MOLECULAR MECHANISMS OF EMBRYONIC INDUCTION 
Richard L. Maas, M.D., Ph.D., Assistant Investigator 
A central theme in vertebrate organogenesis is an 
interaction between two apposed cell layers, result- 
ing in the morphologic transformation of one or 
both cell types. Such inductive interactions are in- 
volved in the development of many vertebrate or- 
gans. Dr. Maas's laboratory' has focused on the role 
that transcription factors of the homeobox and 
paired box classes play in controlling inductive pro- 
cesses in two developmental systems, the formation 
of the vertebrate eye and the kidney. Both organs are 
formed on the basis of inductive interactions, and 
the same classes of gene products are implicated in 
the formation of both organs. A long-term goal is to 
dissect the genetic pathways involved in the forma- 
tion of these two organs by identifying genes whose 
expression is regulated by Hox or Pax genes. An 
ancillary goal is to gain insight into pathogenesis of 
human birth defects involving the eye and the 
kidney. 
Role of PAX6 in Ocular Development 
The vertebrate eye forms from an initial out- 
growth of the forebrain, the diencephalon, which 
forms the optic vesicle and subsequently the optic 
cup. The optic cup comes to lie in apposition to the 
surface ectoderm, causing it to thicken and invagi- 
nate, forming the lens vesicle and then the lens. In 
the mouse, a semidominant mutation called Small 
eye exists in which this basic process is disturbed: 
heterozygotes have microphthalmia, small lenses, 
and absent or missing anterior chambers; homozy- 
gotes have completely absent eyes and noses. The 
ocular phenotype is consistent with a primary de- 
fect in lens induction, since transgenic mice whose 
lenses are ablated during embryogenesis have a phe- 
notype strikingly similar to the Small eye mutation. 
Dr. Maas's laboratory is studying the Small eye 
mutation as a way to study the role of paired box and 
homeobox transcription factors in inductive pro- 
cesses. The murine homeobox- and paired box- 
containing gene. Pax- 6, has been mapped to a re- 
gion near the Small eye locus on mouse 
chromosome 2. To test the hypothesis that Pax- 6 
might be involved in the Small eye phenotype, the 
laboratory cloned the human PAX6 gene in order to 
test its involvement in a disorder of human ocular 
development, aniridia, which has been proposed as 
homologous to the Small eye mutation. One ratio- 
nale for choosing the human gene for study was that 
the number of Small eye alleles is limited to five, 
two of which involve large deletions. In contrast, 
aniridia occurs at a frequency of ~ 1/64,000, and 
new mutations account for about one-third of aniri- 
dia cases. Thus there are a large number of different 
aniridia alleles, which could prove powerful in dis- 
secting parts of the protein important for function. 
To explore the role of PAX6 in ocular develop- 
ment. Dr. Maas and his co-workers cloned the hu- 
man PAX6 gene and localized it to 1 Ipl 3, the map 
location of aniridia. Work from the laboratory of Dr. 
Grady Saunders (University of Texas, Houston) has 
provided evidence that PAX6 is involved in aniridia, 
by localizing PAX6 to a smallest region of overlap of 
70 kb. The human PAX6 cDNA encodes a 422- 
amino acid protein with both homeobox and paired 
box domains, as well as a serine- and threonine-rich 
carboxyl terminus that could function as a transcrip- 
tional activator. This extraordinarily conserved gene 
is 100% identical at all 422 amino acids to the 
mouse protein and 96% identical to the homologous 
zebrafish gene at the amino acid level. Results from 
other laboratories have shown that Pax- 6 is ex- 
pressed in the optic vesicle, lens, olfactory bulb, 
and hindbrain. 
Dr. Maas's group determined the complete geno- 
mic structure of the human PAX6 gene. This gene, 
which spans ~30 kb of genomic DNA, has 14 differ- 
ent exons and is the most complex Hox or Pax gene 
structure determined to date. One exon is alterna- 
tively spliced and, when utilized, inserts 14 amino 
acids into the paired domain, potentially changing 
the DNA-binding specificity of the protein. The se- 
quence of all exon-intron boundaries has been de- 
termined, as well as that of the surrounding flanking 
sequences. This information has permitted the de- 
sign of polymerase chain reaction (PGR) primers 
that amplify each individual exon of the human 
PAX6 gene, making it possible to analyze it effi- 
ciently for mutations by the technique of single- 
stranded conformation polymorphism (SSGP) 
analysis. 
Using genomic DNAs from 10 aniridia patients, 
members of the laboratory identified four indepen- 
dent mutations in the human PAX6 gene in aniridia. 
These mutations include a 1-bp deletion in a splice 
donor site, a 7-bp insertion into the PAX6-coding 
region resulting in a frameshift, and a nonsense mu- 
tation in the homeodomain just prior to the third 
helix, which lies in the major groove of DNA and is 
thought to be primarily responsible for sequence 
recognition. This last mutation would result, at a 
minimum, in a protein that would be unable to bind 
CELL BIOLOGY AND REGULATION 
91 
