Genesis of a Fact: Tay-Sachs Disease as a “Simple Recessive”

Abstract

“Obvious” recessive inheritance of Tay-Sachs disease (TSD; OMIM # 272800) took over half a century to be established. Points now taken for granted were problematic, that: (1) TSD is a biological entity, not an artificial selection of concurrent findings, (2) manifestations have narrow limits, (3) it was not part of a spectrum of disorders, and can be differentiated from other conditions, (4) it will not change to another disease, (5) it is due to a single specific gene, (6) there are no secondary causes, (7) the gene has no apparent clinical effects unrelated to TSD, and (8) the gene is inherited only as a clinical recessive. To a large extent, resolution reflected biochemical understanding that took until mid-20th century, and beyond, to change how physicians viewed diseases in general. With this, biochemical carrier screening and prenatal biochemical diagnosis have become routinely available, and it is a model for carrier population screening, while gene therapy for the disease has been reported with some degree of success. Here, the history of medical ideas about TSD and its inheritance are reviewed to show how it achieved its current status as a distinct recessive disorder.

“All we want are the facts, ma'am.” Dragnet. Sgt. Joe Friday, LAPD (Jack Webb). ¹

Introduction

Facts are obvious—“indisputably the case,” ² but Fleck ³ and Kuhn ⁴ argued for construction by a professional community instead of a simple discovery. In examining this, Tay-Sachs disease (TSD) (OMIM # 272800) ⁵ is now universally considered “a simple recessive,” ⁶ but physicians took over half a century to reach that point. This was not just a lack of awareness of genetics—they consistently, and often spectacularly, got it wrong, and were not always sure about exactly what TSD was. Here, the history of medical ideas about TSD and its inheritance, as well as factors affecting medical uses of genetics, are reviewed to show how it became an obvious recessive disorder.

Background

In storage disorders, biochemicals interfere with function as they accumulate in cells. Findings depend mostly upon how fast and where, and different substances can give similar findings. In TSD, a defective gene for a β-hexosaminidase A enzyme (OMIM HEXA; 606869) on chromosome 15q23 leads to excess GM2 ganglioside in nerve cells, especially in the brain. Problems start early and progress rapidly, as a normal infant deteriorates, loses vision, skills and activity, and dies in a few years. The gene is relatively common with Central/Eastern European Jewish (Ashkenazi) descent, but also occurs in other populations. It is now seen as a standard recessive disorder, with no clinical findings in heterozygotes, a full set of problems if both genes are involved, and a 25% risk for an affected child if both parents are heterozygotes. ⁵

Context

Physicians in 1900 already had a variety of ideas wrapped around findings in relatives. Findings in siblings were familial , across generations hereditary (or inherited ), either direct, from parent to child, or indirect, with “skips.” ⁷ These categories could include infections, cancers, toxins, etc., as well as what we now know as genetic disorders. Heredity was a transmitted force that acted throughout life, directing, organizing, and coordinating structure, development, and physiology. ⁸ That force included diatheses , predispositions to specific problems that interacted with external and internal factors, especially idiosyncrasies and temperaments , often with inherited changes. ⁹

Ideas about what was transmitted arose from, and supported, these concepts. There was p olymorphism , “a unitary something in the neuropathic or psychopathic inheritance that makes itself manifest under many forms,” ¹⁰ so that disorders could vary, and even change into others, with multiple causes operating in a variety of combinations. Neuropathic Inheritance was a “taint” that predisposed to problems of the nervous system, and even nonneurological conditions in some formulations. ¹¹ This was basically incompatible with Mendelism, where one stable unit trait was controlled by one specific gene.

TSD: I. Doctor Bernard Sachs: Sachs, twice President of the American Neurologic Society, and widely respected for his knowledge and skill, made Pediatric Neurology a specialty, and wrote the first text in the field. Still, he is most remembered for reporting on “a peculiar form of idiocy associated with blindness” (and early death), first in 1887, and again, in 1892, when he stressed the significance of a later born sister with the same disorder. ¹² In 1881, Warren Tay, a British ophthalmologist, reported the characteristic “cherry red” spot in the macula of the eye in this disorder, ¹³ later noting sibling recurrences. ¹⁴ Originally termed familial amaurotic (amaurosis = blindness) idiocy, it was later named TSD, in their honor. There were multiple reports of the disorder with familial recurrences early on—one review in 1908 noted 106 cases, with 73 in 23 families. ¹⁵

For a cause, some suggested a toxin in the mother's milk, ^{16 17} but Sachs noted that some affected children were formula fed. ¹⁵ He himself first saw it as “an agenetic condition,” ¹⁸ “a failure of normal embryonic development.” ¹⁹ By 1910, ¹⁵ he saw it as a congenital disease (i.e., present at birth) “even if it is not manifest at the time of birth... [with] a nervous system so inadequate to the demands imposed upon it that its cells, after having performed their function for a few weeks or months, undergo complete disintegration.” He mentioned family predispositions and blood relationships, and noted one family with “innumerable psychoses among the relatives of both parents.” Nothing more was said about cause or inheritance.

Sachs saw it as hereditary, but felt it “impossible to assign any definite reason for the development of this disease in special families,” noting several factors: Predilection for Hebrews, consanguinity in some cases, a family with relatives on both sides with psychoses, and at least one mother injured in pregnancy. ¹⁵

Other factors continued to confuse the issues for decades, to mid-century, and beyond. So, in 1923, at a conference on Heredity in Nervous and Mental Disease , for a TSD family with a taint “so marked that every child inherited the disease,” Sachs still thought “a little less of heredity per se and a little more of the influence of the environment and of extraneous agents upon the ovum during the period of development.” ²⁰ This was standard neuropathic heredity—an inherited predisposition with a variety of subsidiary contributing causes. Mental illness was a classic sign of the taint, and blood relationships, that is, consanguinity, potentiated negative influences. ²¹

In 1923, Sachs also questioned traditional beliefs that prenatal events could affect heredity. So, he accepted “the hereditary character [of TSD], but heredity means something different to me this morning [after the conference] from what it did before. Hereafter I am going to think of heredity as something which is implanted within the germ-plasm, and after that I am going to begin to doubt whether a given disease is truly hereditary or whether it is due to some congenital factor or cause affecting the ovum during the intrauterine period.” ²⁰

Sachs's use of “hereditary” along with familial is worthy of note, since it meant transmitted across generations. ⁷ However, this was appropriate if it involved a taint in the parental generation—and beyond. ²¹

By 1925, ²⁰ Sachs still thought “a little less of heredity per se and a little more of the influence of the environment and of extraneous agents upon the ovum during the period of development.” But now, he was well aware of genetics: In 1926, ²⁰ early in a chapter on “Hereditary and Family Affectations,” he states “the reader should endeavor to acquaint himself with the facts of inheritance and their manner of operation, the significance of the chromosomes and their internal organization, and the various Mendelian characters and laws that are of great importance in a study of genetics. The application of Mendelian laws to the transmission of disease deserves further study.”

He then went on to TSD without a clue, thinking that “the heredity taint is the outstanding factor... that marriages between persons closely related be discouraged; and marriages between persons in whose families there are marked neurotic or psychical traits are also undesirable.” Ultimately, “we must leave it to the biologists, and those especially expert in genetics, to work out the application of the Mendelian law, if it be applicable to the transmission of disease from generation to generation of human beings.” ²⁰

More than 10 years later, in an article on “Heredity in diseases of the nervous system,” he commented on Mendelism again, now, if anything, even more impressed by its potential. And, for TSD, after much study, he concluded that “the chosen taint, unit character of disease, is surely a dominant!” ²²

Other Analyses

Not that others necessarily got it right. Fuchs, a prominent ophthalmologist, felt that “the cause of the disease, which sometimes affects several children of the same family... is unknown,” ²³ sentiments echoed by others. ²⁴ Mott (a pioneer medical biochemist ²⁵ ) saw this as “an inborn failure of the germinal determinants of the nervous system peculiar to the Jewish race... Prevention, as far as we know, resolves itself into discouragement of marriage of blood relations and of persons with an inherited neuropathic and psychopathic taint.” ²⁶ Penrose, an early medical geneticist, remembered medical school lectures in the late1920s discussing occurrences in the same sibship but ignoring any possible mechanism. ²⁷ Even later, Crew ²⁸ noted a comment that “nothing is known of the etiology of the disease apart from its familial and racial incidence.”

Some suggested characteristics of inheritance were wrong, even towards mid-century, for example, “the oldest child of a family is rarely affected, but when the disease appears it may be expected to destroy most of the subsequent children.” ²⁹ Sorsby saw it “at one time regarded as exclusively confined to Jews. This is not the case, and the issue at present is whether there is, in fact, any racial predilection... at all... [it] has been recorded frequently in several sibs, and there is a high rate of consanguinity in the parents. This would suggest recessive inheritance. There are, however, pedigrees in which the affection has occurred in collateral branches... The possibility of intermediate inheritance, with the affection as the expression of the homozygous state cannot be excluded in view of the fact that in families in which Tay-Sachs disease occurs, neurological stigmata are not uncommon. An alternative, and possibly more likely, reading, is that the affection is indeed recessive, with the recessive gene occasionally manifesting itself in the simplex state.” ³⁰ It was described as a dominant, with a recessive mode also present, ³¹ or with occasional irregular dominance, ³² still considered possible in 1961! ³³

Herrmann ³⁴ and Brandeis ³⁵ saw increased parental inbreeding explained by recessive inheritance, but even correct attributions were not always simple or convincing. Barker, once head of Medicine at Johns Hopkins, nicely summarized Mendelian effects with consanguinity, but also noted that, for TSD, “the incidence... among Polish Jews hints at the importance of race in etiology.” ³⁶

Gates called it recessive in 1929, ³⁷ but cited a family update as: “Two normal strong and unrelated German Jews had seven children, five of whom (four boys and a girl) were afflicted and died. The parents of the father were cousins... later history [showed that] the completed family... numbered 12, the later children being all normal with the somewhat doubtful exception of one... [a ratio of 6:6 or 5:7 amaurotic to normal]. A normal daughter has since had two children, one of whom (a girl) died of amaurosis. The evidence indicates that the inheritance is recessive.” Tredgold ³⁸ cited a study of 59 families that also pointed to a recessive mode ³⁹ but, before he concluded the same, stated “that there is a marked hereditary tendency to this disease, and numerous cases have been published showing its appearance in several generations of a family. In a number of these families other forms of mental defect, abnormality and disorder... have also been noted.” ³⁸

In 1933, Slome ⁴⁰ reviewed 130 reported cases in 80 families and found that, of 88 sibships, deducting the child through which they were ascertained, 18% of siblings were affected. There was greatly increased consanguinity, more in non-Jews than in Jews. Kaback ⁴¹ felt that this was the first specific discussion of autosomal recessive inheritance, although this was more the first genetic analysis, and he noted that “several subsequent publications confirmed these findings and provided preliminary estimates of the gene frequencies for this mutation.” Still, his citations are all from 1962 to 1967, with 3 out of 4 from one research group, with which the author of the fourth later collaborated! And, as noted above, recessive inheritance was hardly accepted without reservation before this.

Clinical Issues

What was inherited was not clear either. Originally, there was a distinct picture of rapid early neurological deterioration, specific eye findings, and a high prevalence in Jewish populations. But other cases with similarities and differences raised the possibility of a variable disease with broader findings. Studies in the first decade of the 20th century showed excess nerve cell lipids (certain water insoluble fats) in many of these disorders. ⁴² These conditions still tended to be called familial amaurotic idiocy , which also referred to “classic” TSD. ^{42 43} But, even here, storage was not always seen as the primary problem, for example, “the gene leads to a disharmonious rate of development of the sensory nervous system.” ⁴⁴ In this analysis, the abnormal lipids were a secondary result of another problem, and not primary findings.

Variants were soon proposed—with later onset, or “non-typical” features. In 1905, Vogt ⁴⁵ suggested a single disease of nerve cell lipid accumulation, with clinical types that varied only in age of onset. This often tended to be called by Sachs's original designation of familial amaurotic idiocy, which still also referred to classic cases. Spielmeyer [1907] saw three siblings with Spielmeyer-Sjogren type disease as unrelated to TSD, but later pathological studies made him see this as a TSD variant, with “confusion lasting about 50 years.” ⁴⁶ About a dozen types or variants were included: Bielschowsky-Jansky with late infantile onset, juvenile Spielmeyer-Vogt, possible Batten-Mayou and Sjögren types, and also adult Kufs disease, even though “frequently these patients are not blind and they are never idiotic. Accordingly, one finds such semantic monstrosities as 'family amaurotic idiocy without amaurosis,' and family amaurotic idiocy without idiocy.” ⁴³

This sort of nomenclature was still used in 1969 (and, I suspect, later), mostly by neurologists. ⁴⁷ Pratt, in 1967, in his book, The Genetics of Neurological Disorders , ⁴⁸ still used Amaurotic Family Idiocy as a heading. For what he termed the late infantile form, after Bielschowsky (now known as late infantile neuronal ceroid lipofuscinosis), he noted that “the independence of this form with onset at about 3 years, duration of 6 years and a more pronounced cerebellar atrophy [and no cherry red spot!], is uncertain; some workers maintain that this form is not separate, and that cases may be appropriately allocated either to the infantile or juvenile groups.” He also noted an adult dominant form, which the reference (# 2572) notes is “sans troubles visuels” (without visual problems)!

There was also lipid accumulation in other places. Niemann-Pick disease, which involves liver and spleen, as well as eye and brain, was particularly confusing here. It was first recorded as Gaucher's disease, and sialic aciduria, yet another condition, was also confused with early cases! ⁴⁹ Some saw it as extended TSD, others as a coincidental concordance of findings, perhaps because of “idiotypical affinity!” ⁵⁰ Sorsby felt that “Niemann-Pick disease is the visceral form, and Tay-Sachs disease the neurological form of an affection, which when fully developed shows both these conditions.” ²⁹ There were even a few suggestions of sibships where either disease could develop. ⁵¹ Other conditions were included even without lipids, such as “gargoylism” ⁴⁴ (actually Hurler or Hunter syndromes, mucopolysaccharide disorders) and Batten's disease, with lipofuscin. ⁴⁷

Molecular Advances

Klenk ⁵² showed increased levels of a ganglioside with TSD, and that this differed from the sphingomyelin in Niemann-Pick disease, making it increasingly obvious that they were separate disorders. ³⁰ By 1960, pathways for TSD gangliosides were being constructed, ⁵³ and a complete structure of the stored material was published in 1965. ⁵⁴ TSD, now termed a GM2-gangliosidosis for the accumulated material, was further defined as a deficiency of a specific enzyme, hexosaminidase A, ⁵⁵ and specific gene mutations directly ascertained—Myerowitz ⁵⁶ noted 78 different mutations in 1997. With all this, biochemical carrier screening and prenatal biochemical diagnosis have become routinely available, and it is a model for carrier population screening. ⁵⁷ Recently, with further advances in techniques, adeno-associated virus gene therapy for the disease has been reported with some degree of success. ⁵⁸

Conclusion

In the early 20th century, physicians were aware of Mendelism, and that TSD often occurred in siblings. However, they had problems characterizing TSD as a recessive. Older terms—hereditary, familial, neuropathic taint, and so on—were still used. Even if called recessive, details did not always fit—for example, relatives with minor findings, affected cousins, etc. It was not until the 1950s that the “fact” started to become “indisputably the case,” and even then, many physicians, especially neurologists, were not fully convinced for another decade. And this was not just a few practitioners behind the times—citations here are largely texts or reviews, typically by acknowledged experts.

There were two basic issues here. First, physicians already had a non-Mendelian theory of heredity that fit their needs very nicely. It is worth noting here that ideas such as taints also fit in well with the eugenic thinking so prevalent at least through the 1930s. ²⁷ Second, their concepts of diseases fit poorly with the stable traits and single causative factors of Mendelism. Without a “gold standard” to determine what was or was not TSD, boundaries and definitions were matters of philosophy, without any objective answers.

Overall, several points now taken for granted were uncertain, that: (1) TSD is a biological entity, not an artificial selection of concurrent findings. (2) Manifestations have narrow limits. (3) It can be differentiated from other conditions. (4) It will not change to some other disease. (5) It is due only to a specific gene. (6) There are no secondary causes. (7) The gene has no effects unrelated to TSD. (8) The gene is inherited only as a recessive.

With these issues unresolved, problems in pregnancy, or toxic influences, might very well determine findings with neuropathic inheritance, and psychosis in a relative could be an expression of the taint. So, as Sachs combined Mendelism and older approaches, it was easy to conclude that “the chosen taint, unit character of disease, is surely a dominant.” ²¹

A change from beliefs in multiple causes for diseases to universal, necessary causes ⁵⁹ was crucial to the ultimate success of genetics in medicine, and timing suggests that this was not complete until mid-20th century. This involved a variety of factors, such as the “biochemicalization” of medicine.

The Coris shared a 1947 Nobel Prize for “discovery of the course of the catalytic conversion of glycogen,” showing a series of enzymatic steps, and the association of specific diseases with specific abnormalities; around the same time, Beadle and Tatum (Nobelists in 1958) also linked genes with enzymes. The 1953 Watson-Crick model and the discoveries about DNA (deoxyribonucleic acid) ⁶⁰ that followed were incompatible with older medical ideas and, probably even more importantly, made them old-fashioned.

Neurologists, who often saw nonspecific problems without any clear genetic pattern (e.g., schizophrenia), were the last to accept the new order. However, as disease delineation progressed and genetics continued to advance, older approaches became increasingly outmoded. Biochemical analyses made the identity of TSD clear, and additional or secondary causes were excluded. Certainly, TSD heterozygote screening based on Mendelian ideas ⁴¹ made other interpretations untenable.

As TSD became a clearly defined unit trait, alternatives to genetics were being rejected, and Roberts ⁶ called it “a simple recessive.” The word “simple” suggests one more issue. By definition, with dominance, homozygotes are indistinguishable from heterozygotes. However, this is not true for many single gene traits in humans, for example, sickle cell anemia heterozygotes have blood findings, technically making it a semidominant. Physicians ultimately made dominance the absence of obvious findings in heterozygotes, and TSD became “a simple recessive.”

The more than half century that it took for TSD to become a self-evident recessive supports facts constructed by a professional community. ^{3 4} It also involved a deconstruction of old facts that got in the way. For doctors, TSD fell into a broad spectrum of findings with overlapping etiologies early on instead of Mendelism's distinct findings linked to a single factor. Until it became a causally based unit trait, alternatives to genetics seemingly provided better explanations for its inheritance. Ultimately, integrating genetics into medicine was more than just applying new concepts to old information: It required changes in medical ideas that involved seemingly unconnected areas, and the deconstruction of multiple “facts.”