female PKU mice were completely corrected after 10 weekly administrations of mPAH cDNA. Importantly, there was no apparent liver pathology in mice even after 10 consecutive administrations of the phiBT1 integration system. The results indicate that repeated administration of transgenes in the phiBT1 integration system can lead to their genome-targeted integration in a diverse population of hepatocytes and result in the elevation of transgene expression levels in a cumulative manner, which can be utilized to overcome insufficient transgene expression owing to low genome integration frequencies in a gene therapy paradigm for metabolic disorders. Received 13 February 2007; accepted 18 May 2007; published online 17 July 2007. doi:10.1038/sj.mt.6300257

Introduction In recent years, genome-targeted integration of transgenes utilizing bacteriophage integrase systems has been widely used in genetic manipulation of cells and animals, including the generation of transgenic mice,1 gene therapy for genetic diseases in animal models,2–7 and gene transfer into stem cells.8–10 The ­bacteriophage-encoded integrases expressed during lysogeny11 carry out a recombination reaction between the attachment sites in phage DNA (attP) and those in the bacterial genome (attB), resulting in the integration of phage DNA into the bacterial

genome in a site-­specific manner. The integration is a unidirectional reaction: subsequent release of phage DNA from the bacterial genomes requires the activity of a second phage-encoded enzyme called excisionase.12 A limited number of pseudo­attachment (pseudo-attP) sites also occur naturally in many mammalian genomes.13–16 These pseudo-attP sites can recombine with plasmid DNAs bearing a wild-type attachment site in the presence of the corresponding bacteriophage integrase, leading to integration of the plasmid into the mammalian genome in a site-specific and unidirectional manner. This unique ability has allowed the phage integrase to become a very important tool for a multitude of applications in bacterial and mammalian genetics.4,17–24 To date, five members of this integrase family have been studied functionally: phiC31,25 TP901-1,26 phiFC1,27,28 R4,29 and phiBT1.15,30 In our previously study,15 we reported the identification of eight different pseudo-attP sites of the phiBT1 bacteriophage integrase system, all of which occurred in the intergenic regions of various chromosomes in the mouse genome. The phiBT1 integrase system has been successfully used for genometargeted delivery of reporter genes into mouse hepatocytes by hydrodynamic injection, which led to the persistent expression of transgenes.15 Classical phenylketonuria (PKU)31,32 is an autosomal recessive disorder caused by the deficiency of the hepatic enzyme phenylalanine hydroxylase (PAH), which is responsible for the de novo synthesis of tyrosine from phenylalanine. Affected individuals are predisposed to development of severe and irreversible mental retardation and exhibit phenotypes of hyperphenylalaninemia and hypopigmentation. PKU can be managed by life-long dietary control, although the treatment is often accompanied by poor patient compliance. There is a mouse model of classical PKU that mimics the human conditions33,34 and has been widely used for the development of novel therapeutics. Gene therapy offers an attractive strategy to correct the disorder by introducing a functional PAH gene into the deficient hepatocytes of PKU mice. Various viral vectors have been used for this purpose, including recombinant retrovirus,35 adenovirus,36 and adeno-associated virus.37–40 In addition, non-viral vectors such as the phiBT1 bacteriophage integrase system have been used to correct PKU. In a previous study, we reported the construction of a plasmid

Correspondence: Savio L.C. Woo, Department of Gene and Cell Medicine, Mount Sinai School of Medicine, One Gustave L. Levy Place, Box 1496, New York, New York 10029, USA. E-mail: [email protected] Molecular Therapy vol. 15 no. 10, 1789–1795 oct. 2007

1789

© The American Society of Gene Therapy

phiBT1-PAH Repeated Injections Cure Female PKU Mice

c­ ontaining a mouse PAH complementary DNA (cDNA) expression cassette and the wild-type attB sequence. This plasmid was co-delivered by hydrodynamic injection into tail veins of PKU mice with a second plasmid expressing phiBT1 integrase. After three injections that led to the genetic reconstitution of 2–3% of hepatocytes, complete and permanent correction of the hyperphenylalaninemic and hypopigmentation phenotypes was achieved in male mice, which exhibited 15–20% of normal PAH activity in liver extracts.15 Surprisingly, although the same level of PAH activity was present in the liver extracts of similarly treated female PKU mice, only partial correction of their hyperphenylalaninemic phenotype, and no correction of the hypopigmentation phenotype, was achieved. The biochemical basis of the sexual dimorphism in treatment outcome was shown subsequently to be secondary to a lower hepatic level of the enzyme’s obligate co-factor tetrahydrobiopterin (BH4) in females,41 which became rate-limiting in the 2–3% of genetically reconstituted hepatocytes expressing 15–20% of enzyme activity. This limitation could potentially be resolved by insertion of the murine PAH (mPAH) transgene into a correspondingly higher percentage of hepatocytes in female PKU mice. In this article we report that weekly administrations of the phiBT1 integrase system led to transgene integration into divergent populations of hepatocytes in vivo, with a linearly cumulative effect on the level of transgene expression. The strategy was used to achieve complete correction of the hyperphenylalaninemic and hypopigmentation phenotypes in female PKU mice after 10 weekly administrations of mPAH cDNA in the phiBT1 integrase system.

Results Integration into distinct hepatocytes after two consecutive phiBT1 vector administrations In a previous study using the phiBT1 integration system, we reported that approximately 1% of hepatocytes contained integrated DNA after a single administration by hydrodynamic injection into mice.15 Molecular characterization of hepatic DNA in these mice indicated that more than 95% of the integration events occurred at mpsP3, which is the major site among the eight known pseudo-attP sites in the mouse genome. To evaluate whether repeated administration of transgenes would lead to genome-targeted integration in alternative pseudo-attP sites in the same hepatocytes or the mps3 sites of distinct hepatocytes, two phiBT1 integration vectors expressing green fluorescent protein (GFP) (pCG-B) and red fluorescent protein (RFP) (pCR-B), as well as a phiBT1 integrase–expressing plasmid (pCMV-BTIntNLS), were constructed (Figure 1a). CD1 nude mice were injected with pCG-B plus pCMV-BTIntNLS, and then with pCR-B plus pCMV-BTIntNLS at the same dose 1 week later. The injected mice were killed after two more weeks, when un-integrated DNA is known to have been lost,15 and liver sections were examined with a fluorescence microscope under different wavelengths. Although the number of green and red hepatocytes was similar, there were few orange hepatocytes on the merged photograph (Figure 1b), indicating that the ­transgenes were integrated into the genomes of distinct hepatocytes when administered 1 week apart. 1790

a

CAG

GFP

pA

attB

CAG

RFP

pA

attB

CMV

Int

CAG

lacZ

CAG

mPAH

pCG-B pCR-B pCMV-BT IntNLS

NLS

pCZiS-B IRES

SEAP pA

attB

pA

attB pCmPAH-B

b

Figure 1 Schematic representation and repeated administrations of green fluorescent protein (GFP)– and red fluorescent protein (RFP)– expressing vectors in the phiBT1 integrase system. (a) Schematic repre-­ sentation of recombinant phiBT1 vector structures. CAG, CAG promoter; GFP, green fluorescent protein gene; RFP, red fluorescent protein gene; pA, bovine growth hormone PolyA signal; Int, phiBT1 integrase gene; NLS, SV40 nuclear localization signal; IRES, the internal ribosome entry site from the encephalomyocarditis virus (ECMV); SEAP, secreted human alka-­ line phosphatase gene; mPAH, murine PAH cDNA. (b) Non-­overlapping populations of GFP- and RFP-positive hepatocytes after repeated admin-­ istrations. GFP- and RFP-expressing plasmids in the phiBT1 system were administered 1 week apart into the tail veins of four CD1 nude mice. The mice were maintained for 2 additional weeks to allow the non-integrated plasmid DNA to clear from the hepatocytes. The treated mice were then killed, and liver sections were analyzed by immunofluorescence. As shown in the representative fluorescent images of liver sections from four differ-­ ent mice, transgene-expressing hepatocytes appeared green (left-hand panels) and red (right-hand panels) after GFP- and RFP-expressing plas-­ mid administrations, respectively. Merging of the two images indicates that most of the hepatocytes remain either green or red, and only a min-­ ute fraction of the hepatocytes are yellow (middle panels).

Repeated phiBT1 vector administration led to cumulative elevation of transgene-expression To evaluate whether the level of transgene expression can be elevated by repeated administrations of the phiBT1 vector ­system, another set of CD1 nude mice were repeatedly injected with an integration vector expressing secreted human alkaline phosphatase (SEAP) and lacZ (pCZiS-B) plus pCMV-BTIntNLS15 (Figure 1a), and blood samples were collected weekly from the treated mice for SEAP measurements (Figure 2a). The peak www.moleculartherapy.org vol. 15 no. 10 oct. 2007

© The American Society of Gene Therapy

a

phiBT1-PAH Repeated Injections Cure Female PKU Mice

int+ 1 injection int − 1 injection int+ 10injections int − 10 injections PBS 1 injection PBS 10 injections

10 weekly injections

1,800 1,600 1,400

b

A

B

C

D

SEAP (µg/ml)

1,200 1,000 800 600 400 200 0 0

2

3

4

5

6

7 8 9 10 11 12 13 14 15 Weeks

d

30

Integration per haploid genome %

c

1

Blue hepatocyte %

25 20 15 10 5 0

int−

int+

25

20 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

19.56

1.47

0.52 0.04 mpsP3

mpsP5

0.49

0.04

0.03

0.05

0.04

mpsP7 mpsP4.1 mpsP4.2 mpsP8

0.03

0.04

mpsP9 mpsP10

--

Figure 2  Cumulative elevation of serum secreted human alkaline phosphatase (SEAP) expression levels and percentage LacZ-positive hepatocytes after repeated administrations using the phiBT1 vector system. (a) Serum SEAP curves in CD1 nude mice after a single versus repeated administrations in the presence or absence of phiBT1 integrase. Six CD1 nude mice per group were injected either once at week 0 or with 10 consecutive injections in weeks 0–9, as indicated by the arrows. Blood samples were collected immediately before each injection on a weekly basis and used for SEAP determinations by enzyme-linked immunosorbent assay. Solid squares and triangles: CD1 nude mice after 1 and 10 weekly injections, respectively, of the transgenes-expressing plasmid in the presence of a phiBT1 integrase–expressing plasmid; open squares and triangles: CD1 nude mice after 1 and 10 weekly injections, respectively, of the transgenes-expressing plasmid in the absence of a phiBT1 integrase–expressing plasmid; open diamonds and circles: CD1 nude mice after 1 and 10 weekly injections, respectively, of phosphate-buffered saline. By the end of week 15, serum SEAP levels in all the integrase-negative treatment groups reached background range. In the integrase-positive treatment groups, serum SEAP levels from mice that received 10 injections were approximately tenfold higher than those from mice that received a single injection (P = 0.005). The error bars represent SDs from multiple blood samples from different mice. (b) LacZ staining of the liver sections from CD1 nude mice in the presence and absence of phiBT1 integrase. At the end of week 15, blue hepatocytes were not detected in the integrase-negative treatment groups after (B) 1 or (D) 10 injections. (A) Only a small fraction of the hepatocytes were stained blue in mice that received a single injection of the vector, and (C) the frequency was greatly enhanced in livers of mice that received 10 injections of the vector. (c) Quantitation of blue hepatocytes by morphometry analysis. The liver sections after beta-gal staining were subjected to morphometric analysis to quantify the ratios of blue hepatocytes. As shown in the figure, there were no detectable signals from the integrase-negative treatment groups after 1 (solid bar) or 10 injections (open bar) of the phiBT1 vector system. In the integrase-positive treatment groups, the blue cell ratios were greatly elevated from approximately 1.7% after 1 injection to approximately 27.4% after 10 injections (P = 0.0006). Error bars represent SDs of multiple measurements. (d) Integration frequencies at various pseudo-attachment sites. Integration frequencies at all of the pseudo-attachment sites in genomic DNA samples from the livers of mice after receiving 1 (solid bars) and 10 (open bars) injections of the vector system were determined by polymerase chain reaction analyses using specific primer pairs. After 1 and 10 injections, mpsP3 was the major integration site, with integration frequencies of 1.47 and 19.56%, respectively (P = 0.006). Similarly, integration frequencies at mpsP5 and mpsP7 were also elevated 11–12 times (P = 0.003 and 0.02, respectively). As for the other five minor sites, integration signals became detectable after 10 injections, and collectively they represent approximately 1% of the total integration events. The error bars represent SDs from different samples.

serum SEAP concentrations in the integrase-negative mouse group occurred 1 week after a single administration of the vector. SEAP levels reached approximately 600 µg/ml, followed by a rapid decline to background levels after 2–3 weeks. However, levels in the integrase-positive mouse group reached a plateau at approximately 150 µg/ml. After 10 weekly injections of the vector system, serum SEAP levels in the integrase-negative mouse group remained at approximately 600 µg/ml; they declined to the background level 2 weeks after cessation of injections. Conversely, after repeated injections in the integrase-positive mouse group, serum SEAP levels accumulated linearly over time and reached a steady state of approximately 1,100 µg/ml after cessation of vector injection. Molecular Therapy vol. 15 no. 10 oct. 2007

At week 16, all mice were killed, and the livers were removed for further analyses. After beta-gal staining, there were no blue hepatocytes in the integrase-negative groups after 1 or 10 injections (Figure 2b). Blue hepatocytes were found in the integrasepositive treatment groups, and many more blue cells were present after 10 injections (Figure 2b). The percentages of blue cells on multiple liver sections were analyzed by morphometry, and the results indicated that there was an elevation from 1.7 to 27.4% after 10 injections of the phiBT1 vector system (Figure 2c). Genomic DNA extracted from the liver samples was also used to determine the integration frequencies by quantitative polymerase chain reaction analysis. There were no detectable integrated transgene sequences at any of the pseudo-attP sites in the integrase-negative 1791

© The American Society of Gene Therapy

phiBT1-PAH Repeated Injections Cure Female PKU Mice

a

a 90

70 60

AST (U/l)

80

50 40 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Weeks

30

200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50

100 90 80 70 60 50

30

1,800 1,600 1,400 1,200 Homozygous with 10 PBS injections Homozygous with 10 pDNA injections Heterozygous with 10 PBS injections

1,000 800 600 400 200 0 –1

40 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Weeks

2,000 Serum phenylalanine (µmol/l)

B

ALT (U/l)

b

190 180 170 160 150 140 130 120 110 100 90 80 70 60 50

100

ALT (U/l)

AST (U/l)

A 200

0

1

2

3

4

5 6 Weeks

C

9

10 11

Treated female

Heterozygous

B

D

Figure 3  Apparent lack of liver toxicity after repeated administrations of the phiBT1 integrase vector system. (a) Serum aspartate amino transferase (AST) and alanine amino transferase (ALT) levels in CD1 nude mice after repeated administrations of the transgene-expressing vector in the presence of phiBT1 integrase. To examine the potential liver toxicity in mice after repeated administration of the phiBT1 plasmids, serum AST and ALT levels were measured and further analyzed using a one-­population Student’s t-test. (A) In CD1 nude mice that received a single injection, there were no significant changes in either the AST (open squares, P = 0.81) or ALT (solid circles, P = 0.94) levels over time. (B) In CD1 nude mice that received 10 injections of integrase-positive plasmids, both serum AST (open squares, P = 0.92) and ALT (solid circles, P = 0.99) ­levels also remained stable over time. Similarly, no significant or persistent elevation of AST and ALT in the integrase-negative treatment groups was observed (data not shown). (b) Hematoxylin and eosin staining of liver sections of vector-treated CD1 nude mice. All liver sections looked normal by the end of week 16. (A) Mice with one injection of plasmids with integrase; (B) mice with one injection of plasmids without integrase; (C) mice injected 10 times with plasmids with integrase; (D) mice injected 10 times with plasmids without integrase.

mouse groups after 1 or 10 vector administrations (Figure 2d). In the integrase-positive treatment groups, the integration frequencies at the mpsP3 site increased from 1.47 to 19.56% after 10 administrations of the vector, and detectable levels of integration frequencies were also observed at the minor pseudo-attP sites (Figure 2d). Polymerase chain reaction analysis was performed on total hepatic DNA preparations from the treated PKU mice using internal primers to the LacZ transgene, and no detectable level of random DNA integration was apparent (results not shown).

Apparent lack of liver and other organ toxicities after repeated vector administrations Although the percentages of transgene-positive hepatocytes and the levels of transgene expression can be elevated linearly by repeated vector administration, it is important to determine the extent, if any, of liver pathology after repeated administrations of 1792

8

b Homozygous

A

7

Figure 4  Correction of the hyperphenylalaninemic and hypopigmentation phenotypes in female phenylketonuria (PKU) mice after repeated murine phenylalanine hydroxylase (mPAH) gene administrations in the phiBT1 system. (a) Serum phenylalanine curves. Eight female PKU mice per group were injected weekly with mPAH-expressing plasmid in the phiBT1 integration system in weeks 0–9, as indicated by the arrows. Blood samples were collected from the injected mice on a weekly basis and used for phenylalanine determination by highpressure liquid chromatography. Open circles: homozygous female PKU mice that received multiple plasmid injections; solid squares: homozygous female PKU mice treated with 10 injections of phosphate-buffered saline (PBS); solid triangles: heterozygous mice treated with 10 injections of PBS as normal control. There were no significant differences between the serum phenylalanine levels in the vector-treated female PKU mice and the heterozygous mice at weeks 10 and 11 according to a two-population t-test (P = 0.09). The error bars represented the SDs of multiple samples. (b) Change of fur color of the treated female PKU mice. At the end of week 10, the homozygous PKU mice treated with PBS remained gray (left-hand panel), and the vector-treated female PKU mice had turned completely black (middle panel), which made them look the same as the heterozygous mice (right-hand panel).

the phiBT1 vector system. There was no significant or persistent elevation of liver enzymes in mouse blood after repeated vector administrations (Figure 3a). Livers from all mouse treatment groups were analyzed by histology, and the results showed that there was no apparent liver pathology (Figure 3b). In addition, no pathology was observed in other major organs, including heart, kidney, lung, and spleen (not shown).

Complete correction in female PKU mice by repeated administration of mPAH-expressing vector Female PKU mice were injected weekly with a phiBT1 plasmid expressing mPAH cDNA pCmPAH-B plus pCMV-BTIntNLS (Figure 1a), and blood samples were collected weekly for determination of serum phenylalanine levels. As shown in Figure 4a, serum phenylalanine levels in buffer-treated female PKU mice remained approximately 1,800 µmol/l, whereas those of the heterozygous mice were in the 50–100 µmol/l range. Serum ­phenylalanine levels in the vector-treated female PKU mice started to decrease 1 week after vector administration and continued downward until the normal range was reached after 10 repeated injections (Figure 4a). By week 7 and after eight injections, the gene-treated female PKU mice started to change their fur color, and after one to two more injections, by weeks 9–10, all of them turned black (Figure 4b). www.moleculartherapy.org vol. 15 no. 10 oct. 2007

© The American Society of Gene Therapy

a

phiBT1-PAH Repeated Injections Cure Female PKU Mice

which is comparable to levels in heterozygous mice (Figure 5b). In addition, serum aspartate amino transferase and alanine amino transferase levels were normal in the treated mice, and no apparent pathology was observed in their livers and the other major organs (results not shown).

24

Integration per haploid genome %

23 22

18.75

21 20 19 18 0.55

0.8

Discussion

0.48

0.6 0.4 0.2

0.04

0.03

0.05

0.0 mpsP3

% Normal PAH activity

b

mpsP5

mpsP7 mpsP4.1 mpsP4.2 mpsP8

0.05

0.04

mpsP9 mpsP10

60

50

40 4 3 2 1 0

Heterozygous

Homozygous

Treated females

Figure 5  Plasmid integration and phenylalanine hydroxylase (PAH) activity in the livers of treated female phenylketonuria (PKU) mice. (a) Integration frequency at various pseudo-attachment sites in the livers of treated female PKU mice. At the end of week 12, all mice were killed and the liver samples were subjected to molecular analysis. Genomic DNA extracted from the liver samples was used to determine the frequencies of site-specific genomic integrations by quantitative polymerase chain reaction using primers specifically designed for each of the known integration sites in the mouse genome. The SDs from different samples are shown as error bars. After 10 injections, mpsP3 proved to be the major site of integration events, with a much higher frequency (P = 0.00001), at 18.75%, than mpsP5 at 0.55% and mpsP7at 0.48%. The remaining five pseudo-attachment sites accounted for less than 1% of total integration events. (b) PAH activity in the liver extracts of treated female PKU mice. PAH enzymatic activity in the liver extracts of heterozygous mice was approximately 50% (open bar) of normal, and that of untreated homozygous PKU mice was less than 1% (solid bar) of normal, as expected. PAH activity was approximately 55% of normal in liver extracts of female PKU mice that received 10 plasmid injections (striped bar), which was not significantly different from the heterozygous mice (P = 0.72). SDs from multiple samples are shown as error bars.

All mice were killed at week 11 for further studies. Genomic DNA extracted from the livers of control and vector-treated female PKU mice was subjected to quantitative polymerase chain reaction analysis to determine the integration frequencies at individual pseudo-attP sites after 10 repeated administrations. Similar to the data obtained from CD1 nude mice (Figure 2d), integrations occurring at all pseudo-attP sites increased by approximately tenfold, but the relative frequencies between the sites remained unchanged (Figure 5a). The total integration frequency reached 24% per haploid genome after 10 weekly injections. More than 95% of integration events occurred at the major site mpsP3, with 4% in mpsP5 and mpsP7, and 1% shared by the remaining five minor sites. The liver extracts of the treated mice were also used to measure PAH enzyme activities. Although there was no detectable change in activity in controltreated female PKU mice, PAH activity in the treatment group increased to 55% of normal after 10 weekly vector injections, Molecular Therapy vol. 15 no. 10 oct. 2007

In this article we report genome-targeted integration of transgenes into distinct and non-overlapping populations of hepatocytes after repeated hydrodynamic administration of a vector based on the phiBT1 bacteriophage integrase system, and the application of the methodology to correct completely the hyperphenylalaninemic and hypopigmentation phenotypes in female PKU mice. Genome-targeted transgene delivery mediated by the bacteriophage integrase system has been widely used in many fields of genetic manipulation. A major advantage is its targeted nature of transgene integration into mammalian genomes, where transgene expression will be regulated more uniformly than random integration into a wide spectrum of chromosomal sites. There are, however, limitations to the technology: one limitation is the relatively low transgene integration frequency. Although hydrodynamic injection in mice allowed effective entry of plasmid DNA into the hepatocytes,42 only approximately 1% contain integrated plasmid DNA mediated by phiBT1 integrase, leading to relatively low levels of transgene expression in vivo. While an integration efficiency of 2 × 10–2 (approximately 1 in 50 transfected hepatocytes) appears to be low, bacteriophage integrase–mediated transgene integration actually occurs at frequencies that are three to four orders of magnitude higher than those of non-homologous recombination events at 10–5 to 10–6. Although the elevation in integration frequency mediated by the integrase is impressive, the low percentages of hepatocytes containing integrated transgenes are not optimal for gene therapy applications in metabolic disorders. An example is provided by the PAH gene treatment of PKU mice. In contrast to their male counterparts, the expression of 15–20% of PAH activity in 2–3% of transgene-containing hepatocytes of the female PKU mice was insufficient to correct their hyperphenylalaninemic and hypopigmentation phenotypes completely. The basis of this sexual dimorphism in response to PAH gene treatment is reduced levels of BH4 in hepatocytes of female mice, such that the obligate co-factor for the PAH enzyme becomes the rate-limiting factor in the genetically reconstituted hepatocytes of female PKU mice over-expressing PAH.43 We have further demonstrated that the underlying mechanism is estrogen-mediated suppression, not testosterone-mediated induction, of the endogenous rate of BH4 biosynthesis and/or regeneration in the hepatocytes.43 One strategy to overcome this limitation is effectively to transduce a higher percentage of hepatocytes by phiBT1­mediated transgene integration. This can be achieved if the integration events take place in divergent populations of hepatocytes after repeated administrations of the vector. In this article, we presented unambiguous evidence that CD1 nude mice treated ­successively with a GFP- and a RFP-expressing vector expressed the transgenes in qualitatively distinct populations of hepatocytes. Furthermore, CD1 nude mice treated with 10 weekly injections of a plasmid expressing SEAP and lacZ showed a 1793

© The American Society of Gene Therapy

phiBT1-PAH Repeated Injections Cure Female PKU Mice

l­ inear ­ accumulation of serum SEAP levels and percentages of blue hepatocytes after lacZ staining. Interestingly, the proportion of major versus minor pseudo-attP sites used for transgene integrations in the hepatocytes of mice was maintained after single application and after 10 applications of the phiBT1 integrase vector system. The results indicate that the efficiency of transgene integration can be linearly elevated by repeated vector administrations, which was indeed the case in the phiBT1mediated PAH gene treatment of female PKU mice. Ten weekly injections of the vector led to a complete correction of their hyperphenylalaninemic and hypopigmentation phenotypes. Because the PAH gene defect in this PKU mouse model is a cross-reacting material-positive missense mutation, it is difficult to determine the percentage of hepatocytes in the female PKU mice that became genetically reconstituted after 10 weekly injections of the phiBT1 integrase vector system. Inferring from the CD1 nude mouse experiment, which showed a linear elevation in the percentage of hepatocytes that were transduced after repeated administration of a lacZ-expressing vector in the phiBT1 integrase system, we suggest that a proportionally higher percentage of hepatocytes in the female PKU mice were genetically reconstituted after repeated vector administrations, which led to a much greater level of transgene integration in hepatocytes and production of 55% of normal PAH activity in liver extracts. One potential limitation to repeated administration of transgenes in the phiBT1 integrase system in the treatment of metabolic disorders is toxicity associated with the vector, particularly in light of recent reports of frequent chromosomal translocation events in mammalian cells transfected with a phiC31 integrase– expressing plasmid.13,44 In this regard, there was no evidence of gross toxicity in the phiBT1 integrase–treated animals; they all appeared healthy throughout the course of gene treatment. There was also no apparent liver toxicity in either vector-treated CD1 nude or PKU mice, as their serum aspartate amino transferase and alanine amino transferase levels were within the normal range after repeated vector administrations, and there was no pathology in the liver and other major organs, as revealed by histological examinations. It is also of interest that 10 weekly administrations of a phiBT1 integrase–expressing plasmid in immune-competent mice did not lead to a T-cell-mediated rejection of the genetically reconstituted hepatocytes, making it an attractive system for repeated transgene delivery in the treatment of metabolic disorders.

Materials and Methods Plasmid construction. The construction of pCZiS-B, pCmPAH-B, and

pCMV-BTIntNLS has been described previously.15 The reporter gene integration vectors, pCG-B and pCR-B, were constructed on the basis of pCZiS-B. The lacZ-ires-SEAP fragment was released by XbaI digestion. The GFP gene was digested from plasmid pCX-EGFP45 with EcoRI, and the red fluorescent protein gene was derived from plasmid pDsRed1-1 (Clontech, Mountain View, CA) with EcoRI and XbaI digestion. The two reporter genes were then separately inserted into the vector backbone by ligation with XbaI linker. Animal treatment. All animal experiments were carried out in accordance

with our institutional guidelines. The CD1 nude mice and Pahenu2/J mice were purchased from Jackson Laboratories, and a colony of the latter

1794

was maintained in the Mount Sinai animal facility on a 6% fat diet. The mice used for in vivo gene transfer were 8–10 weeks of age. Plasmid DNA diluted in 3 ml phosphate-buffered saline was injected into the tail vein of each mouse within 6–8 seconds under hydrodynamic pressure. Blood samples were collected as needed and mice were killed at the end of experiments. Liver extracts and frozen tissue sections were prepared for standard pathology analysis as described previously.15 Measurements of serum aspartate amino transferase, alanine amino transferase, SEAP, and phenylalanine. Aspartate amino transferase and

alanine amino transferase concentrations in serum samples were measured in the Center for Comparative Medicine and Surgery Clinical Pathology Laboratory of our institute. Serum SEAP was determined by enzymelinked immunosorbent assay using a method previously reported.15 Serum phenylalanine concentrations were measured by high-pressure liquid chromatography as previously described.15,46 The results were analyzed statistically using Student’s t-test. Characterization of integrated DNA in the hepatocytes of treated mice.

Genomic DNA from liver samples was isolated using the DNeasy Kit (Qiagen, Valencia, CA) for determination of integration frequencies by quantitative polymerase chain reaction (LightCycler; Roche, Indianapolis, IN) as reported previously.15 Primer pairs that amplify the specific junction areas of various pseudo-attP sites were used to measure the corresponding integration events. The integration data in all samples were normalized to percentages per haploid genome by comparison with the unique GAPDH gene in the mouse genome. Standard curves were generated using plasmids containing the corresponding sequences with known concentrations. Measurement of PAH activities in liver extracts of treated mice. PAH

activities in the liver extracts of treated mice were determined using the radioactive thin-layer chromatography method.15 The data were shown as a percentage of activity in the liver extracts of homozygous normal mice, and the results were statistically analyzed using Student’s t-test.

Acknowledgments This work was partially supported by National Institutes of Health (NIH) grant DK-62972. We wish to thank Margaret C Smith (University of Aberdeen) for the kind gift of plasmid pRT801, John Fallon (Mount Sinai School of Medicine) for reviewing the histology slides, and Boxun Xie (Mount Sinai School of Medicine) for technical assistance in the histological studies.

References

1. Hollis, RP, Stoll, SM, Sclimenti, CR, Lin, J, Chen-Tsai, Y and Calos, MP (2003). Phage integrases for the construction and manipulation of transgenic mammals. Reprod Biol Endocrinol 1: 79. 2. Held, PK, Olivares, EC, Aguilar, CP, Finegold, M, Calos, MP and Grompe, M (2005). In vivo correction of murine hereditary tyrosinemia type I by phiC31 integrase-mediated gene delivery. Mol Ther 11: 399–408. 3. Quenneville, SP, Chapdelaine, P, Rousseau, J, Beaulieu, J, Caron, NJ, Skuk, D et al. (2004). Nucleofection of muscle-derived stem cells and myoblasts with phiC31 integrase: stable expression of a full-length-dystrophin fusion gene by human myoblasts. Mol Ther 10: 679–687. 4. Chalberg, TW, Genise, HL, Vollrath, D and Calos, MP (2005). phiC31 Integrase confers genomic integration and long-term transgene expression in rat retina. Invest Ophthalmol Vis Sci 46: 2140–2146. 5. Ortiz-Urda, S, Thyagarajan, B, Keene, DR, Lin, Q, Calos, MP and Khavari, PA (2003). PhiC31 integrase-mediated nonviral genetic correction of junctional epidermolysis bullosa. Hum Gene Ther 14: 923–928. 6. Ginsburg, DS and Calos, MP (2005). Site-specific integration with phiC31 integrase for prolonged expression of therapeutic genes. Adv Genet 54: 179–187. 7. Olivares, EC, Hollis, RP, Chalberg, TW, Meuse, L, Kay, MA and Calos, MP (2002). Site-specific genomic integration produces therapeutic factor IX levels in mice. Nat Biotechnol 20: 1124–1128. 8. Portlock, JL, Keravala, A, Bertoni, C, Lee, S, Rando, TA and Calos, MP (2006). Long-term increase in mVEGF164 in mouse hindlimb muscle mediated by phage phiC31 integrase after nonviral DNA delivery. Hum Gene Ther 17: 871–876. 9. Belteki, G, Gertsenstein, M, Ow, DW and Nagy, A (2003). Site-specific cassette exchange and germline transmission with mouse ES cells expressing phiC31 integrase. Nat Biotechnol 21: 321–324.

www.moleculartherapy.org vol. 15 no. 10 oct. 2007

© The American Society of Gene Therapy

10. Allen, BG and Weeks, DL (2005). Transgenic Xenopus laevis embryos can be generated using phiC31 integrase. Nat Methods 2: 975–979. 11. Thorpe, HM and Smith, MC (1998). In vitro site-specific integration of bacteriophage DNA catalyzed by a recombinase of the resolvase/invertase family. Proc Natl Acad Sci USA 95: 5505–5510. 12. Smith, MC and Thorpe, HM (2002). Diversity in the serine recombinases. Mol Microbiol 44: 299–307. 13. Chalberg, TW, Portlock, JL, Olivares, EC, Thyagarajan, B, Kirby, PJ, Hillman, RT et al. (2006). Integration specificity of phage phiC31 integrase in the human genome. J Mol Biol 357: 28–48. 14. Thyagarajan, B, Olivares, EC, Hollis, RP, Ginsburg, DS and Calos, MP (2001). Site-specific genomic integration in mammalian cells mediated by phage phiC31 integrase. Mol Cell Biol 21: 3926–3934. 15. Chen, L and Woo, SL (2005). Complete and persistent phenotypic correction of phenylketonuria in mice by site-specific genome integration of murine phenylalanine hydroxylase cDNA. Proc Natl Acad Sci USA 102: 15581–15586. 16. Groth, AC, Olivares, EC, Thyagarajan, B and Calos, MP (2000). A phage integrase directs efficient site-specific integration in human cells. Proc Natl Acad Sci USA 97: 5995–6000. 17. Thomason, LC, Calendar, R and Ow, DW (2001). Gene insertion and replacement in Schizosaccharomyces pombe mediated by the Streptomyces bacteriophage phiC31 site-specific recombination system. Mol Genet Genomics 265: 1031–1038. 18. Chen, ZY, He, CY, Ehrhardt, A and Kay, MA (2003). Minicircle DNA vectors devoid of bacterial DNA result in persistent and high-level transgene expression in vivo. Mol Ther 8: 495–500. 19. Groth, AC, Fish, M, Nusse, R and Calos, MP (2004). Construction of transgenic Drosophila by using the site-specific integrase from phage phiC31. Genetics 166: 1775–1782. 20. Lutz, KA, Corneille, S, Azhagiri, AK, Svab, Z and Maliga, P (2004). A novel approach to plastid transformation utilizes the phiC31 phage integrase. Plant J 37: 906–913. 21. Smith, MC, Till, R, Brady, K, Soultanas, P and Thorpe, H (2004). Synapsis and DNA cleavage in phiC31 integrase-mediated site-specific recombination. Nucleic Acids Res 32: 2607–2617. 22. Dafhnis-Calas, F, Xu, Z, Haines, S, Malla, SK, Smith, MC and Brown, WR (2005). Iterative in vivo assembly of large and complex transgenes by combining the activities of phiC31 integrase and Cre recombinase. Nucleic Acids Res 33: e189. 23. Malla, S, Dafhnis-Calas, F, Brookfield, JF, Smith, MC and Brown, WR (2005). Rearranging the centromere of the human Y chromosome with phiC31 integrase. Nucleic Acids Res 33: 6101–6113. 24. Ishikawa, Y, Tanaka, N, Murakami, K, Uchiyama, T, Kumaki, S, Tsuchiya, S et al. (2006) Phage phiC31 integrase-mediated genomic integration of the common cytokine receptor gamma chain in human T-cell lines. J Gene Med 8: 646–653. 25. Calos, MP (2006). The phiC31 integrase system for gene therapy. Curr Gene Ther 6: 633–645. 26. Stoll, SM, Ginsburg, DS and Calos, MP (2002). Phage TP901-1 site-specific integrase functions in human cells. J Bacteriol 184: 3657–3663. 27. Keravala, A, Groth, AC, Jarrahian, S, Thyagarajan, B, Hoyt, JJ, Kirby, PJ et al. (2006). A diversity of serine phage integrases mediate site-specific recombination in mammalian cells. Mol Genet Genomics 276: 135–146. 28. Yang, HY, Kim, YW and Chang, HI (2002). Construction of an integration-proficient vector based on the site-specific recombination mechanism of enterococcal temperate phage phiFC1. J Bacteriol 184: 1859–1864.

Molecular Therapy vol. 15 no. 10 oct. 2007

phiBT1-PAH Repeated Injections Cure Female PKU Mice

29. Olivares, EC, Hollis, RP and Calos, MP (2001) Phage R4 integrase mediates site-specific integration in human cells. Gene 278: 167–176. 30. Gregory, MA, Till, R and Smith, MC (2003). Integration site for Streptomyces phage phiBT1 and development of site-specific integrating vectors. J Bacteriol 185: 5320–5323. 31. Woo, SL, DiLella, AG, Marvit, J and Ledley, FD (1987). Molecular basis of phenylketonuria and recombinant DNA strategies for its therapy. Enzyme 38: 207–213. 32. Scriver, CR (1995). Whatever happened to PKU? Clin Biochem 28: 137–144. 33. Shedlovsky, A, McDonald, JD, Symula, D and Dove, WF (1993). Mouse models of human phenylketonuria. Genetics 134: 1205–1210. 34. Sarkissian, CN, Boulais, DM, McDonald, JD and Scriver, CR (2000). A heteroallelic mutant mouse model: a new orthologue for human hyperphenylalaninemia. Mol Genet Metab 69: 188–194. 35. Peng, H, Armentano, D, MacKenzie-Graham, L, Shen, RF, Darlington, G, Ledley, FD et al. (1988). Retroviral-mediated gene transfer and expression of human phenylalanine hydroxylase in primary mouse hepatocytes. Proc Natl Acad Sci USA 85: 8146–8150. 36. Fang, B, Eisensmith, RC, Li, XH, Finegold, MJ, Shedlovsky, A, Dove, W et al. (1994) Gene therapy for phenylketonuria: phenotypic correction in a genetically deficient mouse model by adenovirus-mediated hepatic gene transfer. Gene Ther 1: 247–254. 37. Oh, HJ, Park, ES, Kang, S, Jo, I and Jung, SC (2004). Long-term enzymatic and phenotypic correction in the phenylketonuria mouse model by adeno-associated virus vector-mediated gene transfer. Pediatr Res 56: 278–284. 38. Mochizuki, S, Mizukami, H, Ogura, T, Kure, S, Ichinohe, A, Kojima, K et al. (2004). Long-term correction of hyperphenylalaninemia by AAV-mediated gene transfer leads to behavioral recovery in phenylketonuria mice. Gene Ther 11: 1081–1086. 39. Ding, Z, Georgiev, P and Thony, B (2006). Administration-route and gender-independent long-term therapeutic correction of phenylketonuria (PKU) in a mouse model by recombinant adeno-associated virus 8 pseudotyped vector-mediated gene transfer. Gene Ther 13: 587–593. 40. Harding, CO, Gillingham, MB, Hamman, K, Clark, H, Goebel-Daghighi, E, Bird, A et al. (2006). Complete correction of hyperphenylalaninemia following liver-directed, recombinant AAV2/8 vector-mediated gene therapy in murine phenylketonuria. Gene Ther 13: 457–462. 41. Chen, L, Thung, SN and Woo, SL (2007). Metabolic basis of sexual dimorphism in PKU mice after genome-targeted PAH gene therapy. Mol Ther 15: 1079–1085. 42. Liu, F, Song, Y and Liu, D (1999). Hydrodynamics-based transfection in animals by systemic administration of plasmid DNA. Gene Ther 6: 1258–1266. 43. Blau, N and Erlandsen, H (2004). The metabolic and molecular bases of tetrahydroblopterin-responsive phenylalanine hydroxylase deficiency. Mol Genet Metab 82: 101–111. 44. Liu, J, Jeppesen, I, Nielsen, K and Jensen, TG (2006). Phic31 integrase induces chromosomal aberrations in primary human fibroblasts. Gene Ther 13: 1188–1190. 45. Okabe, M, Ikawa, M, Kominami, K, Nakanishi, T and Nishimune, Y (1997). ‘Green mice’ as a source of ubiquitous green cells. FEBS Lett 407: 313–319. 46. Dale, Y, Mackey, V, Mushi, R, Nyanda, A, Maleque, M and Ike, J (2003). Simultaneous measurement of phenylalanine and tyrosine in phenylketonuric plasma and dried blood by high-performance liquid chromatography. J Chromatogr B Analyt Technol Biomed Life Sci 788: 1–8.

1795