Enhanced Tumor Formation in Mice Heterozygous for Blm Mutation

See allHide authors and affiliations

Science  20 Sep 2002:
Vol. 297, Issue 5589, pp. 2051-2053
DOI: 10.1126/science.1074340


Persons with the autosomal recessive disorder Bloom syndrome are predisposed to cancers of many types due to loss-of-function mutations in the BLM gene, which encodes a recQ-like helicase. Here we show that mice heterozygous for a targeted null mutation ofBlm, the murine homolog of BLM, develop lymphoma earlier than wild-type littermates in response to challenge with murine leukemia virus and develop twice the number of intestinal tumors when crossed with mice carrying a mutation in the Apctumor suppressor. These observations indicate that Blmis a modifier of tumor formation in the mouse and that Blmhaploinsufficiency is associated with tumor predisposition, a finding with important implications for cancer risk in humans.

Bloom syndrome (BS) is characterized by small stature, immunodeficiency, male infertility, and predisposition to cancer of many tissue types (1). Cells from persons with BS show increased somatic recombination, chromosome breakage, and site-specific mutations (1–3). The BS locus, BLM, encodes BLM, an adenosine triphosphate–dependent, 3′-5′ helicase with homology to the recQ DEXH-box–containing DNA and RNA helicases (4); loss of BLM helicase activity is responsible for the genomic instability of BS cells (5,6). BLM resolves Holliday junctions, suppresses recombination in vitro, and is required for the fidelity of DNA double-strand break repair (7–9).

We have used gene targeting by homologous recombination to disrupt the mouse Blm gene to simulateBLMAsh , a BS-causing mutant allele ofBLM carried by approximately 1% of Ashkenazi Jews (4, 10, 11).BLMAsh contains a frameshift mutation in exon 10 of BLM that results in premature translation termination (4). In contrast to work with two mouse models of BS previously reported (12, 13), we used a gene-targeting construct in which exons 10, 11, and 12 ofBlm were replaced with a hypoxanthine phosphoribosyltransferase (Hprt) cassette (fig. S1A). Germ line transmission of this mutant allele, BlmCin , followed blastocyst injection of targeted embryonic stem cells to generate heterozygous mice (fig. S1B). Crosses to generateBlmCin/Cin mice were unsuccessful, indicating that homozygous disruption of Blm results in embryonic lethality (14). Western blots of protein lysates from Blm +/+ testes, an abundant source ofBlm RNA and BLM protein (12, 13), displayed a specific band of approximately 190 kD when probed with a COOH-terminal antiserum to BLM (fig. S1C). Lysates fromBlmCin/ + testes had an approximately 50% reduction in BLM in comparison to Blm +/+testes. Lysates of heterozygous tissues were similarly evaluated with an NH2-terminal antibody to BLM and revealed no smaller immunoreactive proteins (fig. S1D). This reduction of full-length BLM and the absence of truncated BLM inBlmCin/ + mice confirm that we had generated a null allele. This allele allowed us to examine the biological consequences of Blm haploinsufficiency, that is, a reduction in wild-type (WT) Blm gene dosage and its gene product.

BS somatic cells exhibit increases in chromosome aberrations, sister chromatid exchanges (SCEs), homologous chromatid interchanges, and micronuclei that are a consequence of chromosome breakage (1, 15, 16). Although the cytogenetic analysis of somatic cells from human BLMheterozygotes remains to be completed (17), spermatozoa from two of three obligate heterozygotes have been shown to display excess numbers of chromosome breaks and rearrangements (18). To learn whether Blmhaploinsufficiency affects genomic stability, we cultured primary lung fibroblasts fromBlmCin /+ andBlm +/+ mice with bromodeoxyuridine (BrdU) for two cell cycles. SCE analyses (19) revealed no statistically significant difference between the SCE frequency inBlmCin/ + andBlm +/+ cells (17). However, the number of micronuclei in these BrdU-treated cultures revealed a twofold increase in BlmCin/ + cells as compared to Blm +/+ cells (table S1). A similar effect was observed in untreated fibroblast cultures from two WT and two BlmCin/ + mice (table S1). These results suggest that mouse cells heterozygous forBlmCin have a subtle increase in genomic instability presumably related to the reduced BLM level.

To investigate the effect of Blm haploinsufficiency on tumor formation, we injected 9 Blm +/+ and 15BlmCin /+ newborn mice with murine leukemia virus (MLV). All BlmCin /+mice developed metastatic T cell lymphoma and had an average life-span of 117.3 ± 33.0 days (Fig. 1). Although all of the WT littermates also developed T cell lymphoma, the WT mice had a longer average life-span of 184.4 ± 93.1 days. Histological analyses of four tumors of each genotype revealed no substantial morphological differences between tumors, that all lymphomas were CD8-positive, and that most were CD4-positive (14). Thus,Blm haploinsufficiency enhances T cell tumorigenesis in mice in response to viral challenge.

Figure 1

BlmCin /+ mice die earlier from MLV-induced T cell lymphoma thanBlm+/+ mice do. Three litters ofBlm+/+ andBlmCin /+ newborn littermates (Black Swiss in the F7 generation) were injected intraperitoneally with 100 μl of MLV at a concentration of 1 × 105plaque-forming units per ml and were monitored for lymphoma development. The age in days at the time of death is plotted for eachBlm+/+ (black bars, n = 9) andBlmCin /+ (gray bars, n = 15) mouse. The mean (x̄) age at the time of death is shown for each group of mice; P < 0.05, Student's ttest.

Because the gastrointestinal tract is a common site of cancer in human BS (1, 20), we tested the effect ofBlm haploinsufficiency onApcMin /+-mediated intestinal tumorigenesis. ApcMin /+ mice carry a premature stop codon in one allele of the Apc tumor suppressor gene (ApcMin ), develop multiple intestinal adenomas, and are a murine model of familial adenomatous polyposis coli (21, 22).BlmCin /+ female mice on the 129/SvEv background (backcross generation N3) were crossed with ApcMin /+ male mice (C57BL/6). The progeny were killed at 4 months of age and examined for intestinal tumors (23).ApcMin /+;Blm +/+mice (n = 14) developed an average of 14.2 ± 10.2 gastrointestinal tumors per animal whereasApcMin /+;BlmCin /+mice (n = 8) developed twice that number (31.4 ± 19.1) (Fig. 2A). Tumor size was similar in theApcMin /+;BlmCin /+andApcMin /+;Blm +/+mice (1.3 mm ± 0.2 and 1.4 mm ± 0.1, respectively). Most tumors arose in the small intestine, although 4 of 8ApcMi n/+;BlmCin /+mice and 1 of 14ApcMin /+;Blm +/+mice developed one colon tumor each. No intestinal tumors were observed in any Apc +/+ mice.

Figure 2

Haploinsufficiency of Blmmodifies the tumor phenotype ofApcMin /+ mice. (A) Quantification of tumors fromApcMin /+;Blm +/+(n = 14) andApcMin /+;BlmCin /+(n = 8) mice. Animals were killed at 4 months of age, and the mean number of grossly visible gastrointestinal tumors (≥ 1 mm) per mouse is shown. Statistical analysis was performed with the nonparametric Wilcoxon rank sum test (P < 0.05). (B) Histological analysis of small intestinal tumors (top panels) and colonic tumors (bottom panels) fromApcMin /+; Blm +/+(left) andApcMin /+;BlmCin /+(right) mice. Sections are stained with hemotoxylin and eosin (magnification 1000×). (C) Analysis of Apc and chromosome 18 microsatellite loci D18Mit19, D18Mit17,and D18Mit123 in DNA from microdissected adenoma tissue (T) and adjacent normal tissue (N). The pair on the left is from anApcMin /+;Blm +/+mouse; the pairs in the middle and on the right are from twoApcMin /+;BlmCin /+mice. DNA from C57BL/6 (B6) and 129/SvEv (129) mice are included as controls; bands corresponding to the C57BL/6 and 129/SvEv alleles are highlighted with arrows. (D) PCR/Southern analysis ofBlm in DNA microdissected from adenoma tissue (T) and adjacent normal tissue (N) from anApcMin /+;BlmCin /+mouse. Bands representing the WT alleles (400 base pairs) and targeted alleles (780 base pairs)are shown.

The tumors fromApcMin /+;BlmCin /+andApcMin /+;Blm +/+animals were classified as adenomas with either low- or high-grade dysplasia, based on gland architecture, nuclear/cytoplasmic ratio, amount of interglandular stroma, nucleus location, prominence of nucleoli, and the presence of mucus secretion. Intestinal specimens from mice of both genotypes showed evidence of intraepithelial neoplasia and low-grade adenomas in the small intestine. All colonic adenomas (fourApcMin /+;BlmCin /+and oneApcMin /+;Blm +/+) displayed high-grade dysplasia (Fig. 2B and fig. S2). OnlyApcMin /+;BlmCin /+mice developed tumors with high-grade dysplasia in the small intestine (5 of 223 tumors evaluated, Fig. 2B and fig. S2). Representative histological sections of the stomach, cecum, brain, mammary gland, testis, and thymus, as well as blood smears, were examined inApcMin /+;Blm +/+andApcMin /+;BlmCin /+mice; no neoplasia outside the intestinal tract was observed.

Intestinal adenomas fromApcMin /+;BlmCin /+ mice were evaluated for loss of heterozygosity (LOH) at the Apc locus, a feature typical of ApcMin /+adenomas (24, 25). Five adenomas fromApcMin /+;Blm +/+mice and 20 adenomas from ApcMin /+;BlmCin /+ mice were analyzed (Fig. 2C). The mean ratio of the Apc +to the ApcMin allele in tumors fromApcMin /+;BlmCin /+mice was not different from that in tumors fromApcMin /+;Blm +/+mice. Each was consistent with published values (24,25), demonstrating loss of the WT Apc allele in all tumors from Blm heterozygous mice.

We next investigated the mutational mechanisms responsible for loss of the normal Apc allele in the intestinal tumors by LOH analysis. Apc maps to chromosome 18 and is located on the genetic map at 15.0 centimorgans (26). We used quantitative polymerase chain reaction (PCR) with simple sequence length polymorphism markers (19) for three loci to examine allelic loss on chromosome 18 in our set of 25 tumors (Fig. 2C). Tumors inApcMin /+;Blm +/+mice and 18 of 20 tumors in Apc Min/+;Blm Cin/+ mice were characterized by LOH of Apc and all markers proximal and distal toApc on chromosome 18 (Fig. 2C). Two tumors fromApc Min/+;Blm Cin/+ mice, of which one is shown, remained heterozygous at the proximal markerD18Mit19 (Fig. 2C). These results indicate thatApc loss in ApcMin /+tumors with two WT Blm alleles is characterized by LOH of chromosome 18 but that in someBlmCin /+ tumors, loss of the normalApc allele occurs by somatic recombination.

To test whether the tumors fromApcMin /+;BlmCin /+mice retained the WT Blm allele, we performed a quantitative PCR-based assay on 14 tumors. The ratio of the WT to the targeted allele in each tumor sample fromApcMin /+;BlmCin /+mice was not significantly different from that in adjacent normal tissue (Fig. 2D). Western blots of 10 tumor lysates fromApcMin /+;BlmCin /+mice evaluated with antiserum to COOH-terminus of BLM confirmed that BLM expression had been retained (fig. S3A). Similarly, immunofluorescence of MLV-induced lymphomas fromBlmCin /+ mice demonstrated nuclear BLM staining (fig. S3B). These results suggest that mutation of the remaining WT Blm allele was not required for tumor formation in either T cells or intestinal tissues.

Mutation of one allele of Blm has measurable consequences for the phenotype of murine somatic cells and for the tumor susceptibility of the mouse. Our data demonstrate that Blmhaploinsufficiency is sufficient to affect tumor formation in susceptible mice, and probably alter genomic stability. These effects have not been described in other targeted Blm mice (12, 13). These data also suggest thatBlm haploinsufficiency could promote tumor formation in tissues other than those studied here. Additionally, although none of the tumors inApcMin /+;BlmCin /+mice were invasive at the 4-month end point of our experiments, it is possible that these tumors would progress to malignancy if given more time. Our results are also important for human populations: About 1 in 100 Ashkenazi Jews carry one mutant allele of BLM(25, 26). An accompanying paper by Gruberet al. (27) demonstrates that carriers ofBLMAsh have a more than twofold increase in the occurrence of colorectal cancer. Together, these studies suggest thatBlm/BLM mutation is an important modifier of intestinal cancer predisposition and that individuals carrying one mutant allele of BLM may have one of the noteworthy clinical hallmarks of BS—namely, increased cancer predisposition.

Supporting Online Material

Materials and Methods

Figs. S1 to S3

Table S1

  • * Present address: Department of Surgery, Division of Epithelial Pathobiology, University of Cincinnati College of Medicine, Cincinnati, OH 45267, USA.

  • To whom correspondence should be addressed. E-mail: Joanna.Groden{at}


View Abstract

Navigate This Article