THE JOURNAL OF BIOLOGICAL CHEMISTRY Q 1992 hy The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 267, No. 18, Issue of June 25, pp. 12552-12558,1992 Printed in U.S.A.

Identification and Expression of Five Mutations in the Human Acid Sphingomyelinase Gene Causing Types A and B Niemann-Pick Disease MOLECULAR EVIDENCE FOR GENETIC HETEROGENEITY IN THE NEURONOPATHIC AND NON- NEURONOPATHIC FORMS*

(Received for publication, November 26, 1991)

Tsutomu Takahashi$$, Mariko SuchiS, Robert J. DesnickS, Goro Takadag, and Edward H. SchuchmanST From the $Division of Medical and Molecular Genetics, Mount Sinai School of Medicine, New York, New York 10029 and the §Department of Pediatrics, Akita University School of Medicine, Akita 010, Japan

The deficient activity of the human lysosomal hydro- lase, acid sphingomyelinase (ASM, EC 3.1.4.12), re- sults in the neuronopathic (Type A) and non-neurono- pathic (Type B) forms of Niemann-Pick disease (NPD). To investigate the genetic basis of the phenotypic het- erogeneity in NPD, the molecular lesions in the ASM gene were determined from three unrelated NPD pa- tients and evaluated by transient expression in COS-l cells. A Type A NPD patient of Asian Indian ancestry (proband 1) was homoallelic for a T to A transversion in exon 2 of the ASM gene which predicted a premature stop at codon 261 of the ASM polypeptide (designated L261X). In contrast, an unrelated Type A patient of European ancestry (proband 2) was heteroallelic for a two-base (TT) deletion in exon 2 which caused a frame- shift mutation at ASM codon 178 (designated fsL178), leading to a premature stop at codon 190, and a G to A transition in exon 3 which caused a methionine to isoleucine substitution at codon 382 (designated M382I). Transient expression of the fsL178, L261X, and M382I mutations in COS-1 cells demonstrated that these lesions did not produce catalytically active ASM, consistent with the severe neuronopathic Type A NPD phenotype. In contrast, an unrelated Type B patient of European descent (proband 3) was heteroallelic for two missense mutations, a G to A transition in exon 2 which predicted a glycine to arginine substitution at ASM codon 242 (designated G242R), and an A to G transi- tion in exon 3 which resulted in an asparagine to serine substitution at codon 383 (designated N383S). Inter- estingly, the G242R allele produced ASM activity in COS-1 cells at levels about 40% of that expressed by the normal allele, thereby explaining the mild Type B phenotype of proband 3 and the high residual activity (Le. -15% of normal) in cultured lymphoblasts. In contrast, the N383S allele did not produce catalytically active enzyme. None of these five ASM mutations was detected in over 60 other unrelated NPD patients ana-

* This work was supported by Basic Research Grant 1-1224 from the March of Dimes Birth Defects Foundation, Basic Investigation Grant CD-62521 from the American Cancer Society, Research Grant 1 R 0 1 HD28607 from the National Institutes of Health, and Grant 5 M 0 1 RR0071 for the Mount Sinai General Clinical Research Center from the National Center for Research Resources, National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 To whom all correspondence should be addressed: Division of Medical and Molecular Genetics, Mount Sinai School of Medicine, 100th St. and 5th Ave., New York, NY 10029.

lyzed, nor were these mutations found in over 100 normal ASM alleles. Thus, small deletions or nonsense mutations which trunctated the ASM polypeptide, or missense mutations that rendered the enzyme noncat- alytic, resulted in Type A NPD disease, whereas a missense mutation that produced a defective enzyme with residual catalytic activity caused the milder non- neuronopathic Type B phenotype. These findings have facilitated genotype/phenotype correlations for this ly- sosomal storage disease and provided insights into the functional organization of the ASM polypeptide.

Types A and B Niemann-Pick disease (NPD)l result from the deficient activity of the lysosomal hydrolase, acid sphin- gomyelinase (ASM, EC 3.1.4.12) (1, 2). Type A NPD is a debilitating neurodegenerative disorder of infancy character- ized by failure to thrive, severe hepatosplenomegaly, and progressive neurologic deterioration which generally leads to death by 3 years of age. In contrast, Type B NPD patients have hepatosplenomegaly, no neurologic involvement, and often survive into adulthood (3). Both types of NPD are panethnic; however, they occur more frequently among indi- viduals of Ashkenazi Jewish descent than in the general population (4).

Various artificial and natural substrates have been used to determine the ASM activities in cultured cells and fluids obtained from NPD patients (e.g. Refs. 5-8). In general, Type A patients have less than 5% of normal ASM activity in cultured fibroblasts and/or lymphoblasts when determined in vitro (5 , 6) or in situ (7, 8), whereas Type B patients have from 3 to 10% of normal levels. However, at present it is not possible to reliably predict the subtype or clinical severity of newly diagnosed NPD patients by the level of residual ASM activity. In addition, carrier detection for NPD by enzymatic assay has proven problematic, particularly for carriers of Type B NPD mutations who often have higher residual ASM activities (9). To date, all of the Type A and B NPD patients studied have had cross-reactive immunologic material in crude cell homogenates (10, l l ) , but accurate estimates of the amount of residual enzyme protein have not been made due to the lack of high-titer monospecific antibodies.

Based on the fact that Types A and B NPD patients have remarkably distinct phenotypes, as well as the fact that the

The abbreviations used are: NPD, Niemann-Pick disease; ASM, acid sphingomyelinase; bp, base pair(s); PCR, polymerase chain re- action; PSAll, N-(11-[1-pyrenesulfonyl]amido dodecanoyl); ASOs, allele-specific oligonucleotides.

12552

Acid Sphingomyelinme Gene Mutations 12553

clinical severity among unrelated Type B patients may vary dramatically, investigators have postulated that different mu- tations within the ASM gene are responsible for the variability in this disease (e.g. Ref. 3). The recent isolation of the full- length cDNA encoding human ASM (12, 13) has facilitated identification of the first two mutations resulting in Types A and B NPD (14, 15). In these studies, a severely affected Ashkenazi Jewish Type A patient was found to be homoallelic for an arginine to leucine substitution in residue 496 of the ASM polypeptide (designated R496L). Notably, this mutation occurred in over 30% of the 36 ASM alleles analyzed from other Ashkenazi Jewish Type A NPD patients, while it was detected in only one of 18 non-Jewish Type A patients (14). In addition, the R496L mutation was present as one allele in a mildly affected Ashkenazi Jewish Type B patient (15). The other ASM allele in this Type B patient had a three-base deletion which resulted in the removal of an arginine residue from position 608 of the ASM polypeptide (designated AR608). Although both of the Ashkenazi Jewish Type B patients analyzed were heteroallelic for AR608, this mutation was not identified in 15 other unrelated non-Jewish Type B patients.

In this study, the molecular nature of the ASM lesions causing Types A and B NPD in three unrelated non-Jewish patients was investigated. Five new NPD mutations were identified and characterized by transient expression in COS- 1 cells, providing valuable insights into the molecular basis of the neuronopathic and non-neuronopathic NPD phenotypes and facilitating further investigations of the structure and function of this lysosomal enzyme.

EXPERIMENTAL PROCEDURES

Cell Lines-NPD cell lines GM00112A, GM00165, GM00370, GM00406, GM00559, GM0933, GM02895, and GM03252 were ob- tained from the Human Genetic Mutant Cell Repository (Camden, NJ) . Cell lines 444X.F01,534R.F03,556X.F01,888V.F01,2789X.F01, 4293Q.EO2, 4774Z.FO1, 5113C.LO1, 5115E.FO1, and 6791M.FO1 were obtained from the Service de Biochimie, Hospice de Lyon (Lyon, France). Cell lines DMN 83.126, DMN 84.87, DMN 84.135, DMN 86.49, DMN 87.71, DMN 87.99, DMN 88.12, DMN 88.9, and RNS were provided by Dr. Peter Penchev, National Institutes of Health. Cell line MS1 (proband 1) was obtained from Dr. Emannuel Shapira (Tulane University School of Medicine). For each of the other cell lines (including those from probands 2 and 3), primary cultures of fibroblasts and/or lymphoblasts were established from skin biopsies and/or peripheral blood obtained with informed consent from NPD patients and family members, and from normal individuals. The cells were grown by standard procedures in RPMI 1640 (lymphoblasts) or Dulbecco’s modified Eagle’s media (fibroblasts) supplemented with 10% fetal bovine serum, 1% penicillin, and 1 mg/ml streptomycin (16). The diagnosis of Types A and B NPD was based on clinical criteria (e.g. age a t onset, presence of neurologic involvement, etc.) and by the demonstration of markedly deficient ASM activity in cultured cells (17).

Clinical Findings-Proband 1 was a male Type A NPD patient from Southern India. He presented with severe visceromegaly and progressive neurologic degeneration which led to his death at 3 years of age. Although cultured fibroblasts were established from this patient, no other family members were available for analysis. Proband 2 was a female Type A NPD patient of English, Italian, and German ancestries who had severe neurologic degeneration which led to her demise at 4 years of age. Proband 3 is a mildly affected female Type B NPD patient of English, German, and French ancestries. She was last seen 2 years ago, at which time she was 23 years old and had mild hepatomegaly and pulmonary involvement. Proband 3 also has a sister with Type B NPD who is 15 years old. Neither sibling has neurologic impairment.

Identification of NPD Mutations-To identify the ASM mutations in each proband, genomic DNA was isolated by standard procedures (18) and the complete ASM coding region was PCR amplified in four overlapping fragments. PCR product 1 (469 bp) was amplified using sense and al.tisense primers P1 (5”ATCATCCTCGAGACGGGA-

CAGACGAACCA-3‘) and P2 (5”ATCATCGAATTCATCAGGG- ATGCATTCT-3’), respectively. The sense and antisense primers for PCR product 2 (391 bp) were P3 (5”ATCATCGAATTCTGA- TTTCTCACCATGC-3’) and P4 (5’-ATCATCCTCGAGCCCTCCA- AGGTAGTCATG-S’), respectively. PCR product 3 (561 bp) was amplified using sense and antisense primers P5 (5”TCATCC- TCGAGCACTGACCTGCACTGGG-3’) and P6 (5”ATCATCCTC- GAGCTTTCTTTTCCCTTCCTG-3’), respectively, and the sense and antisense primers for PCR product 4 (1708 bp) were P7 (5’- ATCATCGAATTCTGGCTAGACTGTGAGC-3’) and P8 (5’-AT- CATCGAATTCTAAACCAGCTCCAGGA-3’), respectively. The primers were synthesized on an Applied Biosystems (Foster City, CA) Model 380B DNA Synthesizer using phosphoramidite chemistry (19). PCR amplification was performed for 30 cycles with Taq polymerase (Promega, Madison, WI) as previously described (20). For PCR product 2, a PCR “boost” procedure (14) was required to achieve efficient amplification. Following PCR amplification, the products were isolated from agarose gels and subcloned into the pGEM7Zf(+) vector (Promega, Madison, WI). For each amplified product, from four to eight independent subclones were sequenced by the dideoxy method (21). Computer analyses were performed using the University of Wisconsin Genetics Computer Group DNA Sequence Analysis Software (version 7.0).

Dot-blot Hybridization Analyses-In order to confirm the authen- ticity of the candidate mutations, as well as to determine their frequencies among other NPD patients, their family members, and normal unrelated individuals, dot-blot hybridization analyses were performed. The fsL178 mutation was detected by analysis of PCR product 2, whereas the G242R and L261X mutations were detected in PCR product 3. For dot-blot analysis of the M382I and N383S mutations, sense and antisense PCR primers P9 (5”CAGGA- TTGGAACAAGTG-3’) and P10 (5’-AAGGCTACAGCCAGCGG-3’) were used to amplify an 830-bp ASM genomic fragment, designated PCR product 5.

Dot-blot hybridization was performed using Zetabind nylon mem- branes (AMF-Cuno, Meriden, CT) and a Bio-Rad dot-blot apparatus (Richmond, CA). Hybridizations of the PCR products with normal or mutant allele-specific oligonucleotides (ASOs) were performed for a t least 3 h a t 39 “C. The ASOs were 5”end-labeled with T4 poly- nucleotide kinase and [-y-”’P]ATP (>5000 Ci/mmol) by standard procedures (18). The sequences of the ASOs were 5’- CATCTCTTTGCCTACTG-3’ and 5”AACATCTCTGCCTACTG- 3’, for L178and fs178, respectively; 5”CAGGTGCCGGATACTGG- 3’ and 5‘-CAGGTGCCAGATACTGG-3‘, for G242 and R242, respec- tively; 5’-GAGCCTGTTGAGTGGGC-3’ and 5’-GAGCCTGT& GAGTGGGC-3’ for L261 and X261, respectively; 5”CTCAATAT-

and 1382, respectively, and 5’-CAATATGAATTTTTGTT-3’ and 5’- CAATATGAGTTTTTGTT-3’ for N383 and S383, respectively. Fol- lowing hybridization, the blots were washed at room temperature for 30 min in 6 X SSC (1 X SCC is 0.15 M NaCl, 0.015 M Na3 citrate, pH 7.0) containing 0.1% sodium dodecyl sulfate and then for 30 min in the same solution a t either 50 “C (for the M382I ASOs) or 58 “C.

Site-directed Mutagenesis-To evaluate the effect of the NPD mutations on ASM activity, transient expression was performed in COS-1 cells. Each of the five mutations was introduced into the full- length ASM cDNA, pASM-3, since it had been previously shown that transient expression of this cDNA resulted in a catalytically active ASM in COS-1 cells (13). Prior to mutagenesis, genomic DNA from the three probands was analyzed for the presence of the previously described 322 and 506 ASM polymorphisms (13, 22). Since it was found that each of the three NPD patients had the Thr-322 and Gly- 506 polymorphic alleles, whereas the pASM-3 cDNA was Ile-322 and Arg-506, the first step was to change the sequence of pASM-3 a t codons 322 and 506 to match those found in the patients. For the 506 polymorphism, an 818-bp DraIII-EcoRI fragment was isolated from a human retinal cDNA clone, pASM-lr, which had the Gly-506 codon, and was exchanged as a cassette with the same fragment from pASM- 3, creating a new full-length cDNA designated pASM-3a. For the 322 polymorphism, a 549-bp BstEII fragment was PCR amplified from the pASM-3 cDNA using sense and antisense primers P11 (5’- ATCATCGGTAACCATGAAAGCACACCTGTCAA-3’) and P12 (5’-ATCATCGGTAACCAGGATTAAGG-3’). Sense primer P11 contained the T322 polymorphic sequence (indicated by the under- lined cytosine residue). Following PCR, the amplified fragment was isolated from agarose gels and exchanged as a cassette with the BstEII fragment from pASM-Sa; this modified full-length ASM cDNA was designated pASM-3b. pASM-3b was then subcloned into the

- GAATTTTTG-3’ and 5’-CTCAATATAAATTTTTG-3’ for M382

12554 Acid Sphingomyelinase Gene Mutations pGEM7Zf(+) vector and sequenced in both orientations to confirm that both polymorphic sites had been changed to Thr-322 and Gly- 506 and that no other nucleotide changes had been introduced during the procedure.

To introduce the NPD mutations into the full-length pASM-3b cDNA, the PCR overlap method was used (23, 24). Each round of PCR amplification was performed for 20 cycles and the primers containing the mutated sequences were constructed to be 25 nucleo- tides long and completely overlapping. For the fsL178 mutation, which resulted from a two-base deletion, sense and antisense primers P13 (5’-TGGAGACGCTCAGTGCT-3’) and P15 (5”GGCACAG- TAGGCAGAGATGTTCCAAGATGA-3’) were used to amplify a 108-bp fragment from pASM-3b. P13 overlapped a unique EspI site in pASM-3b and P15 contained the two-base deletion causing the fsL178 mutation. Next, a 278-bp fragment was amplified from pASM- 3b using sense and antisense primers P16 (5”CTTGGAA- CATCTCTGCCTACTGTGCCGAAG-3’) and P14 (5”TATCAA- AAGGGCCGGCTGGGCC-3’). P16 was complementary to P15 and also included the fsL178 two-base deletion. P14 overlapped a unique SfiI site in pASM-3b. Following amplification, each product was isolated from the agarose gels and the combined fragments were then used as templates for the second PCR using primers P13 and P14 to amplify the complete 386-bp EspI-SfiI mutated fragment. Following amplification, the fragment was isolated from the agarose gels, se- quenced in both orientations to confirm the presence of the two-base deletion, and exchanged as a cassette with the EspI-SfiI fragment of pASM-3b, thus introducing this mutation into the full-length ASM cDNA.

The G242R and L261X mutations were introduced into pASM-3b by a similar procedure. For the G242R mutation, sense primer P13 was used with antisense primer P17 (5”CGCCCCAGTATC- - TGGCACCTGGCCG-3’) to amplify a 301-bp fragment. The under-

sense primer P18 (5’-CGGCCAGGTGCCAGATACTGGGGCG-3’) lined thymidine residue in P17 indicates the mutated sequence. Next,

(complementary to P17) was used with antisense primer P14 to amplify a 103-bp fragment, and the complete EspI-SfiI fragment was PCR amplified using the combined fragments as templates and primers P13 and P14. For the L261X mutation, only a single PCR

primer P19 (5‘-ATCATCGGCCGGCTGGGCCCAGCCCACTCxA- amplification was required. Sense primer P I 3 was used with antisense

CAGGCTC-3’) to amplify the 386-bp EspI-SfiI fragment. P19 con- tained the mutant sequence.

The M382I and N383S mutations also were introduced into pASM- 3b by the PCR overlap method, however, a different set of PCR primers and ligation reactions were required. For the M382I mutation, sense primer P11, which contained a BstEII site at the 5‘-end, was used with antisense primer P20 (5”GGAACAAAAATTTATATT- GAGAGAG-3’) to amplify a 213-bp fragment from pASM-3b con- taining the M382 mutation. Sense primer P21 (5”CTCTCAATA- TAAATTTTTGTTCC-3’) (complementary to P20) was then used with antisense primer P12 to amplify a 357-bp fragment containing a BstEII site at the 3’-end. These two fragments were then combined and used as templates for a second amplification with primers P11 and P12. Following DNA sequencing, the fragment was inserted into pASM-3b by cassette exchange with the wild type BstEII fragment.

The N383S mutation also was introduced into the ASM BstEII fragment by the PCR overlap method. For this mutation, P11 was used with the mutated antisense primer P22 (5”CGGGAA- CAAAAACTCATATTGAGAG-3’) to amplify a 232-bp fragment con- taining a BstEII site at the 5’-end. Sense primer P23 (5”CTCT- CAATATGAGTTTTTGTTCCCG-3’) (complementary to P22) was then used with P12 to amplify a 317-bp product containing a 3’ BstEII site, and the combined fragments were used as templates for a second PCR to amplify the mutated 549-bp BstEII fragment. The PCR product containing the N383S mutation was inserted into

pASM-3b by cassette exchange with the wild type BstEII fragment and sequenced in both orientations to confirm the presence of the mutation.

Transient Expression in COS-1 Cells-For transient expression in COS-1 cells, the mutated cDNAs were subcloned into the eukaryotic expression vector, p91023(B) (25). DNA (-20 fig) from the mutant constructs, the wild type full-length ASM cDNA (pASM-3), and the full-length ASM cDNA containing the Tbr-322 and Gly-506 poly- morphisms (pASM-3b) were transfected into COS-1 cells by the method of Chen and Okayama (26). Negative controls for the trans- fections contained either no DNA or vector alone. The transfected cells were harvested at 72 h and the ASM activities were determined using the fluorescent substrate, N-(11-[I-pyrenesulfonyl]amido do- decanoy1)-sphingomyelin (PSAll-sphingomyelin) as previously de- scribed (17). One unit of activity equals that amount of enzyme that hydrolyzes 1 nmol of substrate/h. Protein determinations were per- formed by a modified fluorescamine assay (27).

RESULTS

Identification of the ASM Mutations-To identify the mo- lecular lesions causing Types A and B NPD in the three unrelated probands, the complete ASM coding region from each proband was PCR amplified from genomic DNA and sequenced by the dideoxy method. Five different mutations were identified (Table I). Proband 1, a Type A NPD patient from Southern India and the offspring of a consanguinous mating, was found to be homoallelic for a T to A transversion in exon 2 of the ASM gene (Fig. 1). This nonsense mutation, designated L261X, changed the leucine codon (TTG) to an amber termination codon, TAG. Although this patient’s par- ents were unavailable, the T to A transversion was identified in each of the 12 sequenced subclones, consistent with hom- oallelism for this mutation. This was confirmed by dot-blot hybridization of PCR-amplified genomic DNA from proband 1 with the appropriate ASOs (data not shown). Proband 2, an unrelated Type A patient of English, Italian, and German ancestries, had two different ASM mutations (Fig. 2, A and B ) . One of the ASM alleles in this patient contained a two- base (TT) deletion in exon 2 which caused a frameshift beginning with ASM codon 178 (designated fsL178). This frameshift mutation led to the occurrence of a premature stop at ASM codon 190. The other ASM allele in proband 2 had a G to A transition in exon 3 which predicted a methionine (ATG) to isoleucine (ATA) change at codon 382 (designated M382I). Fig. 2C shows the dot-blot hybridization analysis of PCR amplified genomic DNA from proband 2 and her parents. As expected, proband 2 was heteroallelic at the ASM locus, inheriting the fsL178 allele from her father and the M382I allele from her mother.

Two different missense mutations were identified in pro- band 3, a mildly affected Type B patient (Fig. 3, A and B ) . A G to A transition was found in five subclones from exon 2 and predicted a glycine (GGA) to arginine (AGA) substitution at codon 242 (designated G242R). In the five other sequenced subclones from this patient there was an A to G transition in exon 3 which predicted an asparagine (AAT) to serine (AGA) substitution at codon 383 (designated N383S). Heteroallelism for the G242R and N383S mutations was confirmed by dot-

TABLE I Mutations in the ASM gene causing types A and B NPD

Patient Phenotype Mutation Nucleotide Nucleotide ASM Amino acid change position“ exon change

Proband 1 Type A L261X T-A 782 2 Leu-stop Proband 2 Type A fsL178 Delete TT 532-533 2 Frameshift

M382I G-A 1146 3 Met-Ile Proband 3 Type B G242R G+A 724 2 Gly+Arg

N383S A-G 1148 3 Asn-Ser

” From the full-length ASM cDNA, pASM-3.

Acid Sphingomyelinase Gene Mutations 12555

blot hybridization (Fig. 3C), which indicated that proband 3 inherited these two mutant alleles from her mother and father, respectively. Notably, one of the five ASM mutations (G242R) occurred at a CpG dinucleotide, a known hotspot for muta- tions (28).

In addition to confirming the authenticity of the ASM mutations in probands 1-3, dot-blot hybridization analysis was used to screen over 60 other unrelated Type A and B NPD families. None of the five mutations occurred in any of these NPD families. In addition, these mutations were not identified in over 100 ASM alleles from normal individuals, demonstrating that they were not polymorphisms.

Transient Expression of the NPD Mutations-Residual ASM activity was determined in cultured cells from each of the three NPD patients (Table 11). The two Type A patients, probands 1 and 2, had about 1 and 2%, respectively, of normal ASM activity when determined in vitro using the fluorescent

B’ T G Ser A

G T G T Leu C C

Mutant 5’ A G Ser

+ A Stop

FIG. 1. Identification of the L261X mutation in proband 1. Partial DNA sequence of PCR amplified genomic DNA from proband 1 was performed as described under “Experimental Procedures.” The arrows indicate the position of the T to A point mutation.

A T

Pro C C G

Leu 1: T

Ser C T C

lie T A

B T

Phe

Asn

Met

Asn

T T T A A G T A T A A

natural substrate, PSAll-sphingomyelin. In contrast, proband 3, an unrelated Type B NPD patient, had about 15% of normal ASM activity in lymphoblasts. These results were consistent with the fact that cells from probands 1 and 2 contained about 2-3-fold elevated levels of sphingomyelin compared to normal cells, whereas those from proband 3 had about normal levels of the substrate.

To characterize the effect of each NPD mutation on ASM activity, transient expression studies were undertaken in COS-1 cells. Prior to introducing the NPD mutations into the full-length pASM-3 cDNA, probands 1-3 were analyzed for the two ASM polymorphisms (13, 22) by dot-blot hybridiza- tion analysis (data not shown). Since the pASM-3 cDNA had Ile-322 and Arg-506 at the two polymorphic sites, whereas the genotypes of the three patients were Thr-322 and Gly-506, the first step was to alter the polymorphic sites in the pASM- 3 to agree with those of the patients (see “Experimental Procedures” for details). Following mutagenesis, the effect of these two nucleotide changes on ASM catalytic activity was assessed by transient expression in COS-1 cells. As shown in Table 111, pASM-3b, which contained the Thr-322 and Gly- 506 codons, expressed essentially the same activity in COS-1 cells as the original pASM-3 full-length cDNA. Thus, the polymorphic nature of these 2 amino acid changes was con- firmed.

Next, each of the five NPD mutations was introduced into the pASM-3b cDNA. Following mutagenesis, each of the five mutations was subcloned into the eukaryotic expression vec- tor p91023(B) and transiently expressed in COS-1 cells. When compared to the wild type cDNAs (pASM-3 and pASM-3b), the fsL178, L261X, M3821, and N383S mutations had from 0

C A T C C G T C T C T A

T T

A A

A T A T A A

TYr

Ala

Ser

Ile

Phe

Asn

Ile

Asn

Mutant fsl78 I Normal E M382

Mutant 1382 r

FIG. 2. Identification of the fsL178 and M382I mutations in proband 2. Partial DNA sequences of PCR amplified genomic DNAs ( A and B ) and dot-blot hybridizations ( C ) were performed as described under “Experimental Procedures.” The arrows in paneb A and B indicate the positions of the two-base deletion and the G to A transition in the fsL178 and M382I alleles, respectively. The symbol (0) in panel C indicates an unrelated normal individual.

12556 Acid Sphingomyelinase Gene Mutations

A 3’, G A T C

r 5’

T Phe T

T T

AStl A A G

Met T A T

ASn A A

1

I I

Normal [I ul Mutant

5’

T T Phe T T CI Ser A G T Met A T A ASn A

C

Normal G242 I Mutant R242 n

Normal N383

Mutant S383 U

FIG. 3. Identification of the G242R and N383S mutations in proband 3. Partial DNA sequences (A and B ) and dot-blot hybridizations ( C ) were performed as described under “Experimental Procedures.” The arrows in panels A and B indicate the positions of the G to A transition and the A to G transition in the G242R and N383S alleles, respectively. The symbol (0) in panel C indicates an unrelated normal individual.

TABLE I1 I n vitro levels of ASM activity in cultured cells

Values are means f S.D. Cell line Cell tme ASM activity % of Control

nmol/h/mg

Lymphoblasts 5.8 f 0.61 Skin fibroblasts 80.4 f 0.01

Normal controls MS-1 MS-2

NPD patients Proband 1 Skin fibroblasts 0.34 f 0.01 <1 Proband 2 Lymphoblasts 0.12 f 0.05 2.0 Proband 3 LvmDhoblasts 0.84 * 0.10 14.5

to 2% of normal ASM activity (Table 111). In contrast, the G242R mutation expressed ASM activity at levels about 40% of those obtained with the wild type controls.

TABLE 111 Transient expression of ASM mutations in COS-1 cells

Values are means f S.D.

Enzyme source ASM activity Intracellular

Control constructs COS-1 cells p91023(B) alone pASM-3 (sense) pASM-3 (antisense)

ASM mutant constructs pASM-3b

fsL178 G242R L261X M382I N383S

nrnollhlmg

13.5 f 2.8 14.5 f 3.3 97.1 k 35.9 9.1 f 1.1

105.0 f 47.1

12.3 & 4.9 39.9 f 21.1 14.5 f 4.3 15.7 k 4.2 18.1 f 3.4

DISCUSSION

In 1966, Brady and colleames ( l ) identified the deficient the clinical severity, padicularly for Type B NPD (e.g. Refs. activity of ASM as the primary enzymatic defect in Type A NPD. Schneider and Kennedy (29) soon confirmed this find- ing and described several other NPD patients with the milder The recent identification of the first two molecular lesions T~~ B phenotype who also had Severe ASM deficiencies. causing Types A and B NPD provided initial insights into the During the subsequent 25 years, investigators attempted to nature of the clinical heterogeneity in this disease (14, 15). correlate the levels of residual ASM activity with the pheno- The R496L mutation was originally identified in both ASM typic severity of Types A and B NpD. Artificial and natural alleles from an Ashkenazi Jewish Type A female patient who substrates have been developed to measure ASM activity and expired a t about 3 years of age. More than 60 other unrelated the enzyme levels have been determined in various tissue, Type A and B NPD patients have been screened for this cell, and fluid sources (e.g. Refs. 5-8). However, the amount mutation, and it has been found that the R496L lesion of residual enzymatic activity has not always correlated with accounts for about 32% of the mutant ASM alleles in Ash-

30-32).

Acid Sphingomyelinase Gene Mutations 12557

kenazi Jewish Type A patients. The R496L mutation also was identified in one of the ASM alleles from an Ashkenazi Jewish Type B patient. The other ASM allele in this Type B proband was a three-base deletion which led to the removal of a single arginine near the carboxy terminus of the mature ASM poly- peptide (AR608) (15). Another Ashkenazi Jewish Type B patient which we analyzed also was heteroallelic for this mutation. In contrast, of the more than 40 non-Jewish Type B patients studied, only one had the AR608 mutation, a French Arabic patient who was homoallelic. Thus, these stud- ies provided the first molecular evidence that Types A and B NPD were due to distinct lesions in the ASM gene.

In the present study, we describe five additional ASM mutations that result in Types A and B NPD. In contrast to the R496L and AR608 mutations, which occurred at high frequencies in the Ashkenazi Jewish population, each of these five mutations occurred in only one family. This finding provides further evidence that various genetic factors presum- ably have led to the higher NPD gene frequency in the Ashkenazi Jewish community than in the general population (14).

As illustrated in Fig. 4, the seven NPD mutations identified to date are distributed throughout the ASM gene. Of these, five are due to point mutations, whereas the remaining two (fsL178 and AR608) are due to small deletions. Of the five point mutations, two (G242R and R496L) occurred at CpG dinucleotides, known hotspots for mutations (28). It is intrigu- ing that two of the seven mutations occurred in adjacent codons (M382I and N383S) within exon 3, which is alterna- tively spliced in about 10% of the ASM transcripts (12, 13). Transient expression studies revealed that each of these mu- tations led to severely reduced ASM activity. These findings are consistent with the fact that the alternatively spliced Type 2 ASM transcripts did not express catalytically active ASM in COS-1 cells (13); thus, this region of the ASM polypeptide may be particularly important for the function of the ASM active site.

Transient expression studies in COS-1 cells also revealed that the two mutations which introduced premature stop codons in the ASM gene, fsL178 and L261X, also resulted in no detectable ASM activity. These results are consistent with the fact that these mutations produced trunctated polypep-

Type A NPD

A N383S

Type B NPD

G242R hR6flX

Y i 6

FIG. 4. Schematic representation of the ASM mutations. The locations of the seven mutations in the ASM gene causing Types A and B NPD are indicated. Open boxes indicate ASM exons and straight lines indicate introns. The first in-frame initiation codon (ATG) and termination codon (TAG) also are indicated.

tides which presumably were unstable and rapidly degraded. Intriguingly, the G242R mutation produced ASM activity in COS-1 cells at about 40% of the levels obtained with normal ASM constructs. This result was consistent with the mild Type B phenotype of proband 3, who was heteroallelic for the G242R and N383S lesions, and the fact that cultured lympho- blasts from proband 3 had about 15% of normal ASM activity. It is likely that the higher level of residual ASM produced by the G242R mutation in COS cells compared to cultured lym- phoblasts was due to the fact that multiple copies of the expression plasmid are replicated in this transient expression system (25). These results demonstrate the usefulness of the COS expression system to evaluate Type B NPD mutations and raise the possibility that some of the patients who have been previously diagnosed as having Type C NPD (33) be- cause they had high levels of residual ASM activity in cultured cells may, in fact, have primary lesions in the ASM gene. Clearly, future studies will be directed towards characteriza- tion of the physical and kinetic properties of the residual ASM polypeptides expressed from these mutant constructs using various substrates and inhibitors. In addition, efforts are underway to produce high-titer monospecific anti-ASM antibodies for immunologic characterization of the mutant polypeptides.

Finally, these studies indicate the feasibility of genotype/ phenotype correlations in NPD. To date, each of the patients who were homoallelic for the L261X or R496L mutations had the neuronopathic Type A phenotype. In addition, proband 2, who was heteroallelic for the fsL178 and M382I mutations, also had Type A NPD. These patients had little or no residual ASM activity in cultured cells and when their mutations were evaluated by transient expression in COS-1 cells they pro- duced no catalytically active ASM. Thus, it may be predicted that individuals who are homoallelic for any of these four mutations will have the severe Type A phenotype, while the phenotype of patients who are heteroallelic for these muta- tions will most likely depend on the functional integrity of the enzyme produced from their other mutant ASM alleles. In contrast, all of the NPD patients studied who were either homo- or heteroallelic for the G242R or AR608 mutations have had the non-neuronopathic Type B disease. Thus, the mild phenotype of proband 3 (who was heteroallelic for G242R and N383S) was due to the residual ASM activity produced from the G242R allele, and it is likely that an individual who is homoallelic for the N383S mutation will have a neurono- pathic Type A phenotype.

In summary, five new ASM mutations causing the neuron- opathic and non-neuronopathic forms of NPD have been identified and characterized, permitting genotype/phenotype correlations for this disease. In addition, transient expression of the G242R mutation provided the basis for the mild Type B phenotype of proband 3 and demonstrated the usefulness of the COS expression system for further investigations of the mutant ASM polypeptides in NPD patients.

Acknowledgments-We thank Michael Chase for expert technical

from their NPD patients. assistance and the many colleagues who supplied us with samples

Note Added in Proof-While this manuscript was in review, two additional missense mutations were identified causing NPD: G577S causing Type A NPD (34) and S436R causing Type B NPD (35).

REFERENCES

1. Brady, R. O., Kanfer, J. N., Mock, M. B., and Fredrickson, D. S. (1966) Proc. Natl. Acad. Sci. U. S. A. 55, 366-369

2. Fredrickson, D. S. (1966) in The Metabolic Basis of Inherited Disease (Stanbury, J. B., Wyngaarden, J. B., Fredrickson, D.

12558 Acid Sphingomyelinuse Gene Mutations S., eds) 2nd Ed, pp. 586-602, McGraw-Hill, New York 18. Sambrook, J., Fritsch, E. F., and Maniatis, T. A. (1989) Molecular

3. Spence, M. W., and Callahan, J. W. (1989) in The Metabolic Cloning: A Laboratory Manuel, Cold Spring Harbor Press, Cold Basis of Inherited Disease (Scriver, C. R., Beaudet, A. L., Sly, Spring Harbor, NY W. S., Valle, D., eds) 6th Ed, pp. 1655-1676, McGraw-Hill, 19. Itakura, K., Rossi, J. J., and Wallace, R. B. (1984) Annu. Reu. New York Biochem. 53,323-356

4. Goodman, R. M. (1979) Genetic Disorders Among the Jewish 20. Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, People, pp. 96-100, The Johns Hopkins University Press, Bal- R., Horn, G. T., Mullis, K. B., and Erlich, H. A. (1988) Science timore 239,487-491

5. Gal, A. E., Brady, R. O., Hibbert, S. R., and Pentchev, p. G. 21. Sanger, F., Nickleson, J., and Coulson, A. R. (1977) Proc. Natl. (1975) N . Engl. J. Med. 293,632-637 Acad. Sci. U. S. A. 7 4 , 5463-5467

6. Vanier, M. T., Rousson, R., Garcia, I., Bailloud, G., Juge, M-C., 22. Schuchman, E. H., Levran, 0.9 Suchi, M., and Desnick, R. J. Revol, A., and Louisot, P. (1985) Clin. Genet. 2 7 , 20-32 (1991) Nucleic Acids Res. 19 , 3160

7. Gatt, S., Barenholz, Y., Goldberg, R., Dinur, T., Besley, G., 23. s. N., Hunt, H. D.y Horton~ R. M., J. K., and Leibovitz-Ben Gershon, Z., Rosenthal, J., Desnick, R. J., De- L. R. (1989) Gene (Amst.) 7 7 , 51-59 vine, E. A., Shafit-Zagardo, B., and Tsumki, F. (1981) Methods 24. Horton, R. M., Hunt, H. D., Ho, s. N., Pullen, J. K., and Pease, Enzymol. 72 , 351-375 L. R. (1989) Gene (Amst.) 7 7 , 61-68

L. (1986) Pediatr. Res. 19 , 153-157

96

Suppl. 2, 326 (1978) Nature 274 , 775-780

Lipid Metabolism ZZ (Freysz, L., Dreyfus, H., Massarelli, R., and Gatt, S., eds) pp. 273-299, Plenum Press, New York 209

12. Quintern, L., Schuchman, E. H., Levran, O., Suchi, M., Sandhoff, 30. Sogawa, H., Horino, K., Nakamura, F., Kudoh, T., Oyanagi, K.,

K., and Desnick, R. J. (1989) EMBO J. 8, 2469-2473 Yamanou-Chi, T., Minami, R., Nakoa, T., Watanabe, A., and

13. Schuchman, E. H., Suchi, M., Takahashi, T., Sandhoff, K., and 31, Chan, w. c., ~ ~ i , K. s., and Todd, D. (1977) J, pathol. 121 , Matsura, Y. (1978) Eur. J. Pediatr. 128 , 235-240

Desnick, R. J. (1991) J . Biol. Chem. 266,8531-8539 14. Levran, o., Desnick, R. J., and Schuchman, €3. H. (1991) Proc. 32. Dawson, p. G., and Dawson, G. (1982) Hum, Pathol. 13 , 1115-

Natl. Acad. Sci. U. S. A. 88, 3748-3752 1120 15. Levran, o., Desnick, R. J., and Schuchman, E. H. (1991) J. C h . 33. Pentchev, p. G., Comb, M. E., Kmth, H. S., Vanier, M. T.,

Znuest. 88,806-810 Wenger, D. A., Patel, S., and Brady, R. 0. (1985) Proc. Natl. 16. Bernstein, H. S., Bishop, D. F., Astrin, K. H., Kornreich, R., Eng, Acad. Sci. U. S. A. 82,8247-8251

c. M., Sakuraba, H., and Desnick, R. J. (1989) J. Clin. huest. 34. Ferlinz, K., Honvitz, R., and Sandhoff, K. (1991) Biochem. Bio-

17. Klar, R., Levade, T., and Gatt, S. (1988) Clin. Chim. Acta 176 , 35. Takahashi, T., Desnick, R. J., Takada, G., and Schuchman, E.

8, Lev&, T., Salvayre, R., Bes, J-C., Nezri, M., and Douste-Blazy, 25. Kaufman, R. J., and Sharp, p. A. (1982) Mol. Cell. Bid . 2 ,

9. Gatt, S., Dinur, T., and Barenholz, y. (1980) Clin. Chem. 26,93- 26. Chen, c., and Okayama, H. (1987) cell. Bioi. 7 9 2747-2752 1319

27. Bishop, D. F., and Desnick, R. J. (1981) J . Biol. Chem. 256 ,

10. Jobb, E., and Callahan, J. W. (1987) J. Inherited Metab. Dis. 10 , 28. coulondre, c., ill^^, J. H., ~ ~ ~ ~ b ~ ~ ~ h , p. J., and ~ i l b ~ ~ , w. 11. Rousson, R.7 Vanier, M. Tv and Lousiot, p. (1986) in Enzymes of 29. Schneider, p. B., and ~ ~ ~ ~ ~ d ~ , E. p. (1967) J , Lipid R ~ ~ . 8, 202-

1307-1316

177-181

83,1390-1399 phys. Res. Commun. 179 , 1187-1191

259-268 H. (1992) Hum. Mut., in press