Search

The clinical manifestations of inherited neurodegenerative diseases are often delayed for periods from years to decades. This observation has led to the idea that, in these disorders, neurons die from cumulative damage. A critical prediction of the cumulative damage hypothesis is that the probability of neuronal death increases with age. However, we recently demonstrated, in 17 examples of neurodegeneration, that the kinetics of neuronal death appear to be exponential. These examples include both monogenic disorders, such as photoreceptor degenerations, as well as others that are partly or entirely acquired (such as the clinical phase of parkinsonism and retinal detachment). Exponential kinetics indicate that (i) the risk of death is constant, (ii) death occurs randomly in time and (iii) the death of each neuron is independent of other neurons. We use the term 'one-hit model' to refer to the single catastrophic intracellular biochemical event, analogous to radioactive decay, which leads to neuronal death in the diseases we analyzed. Here, we examine the major features and implications of the one-hit model and provide preliminary evidence that amyotrophic lateral sclerosis also appears to fit this model. We also discuss a testable biochemical hypothesis, the mutant steady-state hypothesis, that we proposed to account for the one-hit model. Finally, we explore six unresolved issues that appear to challenge this model. The one-hit model appears to capture a novel principle underlying many neurodegenerations. Our findings suggest that any consideration of the biochemical basis of neurodegeneration must include a meticulous examination of the kinetics of cell death.

We report the cloning of a novel human cDNA which encodes a 690 amino acid protein with high homology to ubiquitin C-terminal hydrolases. Northern blot analysis shows expression of a 3.3 kb transcript in all tissues examined, with 5- to 10-fold higher levels in retina than elsewhere. We mapped the structural gene to Xp21.2-p11.2. This gene's relatively high levels of retinal expression and recent work showing that perturbations in protein turnover and processing can lead to retinal disease make it an excellent candidate for several X-linked retinal disorders mapping within this interval. Additionally, there is evidence that members of the ubiquitin hydrolase family may play a role in oncogenesis and a locus implicated in ovarian cancer is also located within this region.

We review the current knowledge of transcription factors in mammallan eye development. The 14 transcription factors presently known to be required for eye formation are examined in some detail, incorporating data from both humans and rodents. Aspects of the biochemistry, expression patterns, genetics, mutant phenotypes, and biological insights acquired from the examination of loss-of-function mutations are summarized. The other 32 tissue-restricted transcription factors that are currently known to be expressed in the developing or mature mammallan eye are tabulated, together with the timing and site of their ocular expression; the requirement for most of these genes in the eye is unknown. Contributions to mammallan eye development from the study of the genetics of the Drosophila eye are discussed briefly. Identification of the entire cohort of transcription factors required for eye development is an essential first step towards understanding the mechanisms underlying eye morphogenesis and differentiation, and the molecular basis of inherited eye abnormalities in humans.

To examine the role of ROM1, a homologue of peripherin/RDS, in autosomal dominant retinitis pigmentosa (adRP), we screened 224 adRP and 29 simplex RP probands for ROM1 mutations. Four ROM1 alleles were designated as potentially pathogenic because they were found only in RP patients but not in 50-100 controls nor in 249 other RP probands. The substitutions P60T and T108M were present in a single allele in a subject with typical adRP, and this allele cosegregated with the disease in the small family. The putative null allele L114 [1 bp] was present in an individual with atypical RP but not in three unaffected siblings. This insertion has been previously reported to cause RP only when accompanied by a peripherin/RDS mutation, but no peripherin/RDS mutations were found in any of the four probands reported here. Two substitutions (G75D, R242Q) were present in two other probands with simplex RP. These data suggest that potentially pathogenic ROM1 mutations occur in 1% or less of patients with adRP or simplex RP. The absence of detectable peripherin/RDS mutations in these families suggests either that: (i) mutations in other digenic partners are required for pathogenic ROM1 alleles to cause retinal degeneration; (ii) these ROM1 mutations do not cause RP; or (iii) peripherin/RDS mutations are present but were not identified in these patients.

We determined the molecular basis of cystathionine beta-synthase (CBS) deficiency in three siblings with pyridoxine responsive homocystinuria using a significantly improved mutation screening method in bacteria. The phenotypic expression of the siblings differed even though their CBS genotypes were identical. The paternal allele contained a linked pair of mutations, C233G and G306C, corresponding to P78R and K102N in the polypeptide chain. Together, these inactivated the enzyme; however, expressed separately, they reduced activity by about one half. The single maternal mutation G715A (E239K) effectively abolished CBS activity. Subunits of CBS were absent from patient fibroblast extracts; however, E. coli, transformed with plasmids containing patient CBS cDNA, expressed the subunits, although in reduced amounts. The mother, an obligate heterozygote, was free from all signs of homocystinuria; nonetheless, extracts of her fibroblasts were devoid of CBS protein and activity. We conclude that fibroblast levels of CBS are only partially effective as prognosticators of disease severity and that it is important to test the in vivo response to vitamin B6 in all cases of homocystinuria, including those in which the mutations lead to the absence of the enzyme in cultured fibroblasts.

Rom-1 is an integral membrane protein of the rod photoreceptor outer segment. The ROM1 gene is located on human chromosome 11q13, a region to which the loci of four degenerative retinopathies have been mapped. To identify alleles of ROM1, we have screened the DNA of 57 controls and 180 patients with inherited retinopathies. Six ROM1 polymorphisms were identified: two in non-coding sequences (C-7T, T insertion 966/967), two substitutions (Ala118Gly, Arg223Arg), and two RFLPs outside the transcription unit, detected with BcII and Hind III. One rare sequence variant (Arg229His) was found in two adRP probands; in the one family studied the allele was discordant with the disease. A second rare variant (Ala265Thr) was found in both an adRP family and a control; a third rare variant (Met271Thr) was present only in a control family. These polymorphisms will be useful in the evaluation of ROM1 as a candidate gene in inherited retinal diseases. The recognition of the rare variants will prevent their misassignment as disease-causing mutations.

Rom-1 and peripherin are related membrane proteins of the photoreceptor outer segments. Both proteins are located at the rims of the photoreceptor disks, where they may act jointly in disk biogenesis. Mutations in the gene (RDS) encoding peripherin cause autosomal dominant retinitis pigmentosa, autosomal dominant punctata albescens and butterfly macular degeneration in man, and retinal degeneration slow in mice. To facilitate ROM1 mutation and linkage analysis in inherited retinal diseases, we cloned and characterized the human and murine ROM1 genes. In both species, the ROM1 coding region is contained within approximately 1.8 kb of genomic DNA and is interrupted by only two introns. The structures of the ROM1 and RDS genes are similar, with perfect conservation of the intron splice sites. Putative transcription regulatory regions of the ROM1 locus, 5' to an apparent transcription start site, were identified by cloning the mouse Rom-1 gene and comparing the sequence to the human homologue. Alignment of the human and murine rom-1 predicted protein sequences with the peripherin polypeptides of four species reveals a high degree of conservation (47% overall identity between the six proteins) in the central hydrophilic domain of the two family members. Despite this conservation of sequence, the predicted pI's of only this region of rom-1 and peripherin differ substantially, being 5.2 and 8.2, respectively. The charge difference in this region may mediate the non-covalent association of these two proteins in vivo. The conserved genomic structure and sequence of ROM1 and RDS indicates that these genes evolved from a common ancestor by duplication event.