Novernber/Decernber 2000 . Vol. 2 . No. 6

The GM~gangliosidoses databases: Allelic variation at the HUYA, HEXB, and GM2A loci Paulo Cordeiro, MSC',Peter Hechtmatl, P~D'-~,ard Feige Kaplan, ~hD-7'

The G,, gangliosidoses are a group of recessive disorders characterized by accumulation of GM2 in neuronal cells. The responsible for these disorders are HEXA (Tay-Sachs disease and variants), HEX6 ( and variants), and GM2A (AB variant of GM2 ). We report the establishment of three relational -specific databases recording allelic variation at the HEM, HEXB, and GM2A genes and accessed at the GM2 gangliosidoses home page (http://data.mch.mcgill.ca/gm2-gangliosidoses). Submission forms are available for the addition of new to the databases. The databases are available online for users to search and retrieve information about specific alleles by a number of fields describing mutations, phenotypes, or author(s). Genetics in Medicine, 2000:2(6):319-327. Key Words: G,, gangliosidoses, locus-specificdatabases,

The genetic basis of the G,, gangliosidoses is complex. G,,, gangliosidosis have been classified according to age of the Three lysosomal polypeptides, encoded by three distinct genes, onset of symptoms into infantile, juvenile, and adult forms of are essential for normal catabolism of the G,, ganglioside sub- the disease. However, these classifications are often arbitrary. strate in vivo: the a and /.3 subunits of P-hexosaminidase A Recent advances in computer technology have created a new (Hex A) and the G,, activator . Tay-Sachs disease area of research and development "bioinformatics," which in- (OMIM: 278200) and its variants, caused by deficiency of Hex tegrates biology and informatics. This area of study has pro- A activity, results from mutations at the HEXA gene (15q23- vided invaluable information on newly discovered genes: lo- q24), which encodes the a subunit of Hex A. Sandhoff disease calization of genes influencing complex phenotypes, (OMIM: 268800) and variants, characterized by combined de- molecular variation (mutations) influencing human diseases, ficiency of Hex A and Hex B activity, is caused by in and identification of functional DNA variants. These techno- the HEXB gene (5q13) encoding the /3 subunit common to Hex logical advances are likely to have a profound impact on cur- A and Hex B. The AB variant of G,, gangliosidosis (OMIM: rent knowledge of complex disease etiology and may reveal 272750) is caused by mutations in the GM2A gene (5q31.3- novel approaches to disease treatment and prevention. Locus- q33.1) encoding the G,, activator protein. In AB variant pa- specific mutation databases record genetic variation and are tients, Hex A and Hex B are structurally normal, and relevant to human genetic research, genetic epidemiology, and activities are in the normal range when measured in the pres- health care. ence of synthetic, hydrophilic substrates. However, The exponential growth of information regarding allelic of G,, ganglioside is impaired by absent or defective forma- variation accounting for cause of and/or susceptibility to dis- tion of the G,, ganglioside/GM2activator/Hex A complex. All ease mandates the creation of a formal record of "mutation" at known variants of G,, gangliosidosis display autosomal reces- gene loci. The current pace of mutation discovery far exceeds sive inheritance. the capacity to publish the data by conventional means. Data- The clinical phenotypes associated with G,, gangliosidoses bases provide the most efficient format to maintain an up-to- variants vary widely. They range from infantile-onset, rapidly date record of locus-specific allelic variation. A designated cu- progressive neurodegenerative disease culminating in death rator or curator consortium evaluates information. Curators before age 4 years (classical Tay-Sachs and Sandhoff diseases receive and distribute information upon request. and G,, activator deficiency) to late-onset, slowly progressive The Human Organization (HUGO) provides neurologic conditions compatible with long-term survival and guidelines and recommendations for content and structure of with little or no effect on intellect.'-' Phenotypic variants of mutation database^.^ These will provide the means to link lo- cus-specific databases into a centralized human mutation da- tabase and to ensure cohesiveness in a community of users. According to HUGO guidelines, niutation databases should be relational and have a minimum core of content in a shared The core should include ( 1 ) a unique identifier for each allele, (2 a source/report for data, (3)a context for alleles (patient genotype or in vitro expression system used in stud- Cordeiro et a/.

ies), and (4) a description of alleles. The allele description Table 2 should be verified by a group or consortium. Relational data- Phenotype search page bases enable users to perform queries on any aspect of the Fielda Description database content. Clinical phenotype Acute, subacute, chronic, benign, asymptomatic We report herein the establishment of three relational data- (none),unknown bases for the HEXA, HEXB, and GM2A genes accessed at the Age of onset Infantile, late infantile, juvenile, late juvenile, and adult GM2 gangliosidoses home page (http://data.mch.mcgill.ca/ Biochemical phenotype Biochemical variants (Bl, hexosaminidase Paris, -gangliosidoses/). HEXAdb (http://data. mch. mcgill.ca/ neutral polymorphisms); information hexadb/), HEXBdb (http://data.mch.mcgill.ca/hexbdb/), and regarding enzyme activity, processing, and GM2Adb (http://data.mch.mcgill.ca/gm2a) are relational data- submit association bases and are currently divided into four tables each: Mutation, Genotype Information regarding the second allele, in patients where identified Phenotype, Reference, and Association (Tables 1 and 2). The Diagnostic Method used to identify the presence of a Mutation table contains descriptions of the following fields: mutation (gain or loss of restriction enzyme mutation name, alternative names, mutation type, nucleotide sites, heteroduplex analysis, allele-specific amplification or allele-specific hybridization) changes(1ocation and length), as well as a description of the Ethnic origin The ethnicity of patients associated with a mRNA and the protein size and abundance. The Phenotype certain mutation table contains details on clinical and biochemical phenotype, Also includes mutation name and nucleotide number (from Mutation table), age ofonset, diagnosis, genotype, and ethnicity of patients. The submitter (from Association table) and short citations (from Reference table). Reference table has information on types of publication, au- "Structure of HEXAdb, HEXBdb, and GM2Adb tables showing fields that can thors, journals, titles, etc. It also includes PubMed IDS to en- be used when interrogating using the Phenotype Page search engine. able the user to look up references in MEDLINE. Finally, the Association (also called Resource) table links, i.e. relates, the tables together and provides information on the submitter (the individual who identified the mutation). The purpose of these Databases: software and search engine databases is to collect and distribute information on mutations HEXAdb, HEXBdb, and GM2Adb were constructed using in the genes responsible for G,, gangliosidosis. HEXAdb cur- mySQL v3.21.33~8(T.c.X. DataKonsultAB) on the Red Hat ' rently records 100 alleles. Among 12 alleles for which no asso- Linux v5.1@ (Red Hat, Inc.) operating system. MySQL is a true ciated clinical phenotype has been described, 6 are predicted to multiuser, multithreaded SQL database server that enables one to be disease-causing. Such predictions are possible because in- access, store, and update information. The Web connection exists formation in the Phenotype table includes any in vitro gene through a Perl DBI interface. Modifications and addtions to the expression studies that have been performed on the mutant databases are immediately undertaken by the curator using Mi- allele. HEXBdb currently records 25 alleles; GM2Adb, 5 alleles. crosoft Access 978 (Microsoft Corp.) through an ODBC (Open Database Connectivity) interface or directly using mySQL com- mand scripts. The Perl scripts that make up the search engine were written in Wordpad 4.0@ (Microsoft Corp.) and then uploaded Table 1. to the Debelle server (http://data.mch.mcgill.ca) using WS-FTP95 Mutation search page LE 4.60 (Ipswitch, Inc., 1992-1998). WS-FTP LE is a File Trans- Field" Description fer Protocol (FTP) client application that can connect to any sys- - Mutation name The ofticial name for the mutation given current tem that has a valid Internet address and contains an flPserver nomenclature rules program, allowing one to transfer files between a wide variety of Alternative name A common name, usually the codon change, used to systems, includmg Windows, OS\2, and UNIX systems. refer to the mutation The Web pages were written in HTML 4.0 (Hyper Text Nucleotide change Substitution, insertion, deletion Markup Language) programming code, by using a normal text editor, Wordpad 4.08 (Microsoft Corp.). HTML files were Mutation type Missense, nonsense, abnormal splicing, silent (polymorphism), frameshift then uploaded to the Debelle server (http://data.mch.mcgill.ca) using WS-FTP95 LE 4.60 (Ipswitch, Inc., 1992-1998). Nucleotide number cDNA position where mutation occurs Logos and graphics for the Web pages were designed in Mi- mRNA description Information regarding the mRNA transcribed, if any, crosoft Powerpoint 978 (Microsoft Corp.) and then edited in as a result of the mutation Microsoft Photo Editor 3.08 (Microsoft Corp.). Protein description Information regarding the protein translated, if any, as a result of the mutation Region (e.g. EOl), intron (e.g. IVS-2) RESULTS AIso includes submitter (from Association table) and short citations (from Mutatii table reference table). The Mutation table (see Table 1) provides specific molecular "Structure of the HEXAdb, HEXBdb, and GM2Adb tables showing fields that can be used when interrogating using the Mutation Page search engine. information on mutations identified. Each mutation has an ID

*S *S IN Medicine G, G, gangliosidoses databases

(identification number, the primary key), which is assigned by A table with all references on record is available on the Web its order of entry, a systematic name (usually referring to the page. Currently 132 references are listed for HEXAdb, 32 for cDNA mutation), an alternative name (the corresponding HEXBdb. and 7 for GM2Adb. codon change), as well as information on the nucleotide change (substitution, insertion, deletion), the mutation type Association table (nonsense, missense, silent, abnormal splicing, frameshift), the Finally, the Association (also called Resource) table links nucleotide position, the region of the gene involved (exon, (i.e., relates) the main tables together and records the "submit- intron, UTR) and a description of mRNA and protein pheno- ter" who identified the mutation(s) either by publication or by types. Also provided are information on the submitter, which personal submission. refers to the investigator(s) who identified the mutation either in a peer reviewed publication or by direct submission. Cita- Submission fonn tions of published articles are indicated. All three databases permit online submission of new muta- The search engine also allows one to query the database in tions (see Fig. 3). The submitters are requested to provide data any of the listed fields. For example, Figure 1A illustrates the for all fields. The aim is to obtain as complete a record as pos- mutation data retrieval search page for HEXBdb. Figure 1B sible, including clinical and biochemical data for each new sub- shows the output of a query on the "Mutation type" field for mission. The submission form, which will permit a faster and the value "frameshift" in the HEXB gene. The user can also more efficient means to regroup information, is directed to the select a table that displays mutation information for all the curator, whose role it is to review the information and deter- records in the database. The HEXA database records 100 mu- mine whether it is sufficient for inclusion. tations (51 missense, 19 abnormal splicing, 12 frameshifts, 8 The clinical descriptions requested in the submission form nonsense, 6 deletions, and 4 silent mutations). HEXBdb are limited to two categories, which lend themselves to stan- records 25 mutations ( 10 missense, 6 frameshifts, 6 abnormal dardization: age of onset and an index of severity with only six splicing, 1 nonsense, 1 deletion, and 1 silent) and GM2Adb gradations. The criteria for application of these terms are records 5 mutations (2 missense, 2 frameshifts, and 1 found on the information page "Tay Sachs Disease, Gene Clin- nonsense). ics, NIH." A direct link to this page is found at the top of the Phenotype Information section of the submission form. This Phenotype table form also contains a direct link to the email address of the The Phenotype table (see Table 2) is designed to provide curators, permitting the submitter to request assistance if details on clinical and biochemical phenotype, age of onset, necessary. and genotype and ethnicity of patients harboring specific mu- Additional clinical terms such as ataxic, dystonic, psychotic, tations. Each record in the Phenotype table has an associated muscle-wasting, etc., frequently appear in the clinical literature mutation ID number. Also included in this table (where avail- describing the presentation of nonclassical variants of these able) is a molecular diagnostic method used to detect the mu- disorders. Submitters may wish to include this information in tation. Abnormalities of enzyme properties or specific activity the "Comments" field. Terms used in this field are also avail- are also recorded. able for search. The Phenotype table can be searched on any of these fields, and a complete table of phenotypes is available online. Figure Other features 2A and B illustrate data retrieval following a search under Relevant information pages, provided at the database Web site, "Clinical phenotype" for mutations leading to subacute HEXB include clinical ~nformationon Tay-Sachs and Sandhoff disease and (Sandhoff) variants. To date, clinical phenotypes are known hks to other sources of information on the Web includmg Online for 88 of the 100 entries in HEXAdb. There are 56 acute, 15 Mendelian Inheritance in Man (http://www.ncbi.nlm.nih.gov/ subacute, and 7 chronic TSD variants. Three Hex A pseudo- Omim/),LocusL~nk(http://www.ncbi nlm. nih.gov/LocwLinkl),Gene- deficient alleles and seven benign alleles are also recorded. cards (http:/hoinformatics.weizmann.~il/~,and GeneChcs Twenty-four of 25 entries in HEXBdb are associated with (http://wwwgeneclinics.org/). known clinical phenotypes (15 acute, 4 subacute, 3 chronic, 1 asymptomatic, 1 benign). All five of the GM2Adb entries are Modeling of HEXA mutant alleles: genotype/phenotype correlations associated with an acute phenotype. To provide information concerning the effects of mutations on the manifestations of disease, a feature for viewing mapped Reference table positions of selected HEXA mutations onto the homology- The Reference table provides a listing of articles reporting modeled hexosaminidase structure was developed. The struc- identification or analysis of specific mutations. In the case of tural model of the Hex A a subunit was restricted to the cata- published articles, the full reference to the article is specified as lytic domain (region of the protein containing a (pa), barrel well as the PubMed ID, enabling the user to link to references motif, residues 160-487) of the wild-type protein. This is the containing the abstract in MEDLINE. The Reference table of- region that has been modeled onto the template of the crystal- fers a search tool by author(s), journal, title, and year of lized Serratia marcescens chit~biase.~Theoretical models of publication. HEXA rnissense mutations are being added to the Web page. Cordeiro et a/.

A.

WEED& Mutation Data Retnwieval

ARernaSve Name Pleane choose .each field: Nudeotide Change Mutabon Type Eater hxt m a "he: mRNADesffl~on RminDssmpllon Nudeotide Number mpwA ReglonMedsd Submiuer ~stsrenm HELP

Mutation n-c: For a nucleotidc replacrmmt, eg c.379.C->T. the notation consists of c (indicates cDNA) followcd by a nuclcotide number (preceded ntld followed by a period). chm the oliginat nucleotide and &allythe mutant nuclcotide. For a deletion or mscmon, cg. s.547.ins4 the notation oandsts of c followed by a nudcotide nwnbcr @receded and followed by a pmod). Qcn bs or dc1 and 6naUy the nucleofidc(s) that arc msettcd or deleted tUtarmate -: For an Pmino ndd h.eg. G269S (GIyZ69Ser). the notdon coluists of& add followed by mc numb- ntld then thc mvtpnt ntrdno acid Onc or three lmu abbreviltionsfor Lcamino acid can be wed For an intronk don.og IVSZ+lera. the notation consst. ofmfollowed by inhm mnnbcr, &inhonk podtion (can be +I-).the oliginat rmcleotide und6naUy the rmrtpnt nudcotide. Nucleotide champ: Substitution insdon and delmon

Mutarha me: Masaue. nonsense. abnormal spLciu& dent @oly~~orphirm)and 6ame&W.

mRNA dcsctiptimr: Informphon rsg~@the mRNA ayntaesimd, if my, as a ram& of ths ndoa Protein descmolc Iafomratlon reg~dingthe pro6synthesized. it--. as a redof tbr Pnrntion Nutleatide mnnbex: Pordtion whnr mmDPton occurs. Region tticcted: Enon or &on. eg. E01 or IVS-04. rsfisrs mu!Aon occurs.

sdxdtter: Consists of the &st author's anme and date of publication, ifpublished

References: Condm ofthc short &+ion (&st author's anms md das ofpublicaPion) for articles contab& infomution on this mutlrion

mg. 1 A: Example of a mutation search performed on HEXBdb. The mutation data table has been selected for query. Fields are selected from a drop down menu and field valuesare typed in by the user. Only one field can be searched in each query. Note that paragraph 4 of the Help section provides value names for the "Mutation Type" field. B: Mutation Q~~~ output,~h~ table shown has been returned in response to the query in A. Note that all fields are displayed. Reference table must be accessed in order to obtain reference Links.

en'*t 'CS IN Medicine G,, G,, gangliosidoses databases

A.

HE.XBd!bPhenotype Data Retrieval

Nudeotide number - Rense choose rsarchfield: Iainical phenotype &-Dainical phen-e Onset Enter text m a value: lsubncute BioChemicd phenotype Diagnostic Genotype Ethnic origin Subminer References HELP

BlmUfionnmns: For an andno and chnngt. eg. G269S (Gly269Ser). the no~monconsists of the -0 acld followed by the number nnd then the mutmt doamd Om or three letter abbreviations for hoacid can be wed For nn intronrc mrrtiltioq eg. IVSL+lp>a, !Axe nomon consists of IVS fonowed by mtson number. then iniromc posmon (cm be +I-). the ownucleotidn and 6ualiy the mutmt nucleotide. For a deletion or msemon. e& r-%?.iusA. the notation consists of c fonowed by a mrclcotidc numba (preceded and followed by a paiod). then inn or del nnd finally the nucleotide(s) that arc in?rcrkd or dclkd. Nuclcotide nrmrber: Podtion where mwmoccur#. ClLdcd phmmlype: Acute. subacute. Ehronic. mild chronic, benign, asyrnptormhc (none) and uuknoam Ornee krfmlrle. Late idauble, juvenile. late juvcmle and adult BiDchsmicd phcnmype: Infonnaaon regardrng protcm aaivxiy. processing nnd subtnrit asrocinam Variants (Hexoaamintdase Pans) md Neuhal Polymorp~nre also mduated.

Genotype: Information regarding the second able. in people where'the muration WPP idcded if mailable. Dhportic: &in or loss of nmiction mspme sites. hmroduplu endpis, dele-specib -Scation or anele-spec& hy6ridromoa Etbit & Ethnic$y of patients or cmrias.

Suhudttm: Consists of tha &st authds last nama and date of pubhanos ifpub-d. Referrnces: Consists of the skort citation @rst &or's name and date of publicanon) for miicles conk* inf0Smmic.n on tfiis nnJMioa

fig. 2 A: Example of a phenotype search performed on HEXBdb. The phenotype data table has been selected for query. Fields are selected from a drop down menu and field values are typed in by the user. Only one field can be searched In each query. Note that paragraph 3 of the Help section provides value names for the "Clinical phenotype" field. B: Phenotype Query Ourput. The table shown has been returned in response to the query in A. Note that all fields are displayed. Reference table must be accessed in order to obtain reference links. Cordeiro et a/.

Currently, the 3D structures of the catalytic domains of both the human 13-hexosaminidase a and 13 subunits can be viewed online, as can the localization of putative active site residues and some missense mutations (only for the a subunit). Prese­ HE...\Adb Mutation Submission Form lected groups of residues at which mutations lead to similar clinical phenotypes or functional impairments (i.e., catalytic PloueliSedasf_ta ___ ....o(H-). site) can be viewed or the viewer may place the arrow on a J•· _.. ill(o.naaliaa pl.a.• c..Ud d. £!!!:!!!!!.. section of the protein backbone and left -click the mouse to find the name and position number displayed on the status line. Viewers may also select different rotational views of N-. the protein and different modes of presentation of the peptide

backbone. No theoretical models exist, as of yet, for the GM 2 activator protein. Figure 4 illustrates the mapping of a subset of infantile acute

and adult HEXA mutations onto the hexosaminidase model. (4 c.1136. A·>G) Mutations leading to acute TSD cluster at or near the active (q. GNISJ site. Adult mutations more often involve residues on the pro­ -- tein surface. It is important to note that not all mutations con­ form to these general principles. It is anticipated that correla­ r tion of new data on the biochemical phenotypes and structural localization of specific mutant residues may lead to the defini­ I tion of new functional domains on the Hex A protein such as _,_, the activator binding or the dimerization regions.

TSD screening mutation database development (q.... ••,., qf-w:tiM .u.s)

lllllNA 4aa¥tln; This table will represent a collaborative effort and will in­ (d111<4aa.U.W•f-,.pf] clude information regarding allele frequencies and numbers of alleles screened annually according to ethnic/geographic ori­ (fot..-..-uo/IOW.tU./1/J gin, number of affected fetuses identified by prenatal testing according to ethnic/geographic origin, screening tests for spe­ cific alleles, NTSAD-approved screening centers, etc. The planned format is illustrated in Figure 5. Mutation data will be Foriol'ormllioo oolb< olu1i&c10ioD ofTSD 1111< ofoamor oliai10 recorded separately for and other individuals. Tl!·S.W !);seue Gm:Ciaog NIH The form is intended for use by laboratories conducting com­ ,....,...... , l· ..iod- jJ munity-based Tay Sachs carrier screening. Only those alleles Oi.Uarl,_,, l·ooloct· iJ known to be relatively frequent are listed. Information re­

quested includes allele frequency data and ethnic origin of car­ f!'lh>IA-If-J riers. This information is expected to be useful in selecting appropriate diagnostic tests for local populations/families.

c._, DISCUSSION

/Ethieo,...(f/: The characterization of HEXA, HEXB, and GM2A gene mu­

tations contributed to the understanding of the clinical heter­ • PaWic.an...m.r l·tolod• iJ ogeneity of the GM 2 gangliosidoses, the ethnic distribution of ·N- mutant alleles, and the structure-function relationships of the Hex isozymes. Information about HEXA, HEXB, and GM2A mutations has also provided for more accurate carrier detec­ tion, pre- and postnatal diagnosis, and prediction of outcome for affected individuals. The inclusion of this type of compre­ hensive information is only possible in databases which seek to IDo-l provide in-depth characterization of one or a limited number HIJ:J(AA_...... ,...,,rn.F. ,_,_ ,.. HlfXMj .,.__. JJID __ _,...,. Dthflr/)W G,.,.,. of gene loci. These locus-specific databases can also include detailed information on diagnostic tests and ethnic distribu­ tion of alleles. By contrast, databases such as the Human Gene Fig. 3 Mutation submission form. Mutation Database (HGMD) provide more limited informa-

324 IN Medicine G,, G,, gangliosidoses databases

Select A

* Rease Select

Beta-&ha Bmcl Actme Sde Infantile Mutabons Late Infade Mutahons Juverulc Mutahons Aduit Mutabons BemerpMutahons All above Mutahons HexaInf6mabon II gntmrUt17S5 61ow~bl381 a-

flg. 4 Viewing of Modeling Data. Selection box (lower right) permits selection of display of mutation groups. Further scrolling reveals controls for rotation of model and choices of backbone views (not shown). Following placement of arrow on pept~dechain, left-clicking on mouse displays name and position of ammo ac~din status line.

tion on all known human mutations and are useful to physi- accounts for 20% of heterozygote alleles in non-Jewish Euro- cians or teachers who require only identification of a specific pean-derived populations. Carriers of TSD from other geo- gene locus. HGMD is limited to five useful fields: codon, nu- graphic isolates often harbor a specific subset of mutations cleotide change, amino acid change, name of associated disor- even if the disease gene frequency is not elevated in the popu- der, and linkage to the abstract of a reference. lation. An IVS 9 + 1 G>A mutation is the most frequent TSD allele in carriers of northern European origin,15 while in the Mutations in the HEM gene Japanese population, 79% of mutant alleles result from a The majority of HEXA mutations cause complete loss of C178Y substitution not found in other populations.I6 The two Hex A and lead to severe, infantile-onset neurodegenerative pseudodeficiency mutations account for 36% of alleles in non- disease culminating in death by age 4. Later onset GM2gangli- Jewish Americans of European descent.17 The remainder of ; osidoses are associated with 20 alleles producing varying TSD mutations are considered to be "rare" or "private." ' amounts of residual enzymatic activity. Patients displaying the Genotype/phenotype correlations I B1 variant phenotype are believed to be deficient in a-subunit catalytic activity.1° Two mutations cause Hex A pseudodefi- HEXA gene mutations associated with variant forms of TSD ciency. Of particular interest to laboratories engaged in allele have been mapped onto a modeled structure of the catalytic testing are the six alleles identified only in carriers, which are domain of Hex A.9 Mutations leading to acute TSD were iden-

I presumed to be disease-causing based on the effect of mutation tified in two clusters: one located at or near the active site and

I on enzyme activity in in vitro assays. Late- or adult-onset TSD including mutations associated with the B1 phenotype, the sec- 1. , 1s characterized by extreme variability of phenotype. This vari- ond cluster involving residues far from the active site which are I ation occurs within the patient population expressing the com- associated with effects on enzyme stability and protein pro- I mon G269S allele (variable expressivity) but also among pa- cessing. Mutations leading to a less severe phenotype were also ' tients expressing five other alleles leading to a late-onset divided into two groups. Benign and adult mutations most ' (chronic) phenotype (heteroallelism). Knowledge of the fre- often involved residues on the protein surface where a change quency and distribution of these alleles should improve the does not exert a significant effect on protein stability. More efficacy of TSD carrier screening. severe late infantile and juvenile mutations affected residues which are located closer to the protein core. Not all phenotypes Ethnic distribution of TSD alleles could be interpreted based on these general principles. ' A 7.6-kb deletion in the first exon and flanking sequences is the most common allele in French Canadians.llJ2 All other Mutations in the HEXBgene HEXA mutations are nucleotide substitutions: (n = 81), small Like TSD, the clinical severity of Sandhoff disease correlates insertions (n = 4), or small deletions (n = 14). Three alleles with residual activity of Hex A. The most common Sandhoff account for 92-98% of HEXA gene pathology in Ashkenazi mutation is a 16-kb deletion including 1-5 and flanking Jews.13.14 The most common, the 1278+TATC insertion, also sequence.I8 All point mutations are associated with subacute Cordeiro et a/.

or chronic forms of disease. Mutations with unusual splicing abnormalities have been identified.19 Asymptomatic individu- Tay-Sachs Disease Screening Data Submission Form als usually show some Hex A activity and no Hex B activity, a ------biochemical phenotype referred to as Hexosaminidase Paris. R*u. - &A ,..-n d I- 8wh. &*- scx*-ddd d &* %Xlnnau-Darsbs.. IW~, Benign mutations resulting in heat labile Hex B have the po- z.a u...d-...",*... r.ar,d,.w ------tential to compromise the accuracy of carrier detection and SubmiUer Information prenatal diagnosis. N- I w-. t------Mutations in the GM2A gene Screenlng Information Only five mutations have been documented in the GM2A >ru r- CdY I gene encoding the G,, activator protein, and all patients have B1.n Re,-. 7 rm I been homozygous for individual mutant alleles associated with complete absence of G,, ganglioside cleavage. The range of prur4.r.h hm,&.h p...n.n.dxyL -.- "..m I&.-- "l.r defects includes defective Hex A binding,'O-" enzyme instabil- - =& NZ/ N.=N.=z =/ ity," and degradation in the ER or Golgi.'" .=..ithuVUld.&id

-*uw r--- II I

Gm-4 I--Conclusion E-a+IAIC 1- - - - The HEXAdb, HEXBdb, and GM2Adb online databases as- nau+la-i.c 1- 77 I sist in regrouping scattered information into a centralized lo-

m1 IIII cation that is easily accessible to the scientific community. The ability to submit new mutations online provides an efficient c-4.T IIII method for collecting and disseminating new information NPPIQA r--- I7I about these genes and their associated phenotypes. Protein - ,----- )7 ( d.1.n.m modeling of individual mutations may also assist in developing -- 777- a better understanding of genotypelphenotype correlations. ? The GM2gangliosidoses databases and others like it serve the &dl*.--? 'I" rxa ~-~~--=.J rh-bmm interests of genomics, human genetic diversity, medicine, and the needs of patients, families, and communities. *pr.u.-l w m.-d =z:*d m dlm*mrwr -w 0tb.r - InlMM N-0, N"w.l~rlNbd irdrUL.n(. & 4- - Acknowledgments dUa(lu@Ua(lu - 7I Ir--- This work was supported by grants from the Medical Re- r--IIl-'jj search Council of Canada to Peter Hechtman and Feige Kaplan and from the Reseau Genetique Humaine Appliquee to Feige ua +TAT< 77 I I 1-A Kaplan. The GM2gangliosidoses home page is located at the n8u+mrc 1-7 II Debelle server of the Charles R. Scriver Biochemical Genetics Unit in the Montreal Children's Hospital. n IIII)j

c-4.1 III I -y References nss*I.a* IITIT-Z 1. lohnson WG. The illnical spectrum of hexosaminldase deficiency diseases. Neurol- cyy 1981;31:1453-1456. -.5,,., -.5,,., ,---r----r----r--~~ d.1- 1. Kolodny E. Raghavan S. G,,,-gangliosidos~s. Hexosamlnidase mutations not of the Tay-Sachs t).pe produce unusual illniial variants. Tro~dsNr~iroso 1983;6:16-20. oa"mId.5" 1r--- Ir--- 3. Navon R. Argov 2. Frisch A. Hexosaminidase A deficiency in adults.Am IMed Genet ? I 1986;21:179-196.

hdr).--r "L.. rtTa PDalb 4. Navon R. Aloleculdr and clinical heterogeneity of adult GbILgangliosidosis. Dn, lnw dralaala--=I Ne.lrroscl 1991;13.295-298. 5. Federiio A. Palmeri S. Malandrini A. Fabrizi G. Mondelli M. Guazzi GC. Theclinical , I-] aspects of adult henoadminldase deficiencies. Dui, Neuross~199 1;13:280-287. ------6. Cotton RG. McKus~ckV, Scr~verCR. The HUGO Mutation Database Initiative. IKsGk&l€3B3kl Sclcnr-c 1998279: 10-1 I -..#.lwrh--~~-r- m~~~~-drvasd--~b~~ 7. Sirlver CR. Nowackl PM. Lehvaslaiho H. Guidelines and recommendations for

.Am=---. content, structure. and deployment of mutation datdbases. Hum Miitat 1999i1.1: 4 344-350. 8. Scriver CR. Nowacki Phi. Lehvaslaiho H. Guidelines and recommendations for Fig. 5 Tay-Sachs d~seabe\creening data suhmiaa~onform. This form is intended for use content, structure, and deployment of mutation databases, 11: journey in progress. hy laboratories conduct~ngcommunity-based Tny Sachs carrier screening. The form lists Hiin1 Mutot 2000;15:13-15. only those allele\ known to he rclatlvely frequent and aska for allele frcquency data and 9. Tews 1. Perrakis A, Oppenheim A, Dauter 2, Wilson KS. Vorgias CE, Bacterial ethn~corigin of carriers. Thi\ information IS expected to be useful in selecting appropriate chitobiase structure provides Insight into catalytic mechanism and the basis of T~~. diagnobtlc tests for local populations/f~nl~lies. Sachs d~sease.Not Struct Btol 1996;3:638-648.

G% tics IN Medicine G, G, gangliosidoses databases

10. Bayleran I, Hechtman P, Kolodny E. KabackM. Tay-Sachs disease with hexosamlni- 18. Neote K, McInnes B, Mahuran DJ. Gravel RA. Structure and distribution of an dase A: characterization of the defective enzyme in two patients. Am J Hum Genet Alu-type deletion mutation in Sandhoff dlsease. J Clin Ini,ect 1990;86:1524- 1987;41:532-548. 1531. 11. Myerowitz R, Hogikyan ND. Different mutations in Ashkenazi Jewish and non- 19. Wakamatsu N, Kobayashi H. Miyatake T, Tsuji S. A novel exon mutation in the Jewish French Canadians with Tay-Sachs disease. Science 1986;232:1646-1648. human beta-hexosaminidase beta subunit gene affects 3' splice site selection. I Blol 12. Myerowitz R, Hoglkyan ND. A deletion involving Alu sequences in the heta-hex- Chem 1992;267:?406-2413. osaminidase alpha-chain gene of French Canadians with Tay-Sachs disease. J Biol 20. Schroder M, Schnabel D. Suzuki K. SandhoffK. A mutation in the gene of a glyco- Chem 1987;262: 15396-15399. lipid-binding protein (G,,, activator) that causes G,,,-gangliosidosis vanant AB. 13. Triggs-Raine BL, Feigenbaum AS, Natowicz M, Skomorowskl MA, Schuster SM. FEBS Len 1991.290: 1-3. Clarke IT, Mahuran Dl, Kolodny EH. Gravel RA. Screening for carriers ofTay-Sachs 21. Xie B. Wang W. Mahuran Dl. A Cysl38-to-Arg substitution in the G,? activator disease among Ashkenazi Jews: a comparison of DNA-based and enzyme-based protein is associated with the AB varlant form of GM2 gangllosidosis. Am I HII~ tests. N Engl J Med 1990;323:6-12. Genet 1992:50:1046-1052. 14. Fernandes MI. Kaplan F, Clow CL, Hechtman P, Scriver CR. Specificity and sensi- 22. Xie B, Rigat B, Smiljanic-Georgijev N. Deng H, Mahuran D. Biochemical character- tivity of hexosaminidase assays and DNA analysis for the detection of Tay-Sachs ization of the Cysl38Arg substitution associated w~ththe AB variant form of GM2 disease gene carriers among Ashkenazic Jews. Gen Epidemiol 1992;9:169-175. gangliosidosis: evidence that Cys138 is required for the recognition of the G,, 15. Akerman BR. Zielenski I, Triggs-Raine BL, Prence EM, Natowicz MR, Lim-Steele IS, Kahack MM, Mules EH, Thomas GH, Clarke JT. A mutation common in non- activator/G,,, gangl~osidecomplex by beta-hexosaminidase A. Blochemistry 1998; Jewish Tay-Sachs disease: frequency and RNA studies. Hum Mutat 1992;1:303-309. 37:814-821. 16. Tanaka A. Sakazah H, Murakami H, Isshiki G. Suzuki K. Molecular genetics of 23. Schroder M, Schnabel D, Hunvitz R. Young E. Suzuki K. Sandhoff K. Molecular Tay-Sachs disease in Japan. 1 Inherzt Metab Dis 1994;17:593-600. genetlcs of G,,-gangl~osidosis AB variant: a novel mutation and expression in BHK 17. Kaback M, Lim-Steele J, Dabholkar D. Brown D, Levy N, Zeiger K. Tay-Sachs dis- cells. Hum Genet 1993;92:437-440. ease-carrier screening, prenatal diagnosis, and the molecular era: an International 24. Schepers U, Glombitza G. Lemm T, Hoffmann A, Chabas A. Ozand P, Sandhoff K. perspective, 1970 to 1993. The Internat~onalTSD Data Collection Nehvork. JAMA Molecular analysis of a G,,,-act~vator deficiency in two patients with G,,-gangli- 1993;270:2307-2315. osidosis AB variant. Ain I Huin Genet 1996;59:1048-1056.