Molecular Genetics and Metabolism

Molecular Genetics and Metabolism 95 (2008) 142 151 Contents lists available at ScienceDirect Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme Radiographic evaluation of bones and joints in mucopolysaccharidosis I and VII dogs after neonatal gene therapy Ramin Sedaghat Herati a, Van W. Knox b, Patricia O Donnell c, Marina D Angelo d, Mark E. Haskins c, Katherine P. Ponder a,e, * a Depart ment of Inter nal Med i cine, Wash ing ton Uni ver sity School of Med i cine, St. Louis, MO 63110, USA b Depart ment of Clin i cal Stud ies, Uni ver sity of Penn syl va nia, School of Vet er i nary Med i cine, Phil a del phia, PA, USA c Depart ment of Patho bi ol ogy, Uni ver sity of Penn syl va nia, School of Vet er i nary Med i cine, Phil a del phia, PA, USA d Cen ter for Chronic Dis or ders of Aging, Phil a del phia Col lege of Oste o pathic Med i cine, Phil a del phia, PA, USA e Depart ment of Bio chem is try and Molec u lar Bio phys ics, Wash ing ton Uni ver sity School of Med i cine, St. Louis, MO 63110, USA article info Article history: Received 13 May 2008 Received in revised form 1 July 2008 Accepted 1 July 2008 Available online 15 August 2008 Key words: Gene ther apy Lyso somal stor age dis ease Muco pol y sac cha ri do sis a-l-idu ron i dase b-glu cu ron i dase Gly cos ami no gly can Dys os to sis mul ti plex abstract Muco pol y sac cha ri do sis I (MPS I) and MPS VII are due to defi cient activ ity of the gly cos ami no gly candegrad ing lyso somal enzymes a-l-idu ron i dase and b-glu cu ron i dase, respec tively, and result in abnor mal bones and joints. Here, the sever ity of skel e tal dis ease in MPS I and MPS VII dogs and the effects of neona tal gene ther apy were eval u ated. For untreated MPS VII dogs, the lengths of the sec ond cer vi cal ver tebrae (C2) and the femur were only 56% and 84% of nor mal, respec tively, and bone dys pla sia and artic u lar ero sions, and joint sub lux a tion were severe. Pre vi ously, we reported that neo na tal intra ve nous injec tion of a ret ro vi ral vec tor (RV) with the appro pri ate gene resulted in expres sion in liver and blood cells, and high serum enzyme activ ity. In this study, we dem on strate that C2 and femurs of RV-treated MPS VII dogs were longer at 82% and 101% of nor mal, respec tively, and there were partial improve ments of qual i ta tive abnor mal i ties. For untreated MPS I dogs, the lengths of C2 and femurs (91% and 96% of nor mal, respectively) were not sig nifi cantly dif fer ent from nor mal dogs. Qual i ta tive changes in MPS I bones and joints were gen er ally mod est and were par tially improved with RV treat ment, although cer vi cal spine dis ease was severe and was di f icult to cor rect with gene ther apy in both mod els. The greater sever ity of skel e tal dis ease in MPS VII than in MPS I dogs may reflect accu mu la tion of chon droi tin sul fate in car ti lage in MPS VII, or could relate to the spe cific muta tions. Neo na tal RV-med i ated gene ther apy ame lio rates, but does not pre vent, skel e tal dis ease in MPS I and MPS VII dogs. 2008 Else vier Inc. All rights reserved. The mu co poly sac chari dos es (MPS) are lyso somal stor age diseases that result in accu mu la tion of gly cos ami no gly cans (GAG). MPS I is due to defi cient activ ity of a-l-idu ron i dase (IDUA; EC 3.2.1.76) and results in the accu mu la tion of hep a ran and der ma tan sul fate. MPS VII is due to defi cient activ ity of b-d-glu cu ron i dase (GUSB; EC 3.2.1.31) and results in the accu mu la tion of chon droi tin, hep a ran, and der ma tan sul fate. Clin i cal man i fes ta tions of MPS can include bone and joint dis ease, visual and audi tory defi cien cies, car diac dis ease, upper air way obstruc tion, hepa to splen o meg aly, and men tal retar da tion [1,2]. The skel e tal dis ease of MPS is referred to as dys os to sis mul tiplex, and man i fests as short ened bones, thick ened bones, bones that are abnor mally shaped, and bones with artic u lar ero sions. In humans, MPS I results in facial dys mor phia, kypho sco li o sis, stiff joints, and hip dys pla sia with sub lux a tion that causes a wad dling * Cor re spond ing author. Address: Depart ment of Inter nal Med i cine, Wash ing ton Uni ver sity School of Med i cine, 660 South Euclid Ave nue, St. Louis, MO 63110, USA. Fax: +1 314 362 8813. E-mail address: kpon der@im.wustl.edu (K.P. Ponder). gait [3 8]. MPS I dogs dis play col lapsed inter ver te bral spaces and ver te bral disc her ni a tion, ver te bral anky lo sis, oste o pe nia, focal artic u lar ero sions, degen er a tive joint dis ease, and joint effu sions [9,10]. Although both humans and dogs with MPS I have stunted growth, bone length mea sure ments have not been per formed. MPS I mice have increased bone min eral den sity and bone diam e ter, but rel a tively nor mal bone lengths [11,12]. MPS VII also results in dys os to sis mul ti plex. Humans have small ver te bral bodies, ky pho sis, wid ened ribs, short ened and thick ened long bones, atlan to ax ial insta bil ity lead ing to cer vi cal cord compres sion, and hip dys pla sia with sub lux a tion [13 15]. MPS VII dogs have sim i lar abnor mal i ties to those found in humans, and also have cur va ture of long bones and patel lar lux a tion [16,17]. MPS VII mice have thickening of long bones, and the femurs are»80% of nor mal length [18,19]. The mech a nisms that lead to bone and joint dis ease are not clear. Short bones are likely due to abnor mal i ties in the growth plate, which is very dis or ga nized in MPS VII mice and dogs [19] but is only mildly abnor mal in MPS I mice [20]. Degen er a tive changes may be due to upreg u la tion of pro te ases, as the syn o vium and 1096-7192/$ - see front matter 2008 Else vier Inc. All rights reserved. doi:10.1016/j.ymgme.2008.07.003

R.S. Her at i et al. / Molecular Genetics and Metabolism 95 (2008) 142 151 143 artic u lar car ti lage of MPS VII dogs have increased lev els of matrix metallo pro tein ases (MMPs) [21,22]. Cur rent treat ments for MPS I include hema to poi etic stem cell trans plan ta tion (HSCT) [23] and enzyme-replace ment ther apy (ERT) [24,25]. Human MPS I patients who received HSCT at 1.5 years of age had improved joint mobil ity at 5 years of age, but ky pho sis and hip sub lux a tion pro gressed [8]. HSCT in a 12-yearold patient with MPS VII improved ambu la tion, although radiographic abnor mal i ties were not ame lio rated 1.5 years later [26]. In MPS I dogs that received HSCT at 5 months of age, radio graphic abnor mal i ties were reduced at 20 months but inter ver te bral space nar row ing was still pres ent [9,10]. HSCT to new born MPS VII mice improved bone lengths to 87% of nor mal [27]. ERT to humans with MPS I improved growth [28], but details on the effect on radiographs have not been reported. ERT did not improve radio graph abnor mal i ties in MPS I dogs, but this study was com pli cated by the pres ence of anti bod ies directed against the human IDUA [24]. ERT to MPS VII mice resulted in partial improve ment of bone lengths to 88% of nor mal [18]. Thus, exist ing ther a pies reduce, but do not pre vent, skel e tal dis ease in MPS. Gene ther apy has been stud ied as an alter na tive treat ment for MPS in ani mal mod els. We have pre vi ously injected gamma ret rovi ral vec tors (RV) intra ve nously (IV) into new born MPS I and MPS VII mice and dogs [11,17,19,29,30]. This resulted in trans duc tion of»20% and»3% of hepa to cytes in mice and dogs, respec tively, which secreted enzyme into blood. In addi tion,»1% of blood cells were trans duced and expressed the RV in dogs [31]. Neo na tal admin istra tion of RV nor mal ized bone diam e ters in MPS I mice [11], and resulted in bones that were 86 93% of nor mal in MPS VII mice [19]. Although neo na tal RV treat ment in MPS VII dogs reduced skel e tal dis ease at»1 year of age [17,19], later eval u a tion and quan ti fi cation of the effect on bone lengths has not been reported. In addition, no data have been pre sented on the effect of gene ther apy on skel e tal dis ease in MPS I dogs. Here, we eval u ated the radio graphic man i fes ta tions of dis ease in bones and joints after neo na tal gene ther apy in MPS I and MPS VII dogs. Mate ri als and meth ods Ani mals NIH and USDA guide lines for the care and use of ani mals in research were fol lowed in the ani mal col ony of the School of Vet er i- nary Med i cine, Uni ver sity of Penn syl va nia. MPS I dogs were treated with neo na tal IV injec tion of 3 10 10 9 trans duc ing units (TU)/kg of haat-cid UA-WPRE, which is an RV express ing canine IDUA [11]. Radio graphs from 3 males and 3 females whose clin i cal data were reported pre vi ously [30] were eval u ated: I-99 (female; achieved 23 IDUA U/mL), I-101 (female; 27 U/mL), I-107 (female; 533 U/mL), I-140 (male; 888 U/mL), I-171 (male; 240 U/mL), and I-172 (male; 608 U/mL). A total of 4 males and 3 females were eval u ated for the RV-treated MPS VII group. The RV-treated MPS VII dog des ig nated M1287 (male; achieved 18,039 GUSB U/mL) received hepa to cyte growth fac tor (HGF) to induce hepa to cyte rep li ca tion prior to neo natal admin is tra tion of 12 10 9 TU/kg of haat-cgusb-wpre, which is an RV express ing canine GUSB, as described pre vi ously [17,29]. All other RV-treated MPS VII dogs received neo na tal IV injec tion of 3 88 10 9 TU/kg of haat-cgusb-wpre with out pre ced ing HGF and included M1328 (male; 429 U/mL), M1332 (female; 76 U/mL), M1337 (female; 294 U/mL), M1653 (female; 88 U/mL), M2065 (male; 2089 U/mL), and M2165 (male; 4027 U/mL). Clin i cal data on all ani mals except M1653, M2065, and M2165 were reported pre viously [17]. Untreated MPS I or MPS VII dogs, and het ero zy gous or homo zy gous nor mal dogs from the MPS I and the MPS VII col ony were also eval u ated. There were more females than males for the nor mal groups, as het ero zy gous females are main tained for breeding pur poses, while het ero zy gous males are not main tained due to ani mal space con straints. Sta tis tics All com par i sons were done using Sigm as tat 3.1 sta tis ti cal anal y- sis soft ware (Sy stat Soft ware, Inc., San Jose, CA, USA). Radio graphs Dogs were sedated with pro po fol to effect and posi tioned, the film was placed 40 inches from the emit ter, and radio graphs were obtained. Mea sure ments were per formed on X-ray film using a ruler or on dig i tal radio graphs using the Phi lips I-Site Enter prise 3.5 (March 2006) view ing soft ware. Mea sure ments of the lengths of ver te bral bodies and the inter ver te bral spaces were per formed on lateral views. Mea sure ments of the radius, ulna, femur, and tibia were per formed on lateral views of the limb posi tioned on the table. Lengths of meta car pals and met a tar sals were mea sured on dors opal mar and dor sopl an tar views, respec tively. Bone lengths and inter ver te bral space lengths were com pared using one-way ANOVA with Tu key post- hoc anal y sis. Radio graph scor ing sys tem Radio graphs were eval u ated by a board-cer ti fied vet er i nary radi ol o gist (VWK) who was blinded to the geno type and treat ment of the ani mals. For most cat e go ries, radio graphs were scored on a 0 2 scale, where 0 rep re sented nor mal, 1 rep re sented a mod erate abnor mal ity, and 2 rep re sented a severe abnor mal ity, and the Mann Whit ney U test was used for sta tis ti cal com par i sons. For lux a tion of the cox o fem o ral joint and patella, ani mals received a score of 0 if nei ther side was affected, 1 if one side was affected, and 2 if both sides were affected; this scor ing sys tem did not take into account the sever ity of lux a tion. Car pal bone sub lux a tion was when the front paw devi ated in a val gus (lateral) or varus (medial) direc tion or if the bones were not aligned nor mally. For cer vi cal spine fusion, artic u lar ero sion of the carpi, and radial and ulnar curv ing, abnor mal i ties were clas si fied as 0 if absent and 2 if present and sta tis ti cal com par i sons were per formed with Fisher s exact test. Results Neo na tal gene ther apy for MPS I and VII dogs Six MPS I dogs that received neo na tal gene ther apy [30] achieved 23 888 U/mL of IDUA activ ity in serum. This was 2- to 68-fold the level in nor mal dogs, although high lev els in serum do not nec es sarily result in nor mal enzyme activ ity in tis sues. Six MPS VII dogs that received neo na tal gene ther apy [17,29] achieved 76 4027 U/mL of serum GUSB activ ity, which was 0.3- to 15-fold the level in nor mal dogs. One MPS VII dog who received HGF in addi tion to neo na tal gene ther apy achieved 18,039 U/mL of GUSB activ ity (67-fold nor mal). RVtreated dogs with sim i lar lev els of serum enzyme activ ity had high enzyme activ ity as well as bio chem i cal and his to log i cal evi dence of improve ment in lyso somal stor age in many organs, as was reported pre vi ously [17,19,30]. We have also mea sured urine GAG lev els in the MPS VII dogs, and found that they were nor mal ized in the RV-treated group (nor mal had 13.3 ± 5.5 lg of GAG per mg of cre at i nine, N = 5; untreated MPS VII dogs had 76 ± 52 lg of GAG per mg of cre at i nine, N = 4; and RV-treated MPS VII dogs had 18 ± 9.4 lg of GAG per mg of cre at i nine, N = 4), and there were sta tis ti cally sig nifi cant dif fer ences between untreated and RV-treated dogs (p < 0.05). Urine GAGs were not mea sured in the MPS I dogs. Thus, RV-treated dogs had evi dence of reduced man i fes ta tions of dis ease in many sites.

144 R.S. Her at i et al. / Molecular Genetics and Metabolism 95 (2008) 142 151 Fig. 1. Radio graphs of ver te brae. Radio graphs from male dogs were obtained at the ages in years (Y) shown in the lower left cor ner. The geno type and treat ment sta tus are indi cated above the pan els. RV-treated dogs received neo na tal IV injec tion of an RV express ing the appro pri ate gene. Exam ples of RV-treated MPS I dogs are from I-171 in panels C and H and I-172 in panel M, while all exam ples of MPS VII dogs are from M1328. The scale bar shown in each image is 1 cm, and the cra nial aspect is at the top and the cau dal aspect at the bot tom. (A E) Ven tro dor sal view of the cer vi cal spine. The sec ond cer vi cal ver te bra (C2), C3, and C4 are indi cated. The hor i zon tal white arrows indi cate inter ver te bral spaces. Black ver ti cal arrows indi cate the medial and lateral bor ders of the ped i cle, and the space between the arrows was used to assess width. (F J) Lateral view of the cer vi cal spine. Slanted white arrows indi cate cau do ven tral ver te bral body bea king. The black hor i zon tal arrow indi cates fusion of the artic u lar facet joint. (K O) Lateral lum bar spine. The sixth lum bar ver te bra (L6) and L7 are indi cated. Hor i zon tal white arrows indi cate inter ver te bral spaces. RV-treated MPS I dogs were ra dio graphed at 1 2.2 years of age, and RV-treated MPS VII dogs were ra dio graphed at 1 7 years of age. Con trols were het ero zy gous or homo zy gous nor mal dogs or untreated affected dogs from the same breed ing col ony. It should be noted that these are out bred dogs with substantial genetic vari a tion. Radio graphic eval u a tion of ver te brae As few untreated MPS VII dogs sur vive past 1 year of age, radiographs were obtained at 1 year for nor mal, untreated MPS I, and untreated MPS VII dogs to facil i tate com par i son of the sever ity of skel e tal dis ease between the two mod els. For the RV-treated MPS I and MPS VII dogs, radio graphs were obtained at 1 year to facili tate deter mi na tion of whether or not the bones and joints were improved with RV treat ment. In addi tion, some radio graphs were obtained at a max i mum age of 2.2 years and 7 years for the RVtreated MPS I and MPS VII dogs, respec tively, which allowed the effect of aging upon skel e tal dis ease to be eval u ated. As detailed below, Fig. 1 shows rep re sen ta tive exam ples of radio graphs, while sup ple men tary Figs. 1 3 show many exam ples of radio graphs obtained of the ven tro dor sal cer vi cal spine, the lateral cer vi cal spine, and the lateral lum bar spine, respec tively, at dif fer ent ages for ani mals in each group. Fig. 2 shows aver age ver te bral body and inter ver te bral space lengths ± one stan dard devi a tion (SD) for animals in dif fer ent groups, and Fig. 3 shows subjective scores of a vari ety of param e ters that were eval u ated from radio graphs. Lengths of ver te brae in untreated MPS I and MPS VII dogs The lengths of the cer vi cal ver te brae from untreated MPS I dogs (Figs. 1B and 2A) were only slightly shorter than in nor mal dogs from the same col ony (Figs. 1A and 2A), as the sec ond cer vical ver te brae (C2), C3, and C4 were 37 ± 5 mm (91% of nor mal for the same col ony), 26 ± 4 mm (96% nor mal), and 25 ± 2 mm (93% of nor mal), respec tively. The fail ure to observe sig nifi cant dif fer ences between val ues in untreated MPS I and nor mal dogs may reflect the small num ber of ani mals in each group, and that val ues from dogs of both gen ders were pooled due to the small total num bers of ani mals. For nor mal ani mals, male ver te bral body heights were 107 ± 2% of those found in females. In con trast to the results with MPS I, lengths of C2, C3, and C4 were mark edly reduced in untreated MPS VII ver te brae (Figs. 1D and 2B) at 23 ± 3 mm (57% of nor mal), 15 ± 2 mm (56% nor mal),

R.S. Her at i et al. / Molecular Genetics and Metabolism 95 (2008) 142 151 145 Fig. 2. Axial skel e ton mea sure ments. Mea sure ments of ver te bral body and inter ver te bral space lengths were obtained for the cer vi cal and lum bar spine in post-puber tal dogs at 0.7 3 years after birth, and val ues for males and females were com bined. Some RV-treated MPS VII dogs were eval u ated at sev eral ages, and val ues (which did not vary sub stan tially) were aver aged to give a sin gle value for that ani mal. Val ues in sev eral ani mals were then aver aged and means ± the stan dard devi a tion were cal cu lated. The means for sev eral ani mals in each group ± one SD are shown; ( ) indi cates p of 0.01 0.05 and ( ) indi cates p < 0.01 for com par i son of the groups con nected by a bracket using ANOVA with Tukey post- hoc anal y sis. (A) Ver te bral body lengths for MPS I dogs. Ver te bral body lengths of C2 (sec ond cer vi cal ver te bra), C3, C4, L5 (fifth lum bar ver te bra), L6, and L7 were mea sured for 8 het ero zy gous nor mal (1 male, 7 female), 7 untreated MPS I (5 male, 2 female), and 6 RV-treated MPS I (3 male, 3 female) post-puber tal dogs at 0.6 1 year of age. (B) Ver te bral body lengths for MPS VII dogs. Lengths were mea sured for 12 het ero zy gous nor mal (3 male, 9 female), 15 untreated MPS VII (11 male, 4 female), and 7 RV-treated MPS VII (4 male, 3 female) dogs. (C and D) Inter ver te bral space lengths for MPS I and VII dogs. Inter ver te bral space lengths were mea sured for the same ani mals whose val ues are shown in pan els A and B. Fig. 3. Axial skel e ton radio graphic scores. Radio graphs were scored for abnor mal i ties at 1 year of age for most groups, where 0 indi cates nor mal and 2 indi cates severely abnor mal. Both males and females were eval u ated, means ± SD are shown, and ( ) indi cates p < 0.05 and ( ) indi cates p < 0.01 for com par i son of the groups con nected by a bracket using the Mann Whit ney U test. (A) Radio graphic scores for MPS I dogs. The scores are shown for 8 nor mal dogs at 1 2 years (2 male, 6 female), 6 untreated MPS I dogs at 1 year (4 male, 2 female), and 6 neo na tal RV-treated MPS I dogs at 1 year (3 male, 3 female). Ver te bral body dys pla sia, ped i cle wid en ing, cau do ven tral ver te bral body bea king, artic u lar facet joint fusion, and ver te bral body tip ping were eval u ated. (B) Radio graphic scores for MPS VII dogs. The scores are shown for 8 nor mal dogs at 1 2 years (2 male, 6 female), 6 untreated MPS VII dogs at 1 year (5 male, 1 female), and 5 RV-treated dogs at 1 year (2 male, 3 female), and sta tis ti cal com par i sons were per formed for the 3 groups with the Mann Whit ney U test. Two of the seven RV-treated MPS VII dogs that were alive at 1 year were not eval u ated as radio graphs were not obtained at the appro pri ate age. RV-treated MPS VII dogs were also scored for 5 dogs at 4 7 years (2 male and 2 female at 7 years and 1 female at 4 years), and com par i son of val ues with those in RV-treated MPS VII dogs at 1 year using the Mann Whit ney U test failed to find sig nifi cant dif fer ences. and 16 ± 2 mm (62% nor mal), respec tively (p < 0.01 for com par i son of MPS VII vs. nor mal for all ver te brae). Sim i larly, the fifth lum bar (L5), L6, and L7 ver te bral body lengths were mod estly reduced at 92%, 92%, and 85% of nor mal in untreated MPS I dogs (Figs. 1L and 2A; not sig nifi cant vs. nor mal), but were mark edly short ened at 66%, 67%, and 58% of nor mal in untreated MPS VII dogs (Figs. 1N and 2B; p < 0.01 vs. nor mal). Qual i ta tive eval u a tion of the cer vi cal spine in untreated MPS I and MPS VII dogs A vari ety of param e ters were sub jec tively scored from 0 (normal) to 2 (severely abnor mal) on radio graphs, as shown in Fig. 3. The struc ture of the cer vi cal ver te brae from MPS I dogs appeared nor mal (Figs. 1B and 1G) and received a dys plas tic score of 0 ± 0 (Fig. 3A). In con trast, the cer vi cal ver te brae from MPS VII dogs were mark edly abnor mal (Figs. 1D and 1I) with a dys plas tic score of 1.8 ± 0.4 (Fig. 3B; p < 0.01 vs. nor mal). The cer vi cal spine was assessed for wid en ing, which refers to the thickening of bone around the ped i cle and can be appre ci ated on the ven tro dor sal pro jec tion for the MPS I dog shown in Fig. 1B. Wid en ing occurred in both MPS I (score of 1.1 ± 0.3; p < 0.001 vs. nor mal) and MPS VII (score of 1.5 ± 0.4; p < 0.001 vs. nor mal) dogs. Bea king refers to the cau do ven tral os te o phy to sis that results in a bird-like beak, which can be appre ci ated in the MPS VII dog shown in Fig. 1I, and may be due to her ni a tion of the nucleus pul po sus [5,32]. Bea king was

146 R.S. Her at i et al. / Molecular Genetics and Metabolism 95 (2008) 142 151 pres ent in both MPS I (1.2 ± 0.4; p < 0.001 vs. nor mal) and MPS VII (1.8 ± 0.4; p < 0.001 vs. nor mal) dogs. Fusion refers to artic u lar facet joint space fusion, which can be appre ci ated in the MPS VII dog shown in Fig. 1I. Fusion was severe in both MPS I (2 ± 0; p = 0.01 vs. nor mal) and MPS VII (2 ± 0; p = 0.01 vs. nor mal) dogs. Tip ping refers to a rota tion of the ver te brae so that the nor mal end-toend con for ma tion is lost, and the cra nial sur face of one ver te bra is ele vated rel a tive to the cau dal aspect of the pre ced ing ver te bra. Although tip ping was very mod est in MPS I dogs at 0.5 ± 0.5 (not sig nifi cant vs. nor mal), it was mod er ate in MPS VII dogs (1 ± 0.9; p = 0.04 vs. nor mal). We con clude that both canine mod els of MPS had substantial abnor mal i ties in the cer vi cal spine, although disease was more severe in MPS VII than in MPS I. Effect of gene ther apy on the cer vi cal spine in MPS I and MPS VII dogs A major goal of this pro ject was to deter mine if neo na tal gene ther apy with an RV express ing the appro pri ate gene could prevent bone dis ease in MPS I and MPS VII dogs. The cau dal to cra nial lengths of ver te brae and inter ver te bral spaces were deter mined for post-puber tal ani mals whose ages ranged from 0.7 to 3 years. Some indi vid ual RV-treated dogs were eval u ated at sev eral ages, and the val ues (which were sim i lar) were aver aged. There were no sig nifi cant dif fer ences in the lengths of cer vi cal or lum bar ver tebral bodies in RV-treated MPS I vs. untreated MPS I dogs (Fig. 2A), which may reflect mod est short en ing in MPS I dogs. For RV-treated MPS VII dogs, the heights of C2 (82% of nor mal), C3 (75%), C4 (82%), L5 (93%), L6 (93%), and L7 (83%) were all sig nifi cantly longer than in untreated MPS VII dogs (p < 0.01 for all ver te brae vs. untreated MPS VII), but remained sta tis ti cally shorter than in nor mal dogs for C2, C3, C4, and L7 (p < 0.01), as shown in Fig. 2B. RV-treated MPS I dogs had a sig nifi cant reduc tion in fusion at 1 year of age when com pared with untreated MPS I dogs (Fig. 3A), although the mod est improve ments in wid en ing, bea king, and tip ping were not sig nifi cant. RV-treated MPS VII dogs did not have a sig nifi cant improve ment in any of the param e ters that were scored in Fig. 3B at 1 year of age. In gen eral, the scores at 7 years were sim i- lar to those obtained at 1 year for MPS VII dogs. How ever, fusion appeared to worsen with age, although dif fer ences in the val ues obtained at 1 and 7 years were not sig nifi cant. Eval u a tion of inter ver te bral spaces The inter ver te bral space lengths can be used to deter mine the cau dal to cra nial length of the non-cal ci fied inter ver te bral discs. The inter ver te bral space lengths between C2 and C3 (C2/C3) and C3/C4 were not sig nifi cantly dif fer ent in MPS I or MPS VII dogs from the val ues in nor mal dogs (Fig. 2C and 2D). Although lumbar inter ver te bral space lengths were not sig nifi cantly dif fer ent between untreated MPS I and nor mal dogs, they were greater in untreated MPS VII dogs at 193% and 164% of nor mal at L5/L6 and L6/L7, respec tively (p < 0.01 vs. nor mal). Neo na tal gene ther apy to MPS VII dogs resulted in inter ver te bral space lengths that were reduced to 156% of nor mal for L5/L6 (p < 0.05 vs. untreated MPS VII) and 129% of nor mal for L6/L7 (p < 0.01 vs. untreated MPS VII), respec tively, although the length at the L5/L6 inter ver te bral space remained greater than nor mal (p < 0.01). Three of six untreated MPS I dogs that were eval u ated for 1 year or longer devel oped clin i cal signs of cord com pres sion, which included weak ness, hyper re flexia, and reduced pro pri o cep tion of the limbs, as described pre vi ously [30]. Her ni ated discs can result in cord com pres sion, and can be asso ci ated with inter ver te bral Fig. 4. Eval u a tion of joints. Radio graphs are from male dogs at the ages in years (Y) shown in the lower left cor ner. The geno type and treat ment sta tus are indi cated above the pan els. Exam ples of radio graphs of RV-treated MPS I dogs were from I-172 in pan els C and H, and from I-171 in panel M. All exam ples of radio graphs of the RV-treated MPS VII dog were from M1328. (A E) Car pal joint dors opal mar. The meta car pals are at the bot tom of the image, and the radius (right) and ulna (left) are on the top. The white ver ti cal arrows indi cate radi o car pal bones. The slanted white arrow indi cates ero sions of the 3rd meta car pal bone. (F J) Cox o fem o ral joint ven tro dor sal. The pel vis is on the left side of the image. The hor i zon tal white arrow indi cates lux a tion of the fem o ral head. (K O) Sti fle joint lateral. The femur is at the top of the image and the tibia is at the bot tom. The white arrows indi cate the left aspect of the tri an gu lar-shaped radio lu cent fat pad pres ent at the cra nial aspect of the sti fle joint, which can be visu al ized when no effu sion is pres ent. This tri an gle is absent in an untreated MPS VII dog (panel N) due to an effu sion.

R.S. Her at i et al. / Molecular Genetics and Metabolism 95 (2008) 142 151 147 space nar row ing. Indeed, some untreated MPS I mice had interver te bral space nar row ing, as can be seen at C2/C3 in Fig. 1B. All cer vi cal inter ver te bral spaces were eval u ated for 6 untreated animals, and a total of 9 of 36 (25%) were nar rowed for the untreated MPS I dogs. Although the num ber of nar rowed cer vi cal inter ver tebral spaces of 4 out of 36 eval u ated (11%) from 6 RV-treated MPS I dogs was not sig nifi cantly dif fer ent from the fre quency in MPS I dogs with Fisher s exact test, no RV-treated MPS I dogs devel oped signs of cord com pres sion. Thus, it is pos si ble that cord com pression was due to other causes such as thickening of the dura mater or lig a ments of the spine, or lig a ment lax ity result ing in sub luxa tion of the bones, as has been reported in human patients with MPS [5 7,14,15]. Mag netic res o nance imag ing will be per formed in the future to attempt to iden tify the eti ol ogy of cord com pression in untreated MPS I dogs. Quan ti fi ca tion of the num ber of narrowed inter ver te bral space lengths in the cer vi cal spine of MPS VII dogs was com pli cated by the fact that the marked dys pla sia of the cer vi cal ver te bral bodies in the cau dal regions made it di f icult to define the bound aries of bone and disc. For this reason, inter verte bral space nar row ing was only eval u ated at C2/C3 and at C3/C4. For untreated MPS VII dogs, 5 of 12 (42%) inter ver te bral disc spaces that were eval u ated at 0.5 2 years of age were nar rowed. This frequency was sim i lar at 1 year in RV-treated MPS VII dogs (5 of 12 spaces were nar rowed, or 42%) and at 7 years in RV-treated dogs (4 of 8 spaces were nar rowed, or 50%; see C3/C4 in Figs. 1E and 1J). Clin i cal eval u a tion for cord com pres sion in untreated MPS VII dogs was di f icult, as these dogs did not walk beyond 6 months of age. No RV-treated MPS VII dogs devel oped clin i cal signs of cord compres sion for up to 7 years of eval u a tion. Radio graphic eval u a tion of the car pus The appen dic u lar skel e ton was also eval u ated radio graph i cally. As detailed below, Fig. 4 shows rep re sen ta tive radio graphs, Fig. 5 shows sever ity scores for var i ous abnor mal i ties, and Fig. 6 quan tifies long bone lengths. Addi tional exam ples of radio graphs from dogs of each group are pro vided in the sup ple men tary data sec tion for the cubi tal joint (elbows; sup ple men tary Fig. 4), car pal joint (sup ple men tary Fig. 5), cox o fem o ral joint (hip; sup ple men tary Fig. 6), lateral sti fle joint (knee; sup ple men tary Fig. 7), ven tro dor sal stifle joint (sup ple men tary Fig. 8), and radius/ulna (sup ple men tary Fig. 9). Car pal bones from MPS I dogs were only mildly dys plas tic and lucent, there were no artic u lar ero sions, and there were only mod est effu sions (Figs. 4B and 5A), but these param e ters were all mark edly abnor mal in untreated MPS VII dogs (Figs. 4D and 5B). Fig. 5. Appen dic u lar skel e ton radio graphic scores. (A) Radio graphic scores for MPS I carpi. Radio graphic scores are shown for 8 nor mal dogs at 2 years of age (2 male, 6 female), 6 untreated MPS I dogs at 1 year (4 male, 2 female), and 6 RV-treated MPS I dogs at 1 year of age (3 male, 3 female). Car pal bone dys pla sia, car pal bone lucency, articu lar bone ero sions, car pal joint sub lux a tion, and joint effu sions were scored from 0 (nor mal or absent) to 2 (severely abnor mal). (B) Radio graphic scores for MPS VII fore limb. Radio graphic scores are shown for 8 nor mal dogs at 1 2 years of age (2 male, 6 female), 6 untreated MPS VII dogs at 1 year (5 male, 1 female), 5 RV-treated dogs at 1 year (2 males and 3 females), and 5 RV-treated dogs at 4 to 7 years (2 male and 2 females at 7 years and 1 female at 4 years, these are labeled as 7 years in the fig ure). (C) Radio graphic scores for MPS I hind limb. Radio graphic scores are shown for the same groups that are described in panel A. The cox o fem o ral joint was eval u ated for dys pla sia, sub lux a tion, artic u lar bone ero sions, and degen er a tive joint dis ease (DJD). The sti fle joints were eval u ated for patel lar lux a tion (pat. lux.) and effu sions. (D) Radio graphic scores for MPS VII hind limb. Radio graphic scores are shown for the same ani mals described in panel B. (E) Radio graphic scores for MPS I femur and radius/ulna. Radio graphic scores are shown for the same dogs described in panel A. Fem o ral head dys pla sia, fem o ral ero sions, fem o ral cur va ture, radius/ulna curv ing, and radius/ulna dys pla sia were eval u ated. (F) Radio graphic scores for MPS VII femur and radius/ulna. Radio graphic scores are shown for the same dogs described in panel B.

148 R.S. Her at i et al. / Molecular Genetics and Metabolism 95 (2008) 142 151 Fig. 6. Appen dic u lar skel e ton lengths. Mea sure ments of bone length were obtained for hind limbs for post-puber tal dogs that were 0.7 3 years of age. Val ues did not dif fer in the RV-treated ani mals that were ra dio graphed at 1 3 years of age. ( ) indi cates a p value of 0.01 0.05, and ( ) indi cates a p value <0.01 for the sta tis ti cal com par i son of the groups that are con nected with a bracket. (A) Hind limb bone lengths in MPS I dogs. Mea sure ments of the femur, tibia, and met a tar sals (MT) of the same dogs that are described in Fig. 5A are shown. (B) Hind limb bone lengths in MPS VII dogs. Mea sure ments of the bones of the dogs described in Fig. 5B are shown. RV-treated MPS VII dogs had sig nifi cant reduc tions in car pal bone dys pla sia (p < 0.05), artic u lar ero sions (p < 0.01), and joint effusions (p < 0.05) at 1 year of age. These improve ments were gen erally main tained at 7 years, although artic u lar ero sions appeared to worsen with age. Radio graphic eval u a tion of the cox o fem o ral joint, sti fle joint, and long bones The cox o fem o ral joints, sti fle joints, and long bones were also eval u ated radio graph i cally. For untreated MPS I dogs at 1 year, there was a sig nifi cant but mod est increase in dys pla sia of the cox o- fem o ral joint (Figs. 4G and 5C), the sever ity of sti fle joint effu sions (Fig. 5C), and dys pla sia of the radius and ulna (Fig. 5E) com pared with nor mal dogs (Fig. 4F, 4K, 5A, 5C, and 5E). The sever ity of sti fle joint effu sions was reduced in RV-treated MPS I dogs at 1 year (Fig. 5C) and improve ments were main tained at 2 years. Untreated MPS I dogs that did not develop cord com pres sion had a nor mal gait for up to 1.75 years of age, the old est age of anal y sis. Five of six RVtreated MPS I dogs had a nor mal gait for the dura tion of eval u a tion, although I-99 devel oped lux at ed shoul ders at 13 months of age. For untreated MPS VII dogs, there was severe dys pla sia and luxa tion of the cox o fem o ral joint at 1 year (Figs. 4I and 5D). Indeed, 6 of 7 (86%) untreated MPS VII dogs had mark edly lux at ed fem o ral heads at 1 year, and all were unable to stand or walk by 6 months of age. RV-treated MPS VII dogs had a sig nifi cant reduc tion in dys plasia of the cox o fem o ral joint (Figs. 4J and 5D; p = 0.02 vs. untreated MPS VII). Although the reduc tion in the score for lux a tion in RVtreated MPS VII dogs was not sig nifi cant at 1 year, only one of seven (14%) RV-treated MPS VII dogs had severe sub lux a tion at 1 year (p = 0.029, data not shown). In addi tion, all four RV-treated MPS VII dogs that were eval u ated at 7.5 years could run, as shown in the accom pa ny ing video; M1287 and M1328 did have an abnor mal gait that could be due to the hip dys pla sia that is pres ent to some degree in nor mal dogs from the col ony. Untreated MPS VII dogs also had severe artic u lar ero sions and degen er a tive joint dis ease of the cox o fem o ral joint (Fig. 4I), patel lar lux a tion, sti fle effu sions (Fig. 4N), dys pla sia and artic u lar ero sions of the dis tal femur (Fig. 4N), curv ing of the femur, and dys pla sia of the radius and ulna. Scores for RV-treated MPS VII dogs at 1 year that were sig nifi cantly bet ter than for untreated MPS VII dogs at 1 year included artic u lar ero sion of the cox o fem o ral joint, patel lar lux a tion, sti fle effu sions, artic u lar ero sion of the dis tal femur, and curv ing of the femur (Fig. 5D and 5F). In gen eral, scores at 7 years of age in RV-treated MPS VII dogs were sim i lar to those found at 1 year in RV-treated MPS VII dogs. Although scores of artic u lar ero sions of the cox o fem o ral joint and femur tended to increase in RV-treated dogs at 7 years, these dif fer ences were not sig nifi cantly dif fer ent from those observed in RV-treated dogs at 1 year using the Mann Whit ney U test. Lengths of long bones Lengths of long bones of the hind limb (Fig. 6A and 6B) and the fore limb (sup ple men tary Fig. 10) were deter mined. Lengths of the radius, ulna, meta car pals, femur, tibia, and met a tar sals in untreated MPS I dogs were all >92% of nor mal, and no value was sig nifi cantly dif fer ent from those in nor mal dogs from the same col ony. In contrast, untreated MPS VII dogs had sig nifi cant reduc tions in lengths of the ulna (88% of nor mal; p < 0.01 vs. nor mal), the sec ond meta carpal (MC2) (86%; p < 0.01), MC3 (85%; p < 0.01), MC4 (85%; p < 0.01), MC5 (81%; p < 0.01), femur (84%; p < 0.01), tibia (90%; p < 0.01), the sec ond meta tar sal (MT2) (84%; p < 0.05), MT3 (83%; p < 0.01), MT4 (85%; p < 0.05), and MT5 (83%; p < 0.01). RV-treated MPS VII dogs had lengths that were near-nor mal for all of these bones (not sig nifi cant vs. nor mal), and were sig nifi cantly longer than in untreated MPS VII dogs for the ulna, all meta car pals that were eval u ated, the femur, the tibia, and the sec ond to fourth met a tar sals. Dis cus sion Skel e tal dis ease is more severe in MPS VII than in MPS I dogs The first goal of this study was to com pare the sever ity of skel e tal dis ease in MPS I and MPS VII dogs by eval u at ing skel e- tal radio graphs. MPS VII dogs exhib ited more severe short en ing of ver te bral bodies and long bones, and were more abnor mal in subjective eval u a tion of dys pla sia, artic u lar ero sions, sub lux a tion, and effu sions. There are two pos si ble expla na tions for why bones are less severely affected in MPS I than in MPS VII dogs: (1) the muta tion in MPS I dogs could be rel a tively atten u ated and, therefore, anal o gous to that in a Hurler/Scheie or a Scheie patient; or (2) the accu mu la tion of chon droi tin sul fate in car ti lage and other sites in MPS VII but not MPS I may be par tic u larly impor tant for skel e tal dis ease. It is pos si ble that skel e tal dis ease is less severe in the MPS I dog model because the muta tion allows some resid ual enzyme activ ity to remain. MPS I dogs have a 59 splice site muta tion in intron 1 [33], and liver extracts from untreated MPS I dogs had some activ ity against the arti fi cial sub strate 4-meth yl um bel liferyl a-l-i do pyr ano sid u ron ic acid in vitro, as the fluo res cence of sam ples increased with time dur ing a 3 h incu ba tion (Ping Wang, MEH, KPP, unpub lished data), although the activ ity was low. The enzyme activ ity was cal cu lated to be 1.1 U/mg (5% of the value in nor mal dogs [30]), although it is pos si ble that another enzyme could exhibit some activ ity against the arti fi cial sub strate but not degrade the endog e nous sub strate. Liver from MPS I mice with an inser tion into exon 6 of the 14-exon gene had 0.01 U/mg (0.2% of nor mal), and liver from MPS I cats with a 3 nt dele tion in the coding sequence had 0.6 U/mg (2.4% nor mal) [34]. Since MPS I dogs

R.S. Her at i et al. / Molecular Genetics and Metabolism 95 (2008) 142 151 149 and cats had >60-fold as much enzyme as did knock-out MPS I mice, this raises the pos si bil ity that there may be some resid ual enzyme activ ity in the large ani mal mod els. Although these MPS I dogs do not have detect able IDUA pro tein on immu no blot [33], it is pos si ble that pro tein and cor rectly-spliced RNA are pres ent at low lev els. Our attempts to amplify IDUA RNA using sev eral dif fer ent for wards exon 1 and reverse exon 2 prim ers were not suc cess ful in nor mal dogs (M. Tit ti ger, KPP, unpub lished data), mak ing it di f icult to assess this pos si bil ity. The dog model of MPS VII has a mis sense R166H muta tion and GUSB activ ity is <1% of nor mal [35]. Thus, skel e tal dis ease may be less severe in MPS I dogs than in MPS VII dogs because MPS I dogs main tain a low level of enzyme activ ity. Alter na tively, MPS VII may result in more severe skel e tal disease than MPS I because chon droi tin sul fate accu mu lates in the former but not the lat ter model. Chon droi tin sul fate con tains b- glu cu ronic acid and, there fore, is metab o lized by GUSB, while it does not con tain idu ron ic acid and, there fore, is not metab o lized by IDUA. Chon droi tin sul fate is the major GAG in artic u lar car ti lage and in the growth plate where bone elon ga tion occurs [36,37]. In con trast, lev els of der ma tan sul fate and hep a ran sul fate, which are metab o lized by both GUSB and IDUA, are very low in the car tilage and growth plates [38 40]. Chon droi tin sul fate accu mu lates in the urine of MPS VII dogs [16] and in the growth plate of MPS VII mice but not MPS I mice (Jason Met calf, KPP, unpub lished data). Indeed, MPS VII mice that have <1% of nor mal GUSB activity [41,42] have bone lengths that are extremely short at»80% of nor mal [18,19], while MPS I mice with a knock-out muta tion that results in <0.2% of nor mal IDUA activ ity had long bone lengths that are almost nor mal at >95% of nor mal [11,12,43]. Eval u a tion of skel e tal dis ease in other murine MPS mod els sup ports the hypoth e sis that accu mu la tion of chon droi tin sul fate and/or kera tan sul fate are impor tant for bone dis ease. For exam ple, skele tal dis ease was very severe in MPS VI mice with knock-out of N-acet y lga lac tos amine 4-sul fa tase [44], which is impor tant for deg ra da tion of chon droi tin 4-sul fate. Con versely, skel e tal disease was mild or absent in MPS II mice [45] with defi ciency of idur o nate sul fa tase, which is impor tant for deg ra da tion of the sul fated idu ron ic acid-con tain ing der ma tan and hep a ran sul fate. Skel e tal dis ease was also mild in MPS IIIA [46] and MPS IIIB [47] mice with defi ciency of hep a ran N-sul fa tase and a-n-acetyl-gluco sa min i dase, respec tively, both of which are impor tant for degra da tion of hep a ran sul fate. Inter est ingly although skel e tal disease is severe in human MPS IVA patients that can not degrade chon droi tin 6-sul fate or ker a tan sul fate due to defi ciency of N- acet y lga lac tos amine 6-sul fa tase, MPS IVA mice have bones that are nearly nor mal in length [48]; the lack of skel e tal dis ease in MPS IVA mice was attrib uted to chon droi tin 6-sul fate being <2% of the total chon droi tin sul fate in murine car ti lage [49]. Thus, data from other mouse mod els are con sis tent with the hypothe sis that accu mu la tion of chon droi tin sul fate is asso ci ated with long bone short en ing. It is unclear whether chon droi tin sul fate plays a spe cific role in acti vat ing sig nal trans duc tion path ways, or if other GAGs have the potential to affect bones, but sim ply do not accu mu late in the car ti lage and growth plates. If indeed the reduced sever ity of skel e tal dis ease in MPS I dogs as com pared with MPS VII dogs is due to the spe cific GAGs that accu mulate in each model, it is unclear as to why skel e tal dis ease appears to be more severe in human MPS I-Hurler patients than in MPS I dogs and mice. This could reflect dif fer ences in the amounts of der ma tan and/ or hep a ran sul fate that accu mu late in car ti lage and growth plates in humans as com pared with dogs and mice, or could reflect the fact that humans require more time to develop and GAGs could accu mu late to greater lev els dur ing this period. Although bone lengths in MPS I dogs were not sig nifi cantly shorter than in nor mal lit ter mates, it should be kept in mind that the num ber of ani mals eval u ated was small, and many bones were con sis tently shorter in MPS I dogs at 85 95% of normal. It is pos si ble that anal y sis of more ani mals will iden tify sig nifi cant dif fer ences between the groups. Neo na tal gene ther apy reduces many man i fes ta tions of skel e tal dis ease in dogs A sec ond goal of this study was to eval u ate the long-term effects of neo na tal gene ther apy on the skel e tal dis ease in MPS I and MPS VII dogs. Eval u a tion of the effect of gene ther apy on skele tal dis ease in MPS I dogs was com pli cated by the mild ness of skele tal dis ease in untreated MPS I dogs, the rel a tively short period (at most 2 years) of eval u a tion, and the small num bers of ani mals that were eval u ated. In gen eral, gene ther apy prob a bly reduced but did not elim i nate many aspects of skel e tal dis ease in MPS I dogs, but eval u a tion of more ani mals will be required to achieve sig nifi cance for most param e ters. In the MPS VII model, gene ther apy clearly increased bone lengths and improved many param e ters of skel e tal dis ease over those observed in untreated MPS VII dogs, but bones remained shorter than in nor mal dogs and some param e ters did not improve. The cer vi cal spine was di f icult to cor rect with gene ther apy in both mod els, and remains a site where substantial morbid ity may occur. In gen eral, bones were sim i lar at 7 years in RVtreated MPS VII dogs to those found at 1 year in RV-treated MPS VII dogs, but artic u lar ero sions tended to worsen. Although this study does not eval u ate the path o gen e sis of bone and joint dis ease, pre vi ous stud ies sug gested that car ti lage and syn o vium upreg u lated expres sion of MMPs in MPS. Indeed, our lab o ra tory has recently dem on strated that MMP and cathep sin activ i ties are markedly increased in lig a ments, artic u lar car ti lage, and syn o vium of MPS VII dogs as com pared with nor mal dogs, and that val ues in MPS VII dogs are greater than in MPS I dogs (X. Ma, MEH, and KPP, unpub lished data). These data are con sis tent with artic u lar ero sions and sub lux a- tion of joints being more severe in MPS VII than in MPS I dogs. The role that these destruc tive pro te ases play in bone and joint dis ease in MPS I and MPS VII dogs is cur rently being fur ther eval u ated. The improvement in skel e tal dis ease with RV ther apy could reflect dif fu sion of enzyme from blood, bone mar row, or blood cells into these con nec tive tis sues. Indeed, we pre vi ously dem on strated that RV-treated MPS VII dogs had GUSB activ ity that was 60% of nor mal in syno vial fluid [17]. Impli ca tions This study dem on strates that neo na tal gene ther apy with an RV to MPS VII dogs can increase bone lengths, reduce many qual i ta tive abnor mal i ties of bone and joint dis ease, and enable ani mals to ambulate at 7 years. Nev er the less, skel e tal dis ease was not com pletely corrected, and dis ease in the cer vi cal spine remained severe. Achiev ing higher serum activ ity may not be effec tive, as the skel e tal dis ease in M1287 with 18,039 U/ml of serum GUSB activ ity (see sup ple men tal fig ures) was sim i lar to that in RV-treated ani mals with much lower GUSB activ ity. It is likely that the enzyme does not dif fuse well into the avas cu lar car ti lage, and that it will be di f icult to fully cor rect skele tal dis ease with gene ther apy unless it can be per formed ear lier (i.e. in ute ro) and/or result in trans duc tion of chon dro cytes. An impor tant impli ca tion of this study is that skel e tal dis ease in MPS I is less severe than in MPS VII in dogs, although it remains pos si ble that this reflects a milder muta tion in the MPS I dog than in the MPS VII dog. If indeed skel e tal dis ease is intrin si cally less severe in MPS I than in MPS VII, this may allow gene ther apy to more effec tive in MPS I, as skel e tal dis ease is rel a tively di f icult to treat. Acknowl edg ments This work was sup ported by the Ryan Foun da tion, the National MPS Soci ety, and the National Insti tutes of Health (DK66448

150 R.S. Her at i et al. / Molecular Genetics and Metabolism 95 (2008) 142 151 awarded to KPP, DK54481 awarded to MEH, and RR02512 awarded to MEH). Appen dix A. Sup ple men tary data Sup ple men tary data asso ci ated with this arti cle can be found, in the online ver sion, at doi:10.1016/j.ymgme.2008.07.003. Ref er ences [1] E.F. Neu feld, J. Mu en zer, The Mu co poly sac chari dos es, in: C.R. Scriv er A.L. Beaudet W.S. Sly D. Valle (Eds.), Met a bolic and Molec u lar Basis of Inher ited Dis ease, McGraw Hill, New York, 2001, pp. 3421 3452. [2] B.J. Poor thuis, R.A. We vers, W.J. Klei jer, J.E. Gro en er, J.G. de Jong, S. van We ely, K.E. Nie zen-kon ing, O.P. van Dig ge len, The fre quency of lyso somal stor age diseases in The Neth er lands, Hum. Genet. 105 (1999) 151 156. [3] E.L. Mast er son, P.G. Mur phy, A. O Me ara, D.P. Moore, F.E. Dow ling, E.E. Fo garty, Hip dys pla sia in Hurler s syn drome: ortho pe dic man age ment after bone marrow trans plan ta tion, J. Ped. Ortho. 16 (1996) 731 733. [4] R.E. Ste ven son, R.R. Ho well, V.A. McKu sick, R. Sus kind, J.W. Han son, D.E. El li ott, E.F. Neu feld, The idu ron i dase-defi cient mu co poly sac chari dos es: clin i cal and roent gen o graphic fea tures, Pedi at rics 57 (1976) 111 122. [5] V. Tan don, J.B. Wil liam son, R.A. Co wie, J.E. Wraith, Spinal prob lems in mucopol y sac cha ri do sis I (Hurler syn drome), J. Bone Joint Surg. Br. 78 (1996) 938 944. [6] S.A. Khan, K. Se hat, D. Cal thorpe, Cer vi cal cord com pres sion in an elderly patient with Hurler s syn drome: a case report, Spine 28 (2003) E313 E315. [7] C.B. Bill, J.S. Rose, L. God mi low, S. Sklow er, J. Will ner, K. Hir sch horn, Spas tic quad ri pa re sis due to C1, C2 sub lux a tion in Hurler syn drome, J. Pe di atr. 92 (1978) 441 443. [8] R.E. Field, J.A.F. Buchanan, M.G.J. Cop ple mans, P.M. Aich roth, Bone-mar row trans plan ta tion in Hurler s syn drome, Bone Joint Surg. Br. (1994) 975 981. [9] R.M. Shull, M.A. Walker, Radio graphic find ings in a canine model of muco pol y- sac cha ri do sis I, Invest. Radiol. 23 (1988) 124 130. [10] M.A. Bre id er, R.M. Shull, G. Con stan to po u los, Long-term effects of bone marrow trans plan ta tion in dogs with muco pol y sac cha ri do sis I, Am. J. Pathol. 134 (1989) 677 692. [11] Y. Liu, L. Xu, A.K. Hen nig, A. Kov acs, A. Fu, S. Chung, D. Lee, B. Wang, R.S. Herat i, J. Mo sing er Og il vie, S.R. Cai, K. Parker Pon der, Liver-directed neo na tal gene ther apy pre vents car diac bone ear and eye dis ease in muco pol y sac cha ri do sis I mice, Mol. Ther. 11 (2005) 35 47. [12] L.A. Clarke, C.S. Rus sell, S. Pown all, C.L. War ring ton, A. Borow ski, J.E. Dimmick, J. To one, F.R. Jir ik, Murine muco pol y sac cha ri do sis type I: tar geted dis rup tion of the murine a-l-idu ron i dase gene, Hum. Mol. Genet. 6 (1997) 503 511. [13] W.S. Sly, B.A. Quin ton, W.H. Mc Al is ter, D.L. Rim o in, Beta glu cu ron i dase deficiency: report of clin i cal radio logic and bio chem i cal fea tures of a new mucopol y sac cha ri do sis, J. Pe di atr. 82 (1973) 249 257. [14] R.D. Dick er man, K.O. Colle, C.A. Bruno Jr., S.J. Schnei der, Cra ni over te bral instabil ity with spinal cord com pres sion in a 17-month-old boy with Sly syn drome (muco pol y sac cha ri do sis type VII): a sur gi cal dilemma, Spine 29 (2004) E92 E94. [15] P.D. Piz zu tillo, J.A. Osterk amp, C.I. Scott, M.S. Lee, Atlan to ax ial insta bil ity in muco pol y sac cha ri do sis type VII, J. Pe di atr. Ort hop. 9 (1989) 76 78. [16] M.E. Has kins, R.J. De snick, N. DiF err an te, P.F. Jez yk, D.F. Patt er son, Beta-glu curon i dase defi ciency in a dog: a model of human muco pol y sac cha ri do sis VII, Pe di atr. Res. 18 (1984) 980 984. [17] K.P. Pon der, J.R. Mel niczek, L. Xu, M.A. Weil, T.M. O Mal ley, P.A. O Don nell, V.W. Knox, G.D. A gui rre, H. Maz ri er, N.M. Ell in wood, M. Sleeper, A.M. Magu ire, S.W. Volk, R.L. Mango, J. Zwei gle, J.H. Wolfe, M.E. Has kins, Ther a peu tic neo na tal gene ther apy in muco pol y sac cha ri do sis VII dogs, Proc. Natl. Acad. Sci. USA 99 (2002) 13102 13107. [18] M.S. Sands, C. Vo gler, J.W. Kyle, J.H. Grubb, B. Levy, N. Gal vin, W.S. Sly, E.H. Bir kenme ier, Enzyme replace ment ther apy for murine muco pol y sac cha ri do sis type VII, J. Clin. Invest. 93 (1994) 2324 2331. [19] R.L. Mango, L. Xu, M.S. Sands, C. Vo gler, G. Se il er, T. Sch warz, M.E. Has kins, K.P. Pon der, Neo na tal ret ro vi ral vec tor-med i ated hepatic gene ther apy reduces bone joint and car ti lage dis ease in muco pol y sac cha ri do sis VII mice and dogs, Mol. Genet. Metab. 82 (2004) 4 19. [20] C. Rus sell, G. Hend son, G. Jevon, T. Mat lock, J. Yu, M. Ak luj kar, K.Y. Ng, L.A. Clarke, Murine MPS I: insights into the path o gen e sis of Hurler syn drome, Clin. Genet. 53 (1998) 349 361. [21] C.M. Si mo na ro, M. D Angelo, M.E. Has kins, E.H. Schu ch man, Joint and bone dis ease in mu co poly sac chari dos es VI and VII: iden ti fi ca tion of new ther a- peu tic tar gets and bio mark ers using ani mal mod els, Pe di atr. Res. 57 (2005) 701 707. [22] C.M. Si mo na ro, M. D Angelo, X. He, E. Eli yahu, N. Shtraiz ent, M.E. Has kins, E.H. Schu ch man, Mech a nism of gly cos ami no gly can-med i ated bone and joint disease: impli ca tions for the mu co poly sac chari dos es and other con nec tive tis sue dis eases, Am. J. Pathol. 172 (2008) 112 122. [23] S.L. Sta ba, M.L. Esco lar, M. Poe, Y. Kim, P.L. Mar tin, P. Szabolcs, J. Al li son-thacker, W. Wood, D.A. Weng er, P. Ru bin stein, J.J. Hop wood, W. Kri vit, J. Ku rtz berg, et al., Cord blood trans plants from unre lated donors in patients with Hurler s syn drome, N. Engl. J. Med. 350 (2004) 1960 1969. [24] R.M. Shull, E.D. Kak kis, M.F. McEn tee, S.A. Ka nia, A.J. Jonas, E.F. Neu feld, Enzyme replace ment in a canine model of Hurler syn drome, Proc. Natl. Acad. Sci. USA 91 (1994) 12937 12941. [25] E.D. Kak kis, J. Mu en zer, G.E. Til ler, L. Wab er, J. Bel mont, M. Pas sage, B. Izykow ski, J. Phil lips, R. Doro show, I. Wa lot, R. Hoft, E.F. Neu feld, Enzymereplace ment ther apy in muco pol y sac cha ri do sis I, N. Engl. J. Med. 344 (2001) 182 188. [26] Y. Yam ada, K. Kato, K. S u k e g a wa, S. Tom a tsu, S. Fuk u da, S. Em ura, S. Koj ima, T. Ma tsuy ama, W.S. Sly, N. Kondo, T. Orii, Treat ment of MPS VII (Sly dis ease) by allo ge neic BMT in a female with homo zy gous A619V muta tion, Bone Mar row Trans plant. 21 (1998) 629 634. [27] M.S. Sands, J.E. Barker, C. Vo gler, B. Levy, B. Gwynn, N. Gal vin, W.S. Sly, E. Bir kenme ier, Treat ment of murine muco pol y sac cha ri do sis type VII by synge neic bone mar row trans plan ta tion in neo nates, Lab. Invest. 68 (1993) 676 686. [28] M. Sif u en tes, R. Doro show, R. Hoft, G. Mason, I. Wa lot, M. Di a ment, S. Oka zaki, K. Huff, G.F. Cox, S.J. Swie dler, E.D. Kak kis, A fol low-up study of MPS I patients treated with la ro ni dase enzyme replace ment ther apy for 6 years, Mol. Genet. Metab. 90 (2007) 171 180. [29] L. Xu, M.E. Has kins, J.R. Mel niczek, C. Gao, M.A. Weil, T.M. O Mal ley, P.A. O Don nell, H. Maz ri er, N.M. Ell in wood, J. Zwei gle, J.H. Wolfe, K.P. Pon der, Trans duc tion of hepa to cytes after neo na tal deliv ery of a Mo lo ney murine leu ke mia virus based ret ro vi ral vec tor results in long-term expres sion of beta-glu cu ron i dase in muco pol y sac cha ri do sis VII dogs, Mol. Ther. 5 (2002) 141 153. [30] A.M. Tra as, P. Wang, X. Ma, M. Tit ti ger, L. Sch al ler, P. O don nell, M.M. Sleeper, C. Vite, R. Her at i, G.D. A g u i r re, M.E. Has kins, K.P. Pon der, Cor rec tion of clin i cal man i fes ta tions of canine muco pol y sac cha ri do sis I with neo na tal ret ro vi ral vector gene ther apy, Mol. Ther. 15 (2007) 1423 1431. [31] B. Wang, T.M. O Mal ley, L. Xu, C. Vite, P. Wang, P.A. O Don nell, N.M. Ell inwood, M.E. Has kins, K.P. Pon der, Expres sion in blood cells may con trib ute to bio chem i cal and path o log i cal improve ments after neo na tal intra ve nous gene ther apy for muco pol y sac cha ri do sis VII in dogs, Mol. Genet. Metab. 87 (2006) 8 21. [32] L.E. Swis chuk, The beaked, notched, or hooked ver te bra: its sig nifi cance in infants and young chil dren, Radi ol ogy 95 (1970) 661 664. [33] K.P. Me non, P.T. Tieu, E.F. Neu feld, Archi tec ture of the canine IDUA gene and muta tion under ly ing canine muco pol y sac cha ri do sis. I, Genom ics 14 (1992) 763 768. [34] X. He, C.M. Li, C.M. Si mo na ro, Q. Wan, M.E. Has kins, R.J. De snick, E.H. Schuch man, Iden ti fi ca tion and char ac ter iza tion of the molec u lar lesion causing muco pol y sac cha ri do sis type I in cats, Mol. Genet. Metab. 67 (1999) 6 12. [35] J. Ray, A. Bou vet, C. DeS an to, J.C. Fyfe, D. Xu, J.H. Wolfe, G.D. A g u i r r e, D.F. Patt er son, M.E. Has kins, P.S. Hen thorn, Clon ing of the canine beta-glu curon i dase cdna, muta tion iden ti fi ca tion in canine MPS VII, and ret ro vi ral vec tor-med i ated cor rec tion of MPS VII cells, Genom ics 48 (1998) 248 253. [36] P. Go vin draj, L. West, T.J. Koob, P. Ne ame, K. Do e ge, J.R. Has sell, Iso la tion and iden ti fi ca tion of the major hep a ran sul fate pro te o gly cans in the devel op ing bovine rib growth plate, J. Biol. Chem. 277 (2002) 19461 19469. [37] A. Byers, B. Cat er son, J.J. Hop wood, B.K. Fos ter, Im mu no lo ca tion anal y sis of gly cos ami no gly cans in the human growth plate, J. His to chem. Cy to chem. 40 (1992) 275 282. [38] A.L. Han sen, B.K. Fos ter, G.J. Gib son, G.F. Binns, O.W. Wieb kin, J.J. Hop wood, Growth-plate chon dro cyte cul tures for reim plan ta tion into growth-plate defects in sheep. Char ac ter iza tion of cul tures, Clin. Ort hop. Relat. Res. 256 (1990) 286 298. [39] A.R. Poole, C. Web ber, I. Pi doux, H. Choi, L.C. Rosen berg, Local i za tion of a derma tan sul fate pro teo gly can (DS-PGII) in car ti lage and the pres ence of an immuno log i cally related spe cies in other tis sues, J. His to chem. Cy to chem. 34 (1986) 619 625. [40] T. Nak ano, J.S. Sim, A study of the chem i cal com po si tion of the prox i mal tib ial artic u lar car ti lage and growth plate of broiler chick ens, Poult. Sci. 74 (1995) 538 550. [41] M.S. Sands, E.H. Bir kenme ier, A sin gle-base-pair dele tion in the beta-glu cu roni dase gene accounts for the phe no type of murine muco pol y sac cha ri do sis type VII, Proc. Natl. Acad. Sci. USA 90 (1993) 6567 6571. [42] E.H. Bir kenme ier, M.T. Da vis son, W.G. Be amer, R.E. Gans chow, C.A. Vo gler, B. Gwynn, K.A. Ly ford, L.M. Malt ais, C.J. Wa wrzy niak, Murine muco pol y sac chari do sis type VII. Char ac ter iza tion of a mouse with beta-glu cu ron i dase deficiency, J. Clin. Invest. 83 (1989) 1258 1266. [43] K. Ohmi, D.S. Green berg, K.S. R aj a ve l, S. Rya zant sev, H.H. Li, E.F. Neu feld, Activated microg lia in cor tex of mouse mod els of mu co poly sac chari dos es I and IIIB, Proc. Natl. Acad. Sci. USA 100 (2003) 1902 1907. [44] M. Evers, P. Saf tig, P. Schmidt, A. Haf ner, D.B. McLogh lin, W. Sch mahl, B. Hess, K. von Fig ura, C. Peters, Tar geted dis rup tion of the aryl sul fa tase B gene results in mice resem bling the phe no type of muco pol y sac cha ri do sis VI, Proc. Natl. Acad. Sci. USA 93 (16) (1996) 8214 8219. [45] A.R. Gar cia, J. Pan, J.C. Lamsa, J. Mu en zer, The char ac ter iza tion of a murine model of muco pol y sac cha ri do sis II (Hunter syn drome), J. Inherit. Metab. Dis. 30 (2007) 924 934. [46] M. Bha umik, V.J. Mulle, T. Ro zak lis, L. John son, K. Do bre nis, R. Bhattacharyya, S. Wur zel mann, P. Fin a more, J.J. Hop wood, S.U. Walk ley, P. Stan ley, A mouse