ANALYSIS OF GENES IN HUMAN Xq28 
Jane Gitschier, Ph.D., Assistant Investigator 
Mutations in X-linked genes account for —15% of 
inherited disease in humans. A number of these loci 
genetically map to Xq28, a region of DNA flanked 
by the long-arm telomere and the fragile site associ- 
ated with mental retardation. This laboratory is 
pursuing several lines of research regarding the or- 
ganization, function, and defects of genes in this 
important region. The starting point for these proj- 
ects has been the gene for coagulation factor VIII, 
which is responsible for the bleeding disorder he- 
mophilia A. 
I. Mutations in the Factor VIII Gene. 
Hemophilia A is a relatively common inherited 
disorder, affecting 1 male in 5,000 worldwide. 
About one-third of hemophilia cases appear to be 
due to new mutations. Because of the clinical het- 
erogeneity of hemophilia, a wide variety of mu- 
tations is expected. Thus investigating the origin 
and spectrum of hemophilia A-causing mutations 
should lead to a better understanding of mutagene- 
sis in humans. 
Because the factor VIII gene is very large (186 
kb), a two-step approach has been taken to identify 
mutations in hemophilia DNA samples. A small re- 
gion of interest is amplified from patient DNA by 
the polymerase chain reaction, and the amplified 
DNA is tested for mutations on a denaturing gradi- 
ent gel. Variation in sequence changes the thermal 
stability of the molecule and therefore causes the 
amplified DNA fragment to shift in mobility on this 
type of gel, compared with that of a control sample 
amplified from normal DNA. 
This approach was applied to the search for mu- 
tations in and near exon 8, which codes for one of 
the two acidic domains essential for proper factor 
VIII function. Screening of amplified DNA from 228 
unselected hemophilia A patients revealed two new 
mutations, as well as a new diagnostic polymor- 
phism in intron 7. Although technically more de- 
manding, analysis by "heteroduplexing" amplified 
patient DNA samples with control DNA samples 
prior to electrophoresis should be more sensitive 
than analysis of simple "homoduplexes." Indeed, 
rescreening the same population by heteroduplex- 
ing revealed another mutation. All three mutations 
(two different missense and one 4 bp deletion) 
were observed in patients with clinically severe he- 
mophilia and <1% factor VIII activity. Although 
only three mutations were found, only 3% of the 
mRNA and 0.16% of the gene have been examined. 
Melt-map computer programs (provided by Dr. 
Leonard Lerman of the Massachusetts Institute of 
Technology) were used to calculate the theoretical 
shift in thermal stability for fragments with each of 
the mutations and the polymorphism. Generally 
there is good agreement between the predicted and 
observed mobility changes. These programs are 
now being used to help design a strategy for 
screening coding and flanking sequences through- 
out the factor VIII gene. In time this approach 
should uncover hemophilia-causing mutations in 
many patients. Knowledge of the underlying defect 
will greatly improve genetic counseling and aid in 
mutagenesis studies. 
For example, the genetic origin of mutations was 
investigated in eight families with sporadic hemo- 
philia whose defects had previously been deter- 
mined by Southern blot analysis. Maternal mosa- 
icism was discovered in two cases. In one case the 
mosaicism appeared to be restricted to the germ- 
line, but mosaicism clearly extended to somatic tis- 
sue in the second. This surprising result suggests 
that mosaicism may be a common feature of muta- 
genesis, and this possibility must be taken into ac- 
count for genetic counseling. 
II. Intron-22 Gene. 
One remarkable feature of the factor VIII gene is 
the presence of a small gene completely contained 
within intron 22. This gene, which has no interven- 
ing sequences, is associated with a CpG island and 
a ubiquitous and abundant 1.8 kb transcript. Al- 
though the function of this gene is still unknown, 
several characteristics have come to light within the 
last year. 
Two other copies of the intron-22 gene are pres- 
ent on the X chromosome. Preliminary mapping 
and sequencing suggest that they may be identical 
to the one in the factor VIII gene. Based on physi- 
cal mapping studies (see below), they appear to be 
linked tightly to each other and lie within 1 Mbp of 
the factor VIII gene in the 5' direction. Northern 
analysis of RNA from two hemophilia patients with 
independent deletions has shown that the intron- 
22 gene within factor VIII is transcriptionally active. 
Approximately 60% of intron-22 mRNA derives from 
the factor VIII gene itself, and the remaining 40% 
Continued 
215 
