In Vitro Correction of Iduronate-2-Sulfatase Deficiency By

C Di Francesco1, C Cracco1, R Tomanin1, L Picci1, L Ventura1, F Zacchello1, P Di Natale2, DS Anson3, JJ Hopwood3, FL Graham4 and M Scarpa1 1Department of Pediatrics and Center for Biotechnology CRIBI, University of Padova, Italy; 2Department of Biochemistry and Medical Biotechnology, University of Napoli ‘Federico II’, Italy; 3Lysosomal Disease Research Unit, Department of Biochemistry, Adelaide Children’s Hospital, Adelaide, Australia; and 4Departments of Pathology and Biology, McMaster University, Hamilton, Ontario, Canada

Hunter syndrome is a lethal lysosomal storage disorder response. We describe the generation of a replication- caused by the deﬁciency of iduronate-2-sulfatase and defective adenoviral vector, AdRSVIDS, which is able to characterized by severe skeletal and neurological symp- express in vitro high levels of iduronate-2-sulfatase. After toms. Only symptomatic treatments are available and, infection, accumulation of mucopolysaccharides in treated although bone marrow transplantation has been sug- Hunter cells was normalized. Furthermore, endocytosis of gested, no encouraging results have been obtained so far. the transduced IDS did occur via the mannose-6-phos- Therefore, gene therapy might be a route to be pursued phate (M6P) receptor. Since no animal model for the dis- for treatment of the disease. In this respect, one major goal ease is available, we developed a system based on the to achieve is the generation of an overexpressing vector generation of derma-equivalents which enabled us to verify able to correct, in particular, central nervous system (CNS) the expression of high levels of sulfatase up to 30 days cells. Adenoviruses have been shown to infect CNS cells after infection. efﬁciently with minor or even absent immunological

Keywords: gene therapy; mucopolysaccharidosis type II; adenovirus; derma-equivalents

Introduction age. Facial features, hepatosplenomegaly, short stature, skeletal deformities, joint stiffness, severe retinal Mucopolysaccharidoses (MPS) are a group of 10 dis- degeneration and hearing impairment are coupled with orders caused by the deficiency of lysosomal enzymes an incremental deterioration of the neurological system. (four glycosidases, five sulfatases and one non-hydrolytic Death generally occurs between the ages of 10 and 14 transferase) needed for the catabolism of glycosaminogly- years. cans (GAG): dermatan-, heparan-, keratan- and chondro- No mental impairment is characteristic of the mild itin-sulfate (mucopolysaccharides). Their accumulation in form, however, skeletal deformities can be present to the lysosomes results in cell, tissue or organ dysfunction same degree as in the severe form. Retinal and hearing determining various chronic and progressive patterns of problems are milder than in the severe form and the 1 clinical severity, even within each enzyme deficiency. patient can survive until the fifth or sixth decade. Usually All MPS, except Hunter syndrome, are inherited as cardiac failure or airway obstruction are the cause of autosomal recessive disorders. Hunter syndrome, MPS death. Only symptomatic treatments are available for type II, is a rare X-linked inborn error of metabolism MPS in general. Transient improvement of patient con- characterized by the deficiency of iduronate-2-sulfatase ditions have been obtained with leukocyte and plasma (IDS) (E.C. 3.1.6.13), which removes the sulfate group in infusions,6,7 while fibroblast and amnion transplantations dermatan- and heparan-sulfate. The defect is due to point have not been successful in restoring, even partially, any mutations or deletions in the 24 kb gene, mapping on enzyme activity or improving the objective signs.8–10 2 3,4 Xq28.2. The cDNA was cloned as a 2.3 kb sequence, Bone marrow transplantation (BMT) has been sug- and successfully used to produce a recombinant active gested as a potential method for enzyme supplement for 5 enzyme in CHO cells. MPS patients,11,12 since enzyme replacement therapy is Hunter syndrome occurs in a severe and a mild form. not available yet. In the case of MPSII, however, the value The severe form is characterized by progressive somatic of BMT still has to be investigated. and neurological involvements. The onset of the disease Correction by gene therapy might represent another usually occurs between the second and fourth year of route to be pursued. High levels of recombinant human lysosomal enzymes were obtained in vitro13,14 and in vivo15 by retrovirus-mediated gene transfer in different Correspondence: M Scarpa, Department of Pediatrics and CRIBI, Via MPS as well as a successful metabolic correction of Trieste 75, 35121 Padova, Italy Hunter lymphoblastoid cell lines.16 A phase I clinical trial Received 5 November 1996; accepted 8 January 1997 aimed at increasing the enzyme level in Hunter patients Gene transfer in mucopolysaccharidosis type II C Di Francesco et al 443 affected by the mild form of the disease was also approved.17 In order to improve considerably the life condition of patients affected by the severe form of Hunter syndrome, local production of the enzyme lacking in the brain might be required. In fact, the capability of IDS enzyme to cross the blood–brain barrier still needs to be ascertained. Since adenoviral vectors have been shown to be adequate and safe delivery systems to transfer normal sequences also to nonproliferative cells,18–21 we generated a replication-defective adenovirus vector, derived from human adenovirus type 5, expressing the human IDS (AdRSVIDS). Infection experiments performed on primary Hunter cells showed that AdRSVIDS was able to normalize the intralysosomal GAG accumulation; furthermore, the recombinant enzyme secreted in the extracellular compartment was endocytosed by deficient cells via the mannose-6-phosphate receptor (M6P).22 Because of the lack of animal models to perform long- term expression experiments, we included infected pri- Figure 1 The vector pXCRSVIDS pA used to cotransfect 293 cells with the plasmid pJM17 to generate AdRSVIDS. AmpR: ampicillin resistance; mary fibroblasts from Hunter patients into collagen matr- Ad5: 5′ sequence 1–452 bp; RSV: Rous sarcoma virus LTR; preproins. 23,24 ices (derma-equivalents). This technique enabled us to leader seq.: 5′ rat preproinsulin leader sequence: 48 bp; IDS: IDS cDNA; show expression of the virus-transduced IDS up to 30 SV40 pA: SV40 polyA containing the small intron; Ad5: Ad5 3′ sequence days after infection, confirming that AdRSVIDS might be 3328–5788 bp. a valuable vector for gene therapy of Hunter syndrome.

Results was obtained on non-reverse transcribed total RNA (Figure 2). Construction of pXCRSVIDSpA The 578 bp Rous sarcoma virus long terminal repeat (RSV-LTR) sequence was isolated from pRSVLuc25 by IDS enzyme activity in normal and transduced NdeI–XbaI restriction, treated with Klenow polymerase fibroblasts and cloned as blunt end fragment into the filled in XbaI To show that AdRSVIDS was able to express a functional site of pXCJL1.26 The new plasmid was called pXCRSV. recombinant enzyme, IDS activity was tested on nonin- The SV40 polyA sequence containing the small intron fected and infected Hunter cells compared with normal was isolated as an 876 bp HindIII–BamHI fragment from the pSV23p construct,27 filled in and cloned into pXCRSV after treatment with Klenow of the unique ClaI site. This plasmid was called pXCRSVpA. The 1814 bp IDS cDNA was isolated from the construct pLX-IDS (DS Anson and JJ Hopwood, unpublished) by ClaI–SalI restriction and filled in. The cDNA was cloned into the filled in SalI site of pXCRSVpA. The final plasmid, named pXCRSVIDSpA (Figure 1), was used to generate the viral vector AdRSVIDS.

Molecular analysis of infected Hunter cells The efficiency of adenovirus infection on Hunter cells was first assessed by using AdHCMVsp1lacZ vector at 100 p.f.u. per cell. Nearly 100% of cells were found to express ␤-galactosidase 24 h after infection (data not shown). To assess the transduced IDS activity in the short and prolonged period, primary Hunter fibroblasts were subsequently infected with the vector AdRSVIDS and analysed 48 h and 30 days after infection, respectively. Figure 2 RT-PCR on two different Hunter fibroblasts infected with The amplification of IDS cDNA was performed with IDS- AdRSVIDS. Twenty micrograms of total RNA were retrotranscribed with specific oligonucleotides 5 and 340. A 334 bp band was oligo dT and amplified with IDS-specific oligonucleotides 423a and 424a. − detected only in infected cells (data not shown). A 585 bp product was detected only from Hunter infected cells. Lane C : The amplification of reverse transcribed total RNA amplification of noninfected Hunter cells. Lanes 1 and 3: amplification of total RNA nontreated with reverse transcriptase. Lanes 2 and 4: PCR extracted from Hunter transduced cells, performed with product from retrotranscribed RNA of infected Hunter cells. C+: positive IDS-specific oligonucleotides 423a and 424a showed a control, pXCRSVIDS pA. M: molecular weight marker VI (Boehringer 585 bp band only in infected cells. No amplification Mannheim). B: blank. Gene transfer in mucopolysaccharidosis type II C Di Francesco et al 444 AdRSVIDS. Figure 3 shows that, as expected, a three-fold Table 1 IDS activity detection in infected Hunter cells increased level (286 ± 56 c.p.m.) (bar 1) of accumulation in noninfected Hunter cells with respect to normal ones IDS expression IDS expression ± (U/mg) in cultured (U/mg) in (103 8 c.p.m.) (bar 3) was measured. GAG levels in fibroblastsa derma-equivalentsb Hunter cells were back to normal (95 ± 13 c.p.m.) after infection (bar 2). Hunter fibroblasts Ͻ5 Ͻ5 Control fibroblasts 81 (±10) 111 (±10) Secretion and endocytosis of recombinant IDS in Hunter Hunter cells + cells ± ± AdRSVIDS 1758 ( 120) 2580 ( 85) To evaluate IDS secretion from Hunter-infected cells, IDS activity was determined in IDS-conditioned medium aIDS activity was determined 48 h after infection. ± b detecting 33 8 U/ml. Such levels were 10-fold higher Fibroblasts were included in collagen matrices and IDS activity than those measured in the medium collected from nor- determined 30 days after infection. Ͻ Values represent a mean ± s.d. of three independent experi- mal fibroblasts ( 5 U/ml). Table 3 shows that IDS ments on two different normal primary fibroblasts and four dif- activity was readily achieved in Hunter cells cultured ferent Hunter primary cells. with IDS-conditioned medium, while no enzyme activity was detected when Hunter cells were maintained in medium obtained from normal fibroblasts. Furthermore,

ones (Table 1). No basal IDS activity (Ͻ5 U/mg) was observed in Hunter cells (H130, H423, H435 and H452), while normal cells showed standard IDS activity (81 ± 10 U/mg). In comparison, up to 20-fold higher expression was detected in infected Hunter cells (1758 ± 120 U/mg). In order to assay prolonged IDS expression, ﬁbroblasts were included into collagen matrices and maintained for 30 days. We observed that 30 days after infection Hunter cells were still able to express up to 20-fold higher levels of enzyme (2580 ± 85 U/mg) with respect to normal ﬁbroblasts (110 ± 10 U/mg) (Table 1). Owing to progressive decrease of cell viability, analysis beyond 30 days could not be performed (data not shown).

Expression of lysosomal enzymes in infected Hunter fibroblasts To evaluate whether overexpression of transduced IDS might alter the level of expression of other lysosomal enzymes, activity of ␣-N-acetyl-glucosaminidase, ␤- galactosidase, ␣-l-iduronidase, N-acetylglucosamino-6- sulfatase, galacto-6-sulfatase, acetyl-CoA:␣-glucosaminide-N-acetyltransferase, were measured in Hunter fibroblasts before and after infection. Table 2 shows that the overexpression of IDS induced by AdRSVIDS does not seem to interfere with the expression and activity of endogenous lysomal enzymes. 35 Figure 3 Mucopolysaccharides accumulation analysis by SO4-GAG 35 labeling assay in Hunter cells before and after infection. Before infection SO4-GAG accumulation in Hunter fibroblasts before and after infection with AdRSVIDS Hunter cells (bar 1) are accumulating three-fold more GAG than normal cells (bar 3). After infection the level of GAG is back to normal (bar 2). To evaluate the capability of the recombinant enzyme to Each lane represents a mean ± s.d. of three independent experiments each 35 correct the metabolic defect, SO4-GAG levels were one performed on two different normal primary fibroblasts and four differ- evaluated in Hunter cells before and after infection with ent Hunter primary cells. See Materials and methods for details.

Table 2 Lysosomal enzymes assayed in infected Hunter cells

Cells ␣-N-acetyl- ␤-Galactosidase ␣-l-iduronidase N-acetyl-glucose Galacto-6-sulfatase Acetyl-CoA: glucosaminidase (nmol/mg/h) (nmol/mg/h) amino-6-sulfatase (nmol/mg/17 h) ␣-glucosaminide- (nmol/mg/h) (nmol/mg/17 h) N-acetyl-transferase

Hunter + AdRSVIDS 7.3 ± 1.5 784 ± 85 137 ± 20.7 34.1 ± 5.3 14.4 ± 5.0 49.9 ± 5.2 Hunter 9.3 ± 2.0 602 ± 100 128 ± 25.5 31.0 ± 9.2 15.4 ± 3.0 45.6 ± 4.6

Values represent a mean ± s.d. of four different Hunter primary cells. Gene transfer in mucopolysaccharidosis type II C Di Francesco et al 445 expression from infected cells seemed to be crucial to Table 3 Secretion and endocytosis of recombinant IDS expressed by AdRSVIDS allow cross-correction of noninfected Hunter cells. In fact, the use of IDS-conditioned medium obtained from nor- IDS activity mal fibroblasts was not successful in transferring IDS into Hunter cells. The reason for this might be that IDS, a Conditioned medium for normal housekeeping enzyme, is produced by normal cells fibroblasts Ͻ5 U/ml mostly for intracellular metabolic need. This might be a Conditioned medium from Hunter 33 ± 8 U/ml further explanation to be taken into consideration, cells infected with AdRSVIDS together with antibody generation against implanted Hunter cells cultured with conditioned Ͻ5 U/mg cells, low number of implanted cells and shortness of the medium from normal fibroblasts enzyme half-life, to explain the failure of early clinical Hunter cells cultured with conditioned trials that were unable to ameliorate Hunter patient con- medium from Hunter infected ditions significantly by supplying fibroblasts, amnion or fibroblasts plasma.6–10 However, besides overexpression, the set-up Without mannose-6-phosphate 235 ± 12 U/mg With mannose-6-phosphate Ͻ5 U/mg of a gene therapy protocol for Hunter syndrome also requires a vector able to produce high levels of enzyme Values represent a mean ± s.d. of three independent experi- over a prolonged period. Because of the nonintegration ments on two different normal primary fibroblasts and four dif- of adenoviruses into infected cell genome, prolonged ferent Hunter primary cells. expression analysis is not feasible if replicating cells are used. Experiments performed infecting Hunter fibroblasts with AdHCMVsp1lacZ showed that infected cells IDS endocytosis is blocked by competition with M6P; no were highly positive to X-gal staining for a few days after increased IDS activity was, in fact, detected in Hunter infection (3 days), but totally negative within a few cell cells cultured with IDS-conditioned medium in the passages (10 days). Since Hunter primary fibroblasts presence of M6P. were the only cells available for this study, a reproducible system, based on derma-equivalents, able to allow prolonged expression analysis was developed. It was pre- Discussion viously shown that keratinocytes or fibroblasts mixed Somatic gene transfer is a promising strategy for the with collagen type I contract into a tissue.23 Within the treatment of many metabolic disorders. Essential to any tissue, cells can be maintained for prolonged periods of gene therapy protocol is the expression of the transferred time and can reach a high degree of differentiation exhib- gene which has to be targeted to the major cellular sites iting a bipolar morphology.24 Cell division is blocked at of pathology in sufficient amounts to correct the meta- phases G1 and G2 and the synthesis of macromolecules bolic defect. (collagen, noncollagen proteins, glycosaminoglycans) can For the severe form of Hunter syndrome the target be identified.29 The generation of derma-equivalents is a organ is represented by CNS. Although retroviral vectors well established technique, which has also been exploited have been shown to be the safest and more reliable vec- for medical applications30,31 and for the construction of tors used in the ongoing clinical trials, they cannot be secreting neo-organs for gene therapy purposes.32,33 In used for gene transfer of differentiated neuronal cells. In this study, the use of derma-equivalents allowed us to fact, efficient retrovirus-mediated gene transfer relies detect IDS activity up to 30 days after infection. The strictly on cell replication. On the other hand, adenovirus expression was stable and comparable to the levels meas- vectors have been shown to be reliable tools to infect CNS ured 48 h after infection. To our knowledge, this is the cells in vitro and in vivo18,19,28 and were able to maintain first time that a prolonged term expression analysis has in vivo expression up to 6 months after infection.18,19 been performed in vitro by using adenovirus vectors. In this article, we describe the in vitro correction of Therefore, derma-equivalents can be proposed as a Hunter cells and the overexpression of the human IDS reliable, easy, cheap system for the analysis of adeno- cDNA mediated by the replication defective adenovirus virus-transduced genes in the absence of a suitable vector AdRSVIDS. The vector was able to express a animal model. recombinant IDS protein that could substantially The vector AdRSVIDS was able to correct the metabolic decrease GAG substrate accumulation. Furthermore, it alteration typical of MPSII and therefore it might be con- appeared that recombinant IDS, as the normal enzyme, sidered a valuable tool for gene therapy of patients affec- was endocytosed into cells by the M6P receptor, ted by Hunter syndrome. as shown by the lack of IDS activity in Hunter cells cul- Although our final goal will be the correction of the tured with IDS-conditioned medium in the presence enzyme defect in CNS of Hunter patients, in the absence of M6P. of an animal model, clinical studies aimed to correct the AdRSVIDS has been shown to express high levels of recombinant enzyme, up to 20-fold over normal levels, IDS deficiency in CNS should be preceded by experi- which was maintained for at least 30 days. The over- ments performed in other areas severely affected by the expression of the recombinant IDS might be due to the disease (ie joints). Such experiments should allow evalu- insertion of a 45 bp fragment of the rat preproinsulin ation of the vector safety and hopefully its immuno- leader sequence cloned as substitution of the 5′-non- genicity. In this respect, the next step will be preclinical coding region of IDS. This insertion was necessary since experiments aimed at evaluating the ability of the vector clones generated by stable transfection of unmodified IDS to overexpress IDS, its safety and its immunogenicity in cDNA were producing low levels of enzyme.5 IDS over- areas other than CNS in healthy animals. Gene transfer in mucopolysaccharidosis type II C Di Francesco et al 446 Materials and methods duced IDS was performed with oligonucleotides 340 and 5 by 35 cycles at 94°C for 1 min, 62°C for 45 s, 72°C for ′ DNA cloning 1 min with a 5 extended elongation time at the last cycle. Cloning was performed following standard procedures.34 Amplification of IDS transcript was performed with oligonucleotides 423a and 424a by 35 cycles at 94°C for 1 All enzymes for DNA manipulation were purchased ° ° ′ from Boehringer Mannheim (Mannheim, Germany). min, 59 C for 45 s, 72 C for 1 min with a 5 extended elongation time at the last cycle. Hunter cells infected with AdHCMVsp1lacZ were analyzed 24 h after infection Cell cultures for ␤-gal expression by using X-gal staining procedure.40 Hunter (H130, H423, H435 and H452) and normal primary fibroblasts were obtained from skin biopsies. Collagen preparation Hunter cells were obtained from patients affected by the Collagen type I was obtained from rat tails as described.23 severe form of MPSII; primary cells obtained from three Briefly, rat tails were stored in 70% EtOH until use, ten- healthy subjects were used as a control. Cells were cul- dons were isolated and exposed overnight to UV light. tured in minimum essential medium (MEM) sup- ml The day after, tendons were put in 1% glacial acetic acid plemented with 10% fetal calf serum (FCS), 2 m -gluta- at 4°C for 48 h. Collagen was centrifuged at 15 000 g for mine, Hepes 5 g/l, nonessential amino acids, penicillin 90 min at 4°C and dialyzed against 0.1% acetic acid sol- (50 U/ml) streptomycin sulfate (50 mg/ml) and Fungi- ution for 6 days. Collagen concentration was determined zone (ICN, Costa Mesa, CA, USA) (125 mg/ml) by measuring hydroxyproline content as described.41 (complete medium). Cultures were incubated in a ° humidified atmosphere at 5% CO2 at 37 C. All reagents Reconstitution of derma-equivalents were obtained from GIBCO BRL (Gaithersburg, MD, Normal, Hunter and virus-infected cells were mixed with USA) except for FCS, obtained from Boehringer collagen type I to reconstitute derma-equivalents. As pre- Mannheim. viously described,23 1.2 × 106 cells were plated in complete medium to which collagen and 0.1 n sodium Generation and purification of AdRSVIDS hydroxide were added and incubated in a humidified 35 ° To generate AdRSVIDS, pXCRSVIDSpA and pJM17 atmosphere at 5% CO2 at 37 C. Derma-equivalents were were cotransfected in 293 cells,36 by the Ca-P coprecipi- digested by adding 2 mg/ml Collagenase Type 1A tation technique.37 Purification and titration were perfor- (Sigma, St Louis, MO, USA) from Clostridium histolyticum med by standard procedures.26,38 The virus genome was and incubated for 40 min at 30°C. Cells were counted, analyzed by enzymatic restriction. Correct plaques were pellets resuspended in 150 mm NaCl and stored at −80°C purified and analyzed once more. Medium from the until used. Cell viability was assayed by trypan blue second purification was collected and used to build up a dye exclusion. high titer stock in 293 cells. Determination of IDS enzyme activity Adenovirus infection of Hunter cells Normal, Hunter and AdRSVIDS-infected cells were Hunter primary fibroblasts were infected with 100 p.f.u. tested for IDS activity as previously described.42 Forty- per cell of either AdRSVIDS or AdHCMVSp1lacZ,39 the eight hours after infection, cells were trypsinized, centri- latter expressing the E. coli lacZ gene. fuged and lysed by six cycles of freezing–thawing. Lys- For short-term experiments using AdRSVIDS, cells ates were dialyzed in distilled water for 16 h at 4°C before were infected and analyzed 48 h later. For prolonged assaying. Protein concentration was determined accord- experiments, infected cells were included in derma-equi- ing to Bradford.43 IDS activity was assayed using the valents for 30 days, and analyzed after collagenase diges- radiolabeled disaccharide substrate L-O-(␣-iduronic acid tion as described below. 2-sulphate)-(1-Ͼ4)-D-O-2,5-anhydrol 3H mannitol 6-sul- The infection and IDS transcription was determined by phate. Briefly, 15 ␮l of substrate solution (radioactive di- PCR reaction using the following virus-specific oligonu- sulfated disaccharide, 0.22 mm, 2.6 × 106 c.p.m./min 0.27 ′ m m cleotides: No. 340 forward: 5 TAC GAT CGT GCC TTA sodium acetate buffer pH 4.0/13 m NaN3) were TTA GG 3′, priming within the RSV promoter; and IDS mixed with suitable enzyme aliquots and assayed in 80 cDNA oligonucleotides No. 5 antisense: 5′ ACG TTC ␮l final volume. Reaction was stopped after 24 h incu- ′ ° m AGA GCA TCT GTG GTC GAG TTG GCC 3 , priming bation at 37 C, with 1 ml 1 m Na2HPO4, samples were at position 210–239 of the IDS cDNA;3 No. 423a forward loaded on an ECTEOLA-23 column (FLUKA, Buchs, CAT CAG CAA GCA GGT CAT T, priming at position Switzerland) and the product was eluted by adding 5 ml 20–38 of the rat preproinsulin leader sequence;5 No. 424a 70 mm sodium formiate. Scintillation liquid Pico Fluor 40 reverse: CCA CAG GGC AAA GCA GGT T, priming at (INSTA-GEL Packard, Meriden CT, USA) was added and position 560–578 of the IDS cDNA.3 Taq polymerase was samples were counted by a scintillation counter. Enzy- purchased from Roche Molecular Systems (Branchburg, matic activity was expressed as U/mg of proteins. One NJ, USA). DNA oligonucleotides were synthesized with unit of IDS activity is the amount of enzyme required to a Beckman SM oligosynthesizer (Beckman Instruments, catalyse the hydrolysis of 1% 3H substrate per hour. 35 Palo Alto, CA, USA); H2 SO4 was purchased from Amer- sham (Buckinghamshire, UK). Determination of lysosomal enzyme activities Total DNA was isolated from cells as previously Hunter and AdRSVIDS-infected Hunter cells were described.34 IDS cDNA was obtained by reverse tran- assayed for the following lysosomal enzymes: ␣-N-acetyl- scription of 20 ␮g of total RNA obtained by using RNA- glucosaminidase, ␤-galactosidase, ␣-l-iduronidase, N- Zol B treatment (BIOTECX, Houston, TX, USA) according acetyl-glucosamino-6-sulfatase, galacto-6-sulfatase, ace- to the manufacturer’s protocol. Amplification of trans- tyl-CoA:␣-glucosaminide-N-acetyltransferase according Gene transfer in mucopolysaccharidosis type II C Di Francesco et al 447 to standard procedures.44,45 Activities were expressed as 2 Le Guern E, Couillin P, Oberle I, Boue J. More precise localiz- nmol/mg/h for ␣-N-acetyl-glucosaminidase, ␤-galacto- ation of the gene for Hunter syndrome. Genomics 1990; 7: 358– sidase, ␣-l-iduronidase, as nmol/mg/17 h for N-acetyl- 362. glucosamino-6-sulfatase, galacto-6-sulfatase, and as 3 Wilson PJ et al. Hunter syndrome: isolation of an iduronate-2- U/mg protein for acetyl-CoA:␣-glucosaminide-N-acetyl- sulfatase cDNA clone and analysis of patient DNA. Proc Natl Acad Sci USA 1990; 87: 8531–8535. transferase. 4 Wilson PJ, Meaney CA, Hopwood JJ, Morris CP. Sequence of the human iduronate-2-sulfatase (IDS) gene. Genomics 1993; 17: 35 SO4-GAG accumulation assay 773–775. Studies assessing GAG metabolism were performed as 5 Bielicki J, Hopwood JJ, Anson DS. Recombinant human iduron- previously described.46 Infected fibroblasts were grown ate-2-sulfatase: correction of mucopolysaccharidosis-type II for 48 h in MgSO4 defective medium (Basal Medium fibroblasts and characterization of the purified enzyme. Biochem Eagle Diploid Modified, (BMEDM); ICN Biomedicals, J 1993; 289: 241–246. Costamesa, CA, USA) supplemented with 2 mml-gluta- 6 Di Ferrante N et al. Induced degradation of glycosaminoglycans mine and 10% FCS, the latter dialyzed against sterile in Hurler’s and Hunter’s syndromes by plasma infusion. Proc water and PBS for 1 week and 2 days, respectively. Cellu- Natl Acad Sci USA 1971; 68: 303–307. lar GAG were metabolically labeled by addition of 7 Knudson AG, Di Ferrante N, Curtis JE. Effect of leukocyte trans- fusion in a child with type II mucopolysaccharidosis. Proc Natl H 35SO (Amersham) (4 ␮Ci/ml; 1 Ci = 37 GBq) to culture 2 4 Acad Sci USA 1971; 68: 1738–1741. medium for 48 h. Fibroblasts were then washed with PBS, 8 Dean MF et al. Increased breakdown of glycosaminoglycans and harvested by trypsinization, collected, centrifuged at appearance of corrective enzyme after skin transplants in 1500 g, washed with PBS and centrifuged once more. Hunter syndrome. Nature 1975; 257: 609–612. Two milliliters of 80% ethanol were added. Samples were 9 Dean MF et al. Enzyme replacement therapy by fibroblast trans- boiled for 5 min and centrifuged at 1500 g for 3 min. Etha- plantation in a case of Hunter syndrome. Nature 1976; 261: nol extraction was done twice. 0.5 ml 10% NaOH were 323–325. ° 35 added and samples were heated at 100 C. SO4 incorpor- 10 Adinolfi M et al. Expression of HLA antigens, B2-microglobulin ation was assessed by counting 0.25 ml from each sample and enzymes by human amniotic epithelial cells. Nature 1982; in PicoFluor 40 with the scintillation counter. 295: 325–327. 11 Krivit W, Shapiro EG. Bone marrow transplantation for storage Secretion and endocytosis of recombinant IDS in Hunter diseases. In: Desnick RJ (ed). Treatment of Genetic Diseases. Chur- cells chill-Livingstone: New York, 1991, pp 203–221. 12 Coppa GV et al. Bone marrow transplantation in Hunter syn- To assess whether transduced IDS was secreted, fibro- drome. J Inher Metab Dis 1995; 18: 91–92. blasts were infected with AdRSVIDS for 24 h, washed to 13 Peters C, Rommerskirch W, Modaressi S, Von Figura K. Resto- eliminate free virus and re-fed with complete medium ration of arylsulphatase B activity in human mucopolysacchar- for 24 h more. This medium, called IDS-conditioned idosis type VI fibroblasts by retroviral-mediated gene transfer. medium, was collected, filtered and used to feed Hunter Biochem J 1991; 276: 499–504. cells. IDS activity was measured as described above. 14 Anson DS, Bielicki J, Hopwood JJ. Correction of mucopolysacch- Medium collected from normal fibroblasts was used as aridosis type I fibroblasts by retroviral-mediated transfer of the control. To assess whether recombinant IDS was endo- human ␣-l-iduronidase gene. Hum Gene Ther 1992; 3: 371–379. cytosed via the M6P receptor,19 fibroblasts were plated 15 Moullier P, Bohl D, Heard J-M, Danos O. Correction of lysoso- and allowed to reach confluence, incubated for 24 h in mal storage in the liver and spleen of MPS VII mice by implan- 10 ml IDS-conditioned medium with or without 5 mm tation of genetically modified skin fibroblasts. Nat Genet 1993; M6P. Cell lysates were dialyzed against distilled water 4: 154–159. and then analyzed for total protein content and IDS 16 Braun SE et al. Metabolic correction and cross-correction of mucopolysaccharidosis type II (Hunter syndrome) by retroviral- activity as described above. mediated gene transfer and expression of human iduronate-2- sulfatase. Proc Natl Acad Sci USA 1993; 90: 11830–11834. Acknowledgements 17 Whitley CB. Retroviral-mediated transfer of the iduronate-2-sulfatase into lymphocytes for treatment of mild Hunter syndrome We wish to thank J Rudy for the excellent technical assist- (mucopolysaccharidosis type II). Hum Gene Ther 1996; 7: 537– ance. The work was supported in part by the Italian Muc- 549. opolysaccharidoses Association (IMA), the ‘Salus Pueri’ 18 Akli S et al. Transfer of a foreign gene into the brain using the Foundation and by Regione Veneto ‘Ricerca Finalizzata’, adenovirus vectors. Nat Genet 1993; 3: 224–228. Venice, Italy, 579/01/95. C Cracco is recipient of an IMA 19 Bajocchi G, Feldman FH, Crystal RG, Mastrangeli A. Direct in fellowship. P Di Natale is supported by the Telethon- vivo gene transfer to ependymal cells in the central nervous sys- Italia research grant No. 085. FL Graham’s research was tem using recombinant adenovirus vectors. Nat Genet 1993; 3: supported by a grant from the Medical Research Council 229–234. 20 Ragot T et al. Efficient adenovirus-mediated transfer of a human of Canada. FL Graham is a Terry Fox Research Scientist minidystrophin gene to skeletal muscle of mdx mice. Nature of the National Cancer Institute of Canada. DS Anson 1993; 361: 647–650. and JJ Hopwood are supported by a National Health and 21 Rosenfeld MA et al. In vivo transfer of the human cystic fibrosis Medical Research Council of Australia program grant. transmembrane conductance regulator gene to the airway epi- thelium. Cell 1992; 68: 143–155. References 22 von Figura K, Hasilik A. Lysosomal enzymes and their recep- tors. Annu Rev Biochem 1986; 55: 167–180. 1 Neufeld EF, Muenzer J. The mucopolysaccharidosis. In: Scriver 23 Bell E, Ivarsson B, Merrill C. Production of a tissue-like structure CR, Beaudet AL, Sly WS, Valle D (eds). Metabolic Basis of by contraction of collagen lattices by human fibroblasts of differ- Inherited Diseases, 7th edn. McGraw-Hill: New York, 1995, pp ent proliferative potential in vitro. Proc Natl Acad Sci USA 1979; 2465–2494. 76: 1274–1278. Gene transfer in mucopolysaccharidosis type II C Di Francesco et al 448 24 Bell E et al. The reconstitution of living skin. J Inv Derm 1983; rescue of early region I mutations into infectious human aden- 81: 2–10. ovirus type 5. Virology 1988; 163: 614–617. 25 De Wet JR et al. Firefly luciferase gene: structure and expression 36 Graham FL, Smiley J, Russel WC, Nairn R. Characteristics of a in mammalian cells. Mol Cell Biol 1987; 7: 725–737. human cell line transformed by DNA from human adenovirus 26 Hitt M, Bett AJ, Prevec L, Graham FL. Construction and propa- type 5. J Gen Virol 1977; 36: 59–72. gation of human adenovirus vectors. In: Celis JE (ed). Cell 37 Graham FL, Van Der Eb AJ. A new technique for the assay of Biology: A Laboratory Handbook. Academic Press: Orlando, FL, infectivity of human adenovirus type 5 DNA. Virology 1973; 52: 1994, pp 479–490. 456–467. 27 Huylbroeck D et al. High-level transient expression of influenza 38 Hitt M et al. Techniques for human adenovirus vector construc- virus protein from a series of SV40 late and early replacement tion and characterization. In: Adolph KW (ed). Methods in Mol- vectors. Gene 1988; 66: 163–181. ecular Genetics. Vol. 7B. Academic Press: Orlando, FL, 1995, pp 28 Boviatsis EJ et al. Gene transfer into experimental brain tumors 13–30. mediated by adenovirus, herpes simplex virus and retrovirus 39 Morsy MA et al. Efficient adenoviral vector mediated ornithine vectors. Hum Gene Ther 1994; 5: 183–191. transcarbamylase expression in deficient mouse and human 29 Frey J, Rattner A, Champson A. Synthesis and supermolecular hepatocytes. J Clin Invest 1993; 92: 1580–1586. 40 MacGregor GR, Caskey CT. Construction of plasmids that organization of extracellular matrix components by fibroblasts express E. coli ␤-galactosidase in mammalian cells. Nucleic Acids cultured in a collagen lattice. Path Biol 1992; 40: 859–864. Res 1989; 17: 2365. 30 Maruguchi T et al. A new skin equivalent-keratinocytes prolifer- 41 Woessner JF Jr. The determination of hydroxyproline in tissue ated and differentiated on collagen sponge containing fibro- and protein samples containing small proportions of this imino blasts. Plastic and Reconstr Surg 1994; 93: 537–544. acid. Arch Biochem Biophys 1961; 93: 440–447. 31 Beumer GJ, Vanblitterwijck CA, Ponec M. Biocompatibility of 42 Di Natale P, Ronsisvalle L. Identification and partial characteriz- a biodegradable matrix used as a skin substitute – an in vivo ation of two enzyme forms of iduronate sulfatase from human evaluation. J Biomed Mat Res 1994; 28: 545–552. placenta. Biochem Biophys Acta 1981; 661: 106–111. 32 Enomaa N, Danos O, Peltonen L, Jalanko A. Correction of 43 Bradford MM. A rapid and sensitive method for the quantit- deficient enzyme-activity in a lysosomal storage disease, aspar- ation of microgram quantities of protein utilizing the principle tylglucosaminuria, by enzyme replacement and retroviral gene- of protein-dye binding. Anal Biochem 1976; 72: 248–254. transfer. Hum Gene Ther 1995; 6: 723–731. 44 Di Natale P, Annella T, Daniele A, De Luca T. Biochemical diag- ␣ l 33 Salvetti A et al. In vivo delivery of human - -iduronidase in nosis of mucopolysaccharidoses: experience of 297 diagnoses in mice implanted with neo-organs. Hum Gene Ther 1995; 6: a 15-year period (1977–1991). J Inher Metab Dis 1993; 16: 473–483. 1153–1160. 45 Cerruti-Mainardi P et al. Mannosidosis. Study of two families 34 Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A Labora- and prenatal diagnosis Pediatr Med Chir 1982; 4: 203–214. tory Manual, 2nd edn. Cold Spring Harbor Laboratory Press: 46 Fratantoni JC, Hall CW, Neufeld EF. Hurler and Hunter syn- Cold Spring Harbor, NY, 1989. dromes: mutual correction of the defect in cultured fibroblasts. 35 McGrory J, Bautista D, Graham FL. A simple technique for the Science 1969; 162: 570–572.