MOLECULAR STUDY OF DOWN SYNDROME 
David M. Kurnit, M.D., Ph.D., Investigator 
Etiology 
Analysis by Dr. Kurnit and his colleagues of objec- 
tive molecular polymorphisms on chromosome 2 1 
has already yielded two surprising results: 1 ) 94% of 
meiotic errors are maternal (previous subjective cy- 
togenetic heteromorphism data, which must have 
been biased, indicated wrongly that only 75-80% of 
nondisjunction errors were maternal); 2) crossing 
over is either absent or, less likely, unusually telo- 
meric on a plurality of nondisjoined chromo- 
some 2rs. 
To learn why the risk of having a child with Down 
syndrome increases dramatically with advancing 
maternal age, polymorphisms that abut the centro- 
mere of chromosome 21 are required. For this pur- 
pose, the laboratory constructed a molecular probe 
that detects four independent pericentromeric poly- 
morphisms on 21q. These polymorphisms result 
from variations in the length of oligo d(A,C):d(G,T) 
stretches. A fluorescent detection system has been 
developed and probes tailored so that all four 
polymerase chain reaction (PGR) polymorphisms 
among three family members (a father, mother, and 
child with trisomy 21) can be placed in a single lane 
of a polyacrylamide gel. Analyses of these polymor- 
phisms will allow the inheritance of the centro- 
meres of chromosome 21 to be followed in 600 
Down syndrome families. 
By determining whether the type of meiotic error 
(i.e., meiosis I vs. meiosis II, which can only be 
decided at the centromere) that yields nondisjunc- 
tion shows either no maternal-age dependence or a 
distinct maternal-age dependence, one should be 
able to distinguish whether the increase with mater- 
nal age is due to relaxed selection (whereby older 
mothers lose the ability to select against abnormal 
conceptuses) or to older eggs (whereby older 
mothers release gametes that are more likely to have 
chromosomal errors) . Thus the broader question of 
why older mothers bear more Down syndrome off- 
spring should be answered. Furthermore, it will be 
possible to measure recombination across the cen- 
tromere in the minority of families where short-arm 
markers (cytogenetic heteromorphisms and/or in 
situ hybridization with (3 satellite) are distinct and 
informative. 
Pathogenesis: Recombination-based Assay 
and Genes on Chromosome 21 
Dr. Kurnit's laboratory has developed a recombi- 
nation-based assay for rapid isolation of genes on 
chromosome 2 1 . In this assay chimeric phages or 
phasmids are isolated that have acquired a plasmid 
with supFviz homologous recombination mediated 
by shared homology between the genie (cDNA) se- 
quences in the phages and the genomic sequence in 
the plasmid. This enables one to determine if a geno- 
mic sequence is transcribed in a given tissue at a 
given time of development. The recombination- 
based assay can then be reversed by selecting against 
supFto yield the gene of interest free of the genomic 
sequence in the plasmid that was originally used to 
screen for the transcribed sequence. 
Although these goals can be accomplished via hy- 
bridization, the benefits of the recombination-based 
assay stem from the ability to screen a large number 
of recombinants much more rapidly than by hybrid- 
ization. Through use of the assay, a complex library 
encompassing 10^ recombinants can be screened 
simply by plating on several petri dishes. This speed 
and economy will be required to append a genie 
initiative (i.e., a screen for transcribed sequences) 
onto the genomic initiative, making it possible to 
monitor efficiently whether a few-copy sequence is 
transcribed in a given tissue at a given time of devel- 
opment. If so, the gene of interest can be isolated by 
reversing the recombination-based assay. 
Recombination (5M/>f acquisition) is selected for 
on a lacZam dnaBam/Pl fean-balanced lethal 
host, DM2 1 , that Dr. Kurnit and his colleagues con- 
structed. They also constructed a pMAD supF plas- 
mid based on R6K (not homologous to the ColEI 
replicon, present in pBR322, for example) that al- 
lows screening of cDNA libraries via recombination 
that have sequences homologous to ColEl, includ- 
ing common cDNA libraries constructed in XgtlO 
and Xgtll, and phasmid vectors such as XZAP. 
DM2 1 selects for phages that have incorporated a 
supF plasmid by homologous recombination (due 
to the presence of the dnaBam mutation) yielding 
blue plaques (due to the presence of the lacZam 
mutation). Furthermore, phasmid libraries can be 
screened using DM2 1 (pBR322), astheColEl repli- 
con in DM21 (pBR322) is incompatible with and 
thereby inhibits replication driven by the GolEl 
replicon in phasmid vectors. Excision during rever- 
sal of the selection process (counterselection) 
yields the genie (cDNA) insert free of the genomic 
insert in pMAD. Counterselection of inserts selected 
from phage vectors is achieved by PGR application, 
followed by cloning into a pBLUESGRIPT plasmid. 
Counterselection of inserts selected from phasmid 
212 
