human immunodeficiency virus type 1 gene ex- 
pression via a novel DNA target site. / Virol 
66:3946-3949. 
Tiley, L.S., and Cullen, B.R. 1992. Structural and 
functional analysis of the visna virus Rev-response 
element. / F/ro/ 66:3609-3615. 
Tiley, L.S., Malim, M.H., Tewary, H.K., Stockley, 
P.G., and CuUen, B.R. 1992. Identification of a 
high-affinity RNA-binding site for the human im- 
munodeficiency virus type 1 Rev protein. Proc 
Natl Acad Sci USA 89:758-762. 
Weinberg, J.B., Matthews, T.J., Cullen, B.R., and 
Malim, M.H. 1991. Productive human immuno- 
deficiency virus type 1 (HIV-1) infection of non- 
proliferating human monocytes. / Exp Med 
174:1477-1482. 
MOLECULAR MECHANISMS IN LYMPHOCYTE DEVELOPMENT 
Stephen V. Desiderio, M.D., Ph.D., Associate Investigator 
Assembly of Antigen Receptor Genes 
The antigen receptors of B and T cells exhibit an 
extraordinary range of binding specificities. This di- 
versity results in large part from a peculiarity of 
their gene structure: antigen receptor polypeptide 
chains are encoded in the germline as discrete DNA 
segments and joined during lymphoid development 
by site-specific recombination. Antigen receptor 
gene rearrangement and its regulation remain an 
area of interest for Dr. Desiderio and his colleagues. 
Phosphorylation regulates activity and expres- 
sion of the recombination activator protein 
RAG-2. The ability of cells to support rearrange- 
ment of antigen receptor genes is absolutely corre- 
lated with the expression of the recombination acti- 
vator genes RAG-1 and RAG-2, but the mechanism 
by which these genes activate rearrangement is not 
known. In the past year Dr. Desiderio and his col- 
leagues have found that activity and expression of 
the RAG-2 protein are regulated by multiple phos- 
phorylation pathways. These observations raise in- 
teresting questions concerning the regulation of an- 
tigen receptor gene rearrangement by extracellular 
signals and the coupling of rearrangement to the 
cell cycle. 
To examine the products oi RAG-1 and -2, a series 
of specific polyclonal antibodies were raised and 
used to immunoprecipitate the proteins from cell 
lines expressing their genes. The apparent size of 
the RAG-1 protein was ~ 1 10 kDa, as predicted by 
nucleotide sequence. The RAG-2 protein had an ap- 
parent size of 70 kDa, significantly larger than pre- 
dicted, which suggested that it had undergone post- 
translational modification. Both RAG-1 and -2 were 
found to be phosphorylated and to reside largely in 
the nucleus. 
The RAG-2 protein is 527 amino acids long. By 
site-directed mutagenesis and phosphopeptide anal- 
ysis, the major in vivo phosphorylation sites were 
mapped to Ser356 and Ser392. These sites were 
found to be phosphorylated in vitro by purified ca- 
sein kinase II (CKII). 
To examine the functional importance of these 
CKIl-Iike phosphorylations, mutant RAG-2 proteins 
carrying Ala substitutions at a series of Ser or Thr 
sites were assayed for activation of rearrangement in 
vivo. None of these mutations had any effect on ex- 
pression of RAG-2 RNA or protein. Substitution of 
Ala for Ser at position 356 resulted in a large (5- to 
10-foId) decrease in activity; the other mutations 
had no effect on rearrangement. These experiments 
strongly suggest that the activity of RAG-2 is en- 
hanced by phosphorylation at Ser356 and that CKII, 
or a CKII-like enzyme, regulates RAG-2 function in 
vivo. 
The RAG-2 sequence shows several potential 
p34cdc2 phosphorylation sites, and two of these were 
found to be phosphorylated selectively by p34'''*'^^ 
or a related kinase in vitro. Unexpectedly, a mutant 
carrying an Ala substitution at the major phosphory- 
lation site (RAG-2 T490A) was expressed at a far 
higher steady-state level than wild type. This effect 
was not observed upon mutation of the minor 
p34cdc2 phosphorylation site, nor upon mutation of 
the CKII sites (see above). The relative increase in 
expression of RAG-2 T490A is most likely the result 
of protein stabilization, because 1) mutant and 
wild-type RNA levels were identical and 2) chi- 
meric proteins containing wild-type or mutant 
RAG-2 polypeptide segments exhibited a similar 
phenotype. The activities of p34'^'''^^ and related ki- 
nases vary during the cell cycle; likewise, RAG-2 
protein levels appear to vary, reaching a minimum 
during S phase. These observations have prompted a 
specific hypothesis: that phosphorylation of RAG-2 
by p34'^'''^^ or a related kinase increases its rate of 
GENETICS 179 
