Proc. Natl Acad. Sci. USA Vol. 80, pp. 3074-3077, May 1983 Medical Sciences

Immunological evidence for deficiency in an activator protein for sulfatase in a variant form of metachromatic (crossreacting material/rocket immunoelectrophoresis/GM1 1&galactosidase/arylsulfatase A/lysosomal storage disease) Koji INUI*, MERRILL EMMETTt, AND DAVID A. WENGER*t *Department of Pediatrics and tWebb-Waring Lung Institute, University of Colorado Health Sciences Center, Denver, Colorado 80262 Communicated by David W Talmage, February 14, 1983 ABSTRACT Cultured skin fibroblasts from the patient de- tients with normal hexosaminidase A and B activity toward syn- scribed by Shapiro and co-workers as having a variant form of thetic substrates and normal levels of activator protein but with metachromatic leukodystrophy (MLD) [Shapiro, L. J., Aleck, K. an inability of activator protein to stimulate their hexosamin- A., Kaback, M. M., Itabashi, H., Desnick, R. J., Brand, N., Ste- idase A. Christomanou (8) reported that some patients with a phens, R. L., Fluharty, A. L. & Kihara, H. (1979) Pediatr. Res. 13, juvenile form of Niemann-Pick disease may be missing a re- 1179-1181] were confirmed to have a partial deficiency (25-40% quired activator protein. of controls) of arylsulfatase A activity in vitro and a severe ina- Shapiro et al. (11) described a patient with a variant form of bility to metabolize ['4C]stearic acid-labeled sulfatide presented metachromatic leukodystrophy (MLD) who had only partially in the medium. When 150 ,pg of purified activator protein for GM1 A to the ganglioside (3-galactosidase and sulfatide sulfatase was added in deficient arylsulfatase activity (as opposed 0-10% of 4 ml of medium with the '4C-labeled sulfatide, correction of the normal activity measured in other MLD patients), but whose sulfatide metabolism to the normal range was found. Monospe- cultured cells could metabolize only a small percent of the ra- cific antibodies to this activator protein were prepared in rabbits, diolabeled sulfatide presented in the medium. This small amount and they were used to examine cultured cells for the presence of of sulfatide metabolism could be corrected by the addition of crossreacting material by Ouchterlony double immunodiffusion sulfatide sulfatase activator protein (12). and rocket immunoelectrophoresis. Cell extracts from controls and In this study we demonstrate that sulfatide metabolism in the from patients with GM1 gangliosidosis and MLD were found to cultured cells from the patient of Shapiro et aL (11) can be cor- have a single line of identity. By comparison to known concen- rected to normal by presentation of the activator protein for trations of purified activator protein, cell extracts from controls GM1 ganglioside /3-galactosidase. In addition, by using mono- were found to have 0.76 ± 0.32 ,ig of activator protein (mean + specific antibodies to this activator protein the levels in cul- 1 SD, n = 10) per mg of solubilized protein, whereas those from tured human cells could be quantitated. Extracts from the cells patients with type 1 GM1 gangliosidosis and late infantile MLD of this patient were found to have little or no crossreacting ma- had 1.53 and 1.41 ,ug/mg, respectively. Cell extracts from the pa- terial (CRM) by Ouchterlony double immunodiffusion or rocket tient with a variant form of MLD had no visible precipitin line by immunoelectrophoresis. These studies indicate that the acti- Ouchterlony double immunodiffusion and only a diffuse nonspe- vator proteins for sulfatide sulfatase and GM1 ganglioside 3-ga- cific region of staining by rocket immunoelectrophoresis. These lactosidase may be identical, and that a deficiency of this ac- immunologic studies provide evidence for a deficiency in the ac- tivator protein can be directly detected by using monospecific tivator protein required for normal catabolism of sulfatide in the antibodies. This could lead to another method for diagnosing cells from this patient and possibly provide a method for diagnosis this type of patient with cultured skin fibroblasts. of similar patients. During the past 15 years a number of relatively low molecular METHODS weight proteins that stimulate the enzymatic hydrolysis of Fibroblast Cultures. Skin fibroblast cultures from controls, sphingolipids have been identified (1-6). These include the ac- from patients with late infantile MLD, I-cell disease, and GM1 tivator of sulfatide sulfatase (1), the activator of glucosylcer- gangliosidosis type 1, and from a healthy child with arylsulfat- amide f3-glucosidase and ,3-galactosidase (2, ase A deficiency (pseudodeficient) were grown from forearm 6), the activator that was shown to stimulate GM1 ganglioside skin biopsy samples taken in this laboratory. The cell line from P-galactosidase, trihexosylceramide a-galactosidase, and GM2 the patient with the variant form of MLD was supplied by M. ganglioside ,B-hexosaminidase (5), and the activators that ap- M. Kaback. All cell lines were grown in Eagle's minimal es- pear to stimulate only GM2 ganglioside ,l3hexosaminidase A (3, sential medium supplemented with 10% fetal calf serum, non- 4). The specificities of these activators toward lipids and en- essential amino acids, Penstrep (Flow Laboratories), and glu- zymes have not been adequately defined. tamine (2 mM). Cells were harvested when confluent by using The role these activator proteins play in the etiology of cer- trypsin/EDTA and after washing they were frozen at -20°C tain genetic has been recently described (4, 7- until homogenized for in vitro enzymatic assays and for im- 10). Conzelmann and Sandhoff (4) and Hechtman et aL (7) have munologic studies of activator protein. Reactions were carried described patients who store GM2 ganglioside due to a missing out in vitro as described (13). For immunologic studies, the pel- activator protein for hexosaminidase A. Recently, Li et aL (9) and this laboratory (10) described other GM2 gangliosidosis pa- Abbreviations: MLD, metachromatic leukodystrophy; CRM, crossre- acting material; GM1 ganglioside, Gal,31--+3GalNAcj31--+4Gal- The publication costs ofthis article were defrayed in part by page charge (3E-2aNeuAc),B1--4Glc-; GM2 ganglioside, GalNAc31-> payment. This article must therefore be hereby marked "advertise- 4Gal(3E-2aNeuAc)31--'4Glc-ceramide. ment" in accordance with 18 U.S.C. §1734 solely to indicate this fact. * To whom reprint requests should be addressed. 3074 Downloaded by guest on September 27, 2021 Medical Sciences: Inui et al. Proc. Natl. Acad. Sci. USA 80 (1983) 3075

lets of cells were homogenized in small Duall homogenizers noelectrophoresis was carried out with crude extracts of human (Kontes) in distilled water and centrifuged at 11,000 X g for 15 liver and fibroblast cultures, using the method of Emmett and min. The supernatant was lyophilized, and the dry extract was Crowle (14). Only one major loop of staining was found, al- dissolved in 0.01 ml of immunoelectrophoresis buffer (14). In- though a faint second loop was seen with a similar mobility. This soluble material was removed by centrifugation at 10,000 x g could be another form of the activator protein with a slightly for 5 min. The protein concentrations were determined by us- different isoelectric point (16). The supernatant fraction from ing the method of Lowry et aL (15) with bovine serum albumin fibroblast extracts was subjected to rocket immunoelectropho- as the standard. resis (14, 17). Between 20 and 40 tkg of protein in 1 1ul was ap- 1'4C]Stearic Acid-Labeled Sulfatide Uptake Experiments. plied to the serum-containing gel with known amounts of pu- [1- 4C]Stearic acid-labeled sulfatide was prepared as described rified activator protein. After electrophoresis at 90 V for 2.5 hr (13). The purity was determined to be 99.0%. The medium from the washing and staining procedures of Emmett and Crowle a 25-cm2 flask of confluent cells was replaced with 4 ml of com- (14) were followed. The heights of the loops from the cell ex- plete medium containing 60 nmol of 14C-labeled sulfatide (13). tracts were compared to the heights of the loops from known In some experiments the purified activator protein was in- amounts of purified activator protein. The amount of activator cluded in the complete media along with the '4C-labeled sul- protein in cells is expressed as,ug per mg of protein in the cell fatide. In these experiments the medium containing activator extract supernatant. protein was passed through a 0.22-gim-pore-diameter Millipore filter before addition to the labeled sulfatide. The medium con- 'RESULTS taining the labeled sulfatide and 0-300 ug of purified activator The cells from the patient with the variant form of MLD had protein was left on the cells for 3 days, then the cells were washed a partial deficiency of arylsulfatase A activity when measured and harvested as described (13). The lipids were extracted with with nitrocatechol sulfate and sulfatide (assayed in the presence chloroform/methanol (2:1, vol/vol) and separated by thin-layer of sodium taurodeoxycholate) (13). The values were not as low chromatography, and the radioactive lipids were located after as those found in other patients with MLD or in healthy people exposure to x-ray film (Kodak XR-5) (13). The radioactive re- with low arylsulfatase A (pseudodeficient people), who have 8- gions were scraped from the plate into vials and the radioac- 15% of controls. However, the cells from this patient could me- tivities were measured. The results are expressed as a per- tabolize only 15-20% of the 14C-labeled sulfatide taken up from centage of the labeled sulfatide left unhydrolyzed after uptake the medium (ref. 13; Fig. 1) by day 3, compared to 85-90% into the cultured cells. metabolized by control cells and only 1-2% metabolized by cells Preparation of Activator Protein. The activator protein orig- from patients with late infantile MLD. inally demonstrated to stimulate the enzymatic hydrolysis of As shown on Fig. 1, the intracellular metabolismof '4C-la- GM1 ganglioside byGM1 /3-galactosidase was purified from beled sulfatide was corrected to the normal range by the ad- liver samples from a control and from a patient withGM1 gan- dition of 150pug of activator protein to the medium. Activator gliosidosis, type 1 (5). As described (16), the liver from the pa- protein had no effect on metabolism of 14C-labeled sulfatide tient withGM1 gangliosidosis had about 35 times more acti- when added to the medium of cells from controls and patients vator protein than control human liver did. They both gave only with late infantile MLD. This finding was evidence that the ac- one band on polyacrylamide gel electrophoresis in the presence tivator protein prepared originally as an activator ofGMLI3-ga- and absence of NaDodSO4. The specific activity of the acti- lactosidase also activated the enzymatic hydrolysis of sulfatide vator protein is 9.59 units/mg of protein for purified human and that this patient was, probably, deficient in normal acti- liverGM1 ganglioside 3-galactosidase and 4.17 units/mg of vator protein needed for metabolism of 14C-labeled sulfatide protein for purified human liver sulfatide sulfatase. A unit is taken up from the medium. However, direct evidence for a de- defined as the amount of activator that stimulates the hydro- fect in the structure or function of this activator protein in this lysis of 1 nmol of substrate per min per unit of enzyme. A unit of enzyme is defined as the amount that catalyzes the hydro- - lysis of 1 nmol of natural substrate per min at37°C in the pres- 100 ence of bile salts. to Activator Protein. Preparation of Antibodies Each of two e- 80- young female rabbits was injected with purified activator pro- N tein fromGM1 gangliosidosis liver emulsified in Freund's in- 600 complete adjuvant; ,g10 was injected into each shoulder re- 60- 40 gion. After21/2 weeks another 20ug was injected into the 0 haunches in the same manner. By 4 weeks after the initial in- "0 40- jection precipitin lines were detected by Ouchterlony double 10 immunodiffusion. The rabbits were boosted with,ug20 of ac- *2: tivator protein and 1 week later blood was collected and serum 2u was prepared. In some experiments a crude IgG fraction was made by addition of an equal volume of saturated ammonium sulfate and collection of the precipitate. No precipitin lines were 50 100 150 200 250 300 detected with the preimmunization rabbit serum. Activator protein, Ag/4 ml of medium Studies with Antibodies to the Activator Protein. Ouchter- lony double immunodiffusion was carried out in 1.2% agar gels FIG. 1. Effect of activator protein on the hydrolysis of'4C-labeled prepared in sodium phosphate buffer (0.15 M, pH 7.4). Wells sulfatide taken up by the cultured cells from the patient with a variant were made 5 mm apart, by using a template, and 59-110,ug form of MLD. The flasks ofcells were cultured until confluent, and the offibroblast extract protein was added around a central well media were replaced with complete media containing[140]stearic acid- labeled sulfatide and 0,50, 150, or 300ug of purified activator protein containing antibody. After 48 hr at 40C the gels were washed, per 4 ml. After 3 days the cells were harvested and the lipid was ex- dried, and stained for protein by using Crowle's double stain tracted. The hatched region is the mean ± 1 SD of sulfatide metabolism (14). To check the specificity of the antibody, crossed immu- by control cells (n = 7). Downloaded by guest on September 27, 2021 3076 Medical Sciences: Inui et al Proc. Nad Acad. Sci. USA 80 (1983) had no sharp loop, but only a diffuse, nonspecific region of pro- 1 2 tein staining. With peak heights as indicators of activator pro- tein concentration, controls (n = 10) were estimated to have 0.76 ± 0.32 pug/mg of protein in the soluble fraction of the cell extract. The concentrations of activator protein in the cell ex- 6 3 tracts from the patients with type 1 GM1 gangliosidosis and late infantile MLD were calculated to be 1.53 and 1.41 ,ug/mg of solubilized protein, respectively. It was estimated that the cell 5 4 extract from the patient with the variant type of MLD had less than 5% of the normal amount of CRM, indicating either a se- vere decrease in the concentration of antigenically normal ac- FIG. 2. Ouchterlony double immunodiffusion of extracts from cul- tured skin fibroblasts and the crude IgG fraction from antiserum to ac- tivator protein or a structurally altered activator protein that tivator protein. In the center well is placed the crude IgG fraction. The had lost a major portion of its antigenicity. cell lines and protein concentrations are as follows: 1, control (78 pg of protein); 2, late infantile MLD (100 Mig of protein); 3, pseudodeficient MLD (110 jig ofprotein); 4, I-cell disease (86 pg ofprotein); 5, GM1 gan- DISCUSSION gliosidosis, type 1 (59 ug of protein); 6, patient with a variant form of Cultured skin fibroblasts from the patient said to have a variant MLD (78 pg of protein). form of MLD without arylsulfatase A deficiency (11) were stud- ied. This patient and a sibling had clinical findings consistent patient's cells or tissues has not been reported. Because tissues with a genetic neurodegenerative disease with onset about 41/I2 are not available from this patient (or a similarly affected sib- 6 years of age. Enzyme studies in vitro on the cultured cells by ling) isolation and characterization could not be carried out. Shapiro et aL (11) and this laboratory (13) revealed only a partial Activators of sphingolipid hydrolases have not been identi- deficiency in arylsulfatase A activity with natural and synthetic fied or isolated from cultured skin fibroblasts. Using mono- substrates but a severe inability to degrade either 'S-labeled specific antibodies to this activator protein, we examined cell or [14C]stearic acid-labeled sulfatide given in the medium to the extracts from cultured skin fibroblasts for the presence of CRM. cultured cells. This indicates either that the patient has a more Ouchterlony double immunodiffusion revealed a single pre- severe defect in the arylsulfatase A than indicated by in vitro cipitin line for extracts from controls, a pseudodeficient healthy testing or that the patient may have a defect in an activator pro- child, and patients with MLD and GM1 gangliosidosis (Fig. 2). tein required for reaction between the lipid substrate and the However, cell extracts from the patient with a variant form of enzyme (1). When Stevens et aL (12) gave the cultured cells MLD did not demonstrate a precipitin line corresponding to purified sulfatide sulfatase activator (18) along with the 3S-la- that found in extracts from controls and individuals with ly- beled sulfatide in the medium its metabolism was corrected to sosomal storage diseases. It was either not seen at all or was near normal. This indicated, but did not prove, that a defect in possibly seen as a very faint line close to the well where the a required activator protein was responsible for the decreased extract was placed. In addition, only a very faint line was ob- catabolism of sulfatide in this patient. However, as stated by served in the cell extract from the patient with I-cell disease. Stevens et aL (12), a demonstration of the absence of activator Rocket immunoelectrophoresis was carried out in the pres- is lacking because large samples of tissues are needed and none ence of known concentrations of purified activator protein (Fig. are available. 3). Extracts from cells from controls and from patients with MLD We purified an activator protein from human liver according and GM1 gangliosidosis had distinct loops of precipitation. The to the procedure of Li and Li (5). It was purified to study its extract from the cells of the patient with a variant form of MLD action on GM1 ganglioside f3-galactosidase in normal and patho- logical tissues. It was also found to stimulate the reaction be- tween sulfatide and purified arylsulfatase A, and to be present in a 35-fold excess in liver samples from patients with GM1 gan- gliosidosis (ref. 16; unpublished data). Although statements re- garding the possible identity of the activators of GM1 (3-ga- lactosidase and sulfatide sulfatase have been made (19), no experiments have been published that prove it. Jatzkewitz (20) published the amino acid composition of the sulfatide sulfatase activator and found it to be similar to that of the activator of GM1 f3-galactosidase (5). Peters et al. (21) reported that the ac- tivator of glucosylceramide 83-galactosidase could also stimulate GM1 f3-galactosidase, but we found it could not (6). When 150 ug of activator protein was added in 4 ml of me- dium with 14C-labeled sulfatide to the cells from the patient with the variant form of MLD, correction of sulfatide metab- 1 2 "3 Ta r5 A 6 olism was found (Fig. 1). This is significantly less than the 686 ,ug per 3.5 ml of medium needed by Stevens et aL (12), using FIG. 3. Rocket immunoelectrophoresis of purified activator pro- purified sulfatide sulfatase activator prepared according to Fischer tein and extracts from cultured skin fibroblasts. Extracts were pre- and Jatzkewitz (18). Although both activator proteins were pu- pared and dissolved in electrophoresis buffer. Whole immune rabbit rified from human liver the methods of preparation are dif- serum was used. 1, Purified activator protein (0.038 ug of protein); 2, ferent. However, this study provides evidence that they may purified activator protein (0.076 jpg of protein); 3, cell extract from the be one and the same protein at different stages of purity. of MLD of cell extract patient with the variant form (28 pig protein); 4, to this activator were from a control (37 pg ofprotein); 5, cell extract from a patient with late Monospecific antibodies protein pre- infantile MUD (31 uig of protein); 6, cell extract from a patient with GM1 pared and they were used to estimate the levels in cultured skin gangliosidosis, type 1 (40 ug of protein). fibroblasts. Rocket immunoelectrophoresis produced quanti- Downloaded by guest on September 27, 2021 Medical Sciences: Inui et aL Proc. Natl. Acad. Sci. USA 80 (1983) 3077

tative evidence for a severe deficiency in antigenically normal the patient used in this study. This research was supported in part by activator protein in this patient (Fig. 3). Cell extracts from the grants from the National Institutes of Health (HD 08315, NS 10698, patients with type 1 GM1 gangliosidosis and late infantile MLD and HD 10494). had more CRM than cell extracts from most controls. It is not 1. Mehl, E. & Jatzkewitz, H. (1964) Hoppe-Seyler4s Z. Physiol Chem. yet certain whether this apparent lack of CRM to activator pro- 339, 260-276. tein in this patient reflects a severe structural mutation af- 2. Ho, M. W. & O'Brien, J. S. (1971) Proc. Natl Acad. Sci. USA 68, fecting its function and antigenicity or a lack of production of 2810-2813. quantities sufficient for normal sphingolipid metabolism. 3. Hechtman, P. & LeBlanc, D. (1977) Biochem. J. 167, 693-701. 4. Conzelmann, E. & Sandhoff, K. (1978) Proc. Natd Acad. Sci. USA These studies raise other questions regarding this type of 75,3979-3983. patient and the specificity of this activator protein. Why do the 4. Li, S! C. & Li, Y. T. (1976) J. Biol Chem. 251, 1159-1163. cells from the patient have a partial reduction (40% of controls) 6. Wenger, D. A., Sattler, M. & Roth, S. (1982) Biochim. Biophys. in arylsulfatase A activity measured in vitro if the defect is in Acta'712, 639-649. the activator protein? It is possible that the activator protein 7. Hechtman, P., Gordon, B. A. & Ng Ying Kin, N. M. K. (1982) binds to and stabilizes the catalytic protein in the lysosomal en- Pediatr. Res. 16, 217-222. 8. Christomanou, H. (1980) Hoppe-Seyler's Z. Physiol Chem. 361, vironment. The activator protein ,can bind sulfatide and GM1 1489-1502. ganglioside (unpublished data), but it may also bind enzyme 9. Li, S. C., Hirabayashi, Y. & Li, Y. T. (1981) Biochem. Biophys. with or without substrate present. Why are most of the symp- Res. Commun. 101, 479-485. toms observed in these patients related to a defect in sulfatide 10. Inui, K., Grebner, E. E., Jackson, L. G. & Wenger, D. A. (1983) metabolism and not a defect in GM1 ganglioside metabolism if, Am. J. Hum. Genet., in press. in fact, the same activator protein is required for in vivo stim- 11. Shapiro, L. J., Aleck, K. A., Kaback, M. M., Itabashi, H., Des- nick, R. J., Brand, N., Stevens, R. L., Fluharty, A. L. & Kihara, ulation of both lysosomal hydrolases? Because these patients, H. (1979) Pediatr. Res. 13, 1179-1181. and possibly the one recently described by Hahn et aL (22, 23), 12. Stevens, R. L., Fluharty, A. L., Kihara, H., Kaback, M. M., have a juvenile age of onset of symptoms we may assume that Shapiro, L. J., Marsh, B., Sandhoff, K. & Fischer, G. (1981) Am. a small but significant level of functional activator protein is J. Hum. Genet. 33, 900-906. present. This may be sufficient to handle the slow rate of gan- 13. Kudoh, T. & Wenger, D. A. (1982)J. Clin. Invest. 70, 89-97. glioside turnover but not sulfatide metabolism during the pe- 14. Emmett, M. & Crowle, A. J. (1981) Clin. Immunol Newslett. 2, 99-103. riod of rapid myelination in the nervous system. It will be in- 15. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. teresting to examine the brain ganglioside composition in these (1951)J. Biol Chem. 193, 265-275. patients when, and if, they become available. It is also inter- 16. Inui, K. & Wenger, D. A. (1982) Biochem. Biophys. Res. Com- esting that seizures were a common feature of these patients mun. 105, 745-751. and this tends to be less common in cases of MLD and more 17. Laurell, C.-B. (1972) Scand. J. Clin. Lab. Invest. Suppl. 124, 29, common in disorders of ganglioside metabolism. 21-37. 18. Fischer, G. & Jatzkewitz, H. (1978) Biochim. Biophys. Acta 528, These experiments provide direct evidence for a defect in an 69-76. activator protein required for the metabolism of sulfatide and 19. Li, Y. T. & Li, S. C. (1981) in Lysosomes and Lysosomal Storage GM1 ganglioside in cultured cells from a patient diagnosed as Diseases, eds. Callahan, J. W. & Lowden, J. A. (Raven, New York), having a variant form of MLD. It will be interesting to study pp. 173-180. other patients with proposed defects in activator proteins by 20. Jatzkewitz, H. (1978) in Enzymes of Lipid Metabolism, eds. Gatt, this method. The use of monospecific antiserum will allow for S., Freysz, L. & Mandel, P. (Plenum, New York), pp. 561-571. 21. Peters, S. P., Coffee, C. J., Glew, R. H., Lee, R. E., Wenger, rapid diagnosis of such patients by using cultured cells and small D. A., Li, S. C. & Li, Y. T. (1977) Arch. Biochem. Biophys. 183, tissue samples and will allow for further studies to define the 290-297. role and specificity of activator proteins. 22. Hahn, A. F., Gordon, B. A., Feleki, V., Hinton, G. G. & Gilbert, J. J. (1981) Ann. Neurol 12, 33-36. The authors thank Ms. Harriet McKelvey for her help growing the 23. Hahn, A. F., Gordon, B. A., Gilbert, J. J. & Hinton, G. G. (1981) cultured cells and Dr. M. M. Kaback for supplying the cell line from Acta Neuropathol. 55, 281-287. Downloaded by guest on September 27, 2021