TRANSGENIC APPROACHES FOR STUDYING MOUSE DEVELOPMENT 
Richard D. Palmiter, Ph.D., Investigator 
During the last year this laboratory has contin- 
ued the productive collaboration with Dr. Ralph 
Brinster's laboratory at the University of Pennsylva- 
nia. These projects involve introduction of foreign 
genes into the germline of mice as a means of 
studying genetic phenomena in the intact animal. 
In addition to continuing studies aimed at 1) un- 
derstanding the regulation of animal growth, 2) 
identifying the DNA elements involved in cell-spe- 
cific gene regulation, and 3) analyzing oncogene- 
induced neoplasia of exocrine pancreas and liver, 
there has been progress in three new areas. 
I. Homologous Recombination. 
Targeting foreign DNA to specific chromosomal 
sites by homologous recombination provides an in- 
valuable tool for studying gene function and cor- 
recting genetic defects. Although homologous re- 
combination is the rule in some lower eukaryotes, 
nonhomologous integration of foreign DNA pre- 
dominates in mammalian cells. The frequency of 
homologous events is —0.1-1% of all integration 
events. If the frequency of achieving homologous 
recombination were 1% or better, then microinjec- 
tion of DNA into mouse eggs would represent a di- 
rect way of introducing changes in specific genetic 
loci. Thus experiments were designed to attempt to 
correct a genetic defect in the class II major histo- 
compatibility locus that occurs in some strains 
of mice. 
DNA molecules containing the 5' end of a func- 
tional gene were injected into mouse eggs bear- 
ing mutant E^ genes with a 630 bp deletion that 
encompasses the promoter and first exon. The de- 
letion was corrected in 1 of —500 transgenic mice 
that incorporated the injected DNA, and this cor- 
rected E^ gene was transmitted to progeny that 
were bred to homozygosity. Southern blot analysis, 
polymerase chain reaction amplification of the DNA 
spanning the deletion, and sequence analysis re- 
vealed that the corrected allele resembles the wild- 
type E^ gene. At sites of single-base pair polymor- 
phisms, there was apparently random conversion of 
either the donor or recipient DNA sequence; in ad- 
dition, many point mutations were introduced, sug- 
gesting that the recombination process was error 
prone. Messenger RNAs were produced from the 
corrected allele in a tissue-specific manner, but 
their sizes were different from the wild-type allele 
and they did not produce detectable E^ protein. 
This experiment demonstrates the feasibility of tar- 
geting foreign DNA to a gene that is completely in- 
active in fertilized eggs. However, because there 
was only one correction event after injecting 
> 10,000 mouse eggs, the frequency cannot be cal- 
culated. Several aspects of this experiment that be- 
came evident during its execution suggested that 
the E^ gene may not have been an optimal choice 
for these experiments; hence current experiments 
are aimed at targeting DNA to another gene. 
II. Regulation of Globin Gene Expression. 
For many years, expression of globin genes in 
transgenic mice was either not achieved or 
achieved at very low levels. This was true for several 
reasons: 1) expression of the human (3-globin gene 
was seriously inhibited by plasmid vector se- 
quences that were included with most of the early 
constructs that were tested; 2) the enhancers for 
the human P-globin gene are located within and 
downstream of the gene, features that made analy- 
sis difficult; and 3) optimal expression of the P-glo- 
bin gene depends on sequences located —80 kb 
upstream of the gene. These sequences, now often 
referred to as the locus-activating region (LAR), 
were discovered because there are five sites within 
this region that are DNase I hypersensitive at all de- 
velopmental stages in erythroid cells and because 
in certain thalassemias these sequences are deleted, 
leaving the (i-globin gene and its enhancers intact. 
Frank Grosveld and his collaborators showed that 
when transgenic mice were made that contain the 
LAR along with the human p-globin gene, expres- 
sion of each integrated gene copy was equivalent to 
an endogenous mouse P-globin gene. Subsequent 
experiments, in collaboration with Dr. Tim Townes 
(University of Alabama), revealed that pieces of 
DNA carrying individual hypersensitive sites could 
produce nearly the same result as the intact LAR. 
These sequences could also be used to achieve 
high-level expression of the human a-globin gene, 
a gene that had not been expressed in transgenic 
mice without the LAR. Co-injecting both the a- and 
P-globin genes with LAR sequences produced trans- 
genic mice that express as much human hemoglo- 
bin as mouse hemoglobin, and the human globin is 
functional, as judged by oxygen dissociation mea- 
surements on human globin purified from these 
mice. These experiments demonstrate the feasibility 
of expressing mutant human hemoglobin mole- 
Continued 
279 
