Bloom Syndrome

A number sign (#) is used with this entry because Bloom syndrome (BLM), also referred to here as microcephaly, growth restriction, and increased sister chromatid exchange-1 (MGRISCE1), is caused by homozygous or compound heterozygous mutation in the gene encoding DNA helicase RecQ protein-like-3 (RECQL3; 604610) on chromosome 15q26.

Description

Bloom syndrome is an autosomal recessive disorder characterized by proportionate pre- and postnatal growth deficiency; sun-sensitive, telangiectatic, hypo- and hyperpigmented skin; predisposition to malignancy; and chromosomal instability.

Genetic Heterogeneity of Microcephaly, Growth Restriction, and Increased Sister Chromatid Exchange

See also MGRISCE2 (618097), caused by mutation in the TOP3A gene (601243) on chromosome 17p12.

Clinical Features

Landau et al. (1966) described a patient whose parents were second cousins and who showed low gamma-A and gamma-M serum proteins.

German et al. (1984) collected information on 103 patients. German and Takebe (1989) suggested that differences in skin pigmentation in various ethnic groups may confer a degree of protection against actinic radiation and thus obscure one of the characteristic facial signs of Bloom syndrome, i.e., telangiectasia. As a result, Bloom syndrome may be underdiagnosed in some populations. Legum et al. (1991) described an affected Iranian Jewish male, possibly the first definite non-Ashkenazi Jewish patient. The patient had another unique complication, cardiomyopathy.

Ferrara et al. (1967) described the disease in a 'Chinese-American'; however, the diagnosis was later (Ferrara, 1972) revised to focal dermal hypoplasia (305600).

The 14 Japanese cases reported by German and Takebe (1989) differed somewhat from most cases recognized elsewhere in that dolichocephaly was a less constant feature, the facial skin lesions were less prominent, and life-threatening infections were less frequent. The characteristic predisposition to neoplasia, as well as the probable tendency to diabetes mellitus, was found, however. German (1990) stated that diabetes mellitus of maturity-onset type, developing, however, in the second or third decade, is proving to be a frequent feature. Mori et al. (1990) reported diabetes mellitus in Bloom syndrome. Kelly (1977) observed a case of Bloom syndrome in a black female. German (Szalay, 1978) confirmed the diagnosis of Bloom syndrome in the black female reported by Szalay (1972).

German (1988) stated that the longest survival known to him is that of a man who died of esophageal cancer at the age of 48 years, having survived sigmoid cancer which had occurred 10 years earlier. Almost 150 cases worldwide have been cataloged by German (1990); he has personally examined 96 of these patients. Jewish patients represent 32% of the group, all but 1 of them being Ashkenazi. Complementation studies using sister chromatid exchange (SCE) as the measure of cross-correction indicate that this is one disease. Parental consanguinity was identified in 2 of 36 Jewish cases and in 25 of 75 non-Jewish cases. Heterozygotes do not show increased sister chromatid exchanges. German (1990) has not found an increased frequency of cancer in obligatory heterozygotes.

Passarge (1991) observed 10 patients in Germany during a 20-year period. One patient died at the age of 5 years of acute leukemia, a second at the age of 18 years of pulmonary fibrosis and bronchiectasis, and a third at the age of 21 years of Hodgkin lymphoma and subsequently leukemia.

German (1992) reported that there were 132 cases in the Bloom's Syndrome Registry as of January 1, 1990. One hundred and twenty seven had survived infancy. In all, 93 were still alive. Of the 39 deceased patients, 31 had died of cancer at a mean age of 27.8; cancer had been diagnosed at ages ranging from 4 years to 46 years. Of the 46 cancer patients, 14 had more than 1 primary, 2 had more than 2 primaries, and 1 had more than 3 primaries.

Chisholm et al. (2001) reported a 19-year-old woman with typical clinical features of Bloom syndrome with a successful pregnancy. Because of her small pelvis on clinical examination, the patient underwent computed tomography pelvimetry, which showed adequate pelvic capacity. Preterm labor occurred at 32 weeks' gestation, and the infant was ultimately delivered at 35 weeks' gestation. The infant was less than the tenth percentile for length and weight for gestational age, but was otherwise healthy. Since preterm labor had occurred in this and a previously reported pregnancy in Bloom syndrome (Mulcahy and French, 1981), Chisholm et al. (2001) suggested increased surveillance for preterm labor in pregnancies of women with Bloom syndrome.

Biochemical Features

Vijayalaxmi et al. (1983) found that lymphocytes from patients with Bloom syndrome showed an incidence of cell resistance to the purine analog 6-thioguanine about 8 times the normal. Cells with specific locus mutations have been reported to be present in abnormally great numbers in BS fibroblast cultures, e.g., 6-thioguanine-resistant and diphtheria toxin-resistant cells.

Seal et al. (1991) compared the uracil DNA glycosylase from 2 nontransformed cell strains derived from persons of different ethnic backgrounds, with 2 different, similarly highly purified, normal human uracil DNA glycosylases. For each of the 4, a molecular mass of 37 kD was observed. The Bloom syndrome enzymes differed substantially in their isoelectric point and were thermolabile as compared to the normal human enzymes. They displayed a different K(m) and V(max) and were strikingly insensitive to 5-fluorouracil and 5-bromouracil, pyrimidine analogs that drastically decreased the activity of the normal human enzymes. In particular, each Bloom syndrome enzyme required 10- to 100-fold higher concentrations of each analog to achieve comparable inhibition of enzyme activity.

Other Features

Langlois et al. (1989) used a glycophorin A assay to measure the frequency in persons of blood type MN of variant erythrocytes that lack the expression of 1 allelic form of the protein, presumably due to mutational or recombinational events in erythroid precursor cells. Blood from persons with Bloom syndrome showed 50- to 100-fold increases in the frequency of variants of 3 types, those with a hemizygous phenotype, those with a homozygous phenotype, and those with what appeared to be partial loss of the expression of 1 locus. The high frequency of homozygous variants, indicating altered allelic segregation, could be taken as evidence for increased somatic crossing-over in vivo.

An increased generation of functional hemizygosity and homozygosity in somatic cells may play a role in the high cancer risk of persons with Bloom syndrome. Accumulation of p53 (191170) protein is seen in the nuclei of mammalian cells following DNA damage caused by ultraviolet radiation, x-ray, or a restriction enzyme. Promoters containing p53-binding sites show a dramatic transcriptional response to DNA damage. The p53 response to x-ray is rapid, reaching a peak at 2 hours after radiation, but is very transitory and reduced in magnitude compared with that seen in response to UV. Lu and Lane (1993) found no substantive defect in the p53 response of cells from ataxia-telangiectasia (208900) or xeroderma pigmentosum complementation group A (278700) patients. In contrast, 2 out of 11 primary cultures from Bloom syndrome patients showed complete absence of p53 accumulation following UV irradiation or SV40 infection and a grossly delayed and aberrant response following x-ray.

Van Kerckhove et al. (1988) found a specific defect in the pokeweed mitogen-induced alternative pathway of lymphocyte activation in BS patients.

Krejci et al. (2003) clarified the role of Srs2 in recombination modulation by purifying its encoded product and examining its interactions with the RAD51 recombinase (179617). Srs2 has a robust ATPase activity that is dependent on single-stranded DNA and binds RAD51, but the addition of a catalytic quantity of Srs2 to RAD51-mediated recombination reactions causes severe inhibition of these reactions. Krejci et al. (2003) showed that Srs2 acts by dislodging RAD51 from single-stranded DNA. Thus, the attenuation of recombination efficiency by Srs2 stems primarily from its ability to dismantle the RAD51 presynaptic filament efficiently. Krejci et al. (2003) suggested that their findings have implications for the basis of Bloom and Werner (277700) syndromes, which are caused by mutations in DNA helicases and are characterized by increased frequencies of recombination and a predisposition to cancers and accelerated aging.

Inheritance

Szalay (1963) provided the first evidence of a genetic basis. He described an isolated case in the child of first-cousin parents and 2 affected sibs. Autosomal recessive inheritance was established by German (1969), who maintains a worldwide registry which he periodically reports on (e.g., German et al., 1979). Of the then-known 21 families with Bloom syndrome, 12 were Ashkenazi and in these only 1 parental couple was consanguineous. On the other hand, 6 of the 9 non-Jewish unions were consanguineous. The Jewish gene appeared to have originated in a local area of eastern Europe.

German and Takebe (1989) reported that 14 cases in 12 families had been identified in Japan. Widely separated birthplaces and a frequency of parental consanguinity greater than in the general population suggested that the mutation, although rare, is widely distributed in that country. Complementation studies indicated that the same genetic locus is involved in the Japanese cases as in the Ashkenazi Jewish cases and non-Ashkenazi Jewish cases.

Cytogenetics

Multiple seemingly nonspecific chromosomal breaks have been observed in Bloom syndrome, as in Fanconi anemia (227650), and may be related causally to the high frequency of leukemia (German et al., 1965; German, 1992). (The Bloom and Fanconi syndromes are chromosome breakage or clastogenic syndromes.)

Schroeder and German (1974) showed that chromosomal aberrations were more numerous in Fanconi cells than in Bloom cells. In Bloom syndrome most interchanges were between homologous chromosomes, i.e., sister chromatid exchanges, whereas in Fanconi syndrome they were usually between nonhomologous chromosomes. Sister chromatid exchanges represent a cytologic marker useful for diagnosis including prenatal diagnosis. No test for the carrier state is known; the frequency of sister chromatid exchanges is not abnormal in heterozygotes (German et al., 1977).

Although the nature of the basic defect was not known, the absence of a substance that is supplied by cocultivated normal cells and reduced the rate of sister chromatid exchanges in Bloom syndrome fibroblasts was suggested by the work of Rudiger et al. (1980). Spontaneous SCE, but not mutagen-induced SCE, is inhibited by the Bloom corrective factor present in normal cell-conditioned culture medium. Control cells and cells of Fanconi anemia and xeroderma pigmentosum reduced the rate of sister chromatid exchanges in Bloom cells by about 45 to 50% (Bartram et al., 1981). In contrast, Bloom heterozygous cells reduced the rate of SCE by only 16 to 18%. Bartram et al. (1981) interpreted the findings as indicative of dosage effect. They concluded that the data suggest the existence of a 'corrective factor' which is either inactive or absent in homozygous Bloom cells and reduced in heterozygotes. It may be identical with or closely related to the normal gene product of the Bloom locus.

Thompson et al. (1982) found greatly increased sister chromatid exchanges in a mutant Chinese hamster ovary (CHO) cell line (EM9) with a DNA repair deficiency (see 126340). The deficiency was complemented in human-CHO somatic cell hybrids by human chromosome 19. Is this the Bloom syndrome defect? The mapping of Bloom syndrome to chromosome 15 excludes that possibility. (See Mapping Information.)

Weksberg et al. (1988) addressed the issue of dominance or recessivity of the low-SCE Bloom syndrome phenotype. Although most cells from BS patients exhibit high SCE, lymphoid cells from some patients exhibit dimorphism for high and low SCE. A high-SCE lymphoblast line was mutagenized, and a clone carrying the markers ouabain resistance and thioguanine resistance was isolated to serve as a fusion parent. When fused with low-SCE BS lines, the hybrid was found to have low SCE levels, establishing dominance of the low-SCE phenotype. By the same methodology, Weksberg et al. (1988) did a complementation analysis using high-SCE lymphoblast cell lines derived from patients of diverse ethnic origin: Ashkenazi Jewish, French-Canadian, Mennonite, and Japanese. No correction of the high SCE characteristic of BS cells was seen in any hybrids. Thus, a single gene is responsible for the high-SCE phenotype in BS patients.

Poppe et al. (2001) presented the cytogenetic findings in a Bloom syndrome patient diagnosed with acute myeloid leukemia (AML) of the FAB subtype M1, as well as a review of the literature, which showed the preferential occurrence of total or partial loss of chromosome 7 in BS patients with AML or myelodysplastic syndromes.

Mapping

Lander and Botstein (1987) pointed out that recessive disorders can be efficiently mapped using RFLPs in the study of the DNA of affected children from consanguineous marriages. The method, which the authors called 'homozygosity mapping,' involves detection of the disease locus by virtue of the fact that the adjacent region will preferentially be homozygous by descent in such inbred children. They showed that a single affected child of a first-cousin marriage contains the same total information about linkage as a nuclear family with 3 affected children. They presented calculations to show that it should be practical to map a recessive disease gene by studying DNA from fewer than a dozen unrelated, affected inbred children, given a complete RFLP linkage map. Bloom syndrome was pointed to by Lander and Botstein (1987) as a good candidate for this type of mapping. About 100 living affected persons are known, but there are only 8 known families with 2 living affected members and 1 with 3 affected members. This is insufficient for traditional linkage analysis. By contrast, at least 24 affected persons are children of marriages between cousins. In some more common recessive disorders such as Werdnig-Hoffmann syndrome (253300), multiplex families may be difficult to collect because affected children die at a young age.

Because of the high parental consanguinity rate in non-Ashkenazi families with Bloom syndrome, Ellis et al. (1992) were able to do homozygosity mapping. Tight linkage was found with loci on distal 15q, specifically 15q26.1 (German et al., 1994). A polymorphic tetranucleotide repeat in an intron of the FES gene (190030) was homozygous in 25 of 26 individuals with Bloom syndrome whose parents were consanguineous (German et al., 1994). The location of the BLM gene on chromosome 15 was further supported by the finding of maternal uniparental disomy for that chromosome in a patient reported by Woodage et al. (1994). The patient had features of both Bloom syndrome and Prader-Willi syndrome (176270). Meiotic recombination between the 2 chromosomes 15 derived from the mother had resulted in heterodisomy for proximal 15q and isodisomy for distal 15q. In this individual, Bloom syndrome was probably due to homozygosity for a gene located telomeric to D15S95, which is at 15q25, rather than to genetic imprinting, the mechanism responsible for the development of PWS. This report represented the first application of disomy analysis to the regional localization of a disease gene. Ellis et al. (1994) found a striking association of a specific allele at the FES locus and at the D15S127 locus, both of which are tightly linked to BLM. This linkage disequilibrium constituted strong support for a founder-effect hypothesis to account for the fact that approximately 1 in 110 Ashkenazi Jews carries the Bloom syndrome mutation.

The rarity of Bloom syndrome and the recessive nature of its inheritance limits mapping of the gene by linkage approaches. McDaniel and Schultz (1992) used Bloom syndrome cells as recipients for microcell-mediated chromosome transfer to map a locus that results in complementation of the elevated sister chromatid exchange phenotype. Studying the Bloom cell line GM08505 (Coriell Institute) with a stable frequency of SCEs 10-fold higher than control values, they demonstrated that transfer of human chromosome 15 corrected the defect.

Straughen et al. (1996) described a 2-Mb contiguous map of the 15q26.1 region constructed from P1 clones and yeast artificial chromosomes (YACs) that contains the BLM gene. They also reported a long-range restriction map of this region.

Molecular Genetics

Phenotype/Genotype

Willis and Lindahl (1987) and Chan et al. (1987) independently demonstrated an abnormality of DNA ligase I (126391) in Bloom syndrome. DNA ligase I and DNA polymerase alpha (312040) are enzymes that function during DNA replication; DNA ligase II and DNA polymerase beta (174760) function during DNA repair. That the primary defect resides in the structural gene for DNA ligase I was suggested by the changes in the physical properties of the enzyme, specifically, heat sensitivity (Willis and Lindahl, 1987) and altered aggregation properties (Chan et al., 1987). Experiments with a fibroblast line derived from a Japanese case of Bloom syndrome showed that DNA ligase I from that source was not obviously heat sensitive or present in reduced amounts. Chan and Becker (1988) also came to the conclusion that mutation of the DNA ligase I gene may account for the primary metabolic defect in Bloom syndrome. Their data suggested that the defect in DNA ligase I is not due to a reduction in the number of protein molecules or to inhibitory substances but rather, at least in part, to the ATP binding and/or hydrolytic activity of the enzyme.

Willis et al. (1987) found that all cell lines derived from 7 patients with Bloom syndrome contained a DNA ligase I with unusual properties. In 6 lines the enzyme activity was reduced and the residual enzyme was anomalously heat-labile. In the seventh line, they found a dimeric rather than a monomeric form of ligase I. Several cell lines representative of other inherited disorders had apparently normal DNA ligases. The data were interpreted as indicating that BLM is due to a defect in the structure of DNA ligase I caused by a 'leaky' point mutation occurring at one of at least 2 alternative sites. If the primary defect lies in the structural gene for DNA ligase I, then Barnes et al. (1990) reasoned that the mutation for Bloom syndrome is on chromosome 19, which encodes DNA ligase I.

Since alteration of the DNA ligase I activity is a consistent biochemical feature of Bloom syndrome cells, Petrini et al. (1991) cloned DNA ligase I cDNA from normal human cells. Human DNA ligase I cDNAs from normal and BS cells complemented an S. cerevisiae DNA ligase mutation, and protein extracts prepared from S. cerevisiae transformants expressing normal and BS cDNA contained comparable levels of DNA ligase I activity. DNA sequencing and Northern blot analysis of DNA ligase I expression in 2 BS fibroblast lines representing each of 2 aberrant DNA ligase I molecular phenotypes demonstrated that the gene was unchanged in BS cells. Thus, a factor other than mutation in the ligase I gene must be involved as the basic defect.

Nicotera et al. (1989) suggested that the major biochemical defect in Bloom syndrome is chronic overproduction of the superoxide radical anion. They thought that inefficient removal of peroxide might be responsible for high rates of sister chromatid exchange and chromosomal damage in Bloom syndrome cells. Seal et al. (1988) described a monoclonal antibody, defined by enzyme-linked immunosorbent assay (ELISA), that reacted with normal uracil DNA glycosylase (191525) of human placenta as well as with the glycosylases from normal human cell types and 13 abnormal human cell strains. On the other hand, the antibody neither recognized nor inhibited native uracil DNA glycosylase from any of 5 separate Bloom syndrome cell strains. Lack of immunoreactivity with this antibody, which the authors designated 40.10.09, was suggested as a test for the early diagnosis of Bloom syndrome.

Cairney et al. (1987) described Wilms tumor in 3 patients with Bloom syndrome. Wilms tumor was bilateral in 1 of the 3 patients. Cairney et al. (1987) postulated that the elevated somatic recombination may mediate a high rate of conversion to homozygosity. Somatic recombination leading to homozygosity in Bloom syndrome has been suggested by several findings, including 'twin spots' or areas of hyper- and hypopigmentation on the skin of affected black children (Festa et al., 1979), increased frequency of exchange between the satellite stalks of acrocentric chromosomes (Therman et al., 1981), and increased variant blood group phenotypes in red cells from a patient with Bloom syndrome who was heterozygous for the AB blood group (Ben-Sasson et al., 1985). Petrella et al. (1991) observed autosomal triple trisomy involving chromosomes 2, 8, and 11 in a pregnancy conceived by a couple at risk for an offspring with Bloom syndrome. The SCE rate suggested that the conceptus was either heterozygous for the Bloom syndrome mutation or homozygous normal. They also found the Bloom syndrome gene in a non-Ashkenazi Jew and reported medulloblastoma in a patient with Bloom syndrome.

The hypermutability of Bloom syndrome cells includes hyperrecombinability. Ellis et al. (1995) noted that although cells from all persons with Bloom syndrome exhibit the diagnostic high SCE rate, in some persons a minor population of low SCE lymphocytes exist in the blood. Lymphoblastoid cell lines (LCLs) with low SCE rates can be developed from these low SCE lymphocytes. In multiple low SCE LCLs examined from 11 patients with BS, polymorphic loci distal to BLM on 15q had become homozygous in LCLs from 5 persons, whereas polymorphic loci proximal to BLM remained heterozygous in all low SCE LCLs. These observations supported the hypothesis that low SCE lymphocytes arose through recombination within BLM in persons with BS who had inherited paternally and maternally derived BLM alleles mutated at different sites. Such a recombination event in a precursor stem cell in these compound heterozygotes thus gave rise to a cell whose progeny had a functionally wildtype gene and phenotypically a low SCE rate (Ellis et al., 1995). Ellis et al. (1995) used the low SCE LCLs in which reduction to homozygosity had occurred for localizing BLM by an approach referred to as somatic crossover point (SCP) mapping. The precise map position of BLM was determined by comparing the genotypes of the recombinant low SCE LCLs from the 5 persons mentioned above with their constitutional genotypes at loci in the region around BLM. The strategy was to identify the most proximal polymorphic locus possible that was constitutionally heterozygous and that had been reduced to homozygosity in the low SCE LCLs, and to identify the most distal polymorphic locus possible that had remained constitutionally heterozygous in them. BLM would have to be in the short interval defined by the reduced (distal) and the unreduced (proximal) heterozygous markers. The power of this approach was limited only by the density of polymorphic loci available in the immediate vicinity of BLM. A candidate for BLM was identified by direct selection of a cDNA derived from a 250-kb segment of the genome to which BLM had been assigned by SCP mapping. cDNA analysis of the candidate gene identified a 4437-bp cDNA that encoded a 1417-amino acid peptide with homology to the RecQ helicases, a subfamily of DExH box-containing DNA and RNA helicases (RECQL3; 604610). The presence of chain-terminating mutations in the candidate gene in persons with Bloom syndrome proved that it was BLM. Mutational analysis in the first 13 unrelated persons with BS examined permitted the identification of 7 unique mutations in 10 of them. The fact that 4 of the 7 mutations resulted in premature termination of translation indicated that the cause of most Bloom syndrome is the loss of enzymatic activity of the BLM gene product. Identification of loss-of-function mutations in BLM is consistent with the autosomal recessive transmission, and the homology of BLM and RecQ suggested that BLM has enzymatic activity. In 4 persons with Jewish ancestry, a 6-bp deletion and a 7-bp insertion at nucleotide 2281 were identified, and each of the 4 persons were homozygous for the mutation (604610.0001). Homozygosity was predictable because linkage disequilibrium had been detected in Ashkenazi Jews with Bloom syndrome between BLM, D15S127, and FES (Ellis et al., 1994). Thus a person who carried this deletion/insertion mutation was a founder of Ashkenazi Jewish population and nearly all Ashkenazi Jews with Bloom syndrome inherited the mutation identical by descent from this common ancestor.

The RecQ gene family, of which BLM is a member, is named after the E. coli gene. RecQ is an E. coli gene that is a member of the RecF recombination pathway, a pathway of genes in which mutations abolish the conjugational recombination proficiency and ultraviolet resistance of a mutant strain. RecQL (600537) is a human gene isolated from HeLa cells, the product of which possesses DNA-dependent ATPase, DNA helicase, and 3-prime-to-5-prime single-stranded DNA translocation activities. Ellis et al. (1995) suggested that the absence of the BLM gene product probably destabilizes other enzymes that participate in DNA replication and repair, perhaps through direct interaction and through more general responses to DNA damage.

Ellis and German (1996) reported that the BLM protein has similarity to 2 other proteins that are members of the RecQ family of helicases, namely the gene product encoded by the Werner syndrome gene (WRN; 277700) and the product of the yeast gene SGS1. SGS1 was identified by a mutation that suppressed the slow-growth phenotype of mutations in the topoisomerase gene. These proteins have 42 to 44% amino acid identity across the conserved helicase motifs. In addition, the proteins are of similar length and contain highly negatively charged N-terminal regions and highly positively charged C-terminal regions. Ellis and German (1996) noted that these similarities in overall structure have raised the possibility that the proteins play similar roles in metabolism. Since the SGS1 gene product is known to interact with the products of the yeast topoisomerase genes, they predicted that the BLM and WRN genes interact with human topoisomerases.

Sinclair et al. (1997) showed that mutation of the yeast SGS1 gene causes premature aging in yeast mother cells as demonstrated by shortened life span and the aging-induced phenotypes of sterility and redistribution of the Sir3 silencing protein from telomeres to the nucleolus. Further, in old SGS1 cells the nucleolus was enlarged and fragmented, changes that also occur in old wildtype cells. Their findings suggested a conserved mechanism of cellular aging that may be related to nucleolar structure. The similar effect of the related SGS1 and WRN genes on yeast and human aging, along with age-associated changes in rDNA content reported for several mammalian species, suggested that a common mechanism may underlie aging in eukaryotes.

Men with Bloom syndrome are sterile; women have reduced fertility and a shortened reproductive span. In an immunocytologic study of mouse spermatocytes, Walpita et al. (1999) showed that the BLM protein is first evident as discrete foci along the synaptonemal complexes of homologously synapsed autosomal bivalents in late zygonema of meiotic prophase. BLM foci progressively dissociated from the synapsed autosomal axes during early pachynema and were no longer seen in mid-pachynema. BLM colocalized with the single-stranded DNA-binding replication protein A (see 179835), which had been shown to be involved in meiotic synapsis. However, there was a temporary delay in the appearance of BLM protein along the synaptonemal complexes relative to replication protein A, suggesting that BLM is required for a late step in processing of a subset of genomic DNA involved in establishment of interhomolog interactions in early meiotic prophase. In late pachynema and into diplonema, BLM is more dispersed in the nucleoplasm, especially over the chromatin most intimately associated with the synaptonemal complexes, suggesting a possible involvement of BLM in resolution of interlocks in preparation for homologous chromosome disjunction during anaphase I.

Ellis et al. (1999) described the effects on the abnormal cellular phenotype of BS, namely an excessive rate of SCE, when normal BLM cDNA was stably transfected into 2 types of BS cells, SV40-transformed fibroblasts and Epstein-Barr virus-transformed lymphoblastoid cells. The experiments proved that BLM cDNA encodes a functional protein capable of restoring to or toward normal the uniquely characteristic high-SCE phenotype of BS cells.

In a patient with Bloom syndrome and both high- and low-SCE cell lines, Foucault et al. (1997) identified compound heterozygosity for a cys1036-to-phe (C1036F; 604610.0004) substitution in the C-terminal region of the peptide and an unidentified mutation affecting expression of the RECQL3 gene. Foucault et al. (1997) concluded that somatic intragenic recombination resulted in cells that had an untranscribed allele carrying the 2 parental RECQL3 mutations and a wildtype allele which allowed reversion to the low-SCE phenotype. Topoisomerase II-alpha (126430) mRNA and protein levels were decreased in the high-SCE cells, whereas they were normal in the corresponding low-SCE cells. Foucault et al. (1997) proposed that in addition to its putative helicase activity, RECQL3 might be involved in transcription regulation.

Associations Pending Confirmation

For discussion of a possible association between a Bloom syndrome-like phenotype and variation in the RMI1 gene, see 610404.0001.

For discussion of a possible association between a Bloom syndrome-like phenotype and deletion of the RMI2 gene, see 612426.0001.

Clinical Management

German (1992) commented that BS neoplasms themselves, e.g., leukemic marrow cells, demonstrate a high sister chromatid exchange rate similar to nonneoplastic cells of BS patients. A clinical difference is that leukemia in BS usually presents itself with leukopenia rather than leukocytosis. He also commented on the fact that there is pitifully little that one can do in relation to the proneness to cancer. In contrast to the situation with carcinoma, early diagnosis of leukemia is at present not known to improve the chances of curative therapy.

German (1992) advised against frequent hematologic examinations in children for fear of untoward psychologic effects. Allogeneic marrow grafting has not been carried out in BS. An argument can be made for identifying as soon as possible a potential donor of bone marrow for any person with BS, and the cryopreservation of cord-blood stem cells of HLA-matched sibs who might be born after the BS child is identified can be considered, for possible later transplantation.

Animal Model

Chester et al. (1998) found that mouse embryos homozygous for a targeted mutation in the murine Bloom syndrome gene are developmentally delayed and die by embryonic day 13.5. They determined that the interrupted gene is the homolog of the human BLM gene by its homologous sequence, its chromosomal location, and the demonstration of high numbers of sister chromatid exchanges in cultured murine Blm -/- fibroblasts. The proportional dwarfism seen in the human is consistent with the small size and developmental delay (12 to 24 hours) seen during midgestation in murine Blm -/- embryos. The growth retardation in mutant embryos can be accounted for by a wave of increased apoptosis in the epiblast restricted to early postimplantation embryogenesis. Mutant embryos do not survive past day 13.5, and at this time exhibit severe anemia. Red blood cells and their precursors from Blm -/- embryos are heterogeneous in appearance and have increased numbers of macrocytes and micronuclei. Both the apoptotic wave and the appearance of micronuclei in red blood cells are likely cellular consequences of damaged DNA caused by effects on replicating or segregating chromosomes.

Using embryonic stem cell technology, Luo et al. (2000) generated viable Bloom syndrome mice that were prone to a wide variety of cancers. Cell lines from these mice showed elevations in the rates of mitotic recombination. They demonstrated that the increased rate of loss of heterozygosity resulting from mitotic recombination in vivo constituted the underlying mechanism causing tumor susceptibility in these mice.