The Molecular Genetics of Cancer 
Bryan Williams (Hospital for Sick Children, To- 
ronto). Within this chromosomal region, we 
found a cluster of genes that appear to be in- 
volved in both childhood kidney cancer and nor- 
mal kidney development. In one-half of the tu- 
mors we examined, at least two of these 
candidate tumor-suppressor genes were turned 
oflfr Remarkably, the tumors in which the genes 
were turned off also contained tissues not nor- 
mally found in the kidney, such as muscle, sug- 
gesting that the same genes are involved in both 
tumor development and normal kidney develop- 
ment. We now need to find out what role these 
genes play in kidney development and why they 
are turned off in cancer. 
A surprising result from the cloning and map- 
ping of this large genomic region was the identi- 
fication of an "archipelago" of cytosine-guanine 
islands, or clusters of CG nucleotides that are of- 
ten near genes. Several of the CG islands in the 
Wilms' tumor gene region were methylated, a 
feature normally characterizing the inactive X 
chromosome. This offers a possible mechanism 
for inactivation of tumor-suppressor genes, as 
well as an unexpected link to our research on 
DNA methylation. 
In addition to the known gene on chromosome 
1 1 that predisposes to Wilms' tumor, we have 
discovered another gene at a different location on 
chromosome 11 (band Hp 15) that also predis- 
poses to Wilms' tumor but is involved in bladder, 
breast, and lung cancer as well. Loss of one copy 
of this gene in the germline appears to cause 
Beckwith-Wiedemann syndrome, a disorder of 
organ overgrowth and malignancy. We are now 
attempting to clone this tumor-suppressor/Beck- 
with-Wiedemann syndrome gene. In this eff'ort 
we are using a strategy we have developed to iso- 
late tumor-suppressor genes directly. We frag- 
ment human chromosomes into subfragments of 
2-10 million nucleotides in a way that allows us 
to transfer these into tumor cells. This novel tech- 
nique will allow us to bridge a gap in cloning 
methods between chromosome-size pieces (aver- 
age 100 million nucleotides) and yeast artificial 
chromosomes (average 300,000 nucleotides), 
and it may have general application to cloning a 
wide variety of genes. 
132 
