The Tay-Sachs Disease Gene in North American Jewish

Am J Hum Genet 35:1258-1269, 1983 The Tay-Sachs Disease Gene in North American Jewish Populations: Geographic Variations and Origin GLORIA M. PETERSEN,1 JEROME I. ROTTER,' RITA M. CANTOR,' L. LEIGH FIELD,2 SUSAN GREENWALD,1 JOYCE S. T. LIM,1 CHITRA Roy,' VICTOR SCHOENFELD,' J. ALEXANDER LOWDEN,3 AND MICHAEL M. KABACK' SUMMARY From data collected in a North American Tay-Sachs disease (TSD) heterozygote screening program, the TSD carrier frequency among 46,304 Jewish individuals was found to be.0324 (1 in 31 individuals). This frequency is consistent with earlier estimates based on TSD incidence data. TSD carrier frequencies were then examined by single country and single region of origin in 28,029 Jews within this sample for whom such data were available for analysis. Jews with Polish and/or Russian ancestry constituted 88% of this sample and had a TSD carrier frequency of.0327. No TSD carriers were observed among the 166 Jews of Near Eastern origins. Relative to Jews of Polish and Russian origins, there was at least a twofold increase in the TSD carrier frequency in Jews of Austrian, Hungarian, and Czechoslovakian origins (P <.005). These findings suggest that the TSD gene proliferated among the antecedents of modern Ashkenazi Jewry after the Second Diaspora (70 A.D.) and before their major migrations to regions of Poland and Russia (before 1100 A.D.). Received December 27, 1982; revised March 2, 1983. This research was supported in part by a contract from the Department of Health, State of California (Genetic Disease Branch), a grant from the National Tay-Sachs Disease and Allied Diseases Association, a grant from the Gould Family Foundation, an Ontario Graduate Scholarship (L. L. F.), and National Research Council of Canada Postgraduate Scholarship 1560 (L. L. F.). l Division of Medical Genetics, Departments of Pediatrics and Medicine, Harbor-UCLA Medical Center, Torrance, CA 90509. 2 Division of Medical Genetics, Department of Pediatrics, University of Calgary, Alberta Children's Hospital, Calgary, Alberta T2T5C7, Canada. 3 Research Institute, Hospital for Sick Children, Toronto, Ontario M5GIX8, Canada. 1983 by the American Society of Human Genetics. All rights reserved. 0002-9297/83/3506-0019$02.00 1258

TAY-SACHS DISEASE GENE 1259 INTRODUCTION Tay-Sachs disease (TSD) is a degenerative disease of the nervous system due to deficient activity of the enzyme hexosaminidase A (HEX A) and the consequent intralysosomal accumulation of the natural substrate GM2 ganglioside [1]. TSD is inherited as an autosomal recessive disorder and occurs with the highest-known frequency in Jewish infants of European extraction (Ashkenazim). Based on Hardy-Weinberg estimates from TSD incidence, the frequency of heterozygotes (TSD carriers) in the Jewish population has been estimated to be between 1/25 and 1/45, and that in the non-jewish population to be between 1/300 and 1/400 [2, 3]. It is likely that more than one mutational event resulted in the present worldwide distribution of the HEX A deficiency genes [4-6]. There have been several attempts to investigate the geographic origin of the TSD gene in Jews. Kozinn et al. [2] reported that most of 44 parents of TSD patients came from Eastern Europe. Aronson and Volk [7, 8] determined the origins of grandparents and great-grandparents of infants with TSD in New York City. They found that a greater proportion of TSD ancestors were born in the Polish-Russian provinces of Grodno, Suwalki, Vilno, and Kovno (part of the "Pale"), compared with unmatched "control grandparents," although no statistical computations were reported. Similar findings were obtained by Myrianthopoulos and Aronson [9, 10] in an expanded study. Goldschmidt et al. [11] determined the ethnic distribution of the parents of 41 TSD infants in Israel and reported that 51% originated from Poland. Inferences about the relationship of evolutionary mechanisms and historical events to Jewish TSD geographic ancestry have been attempted. From the above observations, Myrianthopoulos and Aronson, expanding on an earlier suggestion by Kozinn et al. [2] and Aronson et al. [ 12], argued that the TSD gene proliferated among European Jews, and, more specifically, those in Poland and Russia, after the Second Diaspora (dispersion of Jews from the land of Israel after the destruction of the Second Temple by the Romans in 70 A.D.), possibly through the mechanism of heterozygote selective advantage [9, 10, 13, 14]. Chase and McKusick [15] proposed alternatively that founder effect and random genetic drift explain the high TSD gene frequency in this population. Regardless of the mechanism, if indeed the TSD gene has its highest frequency in Jews of Polish-Russian ancestry, this suggests a relatively recent historical increase in the frequency of the TSD allele, that is, around the time of the major Jewish migration to Eastern Europe, which occurred after 1000 A.D., and probably after the Crusades (1200 A.D.) [16-17]. In a detailed analysis of the evolutionary considerations of TSD in the Ashkenazim, Neel [5] rejected genetic drift as an explanation for the high gene frequency because of the substantial size and subdivided nature of the Jewish population in Eastern Europe since the Middle Ages (i.e., drift would have had to occur in parallel in the many small Jewish population clusters). However, there are difficulties with the heterozygote selection hypothesis as well. It has been suggested that the geographic origin of the TSD gene could possibly be resolved by analysis of countries of origin of carriers determined directly by heterozygote testing [5, 18].

1260 PETERSEN ET AL. O'Brien et al. [19] demonstrated that TSD heterozygotes could be detected by partial reduction in HEX A activity in serum and leukocytes. Standardization and automated assays for HEX A activity [20-21 ] soon enabled the establishment of community screening programs directed to the population at highest risk [ 18, 22-24]. Our study was conducted (1) to obtain an estimate of the Jewish TSD heterozygote frequency and its ancestral variation by direct count of individuals screened for Tay-Sachs disease, and (2) to test the hypothesis of a Polish-Russian origin of the TSD gene in the Jewish population by estimating TSD heterozygote frequencies by European country of origin. STUDY SAMPLE AND METHODS The study sample comprised those individuals participating in a voluntary TSD screening program directed at colleges, universities, and Jewish community facilities throughout the United States, primarily California. The background, methods, and organization of this and allied programs have been described [18, 22-24]. Data on 50,546 Jewish individuals were collected over a 9-year period (I1974-1982). In addition, an independent data sample was available from the Tay-Sachs Disease Screening Program in Toronto, Canada [25], based on 13,117 screened individuals (1974-1919). At the time of screening, individuals were administered a questionnaire that included items pertaining to religion and ethnicity, countries of origin of ancestors, and any known blood relationship to individuals who were either TSD patients or TSD gene carriers. Venous blood samples were drawn from each individual for serum screening of HEX A activity using established methods [20, 21]. Based on initial HEX A analyses, individuals were assigned to four possible categories: (1) noncarrier; (2) carrier; (3) inconclusive/ probable noncarrier (I-PNC); and (4) inconclusive/probable carrier (I-PC). In the event that the serum test was inconclusive (categories 3 and 4), individuals were asked to be retested both by serum and white cell HEX A determinations. Bivariate criteria were used for assignment of individuals to each diagnostic category. These included percent HEX A by heat fractionation of serum HEX A, and the specific activity of serum HEX A. TSD carriers had percent HEX A ranges of 40.9% + 5.1% (SD) and specific HEX A activities of 281.2 + 69.7 (SD) nm substrate hydrolyzed/hr per ml serum. Noncarrier individuals had 62.6% + 5.7% (SD) HEX A and specific HEX A activities of 468.1 + 75.9 (SD) nm/hr per ml serum. An inconclusive diagnosis was assigned to those individuals with percent HEX A between the range of 51% to 55%; in addition, the probable carriers in this inconclusive group were those with HEX A specific activity < 350 nm/hr per ml serum. HEX A activity in white cells was measured in individuals with inconclusive diagnoses who returned for testing. With this white cell test, TSD carriers have 20%-55% HEX A, while noncarriers show > 60% HEX A activity. In essentially all instances, diagnoses of serum-inconclusive individuals were resolved by white cell testing. Excluded from analysis for the following reasons were 4,242 individuals: (1) known blood relative of TSD patient (no. = 788); (2) known blood relative of TSD gene carrier (no. = 1,725); and (3) incomplete questionnaire (no. = 1,729). Individuals in the first two categories were excluded because they were not randomly ascertained; their inclusion would inflate carrier frequency estimates. Our estimate of the overall Jewish TSD carrier frequency was based on the remaining 46,304 individuals. To test the origin hypothesis, we made further exclusions on the following bases: (1) origin indicated as U.S.A., Israel, Canada, Australia/New Zealand, or South Africathese were judged by us as representing recent diverse ethnic mixtures that would not provide useful information on origin (no. = 132); (2) origin indicated as multiple countries from a variety of regions for which no proportionate contributions could be estimated (no. = 16,689); and (3) no country or region of origin indicated on questionnaire (no. =

TAY-SACHS DISEASE GENE 1261 1,454). The above exclusions left a sample of 28,029 (suitable for ancestral-origin analysis), consisting of Jewish individuals who specified a single country or region of ancestral origin. We divided the ancestral-origin data into eight world regions reflecting geographic boundaries and/or the historical-geographic divisions of the Jewish people. The number of carriers was adjusted to account for individuals in both I-PNC and I-PC diagnostic categories who did not return for retesting. This adjustment was based on proportions of inconclusive individuals who, on white cell retesting, were found to be carriers. In this group, 11.2% of 250 persons initially diagnosed I-PNC were found to be carriers on retesting, and 64.3% of 140 persons initially diagnosed I-PC were found to be carriers. These adjustments were applied to the groups of inconclusively diagnosed persons who were not retested, and the resulting carriers added to those already definitely identified. The sizes of the inconclusive groups were small enough so that the adjustments did not increase the number of carriers of Western European, Mediterranean, Asian, African, or Latin American origins, but did increase the number of carriers by 9% in the Middle European region and by 12% in the Poland-Russia region. All analyses were performed both with and without these adjustments (i.e., defining the inconclusive individuals as carriers or not). In either case, the results were parallel and consistent. Sitice the adjusted figures more accurately reflect the true TSD carrier frequencies, they are reported here. Diagnoses and country of origin were cross-tabulated; heterozygote frequencies were computed by dividing the total number of carriers by the total tested from the same country of origin. A previous study [6] showed that an estimated 2%-3% of these apparent TSD heterozygotes in the Jewish population are probably not true TSD gene carriers, but rather that they are carriers of variant alleles, which have an aggregate carrier frequency of 1/ 1,000. We corrected our data for these variants by reducing the carrier frequencies in each group by.001. Differences in TSD carrier frequencies were tested both by the chi-square statistic and by the z-approximation for testing equality of proportions [26]. RESULTS Based on direct assessment, we determined the overall TSD carrier frequency in the Jews tested to be.0324 with a standard error of +.0008, giving a 95% confidence interval of (.0308,.0340). TSD carrier frequencies are listed by world region of origin in table 1. No carriers were detected in our sample of 166 Near Eastern Jews. This difference indicates a significantly lower frequency (P <.025) of TSD heterozygotes in this group, relative to the Polish-Russian Ashkenazi Jewish sample. We found a significantly increased TSD carrier frequency in that group of individuals who indicated Middle Europe as their ancestral origin relative to Jews of Polish/Russian origin (the latter comprising 88% of the single country of origin sample). To investigate this observation, we examined the TSD carrier frequencies by individual countries within Middle Europe and in Poland and Russia. With the data sorted by individual countries of origin for all of Europe (see table 2 and fig. 1), the TSD carrier frequencies are significantly higher for Austria (.1092), Czechoslovakia (.0853), and Hungary (.0723), rather than for Poland (.0343), Russia (.0335), or Germany (.0269). In addition, further tests of difference in proportion indicate that the TSD carrier frequency of the Austria/Czechoslovakia/ Hungary region (obtained by combining the results from individual countries) differs significantly from the frequencies both of the Poland/Russia region and of Germany (P <.0001).

1262 PETERSEN ET AL. TABLE 1 TSD CARRIER FREQUENCIES IN U.S. JEWS BY REGION OF ORIGIN Significance relative to Region TSD carrier frequency* (± SE) No. tested Poland-Russia.. Poland-Russia....0327 ±.0012 24,572 Middle Europet....0478 ±.0038 2,928 P <.001 Western Europet....0203 ±.0113 148 N.S t Mediterraneant....0370 ±.0178 108 N.S. Near East....0000 166 P <.025 Africall....0000 78 N.S. Latin America#....0556 ±.0529 18 N.S. Asia**....0909 ±.0849 11 N.S. Mixed originstt....0300 ±.0012 18,275 N.S. Overall Jewish....0324 ±.0008 46,304... * Adjusted for variant alleles; see STUDY SAMPLE AND METHODS. t For individual countries, see table 2. * N.S. = not significant. Includes: Arabia, Armenia, Afghanistan, Iran, Iraq, Jordan, Lebanon, Syria, and Yemen; excludes Israel. Includes: all of African continent, including North Africa. Includes: Caribbean and North and South America; excludes: USA and Canada. ** Includes: Indian subcontinent; excludes: USSR. tt Includes: those who did not specify an origin, or who indicated more than one region of origin, or who specified U.S.A., Israel, Canada, Australia/New Zealand, or South Africa as origin; see STUDY SAMPLE AND METHODS. We were concerned that the observed higher TSD carrier frequency in Austria, Hungary, and Czechoslovakia, even though statistically significant, might be due to sampling variation. Therefore, as an independent second test, we computed the frequency in those groups of individuals who indicated as their origin a mixture of these countries. None of these individuals was included in the above single-country-of-origin analysis. Table 3 shows that the group of mixed origin from the Austria/Hungary/Czechoslovakia region has a TSD carrier frequency (.0928) significantly higher than that of any group of mixed origins from Poland, Russia, or Germany (.0278-.0369). Data from an independent TSD screening program in Canada also show a higher carrier frequency in Jews who had originated from Middle Europe relative to Poland and Russia (table 4). No statistical tests are reported for these data because relatives of TSD patients or carriers were not excluded and because of the smaller sample size. However, the fact that the Canadian data are congruent with the U.S. results further supports these findings, that is, that Jews of Polish and Russian origins generally have a lower TSD carrier frequency than those of Middle European origins. Notably, in this sample, Jews of Austrian origin have the highest TSD carrier frequency. DISCUSSION We observe by direct testing that the TSD carrier frequency in our U.S. Jewish sample is one in 31, which is consistent with indirect estimates based on disease incidence data from other locations [2, 3]. We provide the first analysis of TSD carrier frequencies in an American Jewish population by defined worldwide origins,

TAY-SACHS DISEASE GENE suggesting by inference the frequency of TSD carriers in the migrant Jewish populations from those countries and regions. The low TSD carrier frequency in our sample of American Jews of Near Eastern origins provides the first direct data indicating a significantly higher TSD gene frequency in the Ashkenazi Jewish population than in other Jewish populations, TABLE 2 TSD CARRIER FREQUENCIES IN U.S. JEWS BY EUROPEAN COUNTRY OF ORIGIN Country TSD carrier frequency* (± SE) No. tested 1263 Poland... Russia... Mixed origin: Poland-Russiat....0343 -+.0028.0335 +.0016.0327 ±.0018 3,929 11,268 9,375 Total Poland/Russia... Austria... Bulgaria... Czechoslovakia... Germany... Hungary... Rumania... Yugoslavia... Mixed origin: Middle Europet... Total Middle Europe... Belgium... France... Netherlands... Scandinavia... Switzerland... United Kingdom... Mixed origin: Western Europet... Total Western Europe... Greece... Italy/Sicily/Malta... Spain/Portugal... Turkey/Cyprus... Mixed origin: Mediterraneant... Total Mediterranean... Total European Jewish Sample....0327 ±.0012 24,572.1092 ±.0231 174.0000 8.0853 ±.0241 129.0269 ±.0039 1,597.0723 ±.013911 332.0383 ±.0130 209.0000 5.0781 ±.0121 474.0478 ±.0038 2,928.0000 3.0000 30.0500 ±.0477 20.0000 7.0000 5.0339 ±.0231 59.0000 24.0203 ±.0113 148.0303 ±.0292 33.0476 ±.0456 21.0833 ±.0782 12.0333 ±.0321 30.0000 12.0370 ±.0178 108.0333 +.0011 27,756 * Adjusted for variant alleles; see text. t Consists of those individuals who indicated as origin more than one country from within this region only. f P <.0001, relative to Poland only plus Russia only. P <.004, relative to Poland only plus Russia only. P <.005, relative to Poland only plus Russia only.

1264 PETERSEN ET AL. FIG. 1.-TSD carrier frequencies and nos. tested (n) of U.S. Jews by single country of ancestral origin in Europe, Poland, and Russia. Shaded countries have significantly higher carrier frequencies relative to Poland and Russia. Western Europe and Mediterranean regions are also shown (see table 2 for specific countries in each region). as previously suggested by clinical observations of disease occurrence [6, 11, 27] Ȯf greatest interest is the observed higher carrier frequency in American Jews of Middle European origin relative to American Jews of Polish-Russian origin. The Middle European value appears to be due to the significantly higher TSD carrier frequencies in American Jews of Austrian, Hungarian, and Czechoslovakian origin when compared with those in American Jews of Polish or Russian origin. This was consistent in three independent samples: our single-country-of-origin group, our mixed-origin group, and the Canadian sample. Earlier studies [7-11] reported that the ancestors of TSD infants appeared to derive from predominantly Polish and Russian origins. It was inferred from those findings that the highest frequency of the Tay-Sachs gene occurred in populations from these regions. The results reported here are at variance with this inference. We offer several considerations regarding this discrepancy: (1) The distribution of ancestral origins in the immigrant Jewish population of North America is overwhelmingly weighted toward Poland and Russia [28-311. For example, from 1880 to 1924, the American Jewish population increased 16- fold from 280,000 to 4,500,000, the increase being due to a massive migration of Jews from Eastern Europe [16, 29-31]. In our own single-country-of-origin sample, 88% indicated Polish or Russian origins, and in the Canadian sample, the proportion was 93%. (2) As a consequence, Jews from Austria, Hungary, and Czechoslovakia would be relatively small in number and an increased incidence of TSD would be difficult

TAY-SACHS DISEASE GENE TABLE 3 TSD CARRIER FREQUENCIES IN U.S. JEWS OF MIXED MIDDLE EUROPEAN, POLISH, AND RUSSIAN ORIGINS 1265 TSD carrier frequency* Significance relative to Aus- Countries of origin (+ SE) No. tested tria/hungary/czechoslovakia Austria/Hungary/Czechoslovakia.0928 ±.0288 97 Poland/Russia...0327.0018 9,375 P <.004 Poland/Germany....0369.0072 650 P <.017 Russia/Germany....0278.0036 1,981 P <.002 Poland/Russia/Germany.... 0309.0044 1,487 P <.004 * Adjusted for variant alleles; see text. to detect in this group. Therefore, assessment of TSD patients' ancestral origins by enumeration [5, 11] (i.e., counting and comparing numbers of parents from each country) might be insufficient, and any results must be interpreted with caution. (3) Aronson and Volk [7] attempted to overcome this problem by determining the relative distribution of 165 Jewish TSD patients' grandparental origins compared with those of healthy Jewish controls. Myrianthopoulos and Aronson [3, 8-10] obtained similar results on an augmented sample of 284 Jewish TSD patients; the highest concentration of TSD grandparents appeared to be from certain provinces of Poland and Russia. "TSD ratios" of grandparents of 388 TSD parental sibships vs. 812 control sibships were computed by country of origin [9]. The highest TSD ratios (interpreted as highest concentration of TSD genes, although no statistical tests were done) were found in the southern Kovno, Suwalki, and Grodno provinces of Poland-Russia, and lower TSD ratios were found for the rest of TABLE 4 TSD CARRIER FREQUENCIES IN CANADIAN JEWS OF POLISH, RUSSIAN, AND MIDDLE EUROPEAN ORIGINS Country of origin TSD carrier frequency* No. tested Poland..0674 3,219 Russia..0550 2,147 Mixed origin:. Poland/Russiat....0629 2,800 Total Poland/Russia.0626 8,166 Austria..0976 82 Czechoslovakia..0776 116 Germany..0833 60 Hungary..0723 166 Rumania..0737 190 Total Middle Europe.0782 614 * Relatives of TSD patients and TSD carriers are not excluded from this sample. t Consists of those individuals who indicated as origin both of these countries.

1266 PETERSEN ET AL. Poland, Germany, Hungary, Rumania, and other parts of Russia. No values were given for Czechoslovakia or Austria. This indirect method makes the important assumption that the geographic origins of the American Jewish population are accurately represented in the non-tsd controls. (4) We observe a significantly higher TSD carrier frequency in the Middle European group in this study for two reasons: first, we examined the carrier state directly; and second, we had available an extremely large sample. In addition, it is unlikely that the high TSD carrier frequency in the Austria/ Hungary/Czechoslovakia population is due to nonrandom distribution of non- TSD variant alleles: the reported incidence of classical TSD in infants of these ancestral origins [10, 14] is compatible with the frequencies we report here. If the high frequency were due to non-tsd variants, we would not expect classical TSD phenotypes in Jewish offspring of these origins. Further, it is unlikely that our results are biased because of undetected potential clustering of families of these origins in our data because we observed this finding in three independent data sets: the single-country-of-origin group, our mixed-origin group, and the Canadian single-country-of-origin group, all of which are consistent. For these reasons, we are confident in concluding that there is a significant observed increase in the TSD gene frequency in Jews of Austrian, Hungarian, or Czechoslovakian origins. This unexpected result suggests by inference that the TSD gene had a significantly higher frequency in the Jewish population living in the Austria/Hungary/Czechoslovakia region as compared with that in Poland/ Russia. There are several important implications of this observation. First, since the Jewish migration into Poland and Russia occurred after those into Middle Europe, the increase of the TSD gene frequency in the Ashkenazi Jewish population must have preceded the Polish-Russian migrations (before 1 100 A.D.) [16, 17]. Second, the low TSD carrier frequency in Jews of Near Eastern origin suggests that the increase of the gene occurred after the Second Diaspora (70 A.D.). In this regard, our results are consistent with the possibility that the TSD gene reached its relatively high frequency in the Polish-Russian Jewish population by admixture with a Jewish population in Middle Europe that already had a high TSD gene frequency. This potentially solves the objections to either the drift [15, 32-34] or the selection hypotheses [8-10, 35-37] when applied to the dispersed Jewish population that both migrated to and existed in the Polish- Russian Pale: the gene frequency could have increased by a combination of drift, selection, or admixture in a more defined single population that was confined to a more restricted geographic area in Middle Europe. Third, the Khazar people, a Turkic tribe that lived to the north of the Caucasus Mountains from 400 to 1200 A.D. and who converted to Judaism during this period, have been suggested by Neel [5] (based on Koestler's hypothesis [38]) as a possible source of the TSD "private polymorphism" for the Ashkenazic Jewish population. There is historical evidence of specific migrations of Khazars to Magyar territory (including present-day Hungary). Our results are consistent with this possibility. Further, Lowden, in analyzing the high TSD gene frequency among Canadian Jews of Polish-Russian origin, observed that the Russian Jews in his study tended to come from southern Russia [25].

TAY-SACHS DISEASE GENE In addition, the observation of a higher TSD gene frequency in Austria/Hungary/ Czechoslovakia might resolve a major objection to the selection hypothesis and its proposed agent, tuberculosis (TB). Myrianthopoulos and Aronson, and Myrianthopoulos and Melnick, [9, 10, 14] hypothesized that the TSD gene reached its high frequency in the Ashkenazim by providing heterozygote selective advantage against pulmonary TB because they observed a decreased frequency of TB in Jews who were born in the Polish and Russian regions with the purportedly high TSD gene frequencies. However, if TB were a selective agent for the TSD gene, a higher frequency of TB in areas with a high frequency of the TSD gene would be expected, rather than what was observed. Second, the TSD gene does not occur at such a high frequency that a substantial proportion of the population are carriers, as with the sickle-cell gene in malarial areas. Thus, only a small minority of the Ashkenazi Jewish population (1/30) may have been "protected" TSD carriers, while the remaining 29/30 were still susceptible to TB. It would seem unlikely that this could lead to an observable decreased frequency in TB, particularly in the time period envisioned by Myrianthopoulos and Aronson. Others have argued [5, 33, 34] against any selective advantage conferred on TSD gene carriers in a TB endemic area. However, examination of the Myrianthopoulos and Aronson [10] data on the relative frequencies of Jewish TB patients by country of origin reveals an interesting phenomenon: Jews born in Austria and Hungary constitute a sizable proportion (approximately 35%) of their TB sample. This is much greater than the proportion in our own large sample (2.6%), and still greater than the proportion in their controls (12%). Ironically, this latter observation lends support to their TB selection hypothesis in that a high incidence of TB is found in the same region with a high frequency of the TSD gene. Finally, a clinical implication of this study is that offspring of Jews whose ancestry includes Austria, Hungary, or Czechoslovakia appear to be at even higher risk for TSD than the general high risk for Ashkenazi Jews. In summary, we (1) determined the TSD carrier frequency in a large sample of U.S. Jews to be.0324; (2) observed a lower TSD carrier frequency in Jews of Near Eastern origins relative to Jews of Polish-Russian origins; and (3) observed a higher TSD carrier frequency in Jews of Austrian, Hungarian, and Czechoslovakian origins relative to Jews of Polish and Russian origins. Regardless of mechanisms for the increase of the TSD gene frequency in the Ashkenazi Jewish population, these observations suggest that the period of increase occurred before their major migrations to regions of Poland and Russia (i.e., before 1100 A.D.). ACKNOWLEDGMENTS 1267 We thank Michael Kirk, Beth Kudelsky, and Phyllis Limberg for their contributions to this study. REFERENCES 1. OKADA S, O'BRIEN JS: Tay-Sachs disease: generalized absence of a beta-d-nacetylhexosaminidase component. Science 165:698-700, 1969 2. KoZINN PJ, WIENER H, COHEN P: Infantile amaurotic family idiocy: a genetic approach. J Pediatr 51:58-64, 1957

1268 PETERSEN ET AL. 3. MYRIANTHOPOULOS NC, ARONSON SM: Population dynamics of Tay-Sachs disease. I. Reproductive fitness and selection. Am J Hum Genet 18:313-327, 1966 4. PATRICK AD: Genetic variants of hexosaminidase deficiency, in Tay-Sachs Disease: Screening and Prevention, edited by KABACK MM, RIMOIN DL, O'BRIEN JS, New York, Alan R. Liss, 1977, pp 147-159 5. NEEL JV: History and the Tay-Sachs allele, in Genetic Diseases among Ashkenazi Jews, edited by GOODMAN RM, MOTULSKY AG, New York, Raven Press, 1979, pp 285-299 6. GREENBERG DA, KABACK MM: Estimation of the frequency of hexosaminidase A variant alleles in the American Jewish population. Am J Hum Genet 34:444-451, 1982 7. ARONSON SM, VOLK BW: Genetic and demographic considerations concerning Tay- Sachs disease, in Cerebral Sphingolipidoses, edited by ARONSON SM, VOLK BW, New York, Academic Press, 1962, pp 375-394 8. ARONSON SM: Epidemiology, in Tay-Sachs Disease, edited by VOLK BM, New York, Grune and Stratton, 1964, pp 118-153 9. MYRIANTHOPOULOS NC, ARONSON SM: Reproductive fitness and selection in Tay- Sachs disease, in Inborn Errors of Sphingolipid Metabolism, edited by ARONSON SM, VOLK BW, Oxford, England, Pergamon Press, 1967, pp 431-441 10. MYRIANTHOPOULOS NC, ARONSON SM: Population dynamics of Tay-Sachs disease. II. What confers the selective advantage upon the Jewish heterozygote? in Sphingolipids, Sphingolipidoses and Allied Disorders, edited by VOLK BW, ARONSON SM, New York, Plenum Press, 1972, pp 561-570 11. GOLDSCHMIDT E, LENZ R, MERIN S: Tay-Sachs disease, in Genetics of Migrant and Isolate Populations, edited by GOLDSCHMIDT E, Baltimore, Williams and Wilkins, 1963, pp 290-291 12. ARONSON SM, ARONSON E, VOLK BW: A genetic profile of infantile amaurotic family idiocy. Am J Dis Child 98:50-65, 1959 13. MYRIANTHOPOULOS NC: Some epidemiologic and genetic aspects of Tay-Sachs disease, in Cerebral Sphingolipidoses, edited by ARONSON SM, VOLK BW, New York, Academic Press, 1962, pp 359-374 14. MYRIANTHOPOULOS NC, MELNICK M: Tay-Sachs disease: a genetic historical view of selective advantage, in Tay-Sachs Disease: Screening and Prevention, edited by KABACK MM, RIMOIN DL, O'BRIEN JS, New York, Alan R. Liss, 1977, pp 95-106 15. CHASE GA, McKusICK VA: Founder effect in Tay-Sachs disease. Am J Hum Genet 24:339-340, 1972 16. ANKORI Z: Origins and history of Ashkenazi Jewry (8th to 18th century), in Genetic Diseases among Ashkenazi Jews, edited by GOODMAN RM, MOTULSKY AG, New York, Raven Press, 1979, pp 19-46 17. GRAYZEL S: A History of the Jews. Philadelphia, Jewish Publishing Society of America, 1963 18. KABACK MM, ZEIGER RS, REYNOLDS LW, SONNEBORN M: Tay-Sachs disease: a model for the control of recessive genetic disorders, in Birth Defects, Proceedings of the Fourth International Conference, edited by MOTULSKY AG, LENZ W, Amsterdam, Excerpta Medica, 1974, pp 248-262 19. O'BRIEN JS, OKADA S, CHEN A, FILLERUP DL: Tay-Sachs disease: detection of heterozygotes and homozygotes by serum hexosaminidase assay. N Engl J Med 283:15-20, 1970 20. KABACK MM: Thermal fractionation of serum hexosaminidases: applications to heterozygote detection and diagnosis of Tay-Sachs disease, in Methods of Enzymology, vol 28, edited by GINSBURG V, NEUFELD E, New York, Academic Press, 1973, pp 862-867 21. LOWDEN JA, SKOMOROWSKI MA, HENDERSON F, KABACK MM: The automated assay of hexosaminidases in serum. Clin Chem 19:1345-1349, 1973

TAY-SACHS DISEASE GENE 1269 22. KABACK MM, ZEIGER RS: Heterozygote detection in Tay-Sachs disease: a prototype community screening program for the prevention of recessive genetic disorders, in Sphingolipids, Sphingolipidoses, and Allied Disorders, edited by VOLK BW, ARONSON SM, New York, Plenum Press, 1972, pp 613-632 23. KABACK MM, ZEIGER RS, REYNOLDS LW, SONNEBORN M: Approaches to the control and prevention of Tay-Sachs disease, in Progress in Medical Genetics, vol 10, edited by STEINBERG AG, BEARN AG, New York, Grune and Stratton, 1974, pp 103-134 24. KABACK MM, RIMoIN DL, O'BRIEN JS, EDS: Tay-Sachs Disease: Screening and Prevention, New York, Alan R. Liss, 1977 25. LOWDEN JA: Tay-Sachs screening and prevention: the Canadian experience, in Tay- Sachs Disease: Screening and Prevention, edited by KABACK MM, RIMOIN DL, O'BRIEN JS, New York, Alan R. Liss, 1977, pp 37-46 26. DIXON WJ, MASSEY FJ: Introduction to Statistical Analysis. New York, McGraw- Hill, 1969 27. GOLDSCHMIDT E, LENZ R, MERIN S, RONEN A, RONEN I: Frequency of the Tay-Sachs gene in the Jewish communities of Israel. Abstract, 25th Annual Meeting, Genetics Society of America, August 27, 1956 28. BARON SW: A Social and Religious History of the Jews, vols 2 and 3. New York, Columbia Univ. Press, 1937 29. GAY R: Jews in America. New York, Basic Books, 1965 30. FISHMAN P: The Jews of the United States. New York, Quadrangle, 1973 31. SKLARE M: America's Jews. New York, Random House, 1971 32. MAYO0: On the maintenance of polymorphisms having an inviable homozygote. Ann Hum Genet 33:307-317, 1970 33. FRAIKOR AL: Tay-Sachs disease: genetic drift among the Ashkenazim Jews. Soc Biol 24:117-134, 1977 34. SPYROPOULOS B, MOENS PB, DAVIDSON J, LOWDEN JA: Heterozygote advantage in Tay-Sachs carriers? Am J Hum Genet 33:375-380, 1981 35. MYRIANTHOPOULOS NC, NAYLOR AF, ARONSON SM: Founder effect in Tay-Sachs disease unlikely. Am J Hum Genet 24:341-342, 1972 36. KNUDSON AG, KAPLAN WD: Genetics of the sphingolipidoses, in Cerebral Sphingolipidoses: A Symposium on Tay-Sachs Disease and Allied Disorders, edited by ARONSON SM, VOLK BW, New York, Academic Press, 1962, pp 395-411 37. CHAKRAVARTI A, CHAKRABORTY R: Elevated frequency of Tay-Sachs disease among Ashkenazic Jews unlikely by drift alone. Am J Hum Genet 30:256-261, 1978 38. KOESTLER AC: The Thirteenth Tribe. New York, Random House, 1976