ANNUAL REPORT OF THE
LABORATORY OF BIOCHEMICAL GENETICS
NATIONAL HEART, LUNG, AND BLOOD INSTITUTE
October 1, 1985 through September 30, 1986

  Two ~gtll cDNA libraries from human brain were screened with 3
,:!i,~~ol?e~xynr?cleotide probes .fw recombinant3 coding for a subunits of G signal
,;:':.stiucing proteins, which couple receptors activated by hormones or light to

.: i 'j ect.o,rs such as adenylate cyclase or cGMP phosphodiesterase.  Fourteen of the
,jL. ,300 recombinant clone&screened from a human basal ganglia cDNA library and
':.;-: c-f the 4CK?..OOO-clones screened from a human cerebral cortex library were
.lcCi:cted with:2 or 3 of the 3*P-probes used.  DNA inserts from 13 positive
ciones were   uenced partially; 11 clones were identified as cs cDNA and 2
clones as aj,.  The DNA insert from one of the as clones was sequenced completely
and additional partial sequences were obtained for 10 as clones.  Four species
or cs cDNA were found that differ in nucleotide sequence in the region that
corresgands to .us amino acid residues 71-88.  The clones differ in the codon for
cs amino acid residue 71 (glutamic acid vs. aspartic acid), the presence or
absence of codons for the next 15 amino acid residues, and the presence or
absence of an adjacent serine residue.  A mechanism was proposed for generating
4 species of as mRNA by alternative splicing of precursor RNA transcribed from a
single gene.

   cDNA from one of the two human ai clones was sequenced completely (BG-4),
end a partial sequence was obtained for the second clone.  The first nucleotide
residue of BG-4 ci cDNA corresponds to the 14th residue of the bovine ci coding
sequence and the last residue of BG-4 (1261) is in the 3'-untranslated region.
The amino acid sequence derived from the nucleotide sequence of human DC-4 ai
cDX4 is highly homologous to bovine and rat ci sequences reported by others.   In .-
addjtion, the 3'-untranslated region of BG-4 ci cDNA is highly homologous to the
3*-untranslated regions of bovine and rat ci cDNA. The 3'-untranslated
nucleotide sequences of human, bovine, ar;ld rat as cDNAs also are highly
conserved, but differ markedly from ci 3 -untranslated sequences. These results
suggest that the 3'-untranslated regions of as and ai genes and/or mRNA are
needed for functions that have not been identified thus Far.

   In previous studies we have shown that elevation of CAMP levels of NG108-15
neuroblastoma-glioma hybrid cells or neuroblastoma cells:for several days
results in lo-100 fold increases in the activity of voltage%ensitive calcium
channels, 15-45 fold increases in spontaneous secretion of acetylcholine at
synapses, and 5-15 fold increases in the abundance of synapses with cultured.
striated muscle cells.  In addition, the number of molecules of the
voltage-sensitive CalCiUm channel protein subunit that binds [3H)-nitrendipfne
increases 12-fold.  We previously obtained about 100 cDNA clones that hybridize
tc species of mRNA that are more abundant in NG108-15 or NS20-Y cells that had
been treated with dibutyryl CAMP for several days then in untreated control
cells.  Quantitative studies on the ex,tent of increase in abundance of the
species of mRNA that respond to dibutyryl CAMP were performed using the cloned
cDNA as probes.  Twenty cDN.4 clones were obtained that hybridize to species of
poly A+ RNA that increase in abundance lo-90 fold due to treatment of' cells with
dibutyryl CAMP.  Northern blots also were performed and the number of bands of
poly A+ R?!A that hybridize to each cloned cDNA probe and their chain lengths
were determined.

-l-


:t I "finity purified antibodies to the (x,  B, and Y protein subunits of

vc! t:;S'-s,L G?sitive calcium channels were used to screen a Xgtll cDNA library
v-eprxi
.- f    from poly A+ RNA'from rat skeletal muscle.  Approximately 20
;- ?cyy.>inay-Jt clones were Found that were identified-tentatively as calcium
chz-lr.21  o subunit cDNAs.  Other cDNA clones were obtained that are putative Y
s u'5';`7 1:  clones.              %

   in previous studies a putative cDNA clone for choline acetyltransferase was
found.  Xe now have determined the nucleotide sequence of the 1118 bp DNA insert.
Partial amino acid sequences of several peptides derived from choline
acetyltransferase by the action of peptidases were obtained in collaborative
studies by Lou Hirsh and his colleagues in Dallas.  The hgtll cDNA library was
screened again with 2 new oligodeoxynucleotide probes to different regions of
choline acetyltransferase and cDNA clones were obtained that were recognized by
both probes.  Further studies with these clones are in progress.

   Antigenic molecules termed TOP,  which are distributed in a dorsal > ventral
concentration gradient in chicken retina, are expressed early in development (by
48 hr after fertilization) in the optic cup of chicken embryos and continue to
be expressed in retina thereafter.  35S-labeled-TOP-antibody complexes were
purified by protein A-Sepharose column chromatography and subjected to
NaDodSO4/polyacrylamide gel electrophoresis and autoradiography.   TOP also was
purified from dorsal retina by anti-TOP IgC-Affigel 10 affinity column
chro~etogr aphy -  Both purification methods yielded one major band of protein
with an :<r of approximately 47,000.  A protein of M, approximately 47,000 also
was purified from chicken embryo brain.  Cultured cells dissociated from 8-day
ChiCkS. err.bryo retinas accumulated the amount of TOP expected of cells in the
intact retina,  dependin g on the position of the cells in the retina.  TOP
accuc.JlEtions by cells dissociated From dorsal or ventral retina, mixed in
different proportions and cocultured were additive.  These results show that TOP
is a protein, that the gradient of TOP is established early in development, and
that perpet uation of the gradient does not depend on the continuous presence of
an extracellular gradient of diffusable molecules or on maintenance of
interactions between cells.  Synapses and neurites in the retina of developing
chick embryos were reduced markedly by injection of anti-TOP antibody into the
eye.

    The addition of bradykinin to NG108-15 cells was shown in previous studies
to increase cellular levels of inositol-1,4,5-trisphosphate (IP3) and
diacylglycerol.  The newly synthesized IP3 in turn stimulates the release of
stored calcium ions into t'he cytoplasm, thereby activating calcium-dependent K+
char.2els.  The increased efflux of K+ ionsresults in cell hyperpolarization.
This is followed by cell depolarization due to inhibition of M channels, thereby
de.crea,si r-.g the rate of K+  efflux from cells via M channels.  Additional results
now sh;w that inhibition of M channels is due to diacylglycerol and Ca2+
deps.r.?Etn; activation of protein kinase C.  Several phosphoproteins were detected
by tic; dimensional. gel electrophoresis whose synthesis is dependent upon the
addition of bradykinin to cells.  Whereas,  injection of inosi to1
1,4,5- trisphosphate inside NClO8-15 cells results  in the release of stored
CE?.lCi:K. ir,to the cytoplasm, injection of inositol 1,3,4-trisphosphate or
inositol 1,3,4,5-tetrakisphosphate has little or no effect on calcium
mobilization, but instead results in the activation of nonspecific cation
cha.r.nzls .  Calcium ions are not required for the activation of the nonspecific

-2-


fi,iL i on channels.
..L" -        The nature and significance of these findings warrant further
investigation in light of recent reports that inositol 1,3,4-trisphosphate and
incsitol 1,3,4,5-tetrakisphosphate are present in some tissues and that inositol
l,S,i;,5-tetrakisphosphate is synthesized by phosphorylation of inositol
i,l,5-trisphosphate, catalyzed by an appropriate kinase, and that inositol
l,`$,h-trisphosphate is formed by dephosphorylation of inositol
1,3,4,5-tetrakisphosphate.     *

   Immunofluorescence staining on cryostat sections prepared from embryonic
`brain extract-treated myotubes revealed a precise colocalization of a 43,000 E?I,
cytoplasmic protein (distinct from actin) with newly-formed ACh receptor
aggregates.  This result is consistent with a role for the 43,000 Mr protein in
receptor immobilization, as suggested indirectly by studies from other
laboratories on fish electric organ and the neuromuscular junction.

  We previously showed that partially purified and highly purified fractions
from the extracellular matrix of the Torpedo electric organ induce ACh receptor
aggregation in cultured myotubes with a time course similar to' that of embryonic
pig brain extract.  We now have found that antiserum against a partially
purified ), f-action'from Torpedo (700 units/mg protein) can absorb about 60% of
the receptor aggregation activity of brain extract.  Under the same conditions,
90% of the activity in the Torpedo fraction was absorbed. This result is
consistent with the presence of immunologically related aggregation factors in
electric organ and brain.

  `v/e previously showed that neural factor induced formation of ACh receptor
aggregates on tetrodotoxin-treated myotubes is associated with the localized
deposition of basal lamina.  We now find that embryonic brain extract and
ciliary ganglion explants induce a widespread deposition of basal laaina on
non-tetrodotoxin-treated myotubes.  Ascorbate oxidase blocks this deposition of
basal lamina, suggesting that ciliary ganglion and embryonic brain extract
contain ascorbate-like factors that promote muscle basal lamina formation.   The
extensive induction of ACh receptor aggregate s by ciliary ganglion explants was
only partially inhibited by ascorbate oxidase, and basal lamina deposition still
occurred at the ACh receptor aggregate sites.  These results suggest that the
ascorbate-like factor contributes to, but is not primarily responsible for the
induction of receptor aggregates.  In addition, they suggest that deposition of
basal lamina at receptor aggregates can occur independently of the
ascorbate-like factor.

  We have been studying hormonal and neurotransmitter-dependent mechanisms
that regulate the gene for proenkephalin (pEnk), the precursor of the opioid
peptides methionine- and leucine-enkephalin,  in clonal cell lines of neural
origin,  as well as in rat brain.  NGlOS-15 neuroblastoma-glioma hybrid cells and
C6 rat glioma cells contain pEnk mRNA, quantitated by blot hybridization. C6
cells contain a much higher abundance (3-6 pg:/ug RNA) but lower enke$?alin
content than NC-108-15 cells.  Treatment of C6 cells with compounds that activate
adenylate cyclase and raise the CAMP concentration (e.g. by a beta-adrenergic
receptor agonist such as (-I- norepinephrine or by forskolin) elevate the pEnk
mRXA abundance.  Glucocorticoid hormones such as dexamethasone or cortisol,
while having no effect alone on the pEnk rnRMA level, potentiate the effect of
cA:P elevation, producing maximum elevations of S-fold.  C6 cells contain
proenkephalin but do not process this precursor significantly.  Treatment with

-3-


r;qr,.y;ir.e-`~ -
     ;,.i. i.ne and dexamethasone raises the content of proenkephalin 11-fold.
`;` c' -3 Ii : r, .-z TA F. cf cells with glucocorticoid and forskolin for l-6 hr increases pEnk
;: `s ;1. e t:-anscription at least 2.5-fold.  These results suggest that
`7 ; `,.j `! `7 c TJ " t
E>    ,icoids and neurotransmitters that elevate CAMP transcriptionally
1' e 2 !i 1 i : e enkephalin biosynthesis in enkephalinergic cells.
                         `C
    Sttld'lq
    1u3 have been initiated on the regulation of expression of the gene for
pI~on?Uropeptide Y (pNPY) , the precursor of neuropeptide Y, a putative regulator
in the auton0mi.c nervous system.  pNPY mRZ:A is relatively abundant in NGlOS-1.5
hybyi d cells.  Treatment of these cells with glucocorticoids elevates pNPY mRNA
2-fol d.

  Two novel neuropeptides having anti-analgesic activity were recently
isolated and sequenced by Dr. H. Y. Yangrs group.  Their structures are
Ala-Gly-Glu-Gly-Leu-Ser-Ser-Tro-Phe-Trp-Ser-Leu-Ala-Ala-Pro-Gln-A~~g-Phe-NH2
(Ai8F-:IH2) and Phe-Leu-Phe-Gln-Pro-Gln-Arg-Phe-NH2. A rat hypothalamus Agtll
cD!JA 1 i brar y was screened with 32P-oligodeoxynucleotides corresponding to
portions of these peptides and putative AJ~F-NH~ cDNA clones were obtained.

  Kearly all prokaryotic genes use the translation initiation codon AUG.
However, there are a few examples where GUG or UUG function as initiation codons
in 5. coli.  The gene for E.
   --           coli adenylate cyclase, is one of the genes that
uses t:rle :;nusual UUG initiation codon.  h'e have investigated the effect of this
xluscia 1 initiation codon on the expression of the adenylate cyclase gene by
c-nsy.yir,?
  --a" .c) the DNA sequence coding for the UUG initiation codon to ATG and GTG,
usirlg oligonucleotide-directed mutagenesis.  A comparison of the activities
associated with the three codons was made in three different environments: (1)
in the normal environment, with the adenylate cyclase gene expressed fron! its
own pfo7o:ers, (2) in a transcription fusion with the adenylate cyclase gene
under the transcriptional control of the phage lambda promoter, and (3) in a
gene fusion with the adenylate cyclase gene fused to the E. coli galactokinase
gene to generate a fusion protein with galactokinase activity.  In each of the
three environments, it was observed that the UUG initiation codon had the lowest
efficiency of translation initiation and the AUG initiation codon had the
highest efficiency, while the GUG initiation codon was intermediate.  These
results nay provide a partial explanation for the finding that the cellular
concentration of adenylate cyclase is very low.            A_

    In E. coli CAMP plays a crucial role in regulating the expression of
indclci'ble genes.
catabolite-    The levels of this nucleotide are controlled primarily by a
     dependent modulation of adenylate cyclase activity.  Insight into the
mec:han i s:? of regulation of the activity of this enzyme has come primarily from
studies of permeable cells.  Current information suggests that the
phosphoenolpyruvate:glucose phosphotransferase system (PTS) is intimately
involved in the regulation.  Additionally,
roles in                   potassium and phosphate ions play key
     rrodulating adenylate cyclase activity.  A model for interaction of
adenylate cyclase with PTS proteins and potassium phosphate to form a regulatory
COT,plPX :ias proposed previously by us.  The purpose of the present study was to
test the ;roposed model for.adenylate cyclase regulation using a reconstitution
approach.  Ke found that all of the unique features of adenylate cyclase
&arac';e;n; - ->tic of the re.gulatory complex observed in permeable cells were
reconstit:;te d in cell-free extracts.
t&t a:`,&-".'              The results strongly support the proposal
    -late cyclase activity is regulated by PTS proteins.

-4-