Most human cancer cells are aneuploid, meaning they have an aberrant number of chromosomes. Whether gene mutations are responsible for this aneuploidy is debated1. In yeast, mutations in several genes, including genes involved in chromosome segregation and regulation of the mitotic spindle checkpoint, can result in aneuploidy. In humans, two mitotic spindle checkpoint genes, BUB1 and BUB1B (encoding BUBR1) have been reported to be mutated infrequently in colorectal cancers with chromosomal instability and in other aneuploid tumor types2,3. But convincing functional studies that could establish a causal connection between BUB1B mutations and chromosomal instability have been missing. On page 1159, Sandra Hanks and colleagues4 report hereditary mutations in BUB1B in individuals with a rare recessive disease called mosaic variegated aneuploidy (MVA) and provide such an important piece of evidence.

MVA syndrome

In 1986, Vigfusson et al.5 observed substantial developmental delay and apparent mental retardation in a male pigtail macaque monkey. They found that 75.7% of his lymphocyte nuclei were trisomic, but with the supernumerary chromosome varying from cell to cell. Examination of cultured skin fibroblasts showed a similar unusual karyotype.

A few years later, several humans, including two pairs of siblings, were found to have a similar cytogenetic phenomenon. Cells of these individuals frequently form micronuclei and have abnormal numbers of several chromosomes in multiple tissues, called mosaic aneuploidies (Fig. 1). For example, in 1991, the case of a 17-year-old girl with severe microcephaly and mental retardation was reported6. Karyotype analyses of her phytohemagglutinin-stimulated lymphocytes, cultured skin fibroblasts, direct and cultured bone marrow and Epstein-Barr virus–transformed lymphoblasts all showed that at least 10% of cells were trisomic with respect to many different chromosomes. Trisomies involving all chromosomes except chromsomes 5, 10, 13, 14 and 17 were observed. The existence of this mosaic trisomy in four different tissues and in repeated cultures over a 3-y period suggested that it was due to a genetic abnormality resulting in mitotic instability.

Figure 1: Karyotype of a cell from an individual with MVA.
figure 1

Instead of two copies (normal for a diploid cell), there are three copies each of chromosomes 9, 12, 21 and X.

So far, only 15 cases of MVA have been reported worldwide7. It is not clear whether all cases represent the same condition, as some clinical and cytogenetic differences exist among them. For example, mitoses of some of the individuals have a higher frequency of separated centromeres and splayed chromatids in the presence of colcemid, a cytogenetic phenomenon referred to as premature centromere division8. The two most common clinical abnormalities of individuals with MVA are microcephaly and intrauterine growth retardation. Also, the risk of malignancy for the affected individuals is high. Almost one-third of the young affected individuals reported developed leukemia, rhabdomyosarcoma or Wilms tumor7.

Aneuploidy and mitotic spindle checkpoint

Matsuura et al.9 recently reported that skin fibroblasts cultured from six individuals with both premature centromere division and MVA showed a defect in the mitotic spindle checkpoint. This finding was not unexpected, as others had previously shown that deficiency of the spindle checkpoint components resulted in unequal segregation of chromosomes into daughter cells and thereby led to aneuploidy2. The mitotic checkpoint serves as a surveillance mechanism that ensures the faithful transmission of chromosomes from a mother cell to its two daughter cells during mitosis. It warrants that all chromosomes congress to the metaphase plate before separating into the two newly forming daughter cells, and that the chromosome kinetochores attach properly to the spindle microtubules. Many important members of this checkpoint have been identified, including the genes Bub1, Bub3, Mad1l1, Mad2l1, Mxd3 and Mps1 (ref. 10). Most of these regulatory components seem to function as inhibitors of a multisubunit E3 ubiquitin ligase, known as the anaphase promoting complex or cyclosome (APC/C), which presumably controls the exit from metaphase.

Mutational analysis of one of the mitotic checkpoint genes, BUB1B, in eight families with MVA by Hanks et al.4 showed that five families carry mutations in BUB1B in both alleles. Four families have either nonsense or frameshift mutations in one allele that result in the premature truncation of the encoded BUBR1 protein, deleting its kinase domain and thus inactivating its function. The reported mutations of the second allele were missense mutations. Five (of six) missense mutations were located in the kinase domain of BUBR1. They probably also result in a loss of function, although functional studies are required to prove the case.

The paper by Hanks et al.4 is a perfect illustration of the idea that uncovering the underlying genetic basis of a rare disease with increased cancer risk can contribute to the understanding of general aspects of tumorigenesis.

Chromosomal instability and cancer

Like the supernumerary chromosomes that vary from cell to cell in individuals with MVA, most human cancers are aneuploid, and chromosomal instability presumably drives tumor progression by accelerating the rate of gain or loss of individual chromosomes. The molecular mechanisms of chromosomal instability in human cancers largely remain to be elucidated.

Several years ago, we hypothesized that mutations in human spindle checkpoint genes could cause chromosomal instability in cancers11. We cloned the two human homologs of yeast Bub1, called BUB1 and BUB1B, and found that the two genes were mutated, albeit rarely, in chromosomal instability tumors of individuals with colorectal cancer2. This study provided the first evidence that gene mutations can result in a defective spindle checkpoint in humans and thus might be responsible for aneuploidy in human cancers. One BUB1B mutation in an aneuploid colorectal cancer was a somatic deletion that was predicted to remove part of the protein's kinase domain—analogous to the truncating mutations now found in individuals with MVA.

Mutant mice that express low levels of BubR1 protein develop progressive aneuploidy but form tumors only after being challenged with a carcinogen12. These findings, as well as the low frequency of BUB1 and BUB1B mutations in cancers and the lack of gene mutations found in any other members of the mitotic checkpoint machinery thus far, led to questions regarding the importance and applicability of our original observations. The belief that spindle checkpoint defects cause aneuploidy in human tumors, however, is revived by pinning down BUB1B mutations as the underlying genetic alteration in MVA.