in lysates derived from cells expressing the fusion 
protein but not in lysates from cells expressing the 
glutathionine 5-transferase protein alone. This pro- 
vides strong evidence that the cloned sequence rep- 
resents a PI 4-kinase. 
The cloned cDNA shows no extensive sequence 
similarity to any protein or nucleic acid sequences 
present in computer databases. However, the pro- 
tein does show limited sequence similarity to two 
inositol phosphate-binding proteins, inositol 1- 
polyphosphatase and inositol (1:2) cyclic hydro- 
lase. In addition, the predicted protein contains a 
putative ATP-binding site of the type found in phos- 
phofructokinase. Consistent with this assignment as 
the ATP-binding site, the PI 4-kinase is inactivated 
by treatment with the arginine-specific reagent 
phenylglyoxal, and inclusion of ATP protects en- 
zyme activity. Neither the PI 4-kinase nor the re- 
cently cloned PI 3-kinase shows significant similar- 
ity to serine, threonine, or tyrosine protein kinases 
in this ATP-binding domain, suggesting that the 
lipid kinases are derived from a different ancestral 
kinase than are the protein kinases. 
The PI 4-kinase can also be inactivated by treat- 
ment with fluorosulfonylbenzoyladenosine (FSBA), 
a reagent that selectively labels ATP-binding sites. 
The inactivation is concentration and time depen- 
dent and is prevented by addition of ATP. Isolation 
and sequencing of the FSBA-labeled peptide from 
the PI 4-kinase is currently under way to provide 
definitive identification of the ATP-binding site of 
this enzyme. This information will be used to con- 
struct altered forms of the PI 4-kinase via site- 
directed mutagenesis to study structure-function re- 
lationships within the enzyme. 
Northern blot analysis of RNA derived from a hu- 
man ceil line demonstrated the presence of two hy- 
bridizing species of 3-4 and 4.6 kb. When mouse 
RNA was used, messages of similar size were found 
to cross-hybridize with the human PI 4-kinase cDNA 
probe, indicating that the PI 4-kinase is relatively 
well conserved between these two species. Prelimi- 
nary evidence suggests that the levels of the 3-4-kb 
message are increased in monocytes stimulated with 
phytohemagglutinin, suggesting that PI 4-kinase ac- 
tivity may be regulated at the level of transcription. 
(The project described above was supported by a 
grant from the National Institutes of Health.) 
Phosphatidylinositol- 4- monophosphate (PIP) 
phosphatase. The reactions in the pathway for the 
synthesis of the polyphosphoinositides have been 
presumed to be reversible; however, the enzymes 
responsible for the dephosphorylation of PIP and 
PIP2 (phosphatidylinositol 4,5-bisphosphate) have 
not been extensively characterized. A PIP phospha- 
tase from rat brain has been extensively purified in 
Dr. Pike's laboratory. Physical and biochemical stud- 
ies suggest that it is a glycoprotein with a molecular 
weight of 70,000-90,000. Additional work will 
focus on the complete purification of this enzyme as 
well as the generation of antibodies and protein se- 
quence necessary for cloning of this phosphatase. 
Desensitization of the EGF Receptor 
Previous work in Dr. Pike's laboratory has shown 
that treatment of A431 cells with high concentra- 
tions of EGF reduces their ability to internalize '^'l- 
EGF. Additional studies suggested that several other 
EGF-stimulated responses are also desensitized after 
treatment of A431 cells with EGF. These include 
EGF-stimulated PI turnover and EGF-stimulated tyro- 
sine protein kinase activity. Desensitization was also 
associated with an impairment in the ability of EGF 
to induce receptor dimer formation. The desensiti- 
zation was homologous, or agonist specific, in na- 
ture and did not appear to involve protein kinase 
C-mediated events. 
In several systems in which receptor desensitiza- 
tion has been studied, the underlying molecular 
basis for desensitization has been shown to involve 
phosphorylation of the receptor protein. Therefore 
studies of the phosphorylation state of desensitized 
receptors were undertaken. These studies identified 
a serine-containing phosphopeptide in the EGF re- 
ceptor, the presence of which correlated with the 
condition of desensitization. The identity of the 
peptide was determined by both direct amino acid 
sequencing and radiosequencing of in vivo '^P- 
labeled peptide. 
A peptide was synthesized based on the sequence 
surrounding this site of phosphorylation of the EGF 
receptor. A kinase activity capable of phosphorylat- 
ing the synthetic peptide was identified in A43 1 cell 
cytosol. This activity was shown to be stimulated 
acutely following treatment of the cells with EGF. 
Partial purification and characterization of the ki- 
nase suggested that it had a molecular weight of 
~90,000. The kinase was identified as p34'^'"^^ 
based in part on this molecular weight estimation 
combined with the ability of anti-p34'''''^^ antibodies 
to immunoprecipitate peptide kinase activity. 
Incubation of purified EGF receptors with p34'"*''^ 
demonstrated that the serine kinase directly phos- 
phorylated the EGF receptor. Phosphorylation of 
the EGF receptor by p34'"'''^ led to a reduction in the 
tyrosine kinase activity of the receptor. Future ex- 
periments will examine the effect of this phosphor- 
ylation on the ability of EGF to induce receptor 
CELL BIOLOGY AND REGULATION 
103 
