‘1
Q

A ......
v:

. o
'..r...

. .2.- . {$5

I...) .r. . 3. .1

u:

.7. ..
1....v

 

 

 

 

 

7.1880103

n u m HM“! we: will 11111171 ”

3 1293 00585 3191 ”ERARY
Michigan State
University

.1

 

 

 

 

 

 

 

11

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled
Molecular Characterization of Bovine and Human
Retinal Rod cGMP Phosphodiesterase and Chromosomal

Localization of the Human Gene

presented by

Steven Jay Pittler

has been accepted towards fulfillment
of the requirements for

 

 

Ph.D. degree in Biochemistry

TMDL 61w

Major professor

Date S/m/gy

MS U is an Affirmative Action/Eq ual Opportunity Institution 0- 12771

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return‘on or betore date due.

DATE DUE DATE DUE DATE DUE

JANWZIFZ

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

—fl_

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Inditutlon

 

MOLECULAR CHARACTERIZATION OF BOVINE AND HUMAN
RETINAL ROD cGMP PHOSPHODIESTERASE AND CHROMOSOMAL
LOCALIZATION OF THE HUMAN GENE

BY

Steven Jay Pittler

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Department of Biochemistry

1989

mill’l)

W

MOLECULAR CHARACTERIZATION OF BOVINE AND HUMAN
RETINAL ROD cGMP PHOSPHODIESTERASE AND CHROMOSOMAL
LOCALIZATION OF THE HUMAN GENE

By

Steven Jay Pittler

Retinal rod cGMP phosphodiesterase (cGMP PDE) is a
heterotrimeric enzyme that functions as a signal amplifier in photo-
transduction. This work was initiated to study the molecular biology of
the genes encoding the cGMP PDE subunits. I

Characterization of overlapping cDNA clones encoding the or-
subunit of cGMP PDE revealed an Open reading frame that encodes a
protein of 859 amino acids. Four differing cDNA clones were also identi-
fied that are highly homologous to the a-subunit sequence. All of the
isolated cDNAs contain an identical segment in their predicted protein
coding regions of ~1.6 kb. The sequence of one of these cDNAs differs
from the a-subunit sequence only at the 5’ end encoding a predicted
polypeptide of 500 residues. 'I\vo other cDNAs differ in sequence from
the inc-subunit only at their C-termini, and the fourth cDNA shows
sequence divergence in the 3’ untranslated region. Predominant retinal
RNAs of 4.6 and 4.0 kb were identified with cDNA probes representative

of the common segment, or with probes specific to the a-subunit
sequence. Using probes specific to the a-subunit-like cDNAs, a number
of lower abundance RNAs were observed. Strong evidence from DNA ‘
blotting experiments suggests that a single bovine gene encodes all of the
identified transcripts. The size of this gene is estimated to be at least
160 kb.

Several cDNA clones, including one full length clone were isolated
encoding the human cGMP PDE a-subunit. Overall, the predicted
primary structures of the human and bovine a-subunits showed 94%
identity. Transcripts of ~5.9 and 5.4 kb are identified in human retinal
poly (A)+ RNA. Probes specific for the two bovine cDNAs that differ at
their C-termini do detect human retinal RNAs of similar size and
abundance to that found in bovine retinal RNA. The human a-subunit
gene has been mapped to the long arm of chromosome 5 using somatic
cell hybrid DNA panels.

Similarities were found between the predicted (Jr-subunit amino
acid sequence and that of nonretinal cyclic nucleotide phospho-
diesterases (CN PDEs), indicating that the a-subunit gene is a member of
a conserved gene family of CN PDEs spanning diverse species from yeast
to man. The complexity of the bovine cGMP PDE a-subunit gene is
comparable to other PDE genes of higher eukaryotes and may be a
general feature of the members of this gene family.

I dedicate this work to my family, especially my parents, Floris and Carl Pittler,
who believed in me and supported me even when I gave them every reason not to.
And to my sister, Jean Webb, who I am very proud of for believing in herself. I
also am compelled to extol the contributions of my grandparents, the memories of
grandpas Mickey Pittler and Leonard Finkelstein, and to Mildred and Joseph
Loundy, and Ida Pittler, whom I hope will live forever. I apologize to the rest of
my family that there is not sufficient space to mention all of you for the
cornerstone of Judaism is the camaraderie of the family which has served to

establish the ethics and values that I live by.

AggNoquDGEMENTs

I would like to thank some of the many people that contributed to my understanding of
science and/or my emotional well being. Dr. J. E. Wilson for giving me a chance in his
lab as an undergraduate and instilling in me a deep appreciation for scientific
endeavors. Also for being an excellent role model both as a scientist and as a human
being. Dr. D. G. McConnell for expanding my interests in science to that of the visual
system, and for allowing me to graduate even though I could not beat him on the tennis
court. Dr. R. L. Davis, my thesis advisor, for allowing me to pursue my own interests,
and giving me the skills and knowledge to carry out this pursuit. Dr. T. B. Friedman for
many stimulating and educational conversations that have broadened my scientific
scope. Drs. W. L. Smith and J. B. Dodgson for being reasonable committee members.
Dr. P. Polakis for consistently reminding me of my weak points in tennis and my strong
points in racquetball, for teaching me the appropriate language to communicate with an
HPLC apparatus, and for being a good friend, and a great smelt cleaner. Several other
good friends, in addition to Paul, require acknowledgement: Rich Bean who I "grew
up" with in the dorms, and learned how to party from. Stan Avery, another long term
friend who continued the lessons. Julie Doll who taught me how to live and the science
of professional relaxation. Dr. David Dewitt for a crash course in HPLC and other
aspects of protein manipulations, and for being the bartender in the next lab. Dr. Bob
Schiltz for being the kind of person that is always fun to be around, it’s contagious.
Annette Thelen for inspiring motivation in me. Ted (Theo) Hupp for the Muskegon
bird watching expeditions, and other assorted fun times in the sun. And to Theresa
Krause and Theresa Vollmer for other good times with emphasis on attitude adjustment.
I would also like to thank the members of the Davis lab, both past and present who I
also consider my friends: Sylvia Denome who taught me a lot of techniques in a short
period of time. Chun-Nan Chen for keeping me in line when I had crazy ideas, and for
taking up where Sylvia left off. Also for many enjoyable culinary delights, and for not
serving mai tais or monkey brain. Dr. Marion Healy for a glimpse of Australia through
her eyes, and for her contributions to my writing abilities. Dr. Linda Smith for further
improvements in my writing abilities, and for organizing lab get togethers that always
proved enjoyable. I would also like to thank Mary Eberwine for many things,
including technical assistance, rides to the airport, and her friendship. James Myres for
some technical expertise, but mostly for being my first good friend in Houston.
Yuhong Qiu for her inquisitiveness and enthusiasm. Jenny Tigges for enjoyable and
stimulating conversations. Saira Ahmed for her histological expertise. Liz Richter-
Mann and Allan Nighorn for being the new kids on the block. And also at Baylor
College of Medicine I would like to thank Dr. R. J. Schwartz for an inspiring letter of
recommendation and his support of my abilities. Dr. B. A. French for his friendship
and insight into scientific endeavors. Bahram Varjavand for maintaining my image of
the M.D., Ph.D. candidate, and for many fun times. Maria Capovilla for stimulating
questions and for sharing in some of the fun times. Dr. G. I. Liou for allowing me to
use his computer, providing retinal RNA, and many stimulating conversations. Dr. W.
B. Baehr for his seminal work on the PDE, giving me the space to write this thesis, and
for being a good person to work with. Lastly, I would like to thank my good friend, Al
Smith, for more things than I have space to mention, but mostly for being the best
friend that anyone could hope to have.

iii

was or CONTENTS

List of Tables .
List of Figures
Abbreviations.
Introduction .
Literature Review.
Retinal Morphology .
A Model for Phototransduction
Rhodopsin .
Transducin .
cGMP PDE.

Characterization of Other Retinal PDEs

Other Rod Cell Proteins of Importance in Phototransduction .

lPost Translational Modifications .

Classification of Eukaryotic Cyclic Nucleotide PDEs.

Molecular Genetics of Eukaryotic Cyclic Nucleotide PDEs

Clinical Significance of a Defective cGMP PDE.
Materials and Methods

Materials

Denaturing Polyacrylamide Gels .

Modification of cGMP PDE .

Tryptic Digestion of Modified Protein .

iv

é:

##UJU)

10
11
13
14
16
17
18
22

24
25
25

Results

High Performance Liquid Chromatography of Trypsin Cleaved
cGMP PDE.

Determination of Amino Acid Sequence of Peptide Fragments.
Molecular Cloning Protocols .

RNA Isolation .

Isolation of mRN A .

RNA Blot Analysis .

DNA Isolation .

Isolation of Nucleic Acids From Gels .

Synthesis, Purification, and Radiolabelling of Oligonucleotide
Probes .

Labelling of Nucleic Acids

Screening Human and Bovine cDNA Libraries .

Screening and Analysis of a Bovine Genomic DNA Library .
Determination of Gene Copy Number .

Nucleic Acid Sequence Analysis .

Chromosomal Localization

Computer Analysis .

Section A Bovine cGMP PDE int-subunit

Trypsinolysis and Peptide Sequence Analysis of
Bovine cGMP PDE .

Synthesis of Oligonucleotide Probes

Screening the Bovine Retinal cDN A Library
Characterization of Bovine cDN A Clones .
Blot Analysis of Bovine Retinal RNA .

Sequence Analysis of Bovine Retinal RNA.

25
26
26
26
27
27
28
29

29
3O
30
31
31
32
32

34
34

34
37
38
43
52
57

Isolation of a Bovine Genomic Clone . . . . . . . . 57

Analysis of the Bovine cGMP PDE a-Subunit Gene. . . . . 60
Section B Human cGMP PDE or-subunit . . . . . . . . . 65
Isolation of Human cGMP PDE a—subunit cDNAs . . . . . 65
Characterization of Human a—subunit cDNAs . . . . . . 65
Blot Analysis of Human Retinal RNA . . . . . . . . 68
Chromosomal Localization of the Gene. . . . . . . . 68
Section C Similarities Between Bovine and Human a-subunits . . . . 77
Comparison of Human and Bovine (Jr-subunits . . . . . . 77
Homologies to Nonretinal Cyclic Nucleotide PDEs . . . . . 80
Discussion.. . . . . . . . . . . . . . . . 83
Sequence of Tryptic Peptides . . . . . . . . . . 83
Design of Oligonucleotide Probes . . . . . . . . . 84

Oligonucleotide Screening and Sequence Analysis of

Bovine Retinal cDN As. . . . . . . . . . . . 86
Sequence Analysis of cDNA Clones . . . . . . . . 86
Identification of cGMP PDE (it-Subunit cDN As . . . . . . 87
Identification of a—Subunit Variants . . . . . . . . 88
Possible Functions of Predicted Isozymes . . . . . . . 92
Analysis of the Bovine (it-Subunit Gene . . . . . . . 93
Analysis of the Human cGMP PDE a—Subunit . . . . . . 93
Comparison of CN PDEs. . . . . . . . . . . 95
Appendix.................109

LIST OF TABLES

Table 1: Proteins of the cGMP Cascade of Vision . . . . . . . 8
Table 2: Classification of Mammalian Cyclic Nucleotide PDEs . . . . 20,21
Table 3: Primers Used for Sequence and RNA Blot Analysis . . . . . 48,49

Table 4: Segregation of cGMP PDE a-Subunit Gene in Human-Rodent
Somatic Cell Hybrids . . . . . . . . . . . 76
Table A1: Nucleotide Differences Observed in cDN A Clones . . . . . 118

vii

Figure 1:
Figure 2:
Figure 3:
Figure 4:
Figure 5:
Figure 6:

Figure 7:
Figure 8:
Figure 9:

Figure 10:
Figure 11:
Figure 12:
Figure 13:

Figure 14:
Figure 15:

Figure 16:

gr OF FIGURE—S

The cGMP Cascade of Vision in Retinal Rods .

Analysis of Tryptic Peptides of Bovine cGMP PDE .

Design of Oligonucleotide Probes .

Screening the Bovine Retinal cDNA Library
Schematic Map of Bovine cDNA Clones
Nucleotide and Deduced Amino Acid Sequences of
Bovine cDNA Clones

Location of cDN A and Oligonucleotide Probes .
Analysis of Bovine Retinal RNA .

Direct Chain Termination Sequencing of Bovine
Retinal RNA

Copy Number of the Bovine a-Subunit Gene
Size of the Bovine a—Subunit Gene

Restriction Map of Human cDNA Clones .
Nucleotide and Deduced Amino Acid Sequence
of Human cGMP PDE (it-subunit .

Analysis of Human Retinal RNA .

Blot Analysis of Human-Rodent Somatic

Cell Hybrid DNA

Comparison of Human and Bovine (it-subunit

Primary Structures

36

42
45

51

54

56

59
62

67

70
72

74

79

Figure 17: Similarities Amongst cGMP PDE a—Subunit
and N onretinal CN PDBs.
Figure A1: Complete Sequence of cDNA clone Pct-1

Figure A2: Partial Sequence Determined for cDN A Clone Poe-2.

Figure A3: Complete Sequence of cDNA Clone POL-3 .

Figure A4: Partial Sequence Determined for cDN A Clone Pet-4 .

82
110
112
114
116

ABBREVIATIONS
Abbreviations used without definition were taken from the Proceedings of the National
Academy of Sciences, U.S.A., Information for contributors: Volume 85, pp. iii-v, 1988.
Additional abbreviations used are as follows:
RPE: retinal pigment epithelium
IPM: interphotoreceptor matrix
ROS: rod outer segment
cGMP PDE: bovine rod outer segment 3’,5’-cyclic guanosine
monophosphate phosphodiesterase (EC 3.1.4.17)
CN PDE or PDE: denotes any 3’,5’-cyclic nucleotide phosphodiesterase
that hydrolyzes CAMP or cGMP
TPCK: tosylphenylchloroketone
SDS-PAGE: sodium dodecyl sulfate polyacrylamide gel electrophoresis
MW: molecular weight
PTH: phenylthiohydantoin
DEPC: diethylpyrocarbonate
PEG: polyethylene glycol (molecular weight 8,000)
TMAC: tetramethyl ammonium chloride
DOS: disk operating system
VMS: virtual memory system

pfu: plaque forming unit

INTRD IN

Phototransduction in rod cells is initiated by the absorption of
light. Compelling data from a number of laboratories (25) have shown
that three proteins; rhodopsin, transducin, and cGMP PDE, are central
in the pathway of information flow from the light stimulus to a reduction
in intracellular cGMP concentration. Upon absorption of a photon, the
receptor, rhodopsin, activates a specific G-protein, transducin, which
then stimulates a marked increase in cGMP PDE activity. The ensuing
decrease in cGMP levels produces a hyperpolarization of the plasma
membrane, caused by the closure of cGMP gated ion channels.

In bovine, cGMP PDE is a three subunit enzyme that is disk
membrane associated, with stoichiometric composition 0t(88 kD), [3(84
kD), and 72(10 kD) (26,27). The y—subunit functions as an inhibitor of
cGMP PDE maintaining its activity at a basal level in the dark. Although
the enzyme activity purifies as a heterotrimer, it is not clear that the
enzyme is composed this way in viva. Hurwitz et aL have shown that
both the or and [3 subunits have activity that is separately inhibited by
the y-subunit (28). Unfortunately, the inability to separate and isolate
active or and B subunits in amounts amenable for reconstitution
experiments has made it impossible to investigate their potential
independent regulation.

Recently, Ovchinnikov et al. reported on the primary structure of
the ct-subunit deduced from overlapping cDNA clones (29). A portion of

l

2
this sequence is homologous to an ~27 5 amino acid segment found in
many nonretinal cyclic nucleotide PDEs (CN PDEs) in diverse species
(50). This region, which has been termed the "conserved domain", is
thought to form part of the catalytic domain of this enzyme family.

We initiated a molecular dissection of bovine cGMP PDE to study
the interactions of the a and [3 subunits and the regulation of the enzyme
in the phototransduction cascade. The analysis was extended to the
study of the human a-subunit to probe the potential clinical signifigance
of the enzyme in hereditary retinal dystrophies. In these studies we
confirm the sequence of the bovine cGMP PDE a-subunit by sequence
analysis of several cDNA clones. Additionally, during this analysis we
have surprisingly identified four strikingly similar or-subunit like
variants. These variants, as well as the a—subunit, are encoded by
multiple RNAs from a gene extending over more than 160 kilobases. A
full length human a-subunit cDNA encodes a protein molecule that is
94% identical to the bovine (it-subunit. The gene encoding the a-subunit

has been mapped to the long arm of human chromosome 5.

LITERATUQ REVIEW

W In the eye, light focused onto the retina is

absorbed and then converted to an electrical impulse: a process called
phototransduction (1). The initial transducing events are mediated by
highly specialized cells termed photoreceptors. Most mammalian retinas
contain two morphologically distinct photoreceptor cell types, cones and
rods. There are typically three types of cone cells characterized by the
wavelength of light absorbed by their receptor molecules. The cone cells
function in bright light and mediate color vision (2). Rod cells, on the
other hand, function at lower levels of illumination (3). Rod cells out-
number cone cells in higher mammals by approximately 15 to l, and
because of their abundance and ease of isolation, phototransduction in
the rod cells is better understood.

Rod cells are elongated neural cells that can be subdivided into
four distinct regions (Figure l). Phototransduction occurs in the outer
segment which contains about 2000 membranous discs that are formed
when an invagination of the plasma membrane is pinched off (4). Old
discs are shed at the distal end of the rod cell as new discs are formed.
The shed discs are destroyed by the phagocytic activity of the RPE, a
layer of cells contiguous with the distal part of ROS that extends between

rod and cone cells. The space between rods and cones that is not filled

by the RPE is made up of a filamentous mesh of glycosaminoglycans,
proteins, and glycoproteins termed the IPM (5).

A Model for Phototransducfign, The biochemical basis of signal
transduction in the rod cell has been subjected to intensive study (6-9).
Three proteins, rhodopsin, transducin, and cGMP PDE make up greater
than 80% of the total protein found in ROS. It has been proposed that
the interaction of these three proteins, in large part, mediates the trans-
ducing process (10, summarized in Figure l f and Table I). In its
activated form, rhodopsin interacts with transducin in an undefined
manner to cause the exchange of GDP for GTP on the a-subunit of the
transducin molecule (18). The GDP-GTP exchange brought about by the
interaction of transducin with rhodopsin causes the transducin a-
subunit to dissociate from a complex of the B and 7 subunits. The
dissociated (rt-subunit is then free to interact with and activate the cGMP
PDE. In the dark, cGMP directly interacts with Na+ ion conductance
channels to keep them open (76). Impinging light on the retina leads to
the hydrolysis of cytoplasmic cGMP to 5’ GMP catalyzed by the cGMP
PDE. This is accompanied by the subsequent closure of many Na+
channels which produces a transient hyperpolarization in the rod cell.
Thus, a light stimulus has been converted into an electrical signal that
will be transmitted through other retinal layers finally arriving at the
visual centers of the brain.

Rhodopsin. The seminal work by Wald uncovered the important
role of rhodopsin in the signal transduction process (85), and set the
direction for future experiments. It is now known that rhodopsin is an
integral transmembrane protein that has been localized within the disc

Figure l: The cGMP Cascade of Vision in Retinfl @ds, Shown at the
bottom of the figure is a schematic of a rod cell. Rod cells are

differentiated into inner and outer segments that are connected by a
cilium. The inner segment contains the nucleus, most of the cell's
biosynthetic machinery, and are continuous with the synaptic terminals.
The membranous discs in the outer segment that form by an
invagination of the plasma membrane are pinched ofi to become
autonomous. It is within these discs that the light receptor, rhodopsin,
resides. The primary biochemical events that initiate phototransduction
are shown at the top. A model to explain these events has been proposed
by Stryer (10). Table 1 identifies the symbols used in this model.

1) In the dark, nearly all of the transducin is in the T-GDP form, which
does not activate the cGMP PDE.

2) The chromophore of R, 1 1-cis retinal absorbs a photon which results
in the activation of R to R‘.

3) R’ encounters T-GDP and forms an R‘-T-GDP complex.

4) GTP exchanges for GDP in the R’-T-GDP complex. In essence, the
role of R‘ is to activate transducin.

5) The exchange of GTP for GDP markedly diminishes the affinity of R"
for transducin. In addition, the affinity of To, for TB? is very much
decreased. These changes in binding affinities are crucial for the
efficient operation of the cycle. The released R’ is free to interact with
another T-GDP, enabling R" to act catalytically.

6) Ta-GTP activates the cGMP-PDE by overcoming an inhibitory
constraint (1) imposed by its y—subunit. Activated cGMP PDE hydrolyzes
cGMP at a very rapid rate. This leads to the closure of many Na+ ion
channels and results in a transient hyperpolarization of the plasma
membrane.

7) The GTPase activity inherent in the a-subunit of tranducin converts
Ta-GTP to Ta-GDP in times of seconds to a minute; this in turn results in
the deactivation of the cGMP PDE. Ta-GDP then rejoins TB?

One activated rhodopsin molecule can activate as many as 500
transducins, which in turn activate cGMP PDE molecules that can
hydrolyze as many as 2000 cGMP molecules per second, resulting in an
amplification cascade that yields a 106 gain from a single activated
rhodopsin (l).

a
T @hv ATP ADP
Ir
PDE; Q) n‘ V R'-P
T'GDP \
® / ®
cGMPPDEL-Ta-GDP ' R'.T-GDP

P: GTP
D 3' a
"10 cap
,6... , H20 PPDE'-T.-GTP R‘T- GTP
] ©\

cGMP Cascade of Vision

 
    
 
 

  

 

 

 

  
  
     

(=5
e:
an
El Cytoplasmic
E space
Outer a.
segment :2— Plasma
“5-3 membrane
:9
a
- Ciiium
Mitochondria
inner
segment
Nucleus
. ._ }_ synaptic
, terminal
Figure 1

Table 1: Proteins of the cGMP Cascade of Vision. Some of the major
constituent proteins of the ROS are shown. These proteins comprise
greater than 80% of the total protein of the ROS. A model proposed by
Stryer (10) that integrates the biochemical properties of these proteins is
shown in Figure 1. The sizes of the proteins are based on that observed
from SDS-PAGE analysis.

.coH .adueouue “a .oeeddx daemauonu Ava .aeeuaunedvonmeanm “cod .ddoavedeuu “coca .dfinmov
tank on a» vaueafluee one euddauee Hausa van uoeudu dd addendum ondau ma nuances Hedda any I

 

.xuw>wuon and eunasoau
on sea exec ou sauoouae enan case an: .eaoneuemuu oum>au
ion on huwommno uuw uxaoan use «m voumahuoamnonm ou madam

. m we
definabwuomav any a“ Ue>Ho>dH .dwunauun pawn on uw unwmnedo
.uankuau aawcoauau can define» admwuaaa um «m nonmahuanmuonm

.eumuu xuep an»
3 mam 23 no £53?» 23 an 63335 3 as no 3.333
aduhdmumo can .mawiea an aeuebwuon can: mzuo nauhaouvmm

.aumuu xumu

any ou uneven ou mam ou mam canon nonmaouvaa ea .amcnnoxe
.3613. How confined.» aw £9 .Buom mawida Since 5.... can
350m mowimefi a>wuonaw an caazuan uuuabdooueudw daoavudmua

.aenuouneauonmuonm an»
no doaum>wuon as» aw aunwaaaueucw aewmwamad emu a“ mauiea

.eodmaoxe mowimaw undead
tmumo an dwoaUudnuu maum>wuom .«mv cwnmovonu vauwoxaouanm

Anomauaaum.

 

 

Odom

mv dauneuud
mm owned: daemovonm

on s
cm a Anna exec.
mm 5 annuauneavonmnonm

m >

mm a
an d ARV aflosuncnua
ov Amy Gammovond
Aux. and: ecwauoum

 

 

 

.cowmm> mo evacumo mzwo anu mo ncaauoum ”H cane?

 

9

and plasma membranes of ROS. It consists of a chromophore, l l-cis
retinal, as the prosthetic group, that is covalently attached (lysine 296) to
the apoprotein, opsin (1 1). The molecule contains seven hydrophobic
regions proposed to loop through the disc membrane with the C-
terminus on the cytoplasmic face and the N -terminus within the lumen
of the disc (12). Several potential sites of phosphorylation are located
near the C—terminus that are thought to be involved in the regulation of
rhodopsin activity (13). The initial event of phototransduction is pro-
duced through the chromophore, 1 l-cis retinal. Light absorbed by the
chromophore causes it to be isomerized to all trans retinal, a process
called bleaching (16). Spectroscopic studies have shown that at one of
the transition states (metarhodopsin II) in the isomerization process, the
rhodopsin is converted to an activated form (17). The basis for the
structural changes causing the activation are not understood.

The genes for both human and bovine rhodopsin have been cloned
and their nucleotide sequences determined (14,15). There is marked
conservation in both gene structure and nucleotide sequence. The
coding region in both genes is interrupted by four introns. Both proteins
have the same number of amino acid residues and exhibit 94% overall
homology. Most importantly, three loops thought to be on the cyto-
plasmic face of the disc are perfectly conserved. This is consistent with
the suggestion that the three loops are involved in the binding of
transducin, arrestin, and rhodopsin kinase: however, the stoichiometry
has not been established. Additionally, the 1 l-cis retinal attachment site
and all but one of the phosphorylation sites are identical in the two
proteins. The human rhodopsin gene has been mapped to chromosome

3 (101).

10

Trgsducin. Transducin was so named because of its role as an
intermediate in the transduction process (18). It is composed of three
subunits, (X. (39 kD), B (36 kD), and y (8 kD), and is found peripherally on
the disc membrane, following illumination. Transducin is a member of
the well-characterized family of G-proteins which have been shown, or
are implicated, to be signal mediators in a variety of transduction pro-
cesses (19). Characteristic of all G-proteins is at least one guanyl
nucleotide binding site and intrinsic GTPase activity. These functions
reside on the (it-subunit of transducin (10).

Recently, a transducin or-subunit cGMP PDE y-subunit complex
was identified (98), consistent with the hypothesis that activation of
cGMP PDE occurs by removal of the inhibitory constraint imposed by the
y—subunit. A three dimensional model has been proposed for the a-
subunit based on crystallography data from other related proteins (93).
The region of interaction with the cGMP PDE y-subunit is suggested to be
delimited to amino acids 49-175. This region is predicted to be com-
posed of eight antiparallel B strands.

Corresponding cDNA clones have been isolated for the a, B and 7
subunits of transducin (78-80). Of the several groups that reported the
sequence of the a-subunit, one group reported a sequence (81) that dif-
fered slightly from the others. Subsequently, it was shown that rod and
cone specific transducin molecules are the basis for this difference (82).
The gene encoding the (it-subunit has been mapped to human chromo-
some 3 (94) and mouse chromosome 9 (95). The two genes encoding the
B-subunits that are thought to be common to all G-proteins (99) have

also been mapped. G31 is on human chromosome 1 (94) and mouse 19

l 1

(94). G32 is on human chromosome 7 (94), its mouse location has not
been determined.

cGMP PDE. The cGMP PDE also consists of three subunits: at (88
kD), B (84 kD), and 72 (11 kD) (20.26.27). In the dark, the enzyme
exhibits a very low catalytic activity (50 moles cGMP
hydrolyzed\second\mol), which is increased markedly (3700 moles
cGMP hydrolyzed\second\mol) by light. The observation that limited
tryptic digestion also greatly enhanced the activity suggested that
activation might occur by removal of an . inhibitory constraint.
Subsequent studies showed that this constraint is levied by the 7-
subunit of the cGMP PDE (2 1). Three observations led to this conclusion
(2 1). First, it was shown that limited trypsinolysis rapidly degraded the
y-subunit, while a and B and their association remained unaltered.
Second, when purified y—subunit was added back to trypsin activated
PDE, 99% of the activity was inhibited. The dissociation constant for the
binding of y to or-B is 0.13 nM. Third, during purification of the enzyme,
the inhibitory activity copurified with the catalytic activity.

Complementary DNA clones representing bovine and mouse cGMP
PDE y-subunits have been isolated (22,92). Thirteen of the 87 amino
acids of the y—subunit are tryptic cleavage sites. Interestingly, 10 of these
residues are concentrated in a stretch of 22 amino acids that contains no
acidic residues. This is consistent with the observation that trypsinolysis
increases the activity of the cGMP PDE by removing the inhibitory con-
straint imposed by its y—subunit. The y—subunit gene has been localized
to mouse chromosome 1 1 (97).

Purified cGMP PDE has a measured km for cGMP of ~50-70 M
and 2200 M for CAMP. Recently, the stoichiometry was determined by

12

two groups, using markedly different approaches, to be atBy2 (27,96).
Although the enzyme purifies as a heterotrimer of the three subunits, it
has not been unambiguously demonstrated that the enzyme is composed
this way in viva. Based on the results of biochemical and immunological
analysis, Beavo has suggested that minor isozymic forms of cGMP PDE
exist in the rod outer segment (28, 120). In the report of this analysis, it
was shown that the at and B subunits can be separated by HPLC. It was
then discovered that the subunits retain activity independent of their
presumed association and that the activity could be inhibited by addition
of the y—subunit. Moreover, the inhibition could be removed by addition
of activated transducin, leaving open the possibility that an cry and a By
subunit association may exist in viva representing two distinct enzymes.

Ovchinnikov et al. have also reported on the analysis of cDNA
clones encoding the a-subunit and a portion of the B-subunit (86). A
composite sequence spliced together from overlapping cDNA clones
predicts an at-subunit with a MW of ~99 kD, significantly higher than the
88 kD value predicted from SDS-PAGE analysis (20). The sequence of
two peptides was identified from an at-subunit tryptic hydrolysate, that
are not found in a B-subunit hydrolysate in support of the cDNA
sequence. Additionally, the first 12 amino acids at the N -terminus of the
a-subunit were determined by mass spectrometry. However, only one
cDNA clone was isolated covering a portion of the C-terminus of which no
peptide sequence was presented. Preliminary findings of C-terminal
heterogeneity in retinal cGMP PDE (87) left doubt as to the authenticity
of the reported sequence. It was possible that the overlapping cDNA
clone covering the C-terminus represented a closely related PDE and not
the a-subunit. A portion of the B-subunit sequence was also presented,

13

as deduced from cDNA clones. No peptide sequence was provided in
support of the sequence presented leaving open the possibility that a
PDE polypeptide other than the B-subunit had been cloned. The identity
of their postulated B-subunit still remains unclear. The question of the
number of genes encoding the a and B subunits also remains
unanswered. Even if the reported B-subunit sequence is correct, the
differences in sequence between the subunits can be accounted for either
by alternative splicing of transcripts from a single gene, or two separate
genes.

Characterization of Other Retinal PDEs. While the number of
cGMP specific PDE isozyrnes in the retina has not been firmly estab-
lished, it is clear that multiple forms do exist. A high affinity form
identified only by histological methods has been found in the inner plex-
iforrn layer (100). A cone outer segment PDE has been characterized that
is similar to rod cGMP PDE, but appears to have only one large subunit
of 94 kD and three small subunits of 11,13, and 15 kD (31) . The 11 kD
subunit has been shown to be very similar if not identical to the rod
cGMP PDE y-subunit. Other properties of the cone enzyme further
distinguish it from the rod enzyme. The cone enzyme binds 10 fold more
cGMP/mole than the rod enzyme. This binding occurs at a noncatalytic
site with very high affinity (kD= 1 lnM). The dissociation of cGMP from
this site is slow with a halflife of 10-20 minutes at 370C. Rod transducin
will activate the cone enzyme with half maximal activation requiring 50
fold less (15 nM) transducin than that required for the same activation of
rod cGMP PDE.

A PDE in the IPM, termed 1PM PDE, has also been characterized
which has subunit sizes of 47 and 45 kD (102). The apparent native MW

14

is 350,000 by gel filtration suggesting a multi subunit assembly. The
kinetic properties for this PDE are essentially the same as that observed
for the rod enzyme. The enzyme could be immunoprecipitated with a
monoclonal antibody that blocks activity of the cGMP PDE (28). The
possibility that this PDE is simply a proteolytic product of the rod PDE is
ruled out by the procedure used to isolate the enzyme, which was shown
not to release the cGMP PDE from the disk membrane. The function of
this PDE is unknown, however it may act as a scavenger to map up
excess cGMP, which has been shown to be toxic in high dosage to the rod
cell (84).

Other Rod Cell Proteins of Importance in Photogansdugtion,
Several other proteins have been shown or are thought to participate in
the regulation of phototransduction. Rhodopsin kinase has been shown
to phosphorylate rhodopsin an accessible sites near the C-terminus,
thereby decreasing its activity (23). The enzyme has recently been puri-
fied to near homogeneity (88). It exists as a single polypeptide chain of
MW ~70 kD. Similarities were observed in comparison to the B-
adrenergic receptor kinase (89), suggesting that the two kinases may be
members of a kinase family that also includes the muscarinic acetyl-
choline receptor kinase.

The decrease in activity of rhodopsin requires another protein
called arrestin (24) or S—antigen which will bind to phosphorylated active
rhodopsin but not to unphosphorylated active rhodopsin, thus inhibiting
the binding of transducin (25). S-antigen is a 48 kD protein that appears
to be cytosolic in the absence of light, and associated with the disk
membrane following illumination (90). A cDNA clone encoding arrestin
has been isolated (9 1). Comparison of deduced amino acid sequence for

15

arrestin and the (it-subunit of transducin revealed striking similarities
(83), consistent with the idea that arrestin competes with transducin for
binding sites on activated rhodopsin, thereby inhibiting rhodopsin’s
ability to bind transducin and activate it. It has also been suggested,
based on the results of crosslinking analysis that arrestin binds directly
to cGMP PDE to inhibit its activity (1 03). Perhaps arrestin is responsible
for the reassociation of the cGMP PDE y-subunits and atB, facillitating the
return to the dark state.

Guanylate cyclase catalyzes the conversion of GTP to cGMP, and
therefore is an essential component of the recovery process following a
light stimulus. The instability of the enzyme has made its purification
difiicult, and thus its characterization is based on crude or partially
purified preparations (104). It has been shown, based on microdissec-
tion studies of histologically defined retinal regions that much of the
enzyme activity present in the retina is concentrated at the axoneme of
the rod cell. Only in the plexiform layers has a significant amount of
activity been found outside of the photoreceptor cell (105). More
recently, Koch and Stryer (106) have shown that the enzyme activity is
increased markedly (5-20 fold) in response to very small changes
(200nM-50nM) in the intracellular calcium concentration. It was also
suggested that a new type of calcium binding protein mediates this
effect.

Additional proteins important for guanine nucleotide metabolism
have also been identified and partially characterized in ROS. A 5’
nucleotidase which converts 5'-GMP to guanosine exists in ROS as a
membrane bound enzyme of ~75 kD (107). The km for 5'-AMP was
measured at 1.3 M and for 5’-GMP a km value of 2.3 M was observed.

16

A guanylate kinase which produces GDP from 5’-GMP and a
nucleoside diphosphate kinase which can convert GDP to GTP has been
characterized in frozen retina and from isolated rods (108, 109). The
guanylate kinase has been purified from ROS and it was determined to
have a km of 13 M for GMP and 430 M for AMP. It was suggested that
the enzyme exists in one copy per 800 rhodopsins. These results estab-
lish that the enzymes capable of recycling 5’-GMP to cGMP, namely,
guanylate kinase, nucleoside diphosphate kinase, and guanylate cyclase
are present in ROS, but it is unclear whether the anabolic or catabolic
pathways for the regeneration of 5’-GMP predominate.

The major phosphoprotein identified in ROS has been termed 33k
(1 10). This protein is thought to be specific to visual cells (11 1), and
phosphorylated in a cyclic nucleotide dependent manner (1 12). Inter-
estingly, phosphorylation levels of the protein are highest in dark
adapted retinas and dephosphorylated upon illumination (1 13), in direct
contrast to rhodopsin phosphorylation. The recent finding that 33k
exists as a complex with the B and 7 subunits of transducin (52) suggests
an important role in the regulation of phototransduction, which is yet to
be determined.

Past Translational Modifications. The post translational addition of
palmitate is thought to occur only through a covalent thioester linkage to
cysteine residues. Thioester linked lipidation has been shown to occur
on two cysteines in bovine rhodopsin, but its importance has not been
established (1 14, 1 15). However, palmitylation at a cysteine in two N-ras
and a closely related yeast YPTl protein has been shown to be essential
for normal enzyme function (1 16). Palmitylation has also been observed
in a number of other proteins in which methylation at the C-terminus

17

was also identified, including the ras and ras related proteins (1 17).
Intriguingly, potential palmitylation sites are located near the C-terminus
of the predicted cGMP PDE at-subunit sequences in both bovine and
human (1 18), and methylation has previously been reported to occur on
the bovine a-subunit (159). Indeed, it has been suggested that
methylation of the a—subunit, and potential palmitylation are coupled
and may serve as a membrane anchor by increasing the hydrophobicity
at the C-terminus (1 17).

Classification of Eukaryatic Cyclic Nucleotide PDEs. Character-
ization of CN PDE was reported as early as 1961 by Drummond and
Perrot-Yee (1 19). At that time it was thought that only one form of PDE
existed that hydrolyzed cyclic nucleotides. Currently it is known that
multiple PDEs exist in most, if not all, cell types that are regulated by
very differing mechanisms. It is expected that many more isozymes are
yet to be identified (120).

A pioneer in the study of mammalian PDEs, J. Beavo, has recently
undertaken the effort to classify the numerous isozymes of PDEs that
have been identified (120). This task has proven formidable due to the
large number of recent studies on PDEs, and the problems associated
with studying the biochemistry of the enzymes. Even though many
forms of PDE clearly exist, separation and characterization have proven
difficult because many of the isozymes (even within a single cell type) are
very closely related. Additionally, the problem is compounded by the
relatively low levels of PDE that are present for most isoforms, and par-
tial proteolysis leading to false predictions of heterogeneity.

Nonetheless, certain properties of the various PDEs allow them to
be distinguished and thus separated into separate classes. Shown in

18

Table 2 are the six classes of mammalian PDEs adapted from the review
by Beavo (120). The photoreceptor PDEs exhibit a marked preference for
cGMP PDE and therefore are grouped into the cGMP specific class. The
"low km" PDEs in general are more selective for cAMP as substrate. Ro-
20-1724 is a pharmacological agent that has been shown to selectively
inhibit PDEs in this class. The cCMP and cUMP PDEs are thought to be
very minor species and have not been well characterized.

Several non-mammalian PDEs have also been characterized, most
notably a yeast and a Drosophila cAMP PDE, both of which are most
closely related to the low km class of mammalian PDEs (121,122). Com-
parison of the primary structures of these enzymes deduced from cDNA
clones to mammalian PDEs in classes 1-4 for which sequences were
available has uncovered a significant similarity (50, see also Figure 17).
This establishes the concept that at least some, and perhaps all, mem-
bers of the four classes of mammalian PDEs comprise a conserved
enzyme family that has endured several hundred million years of evo-
lution.

Molecular Genetics of Eukaryotic chlic Nucleotide PDEs, The
best studied PDE gene to date is the Drosophila dunce gene, which
encodes a low km cAMP PDE. The locus resides at chromomere 3D4 on
the X-chromosome (132). While the complete gene structure is yet to be
determined, it is clear that the locus is unusually complex. The gene
spans more than 100 kb of DNA and encodes at least six related RNAs
ranging in size from 9.5 to 4.5 kb (132,133,134). Surprisingly, all of the
RNAs appear to encode an identical protein. The upstream untranslated
region of the gene is split by multiple introns that are spliced
differentially to account for some of the RNA complexity (133, 134). The

19

Table 2: Classification of Mammahan Cyclic Nucleotide PDfi, Cyclic
nucleotide PDEs have been grouped according to enzymylogical,

biochemical, and pharmacological properties adapted from Beavo (120).
The tissue source is noted with its appropriate reference in parentheses.
The term "low km" has been designated arbitrarily to signify any PDE
with a km less than 1 nM. The "low km" forms present in classes 3 and
4 represent distinct isozymes. Subclasses of brain Ca++-calmodulin
activated PDEs have been identified (39,41).

20

 

.uaahuoew heme dunu
race on unuaoau cue uenueaansu
anon .eumuunnau an 320 Mom

 

 

 

huwbwuoeaee Uaou can u nuwnwa Avvv abduuudan thHIcN 0% .n
L8 330 338:8 v2.78 oz 83 «.3335 860 .n :52 to? .v
0:9 .Ava abwuwunee defiance
an on venueueu dean code and
uendonnu vaudnwAdd mzwo ens
.330 35 33 ..au .6?
can»? pass 33333 an... :3 393 83 canon 83833
an acne an unau unused» aw UH .mmv dweun «any need dwasaoaaeoi++eo .m
.onv vauuomeu dean and newuue
.8on unneed— uafiflu 3338 2.9 uouuuao anacoaafi 8m. 9.qu
unau caused Heddnau e ue>oson .mmV enema» enamwve .vmv ueaeunam
.auou Hanna one on on uueommfl .mmv cadence andeuun .va Hebwn vaunasaaum mzuo .~
.emaouunau Avon dd vauoemxe
one ndfianoen .mnouunda «an» Adm. addendum «any need .Q
canvas ca«a«ueou« neuemao Han .Hmv 0:00 «any van .e vaudoemm mzuo .H

tea as» on» deduce «Ham decay

 

 

eudoaaao

neahudu mo uuuemoum

Ho coupon

flam MO auIHU

 

 

 

21

 

.uuom uonuo
on» an vauauouaeunao Add: as

«was oeauooaonz

 

 

 

uoa can acne .mzeo ecu mzuo .ouav expo .mv. mace undone venue .m
.e.«v aoawuoeavaa owaoao henna
mundane»: and: «new ouana
.eeUAuoeHoaa
owaoho Hon owmwuemu you ea van .mvv edmueeudw owuwoemedoz .m
.mzflu neuaaouuha damned «was
euaeaaao naahndu no huuemoum no condom new no eeeao

 

...n.uaoov

"a candy

 

22
most unusual feature of the dunce locus is the presence of multiple
seemingly independent transcription units identified in a 79 kb intron in
the upstream region of the gene (134). At least three genes have been
identified within the locus. The significance of this novel organization
has yet to be determined.

Clinicg Signi_ficgce of a Defective cGMP PDE, A great deal of
evidence has been reported that suggests a link between retinal degener-
ation and defects in phototransduction metabolism. Hollyfield and
coworkers (84) demonstrated that elevated levels of cGMP are toxic to
photoreceptor cells in Xenopus laevis eye rudiments, and in the adult
human retina maintained in culture. The specificity of the degeneration
for rod cells is quite remarkable. Several animal retinal degeneration
mutants have been identified which exhibit very high levels of photo-
receptor cGMP (82, 127, 128, 129), prior to the onset of any
morphological disruption. In one of these mutants, the rd mouse (86),
cGMP PDE has been shown to exhibit abnormal subunit assembly (122),
and to display altered kinetic properties (82). The retinas of these mice
undergo complete photoreceptor layer degeneration within 20 days
postnatal (123), however other retinal layers remain essentially intact.

High cGMP levels were also reported in rod photoreceptors of a 17
year old male with autosomal dominant retinitis pigmentosa (ADRP). The
authors suggested a strong correlation between morphological and
biochemical characteristics in this patient’s retina, and those observed in
the rd mouse (124). indeed, the cGMP PDE gave rise to abnormal kinetic
parameters similar to the rd mouse enzyme. Moreover, the
morphological degeneration observed was shown to be mostly limited to

areas of the photoreceptor layer that are predominantly composed of rod

23
cells. ADRP is one of a group of clinically related retinal dystrophies of
which there is presently no effective treatment and the cause is unknown
(125).

The general objectives of the dissertation research presented here
were to carry out a molecular dissection of rod cGMP PDE. The
availability of the molecular probes provided from this work allows the
opportunity to further analyze the enzyme in ways that have not, up to
now, been possible. A detailed description and analysis of the work is
presented in the pages ahead.

Mm

Bovine eyes for retinal nucleic acid isolation were obtained from
Murco, Inc. of Plainwell, MI. D. G. McConnell in the Department of
Biochemistry, Michigan State University graciously provided purified
cGMP PDE prepared according to Kohnken et al. (26). Further
purification to near homogeneity was achieved by rechromatography of
the final product on Sephacryl S-200 (Pharmacia). y-[32P] ATP and at-[32P]
dCTP were obtained from New England Nuclear. Oligo dT cellulose type
III was purchased from Collaborative Research, Inc. HPLC grade
acetonitrile, trifluoroacetic acid, ammonium acetate and TMAC were
obtained from Aldrich. Proteinase K, DNA polymerase I and Klenow
fragment were purchased from Boehringer Manheim Biochemicals.
Restriction endonucleases were obtained from New England Biolabs,
International Biotechnologies, Inc., or United States Biochemicals. T4
polynucleotide kinase was purchased from Pharmacia. Sodium dodecyl
sulfate was obtained from Pierce. All other biochemicals were purchased
from Sigrnal Chemical Co. Reagents for media preparation were obtained
from Difco, except NZ-amine which was from Sheffield Products.
Bacterial strains used in this study were C600 hfl, JM101, JM103 and
JM109 (51). Cloning vectors used were pUC8, pUC 19, M13mp8,
M13mp9, M13mp18, M13mp19 (51), p38 (Stratagene) and i. EMBL-3
(5 1).

24

25

MM

D n P l c l 1 SDS-PAGE analysis was
performed in 15% acrylamide / 0.08% bis-acrylamide gels as described
(20). Electrophoresis was carried out at 100 V until tracking dye neared
the bottom of the gel. The protein bands were visualized by staining the
gels with Coomassie brilliant blue (53) or with a more sensitive silver
staining method (74).

Modificat_ion of the cGMP PDE, The enzyme was modified by S-
carboxymethylation for sequence analysis as described by Gracy (54).

Emptic Digestion of Modified Protein. Ten uL of TPCK treated
trypsin (100 ug/ ml) was added to the modified protein in an Eppendorf
tube. Digestion was allowed to proceed for 3 hours at 37°C on a roller
drum incubator. An additional ten uL of trypsin solution was added and
the incubation continued overnight. After digestion, the solution was
dried in a speed vac concentrator (Savant SVC 1 00). The protein was
redissolved in 500 BL of water. A 20 BL aliquot was removed for SDS-
PAGE analysis and the remainder was redried and stored at -2000.

High Performance Liquid Chromatography af Inpsin Qieavad
cGMP PDE. Fractionation of the trypsin cleaved protein was performed
on Waters HPLC equipment utilizing reverse phase chromatography.
Initial separation of peptides was achieved by elution from C4 or C18
reverse phase columns using a gradient of H20/ 0.07 5% TFA and
acetonitrile / 0.07 5% TFA as the solvent system. A linear gradient of 0-
70% acetonitrile/ 0.075% TFA was generated over 60 minutes and
approximately one mL fractions were collected manually at timed
intervals throughout the gradient. Most of the collected fractions
containing peptides were purified in two additional chromatography

26

steps. First, the peptides were eluted a second time from a reverse phase
column using the same solvent system as previously described, but with
a narrowed gradient range. Second, each sample was dried,
resuspended and rechromatographed using 25 mM ammonium acetate.
pH 6.0 (buffer A), and 50 mM ammonium acetate, pH 6.0/60%
acetonitrile (buffer B). Elution was achieved with a linear gradient of 0-
70% buffer B generated over 60 minutes. Samples that appeared to be
homogeneous after the first rechromatographic step were not further
purified.

Datarmination of Amino Acid Saguenga of Peptida Fragmants,
Peptide fragments that appeared to be homogeneous after the first
rechromatography step were dried in a speed vac concentrator before
being sequenced. Peptides that were chromatographed in the
ammonium acetate gradients were dried, redissolved in water and dried
again to remove excess buffer before analysis. Sequencing reactions
were carried out on a Beckman System 890 M sequencer, and the
resulting PTH amino acids were determined by two independent HPLC
chromatographs. All sequence analysis was carried out in the
Macromolecular Facility in the Department of Biochemistry, Michigan
State University under the direction of Dr. Y. M. Lee.

Molecular Cloning Protocola. Restriction enzyme digestions were
carried out as previously described (55). Standard ligations into m13
replicative form DNA or plasmid DNA were as described (51). Ligations
directly in nusieve agarose (FMC Bioproducts) were performed according
to Dumais and Nochumson (56). Appropriate strains of E. coli were
transformed as described (55).

27

RNA Iaolat_ian. RNA for sequencing was isolated from bovine
retinas as described by Labarca and Paigen (57) with modification.
Briefly, retinal tissue was homogenized (2-3 mL/ gm retina) in ice-cold
buffer (8 M guanidine HCl, 0.01 M EDTA in DEPC treated water). One-
tenth volume of 2 M NaAc in DEPC water and one-half volume of cold
ethanol were added and the solution incubated for 3 hours at -200C.
The solution was centrifuged at 8000 rpm for 20 minutes in an HB-4
rotor to pellet the RNA. The pellet was redissolved in one-half original
volume of homogenization buffer. The solution was then reprecipitated
as above and two more rounds of resuspension and reprecipitatian were
performed. RNA was extracted by adding one-twentieth of the original
volume of water and heating to 700C for 10 minutes. The solution was
centrifuged for 10 minutes at 10,000 rpm in the HB-4 rotor, the
supernatant recovered and the RNA content determined
spectrophotometrically. Water extractions were repeated until no RNA
remained in the pellet. Yields were typically 0.6 mg/ gm retina (wet
weight). RNA, for northern analysis, was isolated according to Chirgwin
et al. (58).

Isolation of mRNA. Isolation of poly (A)+ retinal RNA was
accomplished by oligo dT cellulose chromatography as described (55),
except that the SDS concentration was reduced to 0.1% and the total
RNA was passed through the column three times. In addition, recovered
polyadenylated RNA was rebound to the column and eluted a second
time, resulting in a higher recovery of poly (A)+ RNA and less
contaminating nonpolyadenylated RNA.

RNA Blot Analfiis. Five to ten ug of twice selected poly (A)"' retinal
RNA was fractionated on 0.8- 1.2% formaldehyde containing agarose gels

28

and directly blotted onto nitrocellulose as described (59). Radiolabeled
DNA probes were prepared by nick translation (55) or by random priming
directly from low melting agarose gels (56) . Hybridizations with cDNA
probes were performed at 42°C in a buffer containing 50% formamide,
0.1 M pipes, 0.8 M NaCl, 5x Denhardts solution (59), 100 rig/ml
denatured salmon sperm DNA and 10% dextran sulfate as described
(59). For Oligonucleotide probes, prehybridization was done in 12 mL of
a buffer consisting of 40% formamide, 5x Denhardt’s, 5x SSC, 50 mM
sodium phosphate pH 6.8, 0.1% SDS and 250 ug/mL alkali denatured
salmon sperm DNA for 2-4 hours at 42°C. Hybridization was performed
in eight mL prehybridization solution, 2 mL 50% dextran sulfate, and
100 ng unpurified end labelled oligomer for 12-16 hours at 42°C.
Washes were one hour with three changes in 2x SSC at room
temperature with agitation, 30 minutes in 0.5x SSC at 50°C, and five
minutes in 0.1x SSC at 50°C.
DNA Isolation.

a) Bacteriophage. Cesium chloride purified and miniprep
bacteriophage DNAs were isolated according to Botstein and Davis (60).

b) Plasmid. Cesium chloride purified plasmid DNA was
isolated according to a scaled up miniprep procedure. Briefly,
transformed cells were grown to saturation, pelleted at 3500 rpm,
washed in 50 mL of 50 mM tris-EDTA pH 8.5. After resuspension by
gentle pipetting the cells were repelleted and resuspended in 18 mL of
solution A (15% sucrose, 50 mM tris pH 8.5, 75 mM EDTA). On ice, 4.5
mL of 5 mg/mL lysozyme in solution A was added, and incubated for 10
minutes with occasional gentle agitation. Thirteen and one half mL of
solution B (Solution A + 0.1% triton x-100) was added and the mixture

29

incubated at 37°C for 15 minutes with gentle and occasional mixing.
The majority of the very viscous solution that resulted was transferred to
a hard walled SW28 tube and spun at 24,000 rpm for 45 minutes at
4°C. The supernatant (cleared lysate) was then purified by two
subsequent cesium chloride equilibrium gradient centrifugations (55) in
vertical rotors. Cesium chloride was removed by dialysis in D83 (10 mM
Tris, 1 mM EDTA, 10 mM NaCl, pH 8.0). Yields were typically 500-1500
ug/ 500 mL culture. Miniprep isolation of plasmid DNA was essentially A
as described (61), except that lysozyme treatment was performed in the
absence of triton X-100, which was added just prior to boiling.
c) Genomic. Genomic DNA from a single bovine brain and

from human placenta was isolated according to Hogan et al. (62).

Isalation af Nuclaig Agids from Gals, DNA fragments were isolated
from agarose gels by lining, with dialysis tubing, a trough cut in front of
the band in the gel. The trough was then filled with buffer and
electrophoresis was continued until the DNA was immobilized against
the membrane. Then, the current was reversed for 30 seconds and the
DNA was removed with a pasteur pipette. Agarose contaminants were
removed by passing the solution through glass wool, and ethidium
bromide was removed by extraction with n-butanol. The DNA was then
precipitated and redissolved at a concentration of 250 ug/mL. Yields
were from 60-90% and the DNA was suitable for all commonly used
molecular biological techniques. In some experiments DNA was isolated
using geneclean milkbeads (Bio 101) according to the manufacturers
protocol.

Smthesia, Purification, and fldiolabelling af gliganaclaatida
Probes. Some oligonucleotides were synthesized on an Applied

30

Biosystems 3803 DNA synthesizer in the Macromolecular Facility at
Michigan State University under the direction of Dr. Y. M. Lee.
Additionally, oligonucleotides were made on an Applied Biosystems 38 1A
synthesizer in the lab. Oligonucleotides were purified for use as
described (63), or by a method of M. Zoller. Briefly, 150ug of oligomer
was fractionated in 14-20% polyacrylamide gels. Products were
identified with short wave UV illumination onto fluorescent TLC plates
behind the gel. Bands were excised, crushed and eluted into a buffer

containing 0.5 M ammonium acetate, 1 mM EDTA, and allowed to difi‘use
overnight at 37°C. Gel pieces were removed from the DNA by passage
through millex-HV filters (Waters). Further purification of the oligomer
was achieved by Sep Pak C 18 (Waters) chromatography.

Labelling of Nucleic Ami

Kination. Oligomers were radiolabeled with y-[32P] ATP using
T4 polynucleotide kinase according to the manufacturer's protocol.
Unincorporated radionucleotide was removed by two rounds of
precipitation using denatured salmon sperm DNA as carrier.

Nick Translation. Labelling was performed as previously
described (59).

Oligolabelling. Random hexamers (pharmacia) were used
with the Klenow fragment of DNA polymerase to incorporate a-[32P] dCTP
essentially as described by Feinberg and Vogelstein (130).

Screening Human and Bovine cDNA Librariea. The human and
bovine retinal cDNA libraries were constructed in the bacteriophage
lambda cloning vector lgtlo (14,15). The human cDNA library was
screened with an oligomer and a cDNA probe. Hybridizations with the
oligomer were performed at 42°C in a buffer containing 0.9 M NaCl, 0.09

31

M NaCitrate, 50 mM sodium phosphate, pH 6.8, 5x Denhardts solution,
100 ug/mL denatured salmon sperm DNA and 10% dextran sulfate as
described (59). Hybridizations with the cDNA probe were done at 42°C
in a bufier containing 50% formamide, 0.1 M pipes (Ultrol), 0.8 M NaCl,
5x Denhardts solution, 100 ug/mL denatured salmon sperm DNA and
10% dextran sulfate as described (59). The bovine library was screened
according to a method developed by E.F. Fritsch, Genetics Institute,
Boston, MA. Briefly, filters containing ~20,000 p.f.u. were prehybridized
at 65°C in a buffer consisting of 3.0 M TMAC, 50 mM sodium phosphate
pH 6.8, lmM EDTA, 0.5% SDS, 5x Denhardt's solution and 100 ug/mL
alkali denatured salmon sperm DNA. After 1-2 hours filters were washed
with kimwipes (Kimberly-Clark Corp.), and then placed into fresh
prehybridization solution for 12- 16 hours at 65°C. Hybridization was
carried out in a large volume (10- 15 mLs per filter) of prehybridization
solution with 0.1 picomoles of kinased oligomer probe added per mL of
solution. Incubation was done at 35-45°C for at least 48 hours, or 72
hours for highly redundant probes. Washes were done in a large volume
of 2x SSC, 0.2% SDS at room temperature for two hours with one change
of solution. Filters were then air dryed and placed at -70°C for
autoradigraphy. 4

Screening and Analysis of a Bovine Genomic DNA library, A
bovine genomic DNA library constructed with it EMBL-3 was purchased
from Clontech. The library was screened as described for the cDNA
library with cDNA probes. An isolated genomic clone was mapped by
restriction analysis, and according to the labelling and partial digestion
method of Baker and Board (64).

32

Detarminatian af Gang Capy Number, The fraction of DNA that a
6.2 kb fragment represents in 10 ug of genomic DNA was calculated by

assuming a genome size of 3x109 base pairs, and 660 g/mole/base pair.
Amounts of the isolated genomic fragment equivalent to 1-6 copies of the
fragment were loaded into an agarose gel adjacent to 10 pg of genomic
DNA digested with HindIII or EcoRI. Comparison of hybridization
intensities allows an estimate of gene dosage (see Fig. 10).

Nucleic Acid Sequence Analfiis.

M13 Single Stranded DNA, M13 single stranded DNA was
isolated as previously described. Nucleotide sequences from inserts

cloned into M 13 vectors were determined using the dideoxy chain
termination method as described (65) with the following modifications.
Buffer gradients (66) were employed using 60 cm gels, or-thio-[35S] dCTP
was used in place of a-[32P] dCTP and reactions were performed with a
sequenase kit (U. S. Biochemicals). These modifications allowed up to
500 nucleotides to be determined on a single 60 cm gel.

Double Stranded Plasmid DNA. Plasmid DNAs were
sequenced exactly as described by Zhang et al. (67).

1 DNA. Inserts in it EMBL-3 were sequenced according to
Manfioletti and Schneider (68) except that DNA isolated from cesium
chloride purified A phage was used as a template.

m_RNA RNA was sequenced essentially as described by
Geliebter (69) except that all dideoxynucleotides were added from a 0.5 '
mM stock.

Chromosomal Localization. The initial location was determined

using fourteen human-rodent somatic cell hybrids. Construction,
karyotyping, and DNA blot analysis of these hybrids has been previously

33

described (70). Confirmation of the assignment and regional localization
was obtained with a panel of 1 1 human-hamster hybrids that contain
only subsets of human chromosome 5 (71, 131). DNA blotting, hybrid-
izations, and washings at high stringency were as described (59).
Computer Analfiis, RNA and DNA sequences were analyzed with
a number of sequence analysis packages run on microcomputers under
DOS, or on a minicomputer under the VMS operating system. Programs
written by R. Staden (72), D. Mount (73), and Bucholtz and Reisner (74),
were used to align and analyze all of the sequence described in this work.
Sequences from autoradiographs were read directly into microcomputers

using a graphbar digitizer (Scientific Accesories Corp.).

Sectian A Bovine cGMP PDE at-Subunit.

figmsinolysis and Paptide Sequence Anflysis pf Bag: QQMP PDE,
One mg of highly purified (see Materials) cGMP PDE was graciously

provided by D.G. McConnell, Department of Biochemistry, Michigan
State University. After SDS-PAGE, cGMP PDE gives rise to three
subunits, a (88 kd), B (84 kd), and y (1 1 kd) which does not stain well
(Figure 2A. lane 1). After modification and tryptic digestion, the sample
appeared as a low molecular weight smear (Figure 2A, lane 2), on a 20%
acrylamide gel. One-half (250 pg) of the total modified and digested
sample was fractionated by reverse phase HPLC. The initial sample was
divided into three aliquots, each of which was fractionated separately.
Figure 2B, profile a, illustrates a representative chromatographic profile.
For two of these fractionations, fractions of one ml were collected at
thirty second intervals. Fractions indicated on the figure were pooled for
rechromatography. However, fractionation of one aliquot gave very high
absorbance readings and so all the material from this final fractionation
were combined with the selected pooled fractions from previous
chromatography steps and rechromatographed. The resulting profile is
shown in Figure 2B, profile b. Fractions were collected at thirty second
intervals and regions were selected for rechromatography. Material

34

35

Figure 2: Analysis of flgzptic Paptidea of Bovina QQMP PDE, One mg of
highly purified protein was modified and digested as stated in Methods.

A. Lane 1: SDS-PAGE analysis of 3 ug purified cGMP PDE protein.
Lane 2: 1 / 25 of the total sample after modification and trypsinolysis. 3.
Initial fractionation of 250 pg of modified trypsin digested sample. Profile
a: Tryptic fragments were eluted from a reverse phase HPLC column.
Peaks designated by letters were pooled for subsequent purification.
Profile b: Pooled fractions from initial chromatographs were recombined
and rechromatographed. Peaks corresponding to peptides which yielded
sequence data are indicated. Profile c: Samples remaining from all
previous chromatographs were pooled and rerun as described. Peaks
corresponding to peptides that yielded sequence data are indicated. C.
Purification of pooled samples. Profile a: Peaks E and F (panel B, profile
b) were combined and fractionated under the same conditions as
previously to yield peptide cPDE6. Profile b: A peak (not shown) purified
from peak D (panel B, profile c) was rechromatographed under the same
conditions as previously. Profile c: Peaks A and B (panel C, profile b)
were fractionated using a linear gradient of ammonium
acetate / acetonitrile as described in Methods. D. Sequence analysis of
peptides. Seven purified peptides were subjected to automated Edman
degradation and analyzed by reverse phase HPLC. Yields of PTH amino
acids are shown for two peptides, PDE3 and PDE6, that proved to be
crucial to this study.

36

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

oz .20 m 383
s o n e n u . e. u n v n u .
u d d u 1 a O u 4 4 a u a O
x 88“. nmoa a. W
A. > 1 on 8 w. M
I 1. 8. l > z r 00 W m
.l U 1 8— r. l p m lo,
a. .. am. a.
p p p p p b 3N F p b p b p nN_l-\ o
3 3388s
on on 8 8 n. on 3 8 n. 3 on 8 om.
a . . . l q
1 1 No.0 1 1 mod 1 1 mod
1 1 5.0 1 1 8.0 1 1 mod W
a M. u
1 ma 1 88 1 oo 1 mod
n . . .
1 1 88 1 N. o 1 a 1 «.0
o 38 n a $2
. . . ..o . . 88 p . n8 0
87.5.8.8
_ ow 8 9. on 8 o. 8 on 9 on 8 o.
O — J a U 0
Bo
: a 1 3o
0 V
80 s. x o m 1 z
u _.o I
moo 9
NS 1 no
..
90 L _ _ _ 3 m

 

 

 

37

remaining from the previous fractionations and any material not used
from the experiment shown in profile b was combined and again
chromatographed with the resulting profile shown in Figure 2B, profile c.
Fractions corresponding to regions A-D were pooled separately. Each
combined fraction was rechromatographed to obtain homogeneous
peptide. Figure 2C, profile a shows the result of rechromatographing
peaks E and F (Figure 2B, profile b), from which peptide PDE6 was
obtained. Similarly, peptides AB 1 and PDE3 were obtained after further
purification of peak D (Figure 28, profile c). In all, eighteen peptides
were purified in sufficient yield to obtain amino acid sequence. Eight
peptides were chosen for sequence analysis in the Department of
Biochemistry, Michigan State University macromolecular facility. Amino
acid sequences were obtained for seven of the peptides, but the eighth
could not be analyzed. No trace of PTH derivative was observed for the
eighth sample indicating the absence of protein. Figure 2D shows the
yield of PTH amino acid for peptides PDE3 and PDE6.

Synthesis of Oligonualeotide Probes. A unique Oligonucleotide
unambiguously reflecting the sequence of each peptide cannot be
fashioned due to the degeneracy of the genetic code. Several alternative
approaches have been used to circumvent this problem. First, unique
oligonucleotides of length greater than 32 bases have been synthesized
using information such as codon preferences and relative instability of
an A:C mismatch compared to a G:T mismatch to create a sequence that
is most likely to correspond to the in viva sequence (75). Second,
mixtures of oligonucleotides have been synthesized that reflect all
possible sequences of a peptide (77). In this case, peptides containing

serine, leucine or arginine residues which have degeneracies at more

38

than one position are avoided. And third, deoxyinosine has been used at
the wobble position (78), since preliminary evidence suggested that the
hypoxanthine base of deoxyinosine acted in an inert fashion, neither
contributing to, nor destabilizing the DNA hybrid (79). Again, peptides
containing amino acids with greater than four degeneracies are not
desirable with this approach. Elements of all three of these approaches
were incorporated into the design of the oligonucleotides described in
this section. A codon usage table was compiled (not shown) from the
deduced amino acid sequences of 12 bovine genes extracted from the
Genbank nucleic acid database (80). From this table, it was found that
leucine residues are encoded by C,T at the first and second positions,
respectively, greater than 75% of the time. Therefore, for the leucines
encountered in the seven peptides, the codon C T1 was used. For amino
acids where there were only two possible codons, both were included in
the Oligonucleotide mixture, and any amino acid where three or four
codons were possible, a deoxyinosine was incorporated as the wobble
base. Figure 3 presents the amino acid sequence for each peptide and
the corresponding oligonucleotides.

Screening the Bovine Retinal cDNA Libra, One hundred and
twenty thousand plaques from a bovine retinal cDNA library (15) were
screened with each Oligonucleotide on duplicate filters. A novel method
of plaque hybridization was used, taking advantage of the properties of
TMAC (see Methods). PYom the initial library screen, 285 hybridizing
clones were identified on both the original and duplicate filters. These
clones were rescreened by scoring, into sixteen sections (Figure 4), an 85
mm petri dish plated with the appropriate bacterial host strain. A drop

of phage suspension was placed in a section and allowed to drip down

39

Figure 3: Desiga pf Oligonucleotide Probes. A. Sequence of seven
tryptic peptides from amino to carboxyl terminus. Peptides PDE3 and

A31 terminate with a tyrosine residue, presumably due to contaminating
chymotryptic activity not quenched by TPCK. Peptide PDE4 is truncated
due to an insufficient amount of sample. No PTH amino acid was
observable for residue 7 in peptide 5. B. Sequence of the
oligonucleotides. Oligonucleotide mixtures containing deoxyinosine
residues and reflecting some codon degeneracy were synthesized (see
text). I is used as the single letter designation for deoxyinosine. An
Oligonucleotide was not synthesized reflecting the sequence of peptide
AB 1 .

 

PDEl
PDEZ
PDE3
PDE4
PDES
PDE6

A81

CPDEl

cPDEZ

cPDE3

cPDE4

cPDES

cPDE6

Figure 3

SO!‘
glu
my
phe

val

glu

glu

3'

3'

3'

3'_

G T

T
C T
C

C C

A A

C A

T
C T
C

gin
lle
thr
leu

esp

leu

_1‘

G)

>

40

3 4 5
asn pro leu
leu phe tyr
asn asn Ieu
asp gln asn
lle tyr lle
esn phe tyr
ale gln tyr
G G l G A I C G
A

G A I A A A T
G
A A

T T T T G A
G G
A T

T T T T
C G G C
A

T A I A T T A
G
A A

T T A A A T
G G

>

5!

3.

T

A

G

ala

lys

tyr

leu

lys

T 5'

T 5'

T 5'

A 5'

T 5'

lys

Q'Y

lys

41

Figure 4: Screening of the Bovine Retinal cDNA mm The primary
screening of the bovine cDNA library resulted in the recovery of 285
positively hybridizing recombinant phages. A. In order to rapidly
rescreen the clones obtained, an 85 mm petri dish was scored into
sixteen sections. Undiluted phage representing each clone was allowed
to drip down one scored slot as described in Methods. B. Number of
positively hybridizing phage for each Oligonucleotide probe obtained in
the original screen. C. Number of positively hybridizing phage after
rescreening each clone obtained in the original screen with each
Oligonucleotide probe. For example, of the thirty-two clones identified
with cPDEl, four of them hybridized to cPDE3. Three more rounds of
hybridization reduced the number of clones to seventeen, all of which
hybridized to cPDE3 and cPDE6.

42

v musmflm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

—N O O an N O

O w O O O O

O O O O O —

mp O O @m p v

N O O m up mp

O O O O O N
nun—mu mwoau vwc..u nwoau Nun—mu .wOLo

case» 352.0 co 62:203.

emcee
mmoeo
emcee
among
«mono

pmomu

ueeraser uo pegjgruepj

 

«N—

an

 

flmeam
22:3:0

ONOQU
mmanu
emomu
nwOmu
Nwan

¢waao

 

one}.

0.

N.

c.

J/

n.

@—

 

 

 

 

 

 

 

 

 

 

 

 

 

43

the agarose surface. Six filters were lifted from each plate and hybridized
as described in Methods. In this manner, all of the initially identified
clones could be rescreened with each Oligonucleotide probe. More than
200 of the original clones gave a hybridization signal using this
procedure. Therefore, it was necessary to find a means of limiting the
subsequent analysis. The criteria used to limit further analysis were
strength of hybridization signal and hybridization to more than one
Oligonucleotide. Twenty-five clones met these criteria and were
rescreened with the Oligonucleotide probes. Of the 25 clones, 20 showed
hybridization with both cPDE3 and cPDE6, three hybridized to cPDE2
only, one hybridized to cPDE5, and one showed no apparent
hybridization. Further characterization focused on 17 of the 20 clones
that hybridized to both cPDE3 and cPDE6 Oligonucleotide probes. The
remaining three clones were omitted from further analysis because they
failed to hybridize in subsequent rescreens. A summary of the results of
the first two screening procedures is shown in Figure 4.

Characterization of Bovine cDNA Clones. DNA isolated from the
seventeen clones described above was digested with EcoRI, fractionated
on a 0.8% agarose gel, and visualized by staining with ethidium bromide
(not shown). All of the clones share 662 and 144 bp fragments, except
one clone where no insert was apparent. Six of the clones designated 8,
15, Pat-1, Pat-2, Pat-3, and Pat-4 (Figure 5) contain more than two EcoRI
fragments. Subsequent analyses concentrated on these clones since it
was thought that they would represent longer cDNAs. Additionally,
another clone designated 7 (Figure 5) was isolated by rescreening the
library with cDNA probe 3 (Figure 7). A schematic map constructed
based on sequence analysis of the cDNA clones is shown in Figure 5.

44

Figure 5: Schematic Map of Bovine cDNA Clones, The approximate
positions of four oligonucleotides (designated 1,3,5 and 6) that were

identified from a cGMP PDE tryptic hydrolysate are shown on the
restriction map at the bottom. Wider rectangles represent protein coding
segments and more condensed rectangles depict upstream or
downstream untranslated regions. A segment that is essentially identical
in all of the clones is shown as an unfilled rectangle. Included in this
region is the "PDE HOMOLOGY“ domain that defines a segment
conserved in a number of CN PDEs (see text). Regions of sequence
divergence are represented by different fill patterns. Arrows at the top
mark the positions of serine (Ser) and glycine (Gly) residues that define
the endpoints of the region common to all of the clones. Predicted amino
and carboxy termini are designated N112 and COO” respectively. The
three cDNAs (numbered 8,7, and 15) that overlap to define the cGMP
PDE a-subunit sequence are shown. Shown at the bottom are the
restriction sites that are common to the clones, and other EcoRI sites
that are clone specific are shown below that clone with the corresponding

size of the unique restriction fragment.

 

 

 

45

whom
a c «8 2:. .z 38“ m E
1 > >d
«3
m. 82.. m. 3.8 . 3 on
VNNVVNNV i—

__ . segue

.nfl
h

 

 

 

 

 

 

\VV‘VV‘NN

NNNNNNNN

 

 

 

 

 

 

 

 

Nidm .
S RN

 

c
u
all u
V‘VVVV“ -
Hid“ 5.5.510: mom i—
N\N\N\\\

 

 

 

.
q
\VVVVVNV
>9393§¥3 (B
m
-
Tl

 

 

 

ma
N:\N\NN

.000

2:80 3:351: .. g _
u

are. .
a assume... %

\NNNNNNN ’

 

 

 

 

 

 

 

46

The sequences of specific Oligonucleotide primers used to complete all of
the sequence analysis described in this study are shown in Table 3.

Three clones (7,8, 15, Figure 5) within one cDNA class overlap to
define the cGMP PDE a-subunit. The composite sequence from these
cDNAs contains a 2577 nucleotide open reading frame that encodes a
sequence of 859 amino acids (Figure 6A). Although the nucleotide
sequence of this cDNA is 99.3% identical to the cDNA sequence of the a-
subunit reported by Ovchinnikov et aL (6), a number of differences were
determined. Discrepancies were identified at fourteen positions of which
three result in amino acid changes (summarized in Table A1).

Four closely related at-subunit like variant cDNAs (designated Pat-1, .
Pct-2, Pct-3, and Pct-4) were also identified (Figure 5). Nucleotide
sequences representative of important residue changes in these cDNAs
are shown (Figure 6B,C,D, Figure A4). Pat-1 and Pat-2 are identical in
sequence with the or-subunit from the start of each clone at nucleotide
914 (see Figure 6A,B,C) through to position 2575. Both cDNA clones
predict 29 additional amino acids following this position before an in
frame stop codon is encountered (Figure 6B,C). These 29 residues show
no obvious relationship to each other or any similarity to the 19 residues
at the C-terminus of the oc-subunit. .

The nucleotide sequence of Pat-3 is identical to the sequence of the
«subunit from nucleotide 941 to the end of the clone at nucleotide 2958
(see Figure 6A,D). The divergent sequence in Pat-3 converts the
methionine codon at amino acid position 296 (Figure 6A) to a valine
codon and introduces an in frame stop codon at nucleotides -5 to -7
(Figure 6D). The methionine at amino acid position 360 (Figure 2A)
would then define the N-terminus yielding a protein of molecular weight

47

Table 3: Primers used for §eguenge and RNA Blot Analysis. The

sequence of all oligonucleotides useful to this study is shown. Base
positions are according to the designated figures. Figure numbers
preceeded by an A signify the Appendix. Position numbers are ascending
for sense primers and descending for antisense primers. Nucleotides
shown in lower case represent mismatches in primers used for human
cDNA sequencing. For example, for sequencing primer sp-35 the human
cDNA contains the sequence CCACAGAGGCACCAATA (mismatches in
bold letters). Primer sp-7 was synthesized according to a published
sequence of a portion of l. gth (126). The letter designations of
Oligonucleotide probes mapped in Figure 7 is shown next to the primer

name.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

angispa o auaoaoonuoauooaoooaoo Hmumu nuuunnu m oaauauonuoooanuaauaau onion
«panama u ooooaauoaaouooaaanuo o¢umn eauuucn m oouoaugdoaaauauaauan onion
nnmupam ma onauoaaanauaauauo amumu mpmuamn u uauauogaoqununuau snug.
poau-m~a« u oauoaauaoaoaaaoao onnmu whoauauoa u «nonnoauauooauauauu mnumu
mmmfiuupnfl w oguaoaomadanuaogoduo ¢mumu «pmaumnna w nouaaoaoadaoaonuqoao nnumu
«crimes 6 oaoaguo¢0¢¢u¢uo¢o¢ anion onuanvaaa w unauaaooaauauauoo autu-
pmmannahu u wooeuaodnaoooonao owns» Hepuupupu .4 oaoo¢4oa¢oaauo4 swim.
mpuwuoauu v4 ooococaaaoondoa salon omnaunmaa u ooooondoaoaooao nuumu
«mauovm m oooaaoooauoaooo «mum. ovauuua u oooaooauooaauoo o~-mu
mammuouam '4 angogdagoaogago panda nupuuavpu v4 aoogaoaaoaaoo¢o wauma
ouuuumpuu v¢ auaaooggdaoaooo maumu uuoa « u oaodaoooaooaoao puma
moannaaon v4 oaoaauaoaaoauge mum. Hoamumoan v: ¢uoaoa¢uqua¢o¢o n-mu
”wouupvcu u auaoaooaaauagmooo vim. oaonunnaa w oaaoooaoauoooan mum.
mmmduaumfl 4d uaaaeaauguoaaoa «non uumu-nmaa c4 aunaooaoanu4a¢u anon
nowuwnom ouauwh coconvom oamz newuwuom Gunman mononvam oadz

 

 

dmwmuwm1¢vuon dfld 6am 00:0 am How v on whoawum "m canda

 

4S)

 

 

 

 

 

 

 

 

 

 

 

 

pm+ umm+ m oaaooooaaaoooeoaoooooaHonda:aooaoonoaoaooaoonoaood .h. «and»
mm+ nmo+ m aoaaaaoaoaoagaduoaoodoo¢¢u¢uaauoaguduoaaoaooogdugn Au. amumu
aw- nan u odanuooao¢U4ooo¢¢a¢aoanadaaooaooaoduaaoanudaaoaao .0. onion
omsuusppu m oaogaueuuooaaguaaanodoaaaoaoaa¢¢daoaddaoa¢o44¢oo .ov anion
«amuuammu m cocoasaucyaooauoooooooooogooadooaao46:4oo:uoo¢ooaa .o. onumu
mvmuuvpmu u ooaoooaaooaaooaoooooaooaaoa .m. pmnmn
on. uv a oooaoaooaaooaoonaoaoaaaoaaoaoaoaoaoaaoao mmumu
mmuom m aduooooaooauaooooaoaooaauduoaoaaugdd .4. mmumu
amniamm m aoooduaggoaaudaooaoopooooo emumo
aowuwnom ouauwh unconvam oaaz
...o.uaoo. "m wands

 

50

Figure6: N l D Amin Ai n

QDNA Clones, Nucleotide numbering is shown on each side of the
sequence and amino acids are numbered above each tenth residue. The
complement of oligonucleotides synthesized for RNA blotting experiments
(see Table 3) are underlined. In frame stop codons that define the open
reading frame are denoted by asterisks. A. Sequence of the cGMP PDE
a-subunit. Peptide sequences obtained from cGMP PDE (see Figure 3)
are boxed. Internal EcoRI sites and the EcoRI" site are underlined. A
solid box is present above residues at the C-terminus that are potential
sites of palmitylation (see text). B. Sequence of cDNA Pa-l and C. Pct-2.
Serine 840 corresponds to the a-subunit sequence. Nucleotide sequence
divergent from the a-subunit is shown in capital letters and numbering
is preceeded with a plus sign. D. Sequence of cDNA Pa-3. Nucleotide
sequence of Pom-3 that differs from the a-subunit sequence is shown in
capital letters and numbering is preceeded with a minus sign. The
proposed amino terminal methionine corresponds to amino acid 360 of

the a-subunit sequence.

51

'h

-A (CI? rut o-oubunui
I II II
. ICIVTAIIVII_VIIAIQ

I mmmmtmmmmmm I II
II II II II

IVILIVIAIVIIILLCIIIAAVI IIVIALIIVIIIIIII

III IMWWW M! - WWWWAW III
II II II III

0 L L I 0 I Q I I L Q A I I C V I I V I I I L C I L L 0 A I I I I L I I V I A I
III WWWWWWAWMWW III
III III III III
I C I A I L A T I L I I V I I I A V L I I C L V A I 0 I I I V I I L 0 I C V V C

 

 

 

 

 

III III III III
I V A L I I I I V I V I I 7 I I 0 I I I C I I V I T I. T I Y Q 7 I I I L I I I I
III ammmmrrrrnn- “ “‘WWWWAMGMWTA I00
III III III III
I I C I 0 V V A I I I V V I I V I C I I I T I I I_ I I I L L I I I. I I A I I. I I
III AMWWCOYMYCAWMIWWM ‘ w. WMWWIOMIUN 720
III III III III

IVIILI'LIICITIICQILLIICIIVIIILIIIIIQIIIA

”I WAMCWIWWWWW III
III III III III

LYYVIAILICDIYIVCLLIIIIQIIIIIVIIVLICIAIIV
“I maumcmmcacmmrmmmmmmmmnx III

 

 

 

III ”I III III
A C I I T I I C I I I I I I I V I D I I L I C I I I I I V I I I I I I 0 I I I I.
III CCTCCTCCCACCACTCCCCAWTCMCTTHAWCAMICATCCTCCA WWWWCHTA IOI0
)II III I? I III
V I C L I 7 V V A Q I C L I C I I I I A I I I I I I A I 0 I I I I. 0 I I C I V I
IOII CTCACCCC'TCTCCCCACHATWCCCCMMTCCCU‘CAWATCA“ ‘ I. ‘- Innlwfii“ ‘ “ mmmn I200
III III III III

IIVLIIIIVIIIIIIVCVAIIYIIIICIIIDIIIITLIII
IIOI MIUCWOAWCCMWCMWWM mmmmmmcmm I330

 

All “I III
LTQILCIIVLIIDTVILIIILIIIIIIIOIIVIIIVIC.I)I
I17] "CACCCACTTCCTCCCCTCCICTCVCWMAYCCTCACACHAWW‘“"'““" - l l IWW‘I’WCMT I660

“I III III III

IIIO?)LIYIEVYCIIPIECIIIILAIILOCILIDADIYI
m: cucwchmcuTmMcmwmanwmtmmmaucmnm mo
in m m Mi
IIIFIFSDLPLYILILVICCIQIYYILIVVDIIIIIQIAL
I III ATT M1MA1'!CCACTTCACCCACCTCCCCCTCACCCMUCCACWWWMMANAWWAYWWUWCUCICCACCCCCTC IIIO
m m m m
VIIIVILIICVIIITVIIIIICIIVCQTIIILLVICILII
m: C‘rcccc'n camancccrcmMeccanccccaccAm.enamwmummmcamacmcmccocmcuccu uoo
am so: III on
YI‘IDLIALAIV'IAAICII IllmolllsglrLallLI
no: tmmmmmm museum nmmtccuwcwmcmmzr 1920
m “I on “I
CSIILIIIILIICITLLIDIILIIIQILIIIOIIIAIIII
IIIIWWMHM ‘ 4.. -‘“-mmurmmroccmmmmmrc mo
m u: on m
pin:iaroiatvrltnturozivoosxrvrrqotvtovuu
nu mmrurmamcmcmamnmu “ I!” “mmmmrmmmmn 2160
m m m m
LDQ‘IIIEIVIAIIIYACDLIAIIIPIIVQSIVALLVAAI'
2m mrmmumammmm - “‘“ ‘ --.-.-e-u_-.mn.mmum 22.0
m m m m
IIOCDLIITVLQQI'IIIIDIIIAIIL'ILOVCIIIIVCT
2a: mwxwmmmummmmaamummm uoo
m to: III In
IVYIIFSIIIEIIYPILDCI‘IIIIIIIIALADIYI‘I’IIIC
no: mamnaccccmccaccwcmrcanccummnmmmammm 2520
m on m - I”
Lt!IIOIQQAAIQAAACIQICCROPCCCPASIICCVQO
IIII €7chAcwucucccacccnccucuccccuccumaxcocmccucmococcmanwmmmrm IIIO
I“ I CACCACCCACCCCCCACCCCCCCCCCACCCTTCACCCACCACACCACCCACCCCWWCCATACCCCACCTCCTWCACAWWWCTCAMCAEM—T 2 7I0
I II I ”WACATCTI'TE'CTACTCCC'TTTCMCCCTCTTH'IC“mmATTCMTCACAACCMCATMWCACICACTCWCAHACTCCCACACTCACCTCC IIIO

 

 

 

 

 

aux 1"!ACCIACAATCTHTC‘IC‘I'TCTTIJACH‘I’CA'I‘ITIATACTACACMATATCCAATCATTCAMACACM‘H "M
I Po-I - C Io-Z
no I II no I H
IICILIVICOIIIIII IPPIIICIICAICIII
In: unmanfimmmmnm on n52 ogucacmmmmmcwcccu on.
21 29 21 29
I'CAACDAVCIALI0 IIICQDICCD‘ITSP'
«s Lqryvccaccrpqrmncunrccmacm on «s uccmawcccucnnmcacucaacmtu on
on cmccuccuécmccmccnacamnaccamac one on nunmncumucaccmaccaumm 0:50
CHI mrmnawcmnmm out on: mmummmamn oiu
i'_r.'-‘:‘—‘" ""” "

- I I 2 “WOMWWIIWWCCWATCCCACTCTCNTTAIACATATAACCICCACTCMCAMMCA - I
'2 egtgggggoggctceoccctocgcIggletc-33.:lccggotggnagggaoolcooctIttoccaggieoiInaciocoiccLgeatggcooogoouocotcogogIcotcccgooicc III!
III I 70

IIAIIIDIIAIQIIIL
IO” Icunag-"nuggetIInugouyginun-cling!“canon-tutu.“ilunacauonoouqcuicogouulluitgcoliccogooogoucut II"

Figure 6

52

58,262. Clone Pen-4 differs from the a-subunit sequence only in its 3'
untranslated region. The sequence divergence begins at residue 1816
(Figure A4). An additional EcoRI fragment of ~ 1500 bp is also present in
the Pom-4 cDNA clone. An artifactual organization caused by the ligation
of two unrelated cDNAs cannot be ruled out. Thus, cDNA sequence
analysis predicts that three proteins exist in the bovine retina with a
striking similarity to the cGMP PDE a-subunit, and one additional cDNA
clone that may differ only in its 3’ untranslated sequence.

Blot Analysis of Bovine Retinal RNA. In order to understand the
relationships amongst the cDNAs specific oligonucleotides were used as
probes for bovine and human retinal poly (A)+ RNA. Several
Oligonucleotide probes specific to the a-subunit sequence. and a bovine
cDNA probe containing sequence common to all of the clones hybridize
predominantly to two relatively abundant bovine retinal RNAs of 4.6 and
4.0 kb (Figure 8, lanes 1-6). Oligonucleotide probes specific to the C-
termini of Pom-1 and Pa-Z hybridize to bovine RNAs of ~4.3 and 4.8 kb
respectively (Figure 8, lanes 7 ,8). Additionally, a Pct-3 specific
Oligonucleotide hybridizes to low abundance transcripts of ~5.8, 4.4, and
3.7 kb (Figure 8, lane 9).

Using Oligonucleotide probes specific for the C-termini of Pom-1 and
Pct-2, transcripts of ~5.3 and 3.5 kb are detected in poly (A)+ from
human retina, respectively (Figure 8, Lanes 10,11). A smear of
hybridization to high molecular weight RNAs was observed (not shown)
with cDNA probe 9 (Figure 7) that is specific for the 2048 bp EcoRI
fragment of Pen-4. A similar but nonidentical smear was obtained (not I
shown) using a probe specific to the 1500 bp EcoRI fragment from Pct-4
(Figure 7). It is possible that these probes are hybridizing to a common

53

Figure 7: Location of cDNA and Oligonucleotide Probes. The cDNA
clones are shown in schematic form as in Figure 5. The a-subunit
cDNAs have been combined to show their composite organization. Maps
for each cDNA clone (7 , 8 and 15) that make up the a-subunit are also
shown. Positions of Oligonucleotide probes are designated by small boxes
(labelled A-G) that are sized to roughly represent the length of the
oligomer (see Table 3). Lines labelled with arable numerals represent
cDNA probes used in RNA and DNA blotting experiments.

54

O.
.D
O
O
N
.4:
.4:
db:
0
d:

Vldm

”lam

 

 

h musmfim

 

 

 

    

 

.‘Q‘OOOG40‘OOOOOQ.O
o+¢o++oooooooooooo.
.I????????P?PP????3

 

\“~w-~‘~‘\

\

 
 

 

 

teammxmnwmx

 

 

     

020: on
\ED\.\

 

>00
.NK

 

 

”€200 UMGflQfl»!5 nu

nmnoum «zoo

unannanad mam mxoo

 

 

 

r;
.P
db

J-

Fit
-
d

.llEllnl .
wmhla llElliEl

DIR“ <
DI .‘\\£C3‘N\. I
2......- g.
\\\\\\\N
N12

.000 zenuum 202260

 

 

 

3m

55

Figure 8: Analysis of Bovine Retinal RNA. Approximately 2 ug poly (A)+
RNA (lanes 1-9, bovine RNA: lanes 10-1 1. human RNA) was fractionated

on a 1.0% agarose, formaldehyde containing gel and transferred to
nitrocellulose (22). Nitrocellulose strips were hybridized as described in
Methods. The probes used for each blot strip shown are as follows: lane
1, probe 5; lane 2, pos. -35-+6 (Figure 9); lane 3, probe A; lane 4, probe
B; lane 5, probe C; lane 6. probe D; lanes 7 and 10. probe E; lanes 8 and
1 1, probe F; lane 9; probe G. All probe locations are shown in Figure 7
unless otherwise noted. Exposure times were three hours without an
intensifying screen for lanes 1-6 and 54 hours with two intensifying
screens for all other lanes. Lane 1 blot strip was hybridized once prior
with probe 10 (Figure 7).

56

I: ; "‘
a: 'W'

 

 

u
..
r.
i H
.H
“;
N, .

m

h

m

m
oc_>om

 

v

m m.

57

set of RNAs, but the data do not unambiguously support such a
conclusion. However it does appear that at least seven related RNAs are
identified using clone specific probes.

Sequence Analysis of Bovine Retinal RNA, The identification of the
a-subunit sequence was based on comparison of the cDNA sequences
described in this study to that reported by Ovchinnikov et al. (29). In
that report the oc-subunit sequence was also determined from a number
of overlapping cDNA clones. Further support for the authenticity of the
sequence was provided by the presentation of several peptide sequences
obtained from purified a-subunit. Two of the identified peptides were
found only in the a-subunit hydrolysate and not in the B-subunit
hydrolysate. However, two regions of the reported sequence were not
supported by sufficient overlap of cDNA clones, nor by peptide sequence.
These regions are roughly defined from nucleotide 175-592 and 2512-3’
end (Figure 6A). In order to confirm the sequence in these regions direct
dideoxy sequencing of RNA was performed. The results of this analysis is
shown in Figure 9. Several specific oligonucleotides (see Table 3) were
used to prime specific regions of interest in the RNA. The sequence
obtained shows complete agreement with the reported sequence (29) and
the sequence identified in this study from cDNAs 7 ,8, and 15 (Figure 6A).
Therefore, the peptide sequences reported for the tat-subunit (29), and
those reported here, together with the RNA sequencing data provides
strong support for the entire predicted oc-subunit primary structure.

Isolation of a Bovine Genomic Clone. A genomic library was
screened with bovine cDNA probe 4 (Figure 7) resulting in the recovery of
a ~13.5 kb clone. A restriction map of this clone is shown in Figure 10A.
The EcoRI digestion pattern is somewhat tentative because of the small

58

Figure 9: Direct Chain Termination Sequencing of Bovine Regal R_NA.
The results of sequence analysis of bovine retinal poly (A)+ RNA is

shown. The location of the sequencing primers used are underlined with
the primers name below the sequence. Numbering is according to figure
6A. Nucleotides shown in bold uppercase letters represent the extent of
the sequence determined with confidence. A 1340 nucleotide portion of
the sequence designated "1340 nt" has been omitted because no RNA

sequence analysis was necessary in this region.

129

189

249

309

369

429

489

549

609

669

729

789

849

909

969

2350

2410

2470

2530

2590

59

rrscmmrcmsmwmrmrcmcrm
mmmrmmrmgmrroomw

ap-55
QQQQQLQQAAQTAGAGAAGTTTCTGGACTCAAAEG!caGCTTTGCCAAACAGWACTACAA

Cctgcgctaccgggccaaggtcatc;cagacgtgctgggagggggggaggcggccgtgga
ap-51

cttcagcaactaccatgcGCTGAACAGCGTGGAAGAGAGEGAAA!TAECTTOGACCTCCT

GCGGGACTTTCAGGACAACCTGCAGGCCGAGAAGEGCGTCTTCAAEGECAEGAAGAAGTT

GTGCTTCCTGCTGcaggccgaccgcatgagcctagggagg;acaggggggggaacggcat
ap-54
cgcagagctggccacccggctcttcaacgtccacaaggatgctgtgctcgaggagtgcct
ggtggcgcctgactcgga9A!CGTGTTCCCCCTGGACAEGGGAGEGGEGGGCCARGITGC
GCTCTCTAAAAAGAICGTCAACGECCCCAACACAGAGGAGGAEGAACAETTCTGEGACT!

TGTGGACACCCTCACAGAGWACCAGACTARGAACAECTTGGCTTCTCCCataatgaatgg

gaaggatgtggtggccataatcatggttgtgaataaagtggatgggccccacttcaccga
gaatgatgaagazzgigttctcaagtacctcaattttgcaaaTCTAAECAIGAAGGTGEI
CCACCTGAGETACCTGCACAACTGTGAGACTCGGCGEGGCCAGREACTGETGEGGECTGG
GAGCAAAGTCTTTGAAGAGCTTACAGACA!TGAGAGGCAGETCCACAAAGCCCTGEACAC
AGTCCGCGCCTTCCTCAACTGTGACAGAEACTCTGEGGGGCTCTTAGACAEGACCAAACA
GAAGGAAITTTTTGAIG!GmGGCCAGECCTGAEGGGGGAGGCTCCACCctacgctggtcc
caggactccggatggaagggaaatcaactttt <----1340 nt----> agaaacaag
ap-40
gcggatgaactccctaagcttcaggtcggcttcatcgACTTTGQTTGCACCTTTGECIAC
AAGGAA!TCTCCCGETTCCACGAGGAGRECACTCCCAIGCTGGAIGGGAITACCAACAAC
CGCAAGGAGTGGAAAGCGCTTGCTGAEGAGTAEGAGACCAAGAEGAAGGGGCTGGAGGAG
GAGAAGCAGAAACAGCAGGCAGCCAACCAAGCAGCCGCAGGAAGTCAACacggggggggg

ggacctqqngcqggcctchtccaaqtcttqctchtccggtagcagagcgaggaccca
sp-58

Figure 9

60

size and large number of inserts. However, the HindIII and BamI-II maps
were determined both by double digestion, and using a method of cos
site labelling (64). This analysis confirms that the 6.2 kb HindIII
fragment identified in the bovine genome is wholly represented in the
isolated genomic clone (see Figure 10C).

Analysis of the Bovine cGMP PDE a-Subunit gene, To assess the
number of genes responsible for these RNAs we carried out several DNA
blotting experiments. The intensity of hybridization of a 1.3 kb cDNA
probe (probe 4, Fig 7) to EcoRI and HindIII fragments of genomic DNA,
compared to EcoRI and HindIII fragments of varying amounts of an
isolated genomic clone representing 0.1 to 4 genome equivalents,
suggests that the gene is present in single copy (Figure 10B, C,
respectively). Additionally, three cDNA probes representative of N-
terminal and C-terminal sequences (probes 1, 3, 6, Figure 7) hybridize to
single genomic DNA fragments in EcoRI, BamHI, or PstI digested DNA
(only the BamHI digests are shown, Figure 1 1, lanes 1, 5, 7), consistent
with the gene copy number data. Therefore, a single locus in the bovine
genome encodes all of the identified transcripts.

A minimum size for the gene was obtained by summing the sizes of
BamI-II fragments identified with representative cDNA probes (Figure l 1).
fiagments that are identified by more than one cDNA probe due to
overlap were only counted once. From such analysis we estimate the
bovine gene to be at least 160 kb and is likely to be much larger. This
estimate is likely a minimum value since any restriction fragments that
contain only intron sequence would not be detected, and the cDNA
clones do not represent the entirety of the RNA molecules identified.
However, we cannot formally rule out the possibility that some of the

61

Figure 10: Copy Number of the Bovine oc-Subunit Gene. A. A restriction
map of the bovine genomic clone used for the copy number
determination is shown. The Sail restrictions sites are provided from the
cloning vector 7t EMBL—3. All other sites shown represent sites found in
the isolated genomic clone. The order of the EcoRI restriction fragments
is tentative due to the number of small fragments. B. Genomic DNA
[lanes R (EcoRI) and H (l-IindIII)] and the )- genomic clone (lanes 0.1-4)
were digested with EcoRI and HindIII, electrophoresed into a 0.8%
agarose gel and transferred to a nylon membrane. Hybridizations were
performed with cDNA probe 4 (Figure 7). Lanes 0.1-4 contain amounts
of the A genomic clone DNA equivalent to 0.1-4 copies, respectively of a
6.2 kb fragment that would be present in the 10 ug of genomic DNA
loaded onto the gel.

62

 

 

A Transcription
13.5
1* fl: Sal I
9 5 4.0

: BamHl

0.7
1.2 6.2 2.1 2.8 -

+ : : : + Hind III

 

3.9 :1.7 : :12: :1.3: :1.3: 1.7 ECORI

 

 

Figure 10

63

Figure l 1: Size of the Bovine a-Subunit Gene. Bovine genomic DNA
isolated from a single brain was digested with BamHI, EcoRI, Pstl or
HindIIl, electrophoresed into a 0.8% agarose gel and transferred to a
nylon membrane. Blot strips containing the DNAs were hybridized with
oligo labelled cDNAs. The results of hybridization with the BamHI
digested DNAs are shown. The probes used for each blot strip were as
follows: lane 1, probe 3: lane 2, probe 5; lane 3, probe 4: lane 4, probe 8:
lane 5, probe 1; lane 6, probe 2: lane 7, probe 6: lane 8, probe 7: lane 9,
probe 9. All probe designations are according to Figure 7.

64

HH muzmflm

 

65

probes are detecting fragments that represent closely related genes.
§§ctlon 3 Human cGMP PDE (IL-Subunit.

Isolation of Human cGMP PDE a-Subunit cDNfi. A human retinal
cDNA library constructed in A gth was screened with a bovine oc-subunit

cDNA probe (probe 4, Figure 7 and 12) representing about one half of the
protein coding region. Five cDNA clones designated HPA-2 through HPA-
6, were recovered that contained only partial sequence for the a-subunit
(Figure 12). Additional small EcoRI fragments in clones HPA-5 and HPA-
6, which presumably hybridized to the probe were not analyzed, since it
was clear that these clones were not full length. In order to isolate a full
length cDNA clone, the library was rescreened with an Oligonucleotide
representing the first 12 amino acids of the bovine a-subunit sequence
(Figure 6A). Twenty clones were recovered from a screen of 2 x 104
recombinants. The complete sequence was determined for all of the
cDNAs shown in Figure 12. The sequence of specific primers used for
sequence analysis is shown in Table 3. No sequence differences were I
observed in any of the overlapping regions of the cDNAs.

Characterization of Human oc-Subunit cDNAs. One of the clones,
designated HPA-l (Figure 12), contains the entire open reading frame of
human cGMP PDE oc-subunit. The sequence of this 2,577 nucleotide
open reading frame and the 78 nucleotides of 5’ noncoding sequence and
249 nucleotides of 3' untranslated sequence is shown in Figure 13. A
poly (A) stretch of 17 nucleotides is present at the 3’ end of the clone,
however no consensus poly (A) site was found. The underlined sequence

near the 3’ end of the clone (Figure 13) may serve as a poly (A) site or it is

66

Figure 12: Restriction Map of Humg cDNA Clones, A schematic map of
the full length human cDNA clone is shown at the top. Wider rectangles

represent protein coding segments and more condensed rectangles depict
upstream or downstream untranslated regions. Included within the
protein coding region is the "PDE HOMODOGY" domain that defines a
segment conserved in a number of CN PDEs (see text). Predicted amino
and carboxy termini are designated NH2 and COO' respectively. Bovine
cDNA fragments (probes 4 and 5, Figure 7) that were used as probes to
isolate some of the cDNAs and for chromosomal localization are shown
as thick darkened lines. An EcoRI restriction map (sites designated R) is
shown with the full length clone designated HPA—l. Other partial cDNA

clones are shown as line segments.

67

. 3E: T

m..<n=._ .r

 

v.55: .F

 

NH musmfim

 

 

d—

 

...-fifim ”WWW-.mufiwr >0040§OI man—
. //////////////// “mm“...mfifi H... a... a ..

 

 

 

-..i AAxx/xx/A//A//A xii:

 

 

 

 

68

possible that the poly (A) stretch represents an internal sequence. An
identical poly (A) segment is found at the ends of clones HPA-5 and HPA-
6.

The cDNA sequence predicts a polypeptide chain of 859 residues
(MW ~ 100,000), significantly larger than that predicted from its mobility
on SDS gels (20). Negative charges are in excess and appear to be
uniformly distributed (calculated pI=5.0). No apparent leader peptide was
detected in the predicted amino acid sequence, consistent with the
existence of the PDE alpha subunit as an unsecreted, soluble protein. It
is possible that the N-terminal Met is cleaved in the mature protein, and
Gly (nucleotide 2, Figure 13) is acetylated, as has been shown for the
bovine oc-subunit (29).

Blot Analysis of Human Retinal RNA. To evaluate the transcription
pattern of the gene in the adult retina we used a 1.4 kb human cDNA
probe (HPA-3, Figure 12) to hybridize to human poly (A)+ RNA.
Transcripts of 5.3 and 4.9 kb were identified consistent with the size of
the predicted open reading frame (Figure 14). The size and relative
abundance of these RNAs is similar to the two predominant RNAs seen in
bovine retinal poly (A)+ with a-subunit probes (Figure 8, lanes 1-6). The
basis for the two transcripts and their differences in abundance in either
organism has not been determined.

Chromosomal Localization of the Gene. A 1.5 kb cDNA probe was
hybridized to a panel of human-rodent somatic cell hybrid DNAs digested
with EcoRI to determine the location of the human cGMP PDE a-subunit
gene. A representative blot is shown in figure 15, and the results of all
blot data are summarized in table 4. We observed 100% concordancy

with chromosome 5 and at least three discordancies were found for all

69

Figure 13: Nucleotide and Deduced Amino Acid Sequence of Human
QQMP PDE a-subunit. Nucleotide numbering is shown on the left of the

sequence and amino acids are numbered above each tenth residue. In
frame stop codons that define the open reading frame are denoted by
asterisks. Internal EcoRI sites and a putative nonconsensus
polyadenylation signal are underlined. A solid box is present above
residues at the C-terminus that are potential sites of palmitylation (see
text).

121

241

361

481

601

721

1081

1201

1321

1441

1361

1681

1801

1921

2041

2161

2281

2401

2521

2641
2761
2881

70

1 11

' I C I V T 4 I E V I I F L D
TCCCTCTTCTCTCCAAAACTCCC4TTCCAACCCCTCCCTCCC4CTCCTTCT4C4C4CT4CCC4CTCCC4TTCCC4CCC4TCCCCCACCTC4CACC4C4CC4CCTCC4C44CTTCCTCC4C
21 31 41 51

8 I 1 C F 4 I Q Y Y I L I Y I 4 I L 1 S D L L C 4 I I 4 4 V 0 F I I Y I 8 P S S
TCC44T4TTCCCTTTCCCAAAC4CT4CT4C44CCTCC4CT4CCCCCCC44CCTC4TCTCCC4CCTCCTTCCGCCC44CC4CCCTCCCCTCC4CTTC4CC44CT4CC4CTCCCCC4CC4CC
61 71 81 91
I I I S E 1 1 F 0 L L I D F Q E I L Q T I I C 1 F I V I I I L C F L L Q 4 0 I I
4TCC4CC4CACCC444TC4TCTTTC4TCTCCTCCCCC4CTTTC4CC4C44TTT4C4C4C4C4C444TCC4TCTTC44TCTC4TC44C446CTCTCCTTCCTCCTCCACCCAC4CCCC4TC
101 111 121 131
8 L F I Y I T I I C I P I L 4 T I L F I V I I 0 4 V L I D C L V I P D Q I l V F
4CCCTCTTC4TCT4CCCCACCCCCAATCCC4TCCC4C4CCTCCCC4CC4CCCTTTTC44TCTCC4C44CC4TCCTCTCCTCC4CC4CTCCCTCCTC4TCCCCC4CC44C4C4TCCTCTTC
141 151 161 171
P L D I C 1 V C I V 4 I 8 I I 1 4 I V P I T I I D E I F C 0 F V D 1 L T E Y I T
CCTTTCC4C4TCCCCATCCTCCCCC4TCTCCC4C4CTCT4404454TTCCT44CCTCCCC44C4C4C4CC4CC4TCACC4TTTCTCTCACTTTCTCC4C4TCCTC4CAC4CT4C44C4CC
181 191 201 211
I I 1 L 4 S P 1 I I C I 0 V V 4 1 1 I 4 V I I V 0 C 8 I F T I I 0 I I 1 L L I Y
44C44C4TCTTCCCTTCCCCC4T44TC44TCCC44CC4TCTCCTCCCC4T44TC4TCCCTCTC44T444CTCC4TCC4TCCC4CTTC4CC44C4C4C4TC44C4C4TTCTTCTC44CT4C
221 231 241 251
L I F 4 I L 1 I I I Y I L 8 Y L I I C I T I I C Q 1 L L I 8 C 8 I V F I I L T D
CTC44TTTTCC444TCT44TCATC44CTCCT4CC4CCTC4CTT4CCTCC4C44CTCTC444CTCC4CCTCCCC4C4T4CTCCTCTCCTCTCCC4CC444CTCTTTC44C44CTT4CCC4C
261 271 281 291
1 I I Q F I I 4 L Y T V I 4 F L I C D I Y I V C L L D I T I Q I I F F D V I P V
4TCC44CC4C4CTTCC4C444CCCCTCT4C4C4CTCCCTCCTTTCCTC44CTCTC4C4C4T4CTCTCTCCCTCTCTT4C4C4TC4CCAACC4C446€44TTTTTTC4TCTCTCCCCCCTT
301 311 321 331
L I C I V P P Y S C P I T P D C I I 1 I F Y I V 1 D Y 1 L I C I E D 1 I V 1 P I
CTC4TCCCTC44CTTCCACCTT4CTCTCCTCCC4CCACTCCCC4TCCAAC4C444TTAACTTTT4C44CCTC4TTC4CT4C4TCCTCC4TCCC444C4CC4C4TC444CTC4TCCCC44T
341 331 361
P P P 0 I I 4 L V S C L P T Y V 4 Q I C L 1 C I 1 I I 4 P 8 I 0 F F 4 F Q I I P
CC4CCTCCTC4CC4TTCCCCTTT4CT4ACCCCTCTCCCC4CTT4TCTTCCCCAC44TCCCCTC4TTTCC44C4TC4TC44CCCCCCTTC4C4CC4CTTTTTTCC4TTTC46444C44CCT
381 391 401 411
L D I 8 C I I 1 I I V L 8 I P I V I I I I I 1 V C V 4 T F Y I I I D C I P F D I
CTCCATCACTCTCCATCC4TCATTAAAAATCTCCTTTC44TCCCC4TTCTC44C44C44CC44C444TTCTTCC4CTCCCC4CATTTT4C44TCC7444C4TCCC44CCCCTTTC4TC44
421 431 441 451
I 0 I T L I C S L T Q F L C I S V L I P D T Y I 8 I I I L I I I I D 1 F Q 0 1 V
4TCC4TC4C4CCCTC4TCC4CTCTTTC4CTC44TTTCTCCCCTCCTCTCTCTT444TCCTC4C4CCT4TC4CTC44TC44T444CTTC444474CC446C4T4TTTTCC4CC4C4T4CT4
461 471 481 491
I Y I V I C 0 I I I 1 Q I 1 L I T I I V Y C I I P I I C I I E I L 4 I 1 L Q 4 I
444T4TC4TCTC44€TCTC4C44TC44C444TTC4C4444TCTTCAAAACC4C4CACCTCT4TCCC4ACC4CCC4TCCCACTCTC4CC44C4CCACCTCCCTC4C4TCCTCCAACCCCAC
501 $11 321 531
L P D 4 D I Y I 1 I I F I F S 0 L P L T I L I L V I C C 1 Q I Y Y I L I V V 0 I
CTCCCACATCC4C4T444T4CC444TT44T444TTTC4CTTC4CTC4CTT4CCCCT44C4C44CTCC4CCTCCT4444TCTCC44T4C4C4TCT4TT4TC4CCTC444CTCCTCC4T444
541 331 561 571
F I 1 P Q E 4 L V I F I Y S L S I C Y I I 1 T Y I I I I I C F I V C Q T I F S L
TTTCACATTCC4C44C4CCCCCTCCTCCCCTTC4TCT4CTCCCTC4CT44600CT4CCCC44C4TCACCT4CC4C44CTCCCCCC4CCCCTTC44CCTCCCCC4C4CC4TCTTCTCCCTC
581 391 601 611
L V T C I L I I Y F T D L E 4 L 4 I V T 4 4 F C I 0 1 D I I C T I I L Y Q I I S
CTCCTC4CCCC444CCTC44CCCCT4CTTC4CCC4CCT4C4CCCCTTCCCC4TCCTC4CTCCTCCTTTCTCCC4TC4C4TTC4CC4C4CACCC4CC44T44CCTCT4CC4C4TC444TCC
621 631 641 631
Q I P L 4 I L I C 8 8 1 L E I I I L I F C I T L L I 0 I 8 L I 1 F Q I L I I I Q
C4C44CCC4CTCCCC44CCTCC4TCCCTCCTCT4TCTTCC444C4C4CC4CTTCCACTTTCCC4444C4CTCCTC4CAC4CC4C4CCCTC44T4TCTTTC4444CCTC44TCCTCC4C4C
661 671 681 691
I I I 4 1 I I I D 1 4 1 1 4 T 0 L 4 L Y F I I I T I F Q I 1 V 0 Q 8 I T Y I S I
C4TC4CC4TCCC4TCC4C4TC4TCC4C4TTCC44TC4TTCCC4C4C4CCTCCCCCTCT4TTTC44644C4CC4CC4TCTTCC4444C4TCCTCC4TCACTCT44C4C4T4TC4C4CTC44
701 711 721 731
Q I I T Q Y I I L I Q T I I l 1 V I 4 I I I T 4 C D L 8 4 1 T I P I I V I S Q V
C4CC4CTCC4C4C4CT4C4TC4TCCTCC4CC4C4C4CCC44CC444TCCTT4TCCCC4TC4TC4TC4CCCCCTCTCATCTCTC4CCC4TC4CC444CCCTCCCACCTCACCACCC4CCT4
741 751 761 771
4 L L V 4 4 I F I I Q C 0 L I I T V L Q Q I P 1 P I I 0 I I I 4 D I L P I L Q V
CCTCTCCTCCTCCCTCCTC44TTCTCCC44C44CCTC4CCTCC4CCCC4CCCTCCTCC44C4C44TCCC4TTCCC4TC4TCC4C4C444C444CCAC4TC44CTCCCT446CTTCAACTC
781 791 801 811
C F 1 D F V C T F V Y I I F S I F I E I 1 T P I L D C 1 T I I I I E I I 4 L 4 D
CCCTTCATTC4CTTTCTTTCC4CCTTCCTCTACAACC44TTCTCCCCTTTCC4CC4CC4C4TC4CCCC44TCTTCC4CCCC4TC4CC44C44TCCC440C4CTCC44CCCCCTTCCTCAT

821 831 841 851
I Y 0 4 I I I V Q I I I I Q I Q Q 8 4 I 8 4 4 4 C I Q P C C I Q P I C 4 T T 8 I
C4CT4CC4TCCC44C4TCAACCTCCACC4CC4C44C44CC46444C4CCACTCCCCC44CTC4CC4CCCCCACC444TCAII- "‘”C4""C‘CCC'TCC44CT4C4TCC44C

 

- a”
s c c 1 a -
mmuacucrccccammamucuccmmmmmwwummmrw
ummmmcmcmflwrmmeccrcccccaummmmurwtamuuanmrwm
MW

Figure 13

71

Figure 14: Analysis of Human Retinal RNA. Approximately 2 ug human
poly (A)+ RNA was fractionated on a 1.0% agarose formaldehyde
containing gel and transferred to nitrocellulose. The results of
hybridization with radiolabelled human cDNA clone HPA-3 (Figure 12) is

shown.

72

 

Figure 14

73

Figure 15: Blot Analysis of Human-Rodent Somatic Cell Hflgrid DEA,
Approximately 15 ug of genomic was digested with EcoRI, fractionated on

a 0.8% agarose gel, and transferred to a nylon membrane. The blot was
hybridized with radiolabelled probe 5 (Figure 7). Lanes designated
human and mouse are from control cell lines that contain only human or
mouse chromosomes respectively. Lanes designated hybrids (see Table
4) contain from left to right the following human—mouse somatic cell
hybrids: NIH-74, NIH-18, P-12, NIH-41. Arrows mark the positions of
human DNA fragments that segregate with the cGMP PDE a—subunit

gene.

74

i . 1.

a; . 5.; .32,

.... 9.41:} CI; .. . ..... .

I

g1. b.5314,
.n _.

 

Figure 15

75

Table 4: Segregation of cGMP PDE oc-Subunit Gene in Human-Men];
§omatic Cell Hybrids. The complement of chromosomes present in each

somatic cell hybrid versus the segregation of the cGMP PDE a-subunit
gene is shown. A plus sign designates the presence of the chromosome
and the minus sign its absence. Abnormal karyotypes identified in some
of the hybrids are explained below the table. The presence of hybridizing
human DNA for each hybrid is denoted by a plus sign in the column
labelled PDE“, and by a minus sign to indicate the absence of
hybridization to human DNA.

76

.uouvnmammfl new nouvnmuwpa mcfimuou ..mfim“m~vvamaupavu

.m.m~vum.mampa mcdmuou

.uouvumavm can nauvsaflmsa mcfimuou ..Hamu-w.hflnm.u

.uouvnmavma can nouvu~.HHmFH maamuou .A~.Hflmumfivv.pflumavu

.uxou on» ca Uonfiuomou an nkunus HAM How com: am: An unaudhv m onoum «zoo

.muflunaa omnoalamaan ohm anon on» new undunan noumamn omocwnolamasn cum hIH nkunam

.Am.m«wum.mam..sfivuawu

MIIOTJO‘d

 

mm Hh mm am am am mm HN mv HN m am am mv mm mm mv mm mm o

I
1
+
l
l
1
+
I
I

I+I+

I
I +-+I
I +-+¥1+

I+++l+
l+l+l+ m
I+++|
u++|

II

II
u+|+u+
l++++l

I+l++l+

04-34-01140
+

l+l+++l+
l+l+l+|

l
I
4.
I
I
I

n+u+n+++++

I
I++0I+++I

I
II+
I++l
l1
1++l
|+
ll
u++|

I
+
l
+
I
l

1
++++++++l
+
1
1
I
+
I
1+|+I

l
I
I
I
I
1
l
I
+
I
+
I
I
I
l
|++1

4.
I

1
1
I
1
I
1
I
l
I
l
1
+
I
l
+

mc mm am am aoamouooudo »

«min: Av“
avumzvfima
«bum: Am”
«Hum .HH
«Hum: .oH
pun: Am
mudm Am
I u u «.na .s
+ u n Imam
Ha.pmz .m
u u u v.H .v
u n n Hm.mm= .m
. H~.H .a
n u u u u ~m-.H .I

|++I
|+l++

|++++I

l
n++++|+
+
I

l+l+ll
I+I+1
l l
l 1
1 I

 

>4
X
N
N
H
N

ON ma ma 5H mH ma vH MH NH HH OH m m
«cacaoeouno amaam

1‘
\O
10

v m u I swam mmuaunam

 

 

 

 

4mfifiuamm Hamu owummdm usanomlmmaam a“ 0:00 uwcsndmld mom

 

mama MoldOMummmmmmw "v wanna

 

77

other chromosomes. In order to delimit the assignment further we
probed 1 1 human-hamster hybrid DNAs (not shown) that contain ‘
various deletions of chromosome 5. Two of these hybrids are deleted for
the majority of 5q. The other nine hybrids contain intact long arms and
various deletions of 5p. Hybridization signals were observed in all of the
hybrid DNAs except the two hybrids deleted for 5q. Therefore the gene
resides on the long arm of chromosome 5 within the region 5q1 l-qter.
Since this is the first CN PDE gene to be localized, the locus
designation CGPR-A is suggested. This designation takes into account
the classification system of Beavo (120) for CN PDEs and is in agreement
with the recommendations of the committee on cytogenetics
nomenclature (47). The locus name denotes the gene encoding the cGMP

phosphodiesterase oc-subunit of the rod photoreceptor cell.

SectLon C Similarities Between Bovine and Human a-Subunits.

Comparison of Human and Bovine a-Subunits. Shown in Figure

16 is a comparison of the predicted amino acid sequences for the human
and bovine a-subunits. The human and bovine cDNA sequences are
88% identical over their matched lengths, and 93% identity was found at
the protein level. The majority of the nonhomologies appear to be
concentrated in the first ~24O amino acids at the N-terminus, and the
last 43 amino acids at the C-terminus of the proteins. The region
identified as the "conserved domain" (see below) in several eukaryotic CN
PDEs (bracketed in Figure 16) is contained within the region that is the
most highly conserved. Interestingly, two cysteine residues at the C-
terminus (residues 856 and 857) are present in both species, suggesting

78

Figure 16: Comparison of Human and Bovine a—Subunit Primm
Mug; Amino acid alignment of the predicted primary structures of

the bovine and human cGMP PDE (II-subunits is shown. The bracketed
region (522-780) roughly defines the region that is conserved in many CN
PDEs (see Figure 17). Dicysteine residues (856-857) that are conserved
in the two sequences and are potential sites of palmitylation are boxed.

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

hu-an
bovine

human
bovine

human
bovine

79

HGEVTAEEVEKFLDSNICFAKQYYNLHYRAKLISDLLCAKEAAVDFSNYBSPSSHEESEI 60

VS R V PR ALI V

lFDLLRDFQENLQTEKCIFNVHKKLCFLLQADRHSLFHYRTRNCHPELATBLFNVHKDAV
D A V A IA

LEDCLVHPDQEIVI-‘PLDHCIVCIIVAHSKKIANVPIITEEDEHFCDFVDILTEYKTIOIILAS
e A s v L v 1' Q

PIHNCKDVVAIIHAVNKVDCSHFTKRDEEILLKYLNFANLIHKUYHLSYLHNCETRRCQI
V P EN VF

LLHSCSKVFEELTDIERQFHKALYTVRAFLNCDBYSVCLLDHTKQKEFFDVHPVLHCEVP
A

PYSCPRTPDCREINPYKVIDYILHCKEDIKVIPNPPPDHHALVSCLPTYVAQNCLICNIH
A

NAPSEDFFAFQKEPLDESCHHIKNVLSHPIVIKKEEIVCVATPYNRKDCKPFDEHDETLH
V

ESLTQFLCHSVLIPDTYESHNKLENRKDIFQDIVKYHVKCDIEEIQKILKTREVYGKEPH
L H T

ECEEEELAEI LQAELPDADKYEI NKFII PS DLPLTELELVKCQQHYYELKVVDKFB I PQE
C

ALVRFHYSLSKCYRKITYHNHRHGFNVGQTHFSLLVTCKLKRYFTDLEALAHVTAAFCHD
l

IDHRCTNNLYQHKSQNPLAKLHCSSILERHHLEFCKTLLRDESLNIFQNLNRRQHEBAIH

mum rammvrmmronvnqsmessomqmmommmmn
79 o

120

180

240

300

360

620

480

540

600

660

720

LSAI‘I'KPIIMSQVALLVAAEl-‘HEQCDLBRTVIQQNPI manoeumven ME 780

or

crmxzrsmemrmncrmnxmmmoemomosusmeu
er c1. 8 A IIQ s

QPCCIQPRCA‘I'TSKSEIQ 859
II K c on v

Figure 16

840

80

that there may be some selective pressure to maintain this sequence (see
Discussion).

Homologies to Nonretinal Cyclic Nucleotide PDEs. Beavo and
collaborators (50) have identified a significant region of similarity that is
present in several CN PDEs from diverse species. This region has been
termed the "conserved domain" and is thought to represent at least part
of the catalytic region of the enzymes. This work establishes that a
number of CN PDEs with very differing properties are likely to have
arisen from a common primordial ancestor. The cGMP PDE a-subunit
also contains the conserved domain, and therefore is included as a
member of the CN PDE gene family. A comparison of amino acid
sequences of the identified CN PDEs is shown in Figure 17. While the
average percentage identity is only about 35% between enzymes
ofdifferent classes, there are a number of amino acids that have been
maintained through several hundred million years of evolution.
Additionally, there is a high degree of conservative substitutions.

81

Figure 17: Similagties Amongst cGMP PDE on-§_ubunit and 1119mm
CN PDEs. Manually optimized alignments for many of the eukaryotic

cyclic nucleotide phosphodiesterase primary structures reported to date
showing only the region of homology and some flanking sequence. Some
of the sequnces have not been determined in their entirety within this
region. Asterisks denote stop codons that define the C-termini in some
of the sequences. Bov Ret on and Human Ret a represent conceptualized
amino acid sequence from the bovine and human cGMP PDE a-subunits
determined in this study, respectively. Bov cG-S, Bov CaM, Yst cA and
Dro cA sequences are from Charbonneau et aL (120). Bov cG-S is from a
cGMP stimulated PDE identified in bovine brain. Bov CaM is from a
Ca++/calmodulin dependent PDE isolated from bovine heart. Rat dnc
sequence is from a homologue of the Drosophila dance PDE (Dro cA)
deduced from cDNA (R. L. Davis, Y. Takayasu, M. Eberwine, and J.
Myres, unpublished results). Yst cA is from a yeast cAMP PDE deduced
from cDNA. Mouse and Human partial sequences were deduced from
genomic clones isolated by virtue of homology to the dunce locus (Y. Qiu
and R. L. Davis, unpublished results). Numbering is according to the
bovine a-subunit sequence (Figure 6A). Gaps denoted by "-" have been
introduced to optimize alignments.

82

Via"
ISII
I..."
DI—LI
ILLI-
SID”
LIIV
ISII
IQQT
III-IP-
LIIL
.1...le
IVYV-
CAI].
CV1"

UCCQQQ

ITS
LLL
III

I
I

A D I V I I I I F I F S D L P L T I L I L V I C C
A D I Y I I I I I I F S D L P L T I L I L V I C C

\I
m
Vial-0.1""
QITQITT
IIIILII
I'BI'II
LIICIII
..-..II
.....V.V.
.....Iulu
IIII‘A‘
”Inn-Ll»..-
DDDDDDD
TTTTTTT
II‘III‘
LII-LL]...
IVVVIII
.A.IUI.I.I_A_A
[DUI-DI]
IIIIIDD
IVVVII"
IIILLHI
.‘IUDI'
.IIILII
.LLLH‘I
.T'D..'
DQQIQl—I
IIIU'u'
IQQDDQQ
IIIDIII
IIIIIII
sass-3'.
IILLILL
Slllull
LCQIDQQ
‘PILF'F
LIIVIII
Ion...”
ICCICLL
”Clan-cs8
Ivor—III!-
III-£100
LQQQIII
LLLILLL
Qlulululalul—
QIIIA—l—l
III-Val“
LACAACC
IVV‘QIP
IAA‘AII
.LLVILL
Hanan.“

COO-.08....

Yet cA

Dre eA

lat due

Dov CaI

Dov eC-I
lov let a
uuaan let a:

ILLK
III:
LII—u
1’."
Il»“
TTTA
VD."
.l.‘
III:
cal-LI
DDDnm
AAAAI
ACCAT
‘fll.°
IVVLD
lat-III
LIILL
I'm-so
Ill-“c
LVVTI
TQQIT
IIIAI
QII‘L
IDDDL

S V

S V
P I C L - - - - - - -
T I I Q I I I L L L C

A
T
Q
I
I Q I I V D Q S I T Y I T Q Q I I T Q T I I L D Q T I I I I V I A I I I T A C D L I A I T I

anmmm
mmmmm

P I I V (730)
.L I I L
P I I V

I Q I I V D Q 8 I T Y I 8 I Q I I T Q Y I I L I Q T I I I I V I A I I I T A C D L I A I T I

Iov let a

lunch let a:

Accccc
CQQQQQQQ

vertornn)

cc

DLQQLSEI

“Fr-PIP.

FIE-Fr???
EIIEIEEE

Inf-“II
Tull“ a
.V.

-...-
1L1]!-
LLIIA
V‘DDI
AVTTT
UUUU'
IIQQI
III."

5'. 0V»!—

A!
A"

LT.
LI
‘0
V8

.'

I
Q
Q 8 Q - V A L L V A A I F I I Q C D L I I T V L Q Q I P I P I I D I I I I D I L P I L Q V C I I D I

IuIII let a:

\l

s

2

a
It... I:
‘0‘ [I
1...: LQ
6" 6'
s!!- ..
IDI— II
“V: I”
QC!- II
II.- TA
uh" ID
III TV.
I" t!
I." an
[IQ IA
VI... lulu
‘11. A‘
III II
II: “U
.10" I!
I'V- II
I." I.

run 0"

III . o'.
Illmmrf
QED—[III
CIIII—cc
Ital-1'00
PTTLLLL
IDDPl-"H
"Isl—1.9...
'lII‘TT
HDDIAII
nunwnv-nrnr.r.
IAIEIII
IUD-III"
I'D-Slur,
DIIDDII
DVVTQSS
PLLLLII
”SDI-LII—
TIASIII
v.0 HWI.V.Y Y
ITTTIVV
VII. . .-
LUH. . .-
ILL. . .-
I’D-PIPI-
QIHIITT
IVVVACC
‘IIIIVV

Figure 17

DIUIN

Sequence of Iggptic Peptides. Seventeen peptide sequences were
obtained from a tryptic hydrolysate of intact purified cGMP PDE. Since

the a and B subunits of the cGMP PDE comprise 94% of the total mass of
the protein, most, if not all, of the peptides were likely to represent
sequences from these subunits. In fact, none of the peptide sequences
could be found in the reported amino acid sequence of the y—subunit (22),
indicating that the sequence of the tryptic peptides originates from either
the a or B subunit. The entire protein, rather than isolated subunits,
was cleaved with trypsin because it was not possible to isolate sufficient
quantities of well separated a and B subunits. Of the seven peptide
sequences obtained (Figure 3A), only three were found (PDEl, PDE3.
PDE6) that showed identity to sequences present in the a-subunit
(Figure 6A).

Peptide PDE5 sequence is found in the (II—subunit sequence,
however residues 2 and 3 are juxtaposed. It is possible that this
anomaly represents a true strain variation reflecting the difference in the
source of cattle used to isolate the protein (26), and the source of cattle
used to isolate the RNA for cDNA library construction (14), but technical
error is also a possible explanation. This is supported by the fact that
the sequence presented in this study predicted from the cDNA is
identical to that published for the a-subunit (29). Additionally, the same
sequence is found in the predicted human a-subunit sequence (Figure

83

84

13). The sequence difference explains the failure to isolate cDNA clones ’
(Figure 4C) with the corresponding oligomer (cPDE5, Figure 7B).

No amino acid determination could be made for residue 7 in this
peptide, however residues 8 and 9 yielded amounts of PTH derivatives
comparable to that obtained for residue 6. The determination of amino
acid sequence depends on the consistency of expected elution profiles for
each amino acid in an HPLC chromatograph. Each PTH derivative is
mixed with a standard amino acid solution. The residue determination is
made essentially by comparison of the HPLC profile of the mixture to a
profile of standards only. For a homogeneous peptide the
chromatograph should show only one difference, corresponding to the
elution time of the determined residue. The failure to identify residue 7
in peptide PDE5 suggests that it is modified in some manner, which
affected the normal elution profile for a histidine residue. This may be of
significance to enzyme function. The participation of specific active site
histidines in catalysis is well documented (135, 136). Moreover, post-
translational modification of a specific histidine involved in catalysis has
been shown to occur (137).

Peptides PDE2 and PDE4 sequences also are closely related to
sequences found in the a-subunit, differing only at residues 3 and 4,
respectively (see Figures 3A. GA). No closely related sequence of peptide
ABl, however, could be found in the oc-subunit sequence. It is possible
that these peptides represent sequence corresponding to the B-subunit.

Design of Oligonucleotide Probes. Six oligonucleotides were
synthesized reflecting the amino acid sequence of the tryptic peptides
(Figure 3B). There are several possible codons for many of the amino

acids in the peptide sequence. Therefore, in order to reduce the number

85

of oligonucleotides necessary to reflect all possible combinations of codon
sequences, deoxyinosine was incorporated into the third position of every
codon with three or four degeneracies. Initial reports suggested that
deoxyinosine behaved inertly as a base analog (79), and reports emerged
that used oligonucleotide probes containing deoxyinosine at the "wobble"
position of each codon in the oligonucleotide (78, 138). The melting
temperature of the hybrid formed with these probes was empirically
determined to be equal to the melting temperature of a sequence without
the deoxyinosine. However, recent thermodynamic studies using model
Oligomers, conducted after the oligonucleotides reported here were
synthesized, unambiguously demonstrated that deoxyinosine does not
behave in an inert fashion, and that surrounding sequence greatly
contributes to the overall stability of a deoxyinosine base pair (157, 158).
This suggests that the melting temperature cannot be predicted for an
oligonucleotide probe containing deoxyinosine even in the presence of
TMAC as the hybridizing salt. Therefore, when screening the cDNA
library with oligonucleotides containing til, it had to be assumed that the
deoxyinosines would behave in a relatively inert fashion and would not
reduce the stability of the hybrid by more than 5°C. The low number of
clones identified with oligonucleotide probe cPDE4 (Figure 4C) may have
resulted if the hybrid formed was unstable, or more likely it was due to
the difference in amino acid sequence observed for the corresponding
peptide compared to that deduced from (II-subunit cDNAs. While this
study provides additional evidence that oligonucleotide probes containing
deoxyinosine residues can be used successfully to obtain clones of
interest from It libraries, their use should be restricted until continuing
studies on their stability (157) are completed.

86

Oli onucleotid reenin an Se n e An 1 f B vin
EMS... Radiolabelled oligonucleotides, described above, were used in
conjunction with the TMAC hybridization method to screen a bovine
retinal cDNA library. Two of the oligonucleotides, cPDE3 and cPDE6,
were successful in isolating cDNAs that encode PDEs. Seventeen clones
were isolated that show hybridization to both oligonucleotide probes (not
shown). All but one of these clones contains 662 and 144 bp EcoRI
fragments. Additionally, six of these clones contain other EcoRI
fragments. This result can be explained in several ways. A number of
cDNA fragments that derive from the same transcript may have been
isolated. Alternatively, the similar DNA fragments may be the result of
alternative processing events (142,143) from a single gene. It is also
possible that a number of transcripts derived from a family of related
genes may be represented. Several experiments were conducted to
distinguish these possibilities.

Sequence Analysis of cDNA Clones. Subsequent restriction
analysis of the six cDNAs demonstrated heterogeneity, suggesting that
sequence analysis would be necessary to uncover the basis of the
heterogeneity. To facilitate rapid sequence analysis the use of specific
oligonucleotide primers was employed using primers that were evenly
spaced (~300 bp) on both strands of the cDNA (Fable 3). Additional ‘
primers were synthesized as needed to sequence through troublesome -
areas. The use of sequence specific primers proved to be a very elegant
and expeditious method to generate nucleotide sequence. Over 20 kb of
DNA sequence was determined manually in less than nine months.
Another major advantage of this strategy is that primers can be used, not
only for closely related bovine cDNAs produced by a single gene, as

87

demonstrated here, but also to analyze genes from other mammals. The
primers were successfully used to sequence the human counterpart of
the bovine cGMP PDE a-subunit (Figure 13). Surprisingly, 17 base
Oligomers with as many as three mismatches were used successfully
(Table 3, sp33, and sp34).

Identification of cGMP PDE a-subunit cDNAs, The sequence
analysis of the bovine cDNA clones predicts the existence of five distinct
classes depicted in Figure 5. The long stretch of identity in the isolated
cDNAs (~1.6 kb), coupled with the fact that the CN PDE conserved
domain (see Figure 17) is contained within this sequence establishes that
all of the isolated cDNAs do represent PDEs. Comparison of cDNA
sequences with a published sequence for the a-subunit (29) suggested
that one of the classes represented the cGMP PDE a-subunit.

However, the results reported did not provide sufficient evidence to
support that the identified overlapping clones did indeed represent the
entire molecule. Only one cDNA clone was found that contained the C-
terminal coding region of the predicted polypeptide. Based on the results
presented here it was possible that the predicted C-terminal region
represented an a-subunit like RNA. Several experiments were conducted
to verify the sequence of the oc-subunit. Results obtained by direct
dideoxy sequencing of bovine retinal RNA (Figure 9) were in complete
agreement with the composite sequence determined from clones 7 , 8,
and 15 (Figure 6A), and the reported sequence of the a-subunit (29).
Oligonucleotide probes representing the C-terminus and the 3’
untranslated region of the a-subunit (probes C, D, Figure 7) hybridize to
RNAs that appear to be of identical size and abundance (Figure 8, lanes
5, 6). The same RNAs are detected using long cDNA probes or other

88

oligonucleotides near the N -terminus (Figure 8, lanes 1-3). Further
verification of the a-subunit sequence is provided from comparison to an
isolated full length human cDNA clone (HPA-l, Figure 12). There is a
marked degree of homology that extends into both the 3’ and 5’
untranslated regions. These results firmly establish the authenticity of
the a-subunit sequence presented here (Figure 6A), and that reported
(29).

Identification of a-Subunit Variang, Four closely related a-
subunit like variants are also predicted to exist based on sequence
analysis (Figure 5). Only one cDNA was isolated representing each of the
variants suggesting they are present in low copy number in the bovine
retina. However, since the cDNA analysis concentrated on only those
clones that hybridized to two oligonucleotide probes, it was possible that
many shorter cDNAs representing the variants went undetected.
Oligonucleotide probes representing clones Pct-1, Pct-2, POI-3 detect RNAs
that are present at much lower abundance. As shown in Figure 8, lanes
1-6 represent a three hour exposure without any signal intensification,
and lanes 7—9 show RNAs detected with Pct-l, -2, -3 specific probes,
respectively, after a 54 hour exposure with two intensifying screens. A 5-
10 fold increase in signal per time results from the use of intensifying
screens. Therefore, the predicted variants do appear to represent lower
abundance RNAs.

The hybridization patterns observed with oligonucleotide probes
(Figure 8, lanes 2-1 1) are thought to represent specific hybridization.
Spurious hybridization to non RNA material is extremely unlikely. Even
in lane 1 l , in which the hybridization appears less convincing, there is a
streak through the hybridizing band where no probe has hybridized.

89

Each lane shown in Figure 8 represents a nitrocellulose strip cut after
blotting. The membrane containing blotted human RNAs was
mismeasured such that each strip contains a portion of two adjacent
lanes of RNA. The lack of hybridization represents the space between
lanes where no RNA is present. On a shorter exposure of the results
shown in lane 10 this streak is also apparent (not shown).

The absence of spurious hybridization does not rule out
nonspecific hybridization to related sequences. This is unlikely, since
even with a 27 base oligomer with a melting temperature significantly
lower than that of the 50 base Oligomers, the hybridization pattern is the
same (compare lane 4 with lanes 2, 3, 5, and 6 or with a cDNA probe,
lane 1). Therefore it is concluded that as many as seven RNAs are
identified using clone specific probes for bovine retinal RNA, and that
analagous RNAs exist in the human retina, representing at least POI-1
and Pct-2.

The ability to detect transcripts in the bovine retinal RNA
population with clone specific oligonucleotides does not provide sufficient
support for the existence of the predicted RNAs. A number of steps were
involved in the construction of the bovine cDNA library (144, 145),
leading to the possibility of isolating artifactually contrived cDNAs. The
most common types of cloning artifacts found are 5’ end rearrangements
(146, 147), or the joining of unrelated cDNAs during the ligation of cDNA
to vector DNA (148). We cannot rule out the possibility that the 1.5 kb
EcoRI fragment in clone Pom—4 (Figure 5) does result from the joining of
unrelated cDNAs. These types of artifacts, however, cannot account for
the organization of Pa-l, POI-2, and POI-4 since the 3' most EcoRI

fragments, where the divergence is found, also contains sequence

90

common to the (II-subunit. Likewise for the 5’ EcoRI fragment containing
the start point of divergence in Pct-3, a-subunit sequence .is present.
Nonetheless, it is still important to demonstrate that the clones do
represent intact retinal RNAs, and not novel exceedingly rare types of
cloning artifacts.

Several lines of evidence support the authenticity of the cDNAs.
The cDNA library was made from polysomal RNA (14). The isolation
method separates very dense polyribosomes from other less dense
cellular material. Only mRNAs that are actively being translated should
be isolated by this procedure. This does not rule out the possibility of
obtaining artifactual cDNAs, but does essentially eliminate the chance of
cloning nuclear unprocessed or partially processed RNAs.

Pa- 1 and POI-2 diverge at the same nucleotide (compare in Figure
6A. B, C,) and 29 amino acids are encoded past this point in both cDNAs
before a stop codon is encountered. It is extremely unlikely that two
clones would be truncated at their 3’ ends at the same nucleotide, and
that both of these clones would pick up unrelated cDNAs presumably by
blunt end ligation. The chances become astronomical that ligation of the
two unrelated cDNAs would create organizations that fortuitously lead to
coding potential for 29 C-terrninal amino acids past the point of
divergence. Moreover, oligomers 50 nucleotides in length, that are
specific for POI-1 and POI-2 detect transcripts in bovine and human retinal
RNA (Figure 8, lanes 7, 10, and 8, l 1, respectively). The POI-1 transcripts
appear to be of very similar size and abundance. Thus, it is expected
that Pom-1 and POI-2 cDNAs do represent true bovine retinal “RNAs, that
can be incorporated into a stable polyribosome complex.

91

Pct—3 specific oligonucleotide probes also detect transcripts in
bovine retinal RNA, however no transcript was detected in human retinal
RNA (not shown). This result may be due to the very low abundance of
the RNAs, which is compounded with the technical problem of high
background unexpectedly occurring during the analysis (Figure 8, lanes
10, 1 1). The best evidence in support of the predicted organization of Pa-
3 is from DNA blotting analysis. A cDNA probe representative of the 5’
end EcoRI fragment (probe 8, Figure 7) hybridizes to fragments of «2.8
kb and 6.0 kb in HindIII and BamHI digested genomic DNA, respectively
(BamHI digest is shown, Figure 1 1, lane 4). Identical size fragments are
observed (autoradiographs are superimposable) using the adjacent EcoRI
fragment in POI-3 as a probe (not shown). Fragments of identical size are
identified in BamHI digested DNA with probes 4 and 5 (Figure 1 1, lanes
2 and 3) consistent with the overlap of the probes (Figure 7). The
isolation of two random cDNAs that hybridize to bands of identical size in
genomic DNA from two separate digests is extremely unlikely. This
results then suggests that the two fragments are linked and therefore
supports the organization of POI-3.

The sequence of POI-4 predicts the possibility of alternative splicing
in the 3’ untranslated region. Whfle alternative splicing is well
documented (reviewed in 142, 143), there are a paucity of reports that
describe its occurrence in this region (149, 150). This may be because
most studies of cDNA analysis have concentrated on the determination of
the open reading frame and such events have therefore gone unnoticed.
Or, it may be that these events occur in relatively few transcripts of
specific genes, and may represent an important mode of regulation. It is
recognized that the 3’ untranslated region contains a signal that specifies

92

polyadenylation (1 56), and it has been suggested that alternative splicing
permits differential use of multiple poly (A) sites (143). Furthermore,
splicing of upstream exons may in part be regulated by downstream
splicing events in genes that exhibit tissue specific expression (143). The
recent finding of a gene that produces a transcript apparently devoid of 3’
noncoding sequence (151) increases the necessity to better understand
the importance of this region. Further analysis of POI-4‘ may yield
important information in this regard.

Possible Functions of Predicted Isms, There are several
potential functions for the predicted isozymes. It is possible that all of
the predicted polypeptides are expressed only in the rod cell as has been
suggested for the a-subunit (30). If this is so, it is possible that some of
the isozymes subserve more housekeeping type functions and may be
localized in other rod cell compartments. The relative abundance of the
predicted isoforms is more consistent with that observed for other CN
PDEs (120). The identification of a COM? hydrolyzing PDE in the IPM
consisting of smaller subunits (102) is intriguing. Perhaps the POI-3
protein product represents this PDE. Another possibility is that one of
the cDNAs represents the B-subunit. While a portion of a putative B-
subunit sequence has been reported (29), confirmation of the sequnce
has not been obtained.

The possibility that the isozymes are expressed in other retinal cell
types or even in tissues outside the retina cannot be ruled out. Other
retinal PDEs that hydrolyze cGMP have been identified. A high affinity
form in the inner plexiform layer (24) and a cone outer segment PDE
have shown to exist (31, 100). Both of these forms are present in greatly
reduced amounts in comparison to the a-subunit. It is interesting to

93

speculate that the oc-subunit gene may be responsible for producing both
highly specialized and housekeeping proteins, a phenomenon that has
not, to my knowledge, been observed.

Analysis sf tips Bgvins g—Subunit ane, In order to understand the
origin of the multiple transcripts, experiments were conducted to identify
the number of genes that encode them. Both genome reconstruction
(Figure 10) and DNA blot analysis (Figure 1 1) are consistent with a single
bovine gene. The ability to detect segregation of only one chromosome in
somatic cell hybrids (see Table 4) is consistent with a single human gene
as well. Alternative splicing and at least two promoters are likely to be
operative at the loci to explain the observed transcript heterogeneity.

The determination of a single locus allowed an estimation of the
size of the gene by simply adding up the sizes of hybridizing restriction
fragments identified with cDNA probes (Figure 1 1). During the analysis
there was one unexpected observation. Probe 1 (Figure 7) does not
hybridize to any fragments identified with adjacent probes 2, 4, or 6.
This is only possible if intron junctions are present near each end of the
region represented by probe 1. Such an arrangement is supported by
the failure to detect any common hybridization pattern in I-IindIII or Pstl
digested DNA also (not shown). Further characterization of the gene is
required to unambiguously verify this hypothesis.

Analysis of me Human cGMP PDE gag-subunit, The implications
that a defective cGMP PDE could be the primary cause of certain forms of
hereditary retinal degenerations in mammals (see Literature Review),
makes the gene a good target for the candidate gene approach. In this
approach one studies a gene in which a defect is implicated as a

causative factor of a disease. The first step then would be to isolate and

94

characterize the gene or a representative cDNA. Once isolated the cDNA
or gene fragment can be used to identify the chromosomal location for
comparison to disease loci already known (139). The second step is to
analyze large kindreds with well established pedigrees for linkage to the
disease of interest. A very high lod score (140) is expected (lod >> 3) with
a recombination fraction close to zero if a defect in the gene is
responsible for the disease. This method of study is particularly
amenable to the study of diseases that occur in well studied, highly
specialized tissue such as the retina. This approach complements the
more conventional linkage studies in which random probes are used in
an attempt to narrow down the location of the disease gene. The
candidate gene approach has the advantage of allowing relatively fast
assessment of the involvement of a defect in a specific gene to hereditary
disease. Additionally, it guarantees that important information will be
obtained at the very least from analysis of the cDNA or gene, which is not
true for conventional linkage analysis.

The completion of the first step for our candidate, the cGMP PDE
a-subunit is presented here. We have isolated a full length cDNA (Figure
12), and we have localized the gene to the long arm of chromosome 5 (5q:
Table 4). Completion of the second step is presently underway in
collaboration with investigators at the Berman Gund and Howe ‘
Laboratories of OphthalmologI at Harvard Medical School. While we
were unable to find any retina specific disorders that have been mapped
to 5q, two disease genes have recently been localized in the region.
Familial polyposis coli (FPC) is a disease that predisposes afiected
individuals to the formation of multiple adenomatous colonic polyps
(152). Interestingly, Gardner Syndrome which is clinically very difficult

95

to distinguish from FPC, has been associated with congenital
hypertrophy of the RPE (153). Additionally, a susceptibility locus for
schizophrenia has been mapped to 5q (154). It is difficult at present to
envision a causal relationship between these disorders and defects in the
cGMP PDE a-subunit gene. It is more prudent to consider the possibility
that the a-subunit gene may be more closely linked than the presently
available markers and thus be of importance in the study of these
disorders.

Comparison of CN PDEs. The high degree of homolog' observed in
comparison of the bovine and human (II-subunits is consistent with
homologies that have been shown for other phototransduction proteins
(15,141). It was somewhat unexpected, however, that the a-subunits
share a region of homology with a number of nonretinal CN PDEs (155).
This homology found in PDEs representing many different classes (120)
suggests that the majority of CN PDEs and perhaps all mammalian PDEs
may make up a conserved PDE gene family.

The molecular genetics of PDE genes has only recently begun to be
unravelled. The best studied gene is the Drosophila dunce gene which
encodes a cAMP PDE. Over the past ten years a wealth of information
has been disseminated about this extraordinarily complex locus (132- I
134). The complexity has made studies of the gene structure a very
arduous task and therefore incomplete. The partial characterization of
the bovine cGMP PDE a-subunit gene presented here exhibits a
strikingly similar complexity. That is, a single large gene appears to
encode multiple RNAs, that are likely to arise via the use of alternative
splicing and multiple promoters. Further studies of PDE genes may

96

indeed show that such complexity is a feature common to all PDEs
within this gene family.

LIST 01" REFERENCES

10.
11.

12.

13.

REFEREN

Kaupp, U. B. and Koch, K.-W. (1986) ’D‘ends Biochem. Sci. 11, 43-
47.

Dartnall, H. J. A., Bowmaker, J. K. and Mollon, J. P. (1983) in
Colour Vision Physiology and Psychophysics, Academic Press, New
York.

Altman, J. (1985) Nature 313, 264-265.

Bok, D. (1982) Methods in Enzymology 81, 763-782.

Johnson, L. V., Hageman, G. S. and Blanks, J. C. (1985) in The
Interphotoreceptor Matrix in Health and Disease, Alan R. Liss, Inc.,
New York.

Schwartz, E. A. (1985) Ann. Rev. Neurosci. 8, 339-367.

Liebman, P. A., Parker, K. R., Dratz, E. A. (1987) Annu. Rev.
Physiol. 49, 765-791.

O’Brien, D. F. (1982) Science 218, 961-966.

Hurley, J. B. (1987) Annu. Rev. Physiol. 49, 793-812.

Stryer, L. (1986) Ann. Rev. NewoscL 9, 87-1 19.

Dratz, E. A. and Hargrave, P. A. (1983) Trends Biochem. Sci. 8,
128-131.

Findlay, J. B. C. and Pappin, D. J. C. (1986) Biochem J. 238, 625-
642.

Aton, B. R. (1986) Biochemisz 25, 677-680.

97

14.

15.
16.

17.

18.

19.
20.

21.

22.

23.

24.

25.

26.

27.

98

Nathans, J. and Hogness, D. S. (1984) Proc. Natl. Acad. Sci. USA
81, 4851-4855.

Nathans, J. and Hogness, D. S. (1983) Cell 34, 807-814.

Dartnall, H. J. A. (1957) in The Visual Pigments, Methuen and Co..
Ltd., London.

Emeis, D., Kuhn, H., Reichert, J. and Hoffmann, K. P. (1982)
FEBS Lett. 143, 29-34.

Fung, B. K.-K., Hurley, J. B. and Stryer, L. (1981) Proc. Natl. Acac.
Sci. USA 78, 152-156.

Gilman, A. G. (1984) Cell 36, 577-579.

Baehr, W., Devlin, M. J. and Applebury, M. L. (1979) J. Biol.
Chem. 254, 1 1669-1 1677.

Hurley, J. B. and Stryer, L. (1982) J. Biol. Chem. 257, 1 1094-

1 1099.

Ovchinnikov, Y. A., Lipkin, V. M., Kumarev, V. P., Gubanov, V. V.,
Khramtsov, N. V., Akhmedov, N. B., Zagranichny, V. E. and
Muradov, K. G. (1986) FEBS Lett. 204, 288-292.

Wilden, U. and Kuhn, H. (1982) Biochemistry 21, 3014.
Zuckerman, R., Buzdygon, B., Philp, N., Liebman, P. and
Sitaramayya, A. (1985) Biophys. J. 47, 37a.

Kuhn, H., Hall, S. W. and Wilden, U. (1984) FEBS Lett. 176, 473-
478.

Kohnken, R. E., Eadie, D. M. and McConnell, D. G. (1981) J. Biol.
Chem. 256, 12510-12516.

P. Deterre, J. Bigay, F. Forquet, M. Robert, M. Chabre (1988) Proc.
NatL Acad. Sci. USA 85, 2424.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

99

Hurwitz, R. L., Bunt-Milam, A. H., Chang, M. L. and Beavo, J. A.
(1985) J. Biol. Chem 260, 568-573.

Ovchinnikov, Yu. A., Gubanov, N. V., Khramtsov, K. A., Ischenko,
V. E., Zagranichny, V. E., Muradov, K. G., Shuvaeva, T. M. and
Lipkin, V. M. (1987) FEBS Letters 223, 169-173.

Hurwitz, R. L., Bunt-Milam, A. H., and Beavo, J. A. (1984) J. Biol.
Chem. 259, 8612-8618.

Gillespie, P. G., and Beavo, J. A. (1988) J. Biol. Chem 263, 8133-
8141.

Beavo, J. A., Hardman, J. G., and Sutherland, E. W. (1971) J. Biol.
Chem. 246, 3841-3846.

Egrie, J. C., and Siegel, F. L. (1977) Biochim Biophys. Acta 483.
348-366.

Hikada, H., and Asano, T. (1976) Biochim. Biophys. Acta 429, 485-
497.

Klotz, U., and Stock, K (1972) Naunyn-Schmiedeberg's Arch.
Pharrnacol. 274, 54-62.

Martins, T. J., Mumby, M. C., and Beavo, J. A. (1982) J. Biol.
Chem 257, 1973-1979.

Franks, D. J ., and Macmanus, J. P. (1971) Biochem Biophys. Res.
Comm 42, 844-849.

Sharma, R. K., and Wang, J. H. (1986) J. Biol. Chem 261, 14160-
14166.

Sharrna, R. K., Adachi, A. M., Adachi, K., and Wang, J. H. (1984) J.
Biol. Chem 259, 9248-9254.

Hansen, R. S., and Beavo, J. A. (1982) Proc. Natl. Acad. Sci. USA
79, 2788-2792.

41.

42.

43.

44.

45.

46.
47.

48.
49.

50.

51.

52.

100

Hansen. R. S., and Beavo, J. A. (1986) J. BioL Chem 261, 14636-
14645.

Conti, M., Kasson, B. G., and Hsueh, A. J. (1984) Endocrinology
114, 2361-2368.

Weber, H. W., and Appleman, M. M. (1982) J. Biol. Chem 257.
5339-5341.

Thompson, W. J ., Epstein, P. M., and Strada, S. J. (1979)
Biochemistry 10, 31 1-316.

Landt, M., and Butler, L. G. (1978) Biochemistry 17, 4130-4135.
Anderson, T. R., (1982) Mol. Cell. Endocrin. 28, 373-385.

Shows, T. B., McAlpine, P. J ., Boucheix, 0., Collins, F. S.,
Conneally, P. M., Frezal, J ., Gershowitz, H., Goodfellow, P. N ., Hall,
J. G., Issitt, P., Jones, C. A., Knowles, B. B., Lewis, M., McKusick,
V. A., Meisler, M., Morton, N. E., Rubinstein, P., Schanfield, M. S..
Schmickel, R D., Skolnick, M. H., Spence, M. A., Sutherland, G.
R., 'h'aver, M., Van Cong, N., and Willard, H. F. (1987) Cytogenet.
Cell Genet. 46, 1 1-28.

Saltiel, A. R., and Steigerwalt, R. W. (1986) Diabetes 35, 698-704.
Marchmont, R. J ., Ayad, S. R., and Houslay, M. D. (1981)
Biochem J. 195, 645-652.

Charbonneau, H., Beier, N ., Walsh, K. A. and Beavo, J. A. (1986)
Proc. Natl. Acad. Sci. USA 83, 9308-9312.

Berger, S. L., and Kimmel, A. R. (1987) in Guide to Molecular
Cloning Techniques, Academic Press, Inc. San Diego, CA.

Lee, R. H., Lieberman, B. S., Lolley, R. N. (1987) Biochemistry 26.
3983-3990.

53.

54.
55.

56.
57.

58.

59.

60.

61.

62.

$83

66.

101

Pittler, S. J ., Kozak, L. P. and Wilson, J. E. (1985) Biochim.
Biophys. Acta 843, 186-192.

Gracy, R. W. (1977) Met. Enz. 47, 195-204.

Maniatas, T., Fritsch, E. F. and Sambrook, J. (1982) in Molecular
Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, NY.
Dumais, M. M., and Nochumson, S. (1987) Biotechniques 5, 62-67.
Labarca, C. and Paigen, K. (1977) Proc. Natl. Acad. Sci. USA 74.
4462-4465.

ChirgWin, J. M., Przybla, A. E., MacDonald, R. J ., and Rutter, W. J.
( 1979) Biochemistry 18, 5294-5299.

Pittler, S. J ., Salz, H. and Davis, R. L. (1987) Mol. Gen. Genet, 208,
315-324.

Botstein, D. and Davis, R. W. and Roth, J. R. (1980) in A Manual
for Genetic Engineering: Advanced Bacterial Genetics, Cold Spring
Harbor, NY.

Holmes, D. S., and Quigley, M. (1981) Anal. Biochem 114, 193-
196.

Hogan, B., Costantini, F., and Lacy, E. (1986) in Manipulating the
Mouse Embryo: A Laboratory Manual. Cold Spring Harbor
Laboratory, Cold Spring Harbor, N.Y.

In Applied Biosystems Model 3808 Operations Manual (1985).
Baker, R. T., and Board, P. G. (1988) Nucl. Acids Res. 16, 1 198.
Sanger, F., Nicklen, S. and Coulsen, A. R. (1977) Proc. Natl. Acad.
Sci. USA 74, 5463-5467.

Biggin, M. D., Gibson, T. J. and Hong, G. F. (1983) Proc. Natl.
Acad. Sci. USA 80, 3963-3965.

67.

68.

69.
70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

102

Zhang, H., Browse, J ., and Sommerville, C. (1988) Nucl. Acids Res.
16, 1220.

Manfioletti, G., and Schneider, C. (1988) Nucl. Acids Res. 16.
2873-2884

Geliebter, J. (1987) Focus 9, 5-8.

Chen, S. H., van Tuinen, P., Ledbetter, D. H., Smith, L. C., and
Chan, L. (1986) Somatic Cell Molec. Genet. 12, 303-306.
Overhauser, J ., Beaudet, A. L., and Wasmuth, J. J. (1986) Am J.
Hum Genet. 39, 562-572.

Staden, R. (1984) Nucl. Acid Res. 12, 521-538.

Mount, D. (1986) Nucl. Acid Res. 14, 443-454.

Wray, W., Boulikas, T., Wray, v. P. and Hancock, R. (1981) Anal.
Biochem 118, 1907-2003.

Jaye, M., Salle, H., Schamber, F., Balland, A., Kohli, V., Findeli, A.,
Tolstoshev, P. and Lecocq, J.-P. (1983) Nucl. Acid Res. 8, 2325-
2335.

Koch, K.-W. and Kaupp, U. B. (1985) J. Biol. Chem 260, 6788-
6800.

Wallace, R. B., Johnson, M. J ., Hirose, T., Miyake, T., Kawashima,
E. H. and Itakura, K. (1981) Nucl. Acid Res. 9, 879-894.
Takahashi, Y., Kato, K., Hayashizaki, Y., Wakabayashi, T.,
Ohtsuka, E., Matsuki, S., Ikehara, M. and Matsubara, K. (1985)
Proc. Natl. Acad. Sci. USA 82. 1931-1935.

Ohtsuka, E., Matsuki, S., Ikehara, M., Takahashi, Y. and
Matsubara, K. (1985) J. Biol. Chem 260, 2605-2608.

80.

81.

82.

83.

84.

85.

86.
87.

88.

89.

90.

91.

92.
93.

103

Bilofsky, H. S., Burks, C., Fickett, J. W., Goad, W. B., Lewitter,
F. I., Rindone, W. P., Swindell, C. D. and Tung, C.-S. (1986) Nucl.
Acid Res. 14, 1-4.

Lochrie, M. L., Hurley, J. B. and Simon, M. I. (1985) Science 228,
96-99.

Lolley, R. N., and Lee, R. H. (1984) Adv. Cycl. Nucl. Prot. Phos.
Res. 17, 315-327.

Yamaki, K., Tsuda, M., and Shinohara, T. (1988) FEBS Lett. 234,
39-43.

Hollyfield, J. G., Raybom, M. E., Ulshafer, R J ., and Fliesler, S. J.
(1985) in Retinal Degeneration: Experimental and Clinical Studies,
Alan R. Liss, Inc., N. Y.

Wald, G. (1968) Nature 219, 800-807.

Bruckner, R (1951) Document. Ophthalm 5-6, 452-554.
Pittler, S. J ., McConnell, D. G., and Davis, R. L. (1987) Am J. Hum
Gen. Suppl. 41, A234.

Palczewski, K., McDowell, J. H., and Hargrave, P. A. (1988) J. Biol
Chem 263, 14067-14073.

Benovic, J. L., Mayor, F., Jr., Staniszewski, C., Leikowitz, R. J .,
and Caron, M. G. (1987) J. Biol Chem 262, 9026-9032.

Kuhn, H., Hall, S. W. and Wilden, U. (1984) FEBSLett. 176, 473-
478.

Wistow, G. J ., Katial, A., Craft, C. and Shinohara, T. (1986) FEBS
Lett. 196, 23-27.

Tuteja, N., and Farber, D. B. (1988) FEBS Lett. 232, 182-186.
Hingorani, V. N., and Ho, Y.-K. (1987) FEBS Lett. 220, 15-22.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

104

Blatt, C., Eversole-Ciri, P., Cohn, V. H., Zollman, S., Fournier, R.
E. K., Mohandas, L. T., Nesbitt, M., Lugo, T., Jones, D. T., Reed, R.
R., Weiner, L. P., Sparkes, R. S., and Simon, M. I. (1988) Proc. Natl.
Acad. Sci. USA 85, 7642-7646.

Danciger, M., Kozak, C. A., and Farber, D. B. (1989) Genomics 4,
215-217.

Griswold-Prenner, L, Young, J. H., Yamane, H. K., and Fung, B.
K.-K. (1988) Invest. Ophthalmol. Vis. Sci. Suppl. 29, 218

Danciger, M., Tuteja, N., Kozak, C. A., and Farber, D. B. (1989)
Exp. Eye Res. 48, 303-308.

Deterre, P., Bigay, J ., Robert, M., Pfister, C., Kuhn, H., and
Chabre, M. (1986) Proteins Struct. Phnc. Genet. 1, 90-99.
Amatruda, T. T., Gautam, N., Fong, H. K. W., Northup, J. K., and
Simon, M. I. (1988) J. Biol. Chem 263, 5008—501 1.

Pannbacker, R. G. and Lovett, K. (1977) Invest. Ophthalmol.

Visual Sci. 16, 166-168.

Sparkes, R. S., Klisak, L, Kaufman, D., Mohandas, 1., Tobin, A. J .,
and McGinnis, J. F. (1986) Curr. Eye Res. 5, 797-798.

Barbehenn, E. K., Wiggert, B., Lee, L., Kapoor, C. L., Zonnenberg,
B. A., Redmond, T. M., Passonneau, J. V. and Chader, G. J. ( 1985)
Biochemistry 24, 1309-1316.

Zuckerman, R. and Cheasty, J. E. (1988) FEBS Lett. 238, 379-384.
Fleischman, D. (1982) Met. Enz. 81, 522-526.

De Vries, G. W., and Ferrendelli, J. A. (1982) Met. Enz. 81, 517-
522.

Koch, K.-W., and Stryer, L. (1988) Nature 334, 64-66.

107.

108.

109.

110.

111.

112.

113.

114.
115.

116.

117.

118.

1 19.

120.

105

Fukui, H., and Shichi, H. (1981) Arch. Biochem Biophys. 212, 78-
87.

Kuhn, H., and Hall, S. W. (1986) Eur. J. Biochem 161, 551-556.
Berger, S. J ., De Vries, G. W., Carter, J. G., Schulz, D. W.,
Passoneau, P. N., Lowry. O. H., and Ferrendelli, J. A. (1980) J. Biol.
Chem 255, 3128-3133.

Lolley, R. N., Brown, B. M., and Farber, D. B. (1977) Biochem
Biophys. Res. Comm 78, 572-578.

McGinnis, J. F., and Leveille, P. J. (1985) Curr. Eye Res. 4, 1127-
1 135.

Lee, R. H., Brown, B. M., and Lolley, R. N. (1982) Biochemistry 21.
3303-3307.

Lee, R. H., Brown, B. M., and Lolley, R. N. (1984) Biochemistry 23.
1972-1977.

O'Brien, P. J ., and Zatz, M. (1984) J. Biol. Chem 259, 5054-5057.
Ovchinnikov, Y. A, Abdulaev, N. G., and Bogachuk, A. S. (1988)
FEBS Lett. 230, 1-5.

Molenaar, C. M. T., Prange, R., Gallwitz, D. (1988) EMBO J. 7, 971-
976. .
Clarke, S., Vogel, J. P., Deschenes, R. J., and Stock, J. (1988)
Proc. Natl. Acad. Sci. USA 85, 4643-4647.

Pittler, S. J ., McConnell, D. G., Baehr, W. and Davis, R. L. (1988)
J. Cell Biology, 107, A3638 (replacement).

Drummond, G. 1., and Perrot-Yee, S. (1961) J. Biol. Chem 236,

1 126- 1 129.

Beavo, J. A. (1987) Adv. Sec. Mess. Phosphoprot. Res. 22, 1-38.

121.

122.

123.

124.

125.

126.
127.

128.

129.

130.

131.

132.
133.

134.

106

Davis, R. L., and Kauvar, L. M. (1984) Adv. Cycl. Nucl. Prot. Phos.
Res.16, 393-402.

Lee, R. H., Navon, S. E., Brown, B. M., Fung, B. K.-K., and Lolley,
R. N. (1988) Invest. Ophthalmol. Visual Sci. 29, 1021-1027.

Noell, W. K. (1958) A. M. A. Arch. Ophthalmol. 60, 702-733.
Farber, D. B., Flannery, J. G., Bird, A. C., Shuster, T., and Bok, D.
( 1987) in Degenerative Retinal Disorders: Clinical and Laboratory
Investigations, Alan R Liss, Inc., N. Y.

Heckenlively, J. (1988) in Retinitis Pigmentosa, J. B. Lippincott,
Co., Pennsylvania.

Kuziel, W. A., and Tucker, P. W. (1986) Nucl. Acid Res. 15, 3181.
Aguirre, G., Farber, D., Lolley, R, Fletcher, R T., and Chader, G.
J. (1978) Science 201, 1 133-1 134.

Woodford, B. J ., Liu, Y., Fletcher, R T., Chader, J. G., Farber, D.
B., Santos-Anderson, R, and Tso, M. O.-M. (1982) Exp. Eye Res.
34, 703-714.

Leon, A., Hussain, A. A., and Curtis, R (1988) Inv. Opthalmol. Vis.
Sci. Suppl. 29, 143.

Feinberg, AP. and Vogelstein, B. (1984) Anal. Biochem 137, 266-
267.

van Tuinen, P., Rich, D. C., Summers, K. M., and Ledbetter, D. H.
(1987) Genomics 1, 374-381.

Davis, R L., and Davidson, N. (1986) Mol. Cell. Biol. 6, 1464-1470.
Chen, C.-N., Malone, T., Beckendorf, S., and Davis, R L. (1986)
Nature 329, 721-724.

Chen, C.-N., Denome, S., and Davis, R. L. (1986) Proc. Natl. Acad.
Sci USA 83, 9313-9317.

107

135. Topham, C. M., and Dalziel, K. (1986) Eur. J. Biochem 155, 87-
94.

136. Tomich, J. M., and Colman, R F. (1985) Biochim Biophys. Acta
827, 344-357.

137. Chen, J. Y., and Bodley, J. W. (1988) J. Biol Chem 263, 1 1692-
l 1696.

138. Miura, N ., Ohtsuka, E., Yamaberi, N., Ikehara, M., Uchida, T. and
Okada, Y. (1985) Gene 38, 271-274.

139. McKusick, V. A. in Mendelian Inheritance in Man. (1988) Johns
Hopkins University Press, Baltimore, MD.

140. Ott, J. in Analysis of Human Genetic Linkage (1985) Johns Hopkins
University Press, Baltimore MD.

141. Van Dop, C., Apone, L. M., and Medinsky, D. C. (1989) J. Cell
Biology 107, 296a.l39.

142. Leff, S.E., Rosenfeld, M.G. and Evans, RM. (1986) Ann. Rev.
Biochem 55, 1091-1 1 17.

143. Breitbart, R E., Andreadis, A., Nadal-Ginard, B. (1987) Ann. Rev.
Biochem 56, 467-495.

144. Buell, G. N., Wickens, M. P., Payvar, F., and Schimke, R T. (1978)
J. Biol. Chem 253, 2471-2482.

145. Wickens, M. P., Buell, G. N., and Schimke, R T. (1978) J. Biol.
Chem 253, 2483-2495.

146. Weaver, C. A., Gordon, D. F., and Kemper, B. (1981) Proc. Natl.
Acad. Sci USA 78, 4073-4077.

147. Shyam, E., Reddy, P., and Rao, V. N. (1985) Gene Anal. Techn. 2.
46-49.

 

 

148.

149.

150.

151.

152.

153.

154.

155.

156.

157.

158.

159.

108

Jones, D. T., and Reed. R R. (1987) J. Biol. Chem 262 14241-
14249.

Wieczorek, D. F., Smith, C. W. J ., Nadal-Ginard, B. (1988) Mol.
CelL Biol. 8, 679-694.

Brown, C., McHugh, K. M., and Lessard, J. L. (1989) J. Cell Biology
107,252a.

Bishop, D. F., Komreich, R, and Desnick, R J. ( 1988) Proc. Natl.
Acad. Sci USA 85, 3903-3907.

Leppert, M., Dobbs, P., Scambler, P., O’Connell, Nakamura, Y.,
Stauffer, D., Woodward, S., Burt, R, Hughes, J ., Gardner E.,
Lathrop, M., Wasmuth, J ., Lalouel, J .-M., and White, R (1987)
Science 238, 141 1-1413.

Lewis, R A., Crowder, W. E., Eierman, L. A., Nussbaum, R. L., and
Ferrell, R E. (1984) Ophthalmology 91, 916-921.

Sherrington, R, Brynjolfsson, J ., Petursson, H, Potter, M.,
Dudleston, K, Barraclough, B., Wasmuth, J ., Dobbs, M., and
Gurling, H. (1988) Nature 336, 164-167.

Barbehenn, E., Wiggert, B., Redmond, T., and Chader, G. (1985) in
The I nterphotoreceptor Matrix in Health and Disease, Alan R Liss,
Inc., NY.

Nevins, J. R (1983) Ann. Rev. Biochem 52, 441-466.

Martin, RH. and Castro, M.M. (1985) Nucl. Acid Res. 13, 8927-
8938.

Kawase, Y., Iwai, S., Inoue, H., Mirua, K. and Ohtsuka, E. (1986)
Nucl. Acid Res. 14, 7727-7736.

Swanson, R. J ., and Applebury, M. L. (1983) J. Biol. Chem 258,
10599-10605.

APPENDIX

109

Figure A1: Complete Sequence of cDNA clone POI-1. The sequence of
cDNA clone Poc-l is shown. Postion 1 in this sequence corresponds to
position 914 of the a-subunit sequence (Figure 6A). Internal EcoRI sites

are underlined so that relative positions can be oriented according to

Figure 7.

1113

FFFUFU<<UH
FU<U<PUOPU
UF<UPFUFU<
UUUDFU<UFU
UPF<U<<U<U
FDFUU<FDDU
<UFU<UU<FF
<0UUU<<U<<
UUUUU<<U<<
UU<<UU<UH<
Uh<UUhUUH<
P0h<b<<<hu
<PU<UUO<U<
UU<<UPFDUU
PU<hUF<U<U
U<<U<<U<<u
<00¢U<<<<U
<Uuuh<<u<h
PU<UO<U<UH
c<<<UUDh<U

h<UUhUhU<u
UHUPDUPUPH
UPUUP<<<<h
UPD<<UH<<U
U<Uhhhuu<<
<hUhUh<UU<
<h<hchu<oh
UU<PF<UUDE
<U<U<UUF<U
UUU<UFDPUH
U<<<<UUFPU
UU<U<UE<U<
U<UU<UPF<U

E085

hh<<UUPUh<
U<UUUE<<UU
th<<<U<<U
PPPUU<UUUO
PUUUOUH<<U
UhUUh<U<HU

u<<UHUUUUF
UU<UUUHUU<
U<<UU<UU<U
DFU<<UU<U<
PPPUOFFU<U
HUUUthUU<
U<<UC<UOUD
<UUFUUh<UD
P<UUUHP<UU
(UPUPUUUDU
F<UU<UU<<<
U<UFFUHU<U
<00<UUUHUH
<<U<UU<PDU
<<Uht¢F00<
UFUUDU<UU<
F<UFPU<UP<
FUOUUPFUPP
H<Uh<u<<uu
<Ufih<UFUU<

hUUhUbU<UU
th<UhUU<U
U<U<U<UUUU
UUFU<<UU<U
<<U<P<U<<D
Utuo<¢uu<<
U<UU<<UU<<
DbD<oh<u<u
UH<<U<U<<U
<Uh<Uh<UH<
<U<<Uhhh<h
<<<<h00hhh
FUUUUUUUU<
<UP<UU<UUD
UUHU<<UUU<
<UUUH<FUHU
hPD<U<UPUU
<<<UU<UU<<
hhh<UhUUDU
<U<hhhho<<

hb<<<h<<<u
UH<UHUU<UH
U<<thuuhu
OOUHFUU<DD
<<h<Uthhh
UUHFUUEPD<
unu<ouu<ou
u<uo<uohhh
UFUUHUUU<U
UUUUH<UHUH
UFPUOUHHUD
U<UUFFU<UU
UPUUh<UUUU
U<FUOUU<<F
UhUUUDUhUU
U<U<U<UD<<
h<<<PhUhUh
U<<U<<Uh0h
H<<<<UUUUH
Uh<<<ouu<<

hu<h<<<<h<
DHFDHHOHD<
<<Uh0<huu<
U<U<UU<OUD
u<<UhUhD<<
UhhththO

uo<huuuohu
<<UHF<UU<U
u<oh<<<<oh
FU<UU<<DHD
0<h<<<uo<<
U<U<<UUh<h

<u<<<u<uomnmu<oo<co<u

UUUUhohh<<
UUU<UUHUU<
h<<<uu<<uu
<U<D<h00hh
<U<U<<<UUH
hhuouu<00h
U<Uhouohh<
<UUU<UPPO<
<<Uhh0h<<u
UFUUEUUUUH
h<<DDUh<hU
h0h<hhu<oo
UUE<UUUDHD

DU<<U<FUEU
UhUU<<U<<U
U<D<U<Dh<u
¢0h<uuuhh<
HUH<OUFUUH
UD<D<U<UFP
huh<0hhuuo
00hh<<<h<<
<U<Uoh<<<°
OthU<UUU<
hthUhUh<<
DUHUHOUUU<
<UU<UUUHUU

HHDU<<FHU<
PUUUUHHUUU
UHH<UHU<<U
U<<Oh<9hoh
UOU<UHU<UO
<09h<0hohh
UHUUUUU<<U
PHHUU<UOHH
UUPUHH<<UH
UHDUH<UH<U
U<Uok<0h<u
UUUF<UUHUU
U<F<uuu<<u
UhuUhDUUUU
hh<<<Uh<h<
<<Uh<<u<0h
thhohu<uu
<<<hh<0h<o
Uhvchhhouv
huuu<houu<

 

u<<oh<u<0h
U<FUUH<UU<
HU<UU<<UHB
UU<U<DPOU<
UOFUUU<UHD
UDPUU<UUF<
h<u<<uu<u<
huhhho<00h
DUEDUHPUUH
<PU<UUU<UU
<UUP<<UUE<
<<UUUUHU<U
PUU<<<UUUU
<UU<UhuUhh
<<U<D<UUF<
UHU<<UPUU<
h<UhUUU<U<
Uh<UUhUhU<
Uhu<uu<u<u
UUHUUU<UUU

HHUUUU<UB<
o<<uohuu<u
<U<UH<UUUU
uou<u<<0hh
FOHUHHUDU<
thhU<U<h<
Uh<hu<ob<u
<UPFUUUUhU
DUPUFDOUHU
FU<UU<UU<D
DU<UF<UU<U
UU<<U<DUUh
(UFUUHUUPF
<U<UFHH<<<
U<UUOHUU<U
uo<h<<<uhu
UH<UUE<<<U
UH<UUFUFUU
UHUUHUUH<<
uuh<0huohu

s<ueeueh<o
u<hu<<uuuh
ssooouo<oe
<uuh<uu<ue
auas<uoueu
<u<<<ueueu
houseuuuuc
c<upsuo<<p
u<<uu<u<uu
eu<u<oo<p<
uu<uuue<<u
<<<oe<u<ou
uuhoeaue<u
e<uusouhu¢
<ou<<oueoo
ue<o<u<<uo
h<uee~uou<
u<u<<<u<ou
oouoe<uhu<
<ouousupuh

H4 wusmflm

Uhh<<u
UUUhUDU<bc
hh<UUthhh
thbCh<hU<
U<Uhuu<<u<
h<U<U<<UU<
cuc<UUUHUU
<<UUPD<UU<
UUUhU<<Uh<
PUU<UUUHPD
U<U<<PUHU<
Huu<<U<Uhh
<hUHUh<<U<
U<U<UUUUUh
<<UPUU<UP<
h<D<UhUUUh
hhhh<h<uu<
<<UUUF<U<<
hh<00hhhhh
<CUH<U<U<<
<9hhhhhh<<

.ch
pawn
dean
neon
-a~
doc—
~oo~
ncmq
nec—
nun—
ucwu
«co—
~00
nee
-~
uoo
«we
non
new
mun
n

u~Iem nee—u

 

 

111

Figure A2: Partial Sequence Determined for cDNA Cloris Psi-2. The
partial sequence of cDNA clone POI-2 is shown. The ~1.5 kb EcoRI

fragment at the 3’ end (Figure 5) was not completely sequenced in the
determined 3' untranslated region. Postion 1 in this sequence
corresponds to position 914 of the a-subunit sequence (Figure 6A).
Internal EcoRI sites are underlined so that relative positions can be
oriented according to Figure 7.

1112

<<<hUQHth
<00hhfl<uu<
<UUOU<<U<<
800050:
00<<UU¢UH<
OH<UUHUOH<
hUh<h<<<Ho
<h0<000<u<
DO<<UHHUOU

<U<OU<UHHU
Uhhhuu<00h
oo<hh<UUUh
<U<U<UUH<U
UUU<OhUhUh
u<<<<uohko
OU<U<UF<O<
U<UU<UPH<0
U<UUUUUUFO

huepuh<u<muma<<oopup<

u<<u<<u<<o
<UQ<U<<<¢U
<000h<<0<h
hU<UO<U<0h
u<<¢uuoh<0

Udoooh<<uu
hhb<<<u<<o
hhhuu<uooo
huuuouh<<u
9h00h<u<hu

u<u<<h0hhh
u<<uo<uouu
¢Uuh00h<uo
hduoohh<uo
<UHUFUUUUU
h<00<oo<<<
UdOhh0h0<o
<Uo<000h0h
<<U<00<h00
<40h00h0¢<
UHUUUU<UU<
h<0hhu<0h<
houuuhhohh
h<0h<u<<uo
duhh40huu<

...<<<<h
h0<00hhhhh
U<UU<<UU<<
UHUCUH<U<U
UH<<U<U<<U
<0h<oh<ch<
<U<<Uhhh<h
<<<<HO0Phh
BUOUUUUUU<
<Uh<00<000
ouho<<uou<
<UUOE<HUHU
hhu<ucohuu
<<<UU<UU<<
hhh<uhouuo
<0<hhhh0¢<

u<<hhhauu<
<h<<h<0<h<
UUU<UUU<UU
U<UU<UUHHH
UHUUPUUU<U
UUOUF<UHUF
Ubbuuurhoo
0<OUFFU<UU
Uh00h<uooo
O<houuu<<h
UhUUUUUhoo
U<U<U<UO<<
h<<<hh0h9h
U<<U<<UHUH
H<<<<UUUQB
Ub<<<uuo<<

<0h0h<<fi<h
UHHUU<UHUO

<0<<<U<UU<

UUUUHUHH<<
UOU<UUHUU<
h<<<uu<<uo
<U<U<900Hh
<0<U<<<UUH
hhuuuu<00h
U<UHUUUHH<
<Uuu<ofiho<
<<uhhoh<<u
UHDUHUUUUH
h<<UUUh<HU
hDh<hhu<ou
00h<0000h0

hhh¢<<uhhu
<00<U<Uh00
<U<UU¢UU<U
UU<<U<PUHD
UHUU<<U<<U
U<U<U¢0h<o
<Uh<uoohh<
BUH<UUHUOR
UU<D<O<Uhh
hDh<Uthou
UUHH<<<F<<
<U<Qoh<<<o
UUHHU<UUU<
thUUhUh<<
UOHUHUUUU<
<UU<UUUHUU

<0<UH<UU<<
<00<h<h<u<
UHUUUUU<<U
hthUduohh
UUHUHH<<UH
UPUUH<UH<U
U<UUH<UH<U
QUUH<UUHUU
U<h<uuo<<u
UHOOHUUUUU
hh<<<0h<h<
<<Uh<<o<0h
OHFPUHU<UU
<<<hh<oh<0
Dhu<hhhuuo
houu<hoou<

 

Uhhh<uh<hh
<UUUOU<UHU
B<U<<UU<U<
h0hhh0<00h
UUhUUhhuUH
<hU<UUU<UU
<00h<<uoh<
<<Uuoobo<0
PUU<<<OOUU
<oo<oh00hh
<<U<U<UQH<
OHU<<UPUU<
h<0hooo<u<
Dh<00h0hu<
UPO<UU<U<U
UUhUUU<UUU

th<Uh<<Uh
UU<<<UQUU<
Uh<hu<0h<u
<UthUUUhU
UUthUUUHU
PU<UO<UU<U
uu<0hcuu<u
UG<<U<UUUH
(UHOUHUUHH
<o<0hhh<<<
U<UUUHUU<U
UU<h<<<UhU
Uh<UUh<<<U
Oh<uohohuu
choohooh<<
UUH<QHUUHU

<<hhhhoh0h
<UU<UUU<UH
BUUHPUUUU<
U<UHPUU<<h
U<<UU<U<UU
HU<U<UU<R<
uo<oooh<<o
<<<UH<U<UU
UUBOFHUH<O
H<UOHUUBU<
<UU<<UUHUU
UH<U<U<<UO
h<UbkbUUU<
D<U<<<U<UU
OUUUH<UHU<
(UUUUHUPUH

m4 magmas

<00<Uhhh<<
UU<UDH<<<<
<<UUHU<UU<
UUUFU<<UH<
hUU<UUUhho
U<U<<Hubu<
hou<<o<UHh
<HUHUH<<U<
U<U<UUUUUH
<40h00<0h<
H<U<Uhuuoh
PhhhdhdUDC
<<UUUH<U<<
hh<00hhhhh
<<UH<U<U<<
<Uhhhhhh<<

acc—
qwoq
~0n~
«cc.
-n~
~o-
doo—
uoo
men
~N~
«on
~o¢
qon
new
-~
~

”~Ia. «aouu

113

Figure A3: Complete Sequence of cDNA Clone POI-3, The sequence of

cDNA clone POI-3 is shown. Postion 932 in this sequence corresponds to
position 941 of the a-subunit sequence (Figure 6A). Internal EcoRI sites

are underlined so that relative positions can be oriented according to

Figure 7.

1.141

0H00<0h0<0
<<0<0<<<0h
0<0<00<H<<
000<<0h<0<
0<00Hhh08h
hh<<0h00h0
h<08<0<h0<
h<0h<0<00h
(008004400
000<<0900<
h00000<00<
¢0h<h<<<0<
<<0<0h0h0<
0h0<008<0h
h<0h<00b<0
HHH0000H0<
<h000<00h0
h0h0090<<0
<<<h00<h<0
h<005b<ac0
<00<<<00hh
<<<0H0hh00
<Hh<90<hhh
0h00<<<0h<

 

<000B0<hh<
0<<<<0hhhu
00k000900h
<00<U<0h<h
H01009<0hh
0Fh0050080
000<00h0<0
<<00h<00<0
00h0<000<<
<<0000<090
0%00Hh<0<0
0(00h<0<00
<0h00<00<h
000<0<0h<0
090+0<0h<0
00<0<00b00
00<00000h0

0<<<0h0<0<
000h0hh0<0
h0h0<<00h<
0<0h<0h00h
0000h00<0h
#000800<<0
0<00<0h0<0
h<00<000<0
0<000h<<<0
0%008h00h0
Hhh<<<b<00
0900<0<00<
<<<0h00h<0
0h<<<0h<0h
0h0h000<0<
hUUh<<0000
<<0<<h0h<0

 

0hh0h0hh<<
<<<<H0<<<<
H<0<00000<
00<<0<90hh
hUhhh0<0h0

0h000hh<00
<h0h0<<<<<
(RUHUUHUHU
00hh<fi00<h
000H0h<0h0

<h0h<h0<<< h<H<0<0<00
0hhhh<h09< 00<0<0<hhh

0409Bh00<k
<090<<0<00
00h0000000
h0000<<<00
h00<<h000h
0<0<00h00<
<00<h<0<0<
00H<<0h00<
h<0<00<b09
HHUH<00<0<
h00h0<<<0h
<00900h<0¢
<0<<00hhhh
hh000<<<00
<<0<00hh<0
h<090<<<06
<4090<00<0
090H<0h<hh
00h<0<<<0<
08¢00<0080
hhhohhhhoh
<h<<<0<0hh
h09<h0<0<h
h<0HHHUH<0

0hh<<0<0<h
hhh<0<<00<
800H00<000
0000900<<0
h0<00<¢000
0<<0h<0000
000hh000<<
<90h0<0h<0
<0<0hhh0h<
0h<<0<<h0<
00000h00<<
00<0h<h0<8
0h000h0<<0
<h<00<<00<
Ub<0<<4000
0hhhhhh0<0
<0<0<<0<<<
0<<h<h<u<h
h0hh00hh00
<Hh<0<000h
Uh<<<0<0<0
BHHUHHHHUU
<0<h<<<<¢<
00<00<<000
h0<0¢0hh<0

0<<<<0<0hh
<0<0h<<0hh
0<h<00<<<0
0<<<000000
0<<0<<00<h
0<<0<<<0<0
00<0h<000<
0h00h<0<<<
h<<<h000<0
000<0<0<00
0Hh0000<00

 

<0h<<00h<h
<hhhhh0<<0
00h0<0000<
<0<<0H0<<0
H<000h<00h
<00h<0h<00
0h0h0h<0h0
<00h90h<00
<0h<0<0<0h
<0hh<0<00<
0000h0<<0<

u~<o<oss<<Imusue<<<ua

<<0<<00<00
0<<<¢0hh0<
h<<0<hhhh0
0(0409h000
0009(00h00
<hh000h0h0
0<0hh00h0h
090<h0hhhh
000Fh80h0<
<0<08hh0<0
<090<h<0h<
h00000h<0h
00H00<0h0<

0b<<000h00
<<0<<09<09
0<0000h000
009<<0h<0h
h<0<h0<0hh
<0009<80H0
H0<00<0905
00<0h00<0h
0<<0H0h00h
H¢¢<0080<<
0(000‘hhh0
0<0h0<0<hh
0000h<0<0h

<<<0<0<h0<
<0h0h00¢0h
000<00<0<0
0<00000<00
00h<000h0<
0hh<000h<<
h00000<0h<
<00<<<<0<<
H0h0<0<<<<
000h095000
00<h00<0h<
00h00<00h0
00<00<<000
h0<0h<hh0<
0Hh0hh<<<0
<0<<00hhh<
<0h00<<0<h
h00<<<000h
000<<0090<
hh0<0h0h0h
0hh<000<<<
h<00000000
H0000<<<<H
<h<0<hh0h<
h0hhhhh0<<

h<hhhh<0hh
hhhh0h000<
0(000<0hh0
<<00<<000<
0h<0<00<00
0<0<<00800
0h<0h<0000
0hhh<k0hh0
h00hhh0<00
00000<0h00
00¢0000<h0
<<000<0h00
H<h0h00<0<
0<0h00b<<<
0<00<<0<<0
090000h<<<
hhh0<<0h<<
h000<<0h0h
H0<<h0<00h
0hh<k<hh<h
<<0900<00h
<0hohhh<0h
0h0h<<u<h<
<<00<0000h
h<k<h0h<<<

90<h<hh0hh
<0hhh00050
00(0000000
000<UU<0<<
<009hh0000
H000<0000<
h<0hohh<<<
00Hh00<0<0
hh0<00<0<0
h<00009h00
000<<h0<0h
000b00<000
0<00<<<<0h
hh0h0h0500
<<0h0hh<<0
<00009P0h<
<000<<009<
0<0hb0<hdh
00¢00hhh0h
0<<hh0000h
hhhh00<H00
UUHPFUHUUU
<00<hh<0h<
00<hhh<0h0
<h<0<h<000

UPUHHPHUF<
<h00hhh0h<
000(000000
<0<00<<0<0
h0hh<<00<¢
00b00<0h00
00<<000<0<
<h009h<0h<
<<<00hh0h<
00<00h00<0
00089<80H<
<0??0<00%h
h0h<<0<0<0
hooouhUUHh
00h<h0hfih0
PPU<0000F0
000090<00<
Uhhh<0<<h0
Hh00<0<0h<
0<<0hh<009
Hhhhh00<<0
UHUUHFUPUH
<0000h<000
<<<<<0050<
<00h000h00

 

m4 musmflm

<o<hou<hhu
0<hhh0<hhh
eooo<oo<uo
o<ou<oohoo
u<eoeohshm
<<o<<oooeo
o<oe<uosuo
uooes<o<uo
ooaoueoooh
<o<ueeo<a<
oesooousoo
<<<a<<hs<<
osc<<u<<uw
u<ooo<ohhe
ueoe<<<<<h
hoouo<opu<
ouosuueuou
ohe<006hue
<<so<e<o<<
shoo<<<ueu
oeseo<ua<a
heh<oehueo
0h<0hhh<<<
u<<oouohh<
HOH00hhh<<

~mn~
~o~n
~¢o~
-n~
uovu
~a-
~o_~
dean
uuaq
uoc~
~ao~
«and
uqca
amn—
~c-
aco—
«ca
new
~N~
goo
now
mom
~v~
-~
~

unIaa uao_u

115

Figure A4: Psmal Sequence Determined for cDNA quns Pq-4, The
partial sequence of cDNA clone POI-4 is shown. The ~1.5 kb EcoRI

fragment at the 3’ end (Figure 5) was not completely sequenced. Postion
1 in this sequence corresponds to position 914 of the a-subunit sequence
(Figure 6A). The sequences are identical up to position 1816 where the
divergence begins. lntemal EcoRI sites are underlined so that relative

positions can be oriented according to Figure 7.

 

1.163

vfi musmflm

...<<00<0<‘ h0h0h<h000 ~oam
00h000<h00 0<<h0<0hhb 00h0<<b00< 0<<<h0<000 0h<h0<0h<h H0hh<h00hh <<HOHhh0h0 <h0<0h¢hh0 00<0<<00h0 H<<<<<hh<< 0<h0h00<00 0(0h0h00hh «can
00<0h00h00 0H0h00b<<h 0h0hhh0hh0 h0<0<<hh0h Hh<h<00h00 00h<h<0hhh 0<0h00hh0< H0<<<0hhh0 <<0h<<h0h0 h0<0h00hhh h0thh00<h h<0<<00000 -~n
h<<0<090hh 0<h00h<<<0 0<0000<h0h H<<0F<<<<0 0<<0h<<0h0 00h<h00000 <<0<0800<0 00<0H0<00< Hhh0hh<000 04000h0<00 H00h000h0h 00h0h<0h00 «can
h00<<00h<0 <0<000<009 0Hh0h<0000 <00900h000 0h<000hh00 0h<<00h0<h <<0<<000<0 0h0hh<<00h <908000H00 h00k<0000< h0h0<00h0h 000(090hhh «men
0h0<080h0h <0F0<09h<0 h00h<0h<<0 hh00h<00h0 #001500990 h0hh0h0<<fi hh0h00hh<0 0b¢<<0h0<h <00h0h<<0< h<h<00hh00 Ph00<0h0h0 <0<H0h<h0h gonn
hhhh0<uu00 <<h0<h<09< 0h00k<000< 00<0h0<0<< <00hhfih<00 <h<0h<k00h <0<H<h<0$0 h0h0<0hh0< ¢<00hh<00< hhhh¢<<hhh hh0<00h0<0 h<0hh0h<h0 ucwn
00h<00<00< h00h<0<090 h<00<<0<<< h0<000<k<< 0hhh<0h000 000H<<h<H0 0Hhhh00<00 H<<h00hh<0 hh0h0<0h00 00<h0h<hhh h<P00<<<0h 0<0h0<0hh0 "aqn
hh<0h0<000 00<000h0h< 0h0hh0<0<h 0h00<00hh0 h0hh0hh00h <0090<000< 00hhh<<0<h 0h80h<00hh Phh0<0<<0< <<<00h<<00 h0<0h00<<< <000<000§0 «can
hh<00h0h<0 <hhhh<0h<< 000<<0<0<H <0hh0<00<< H<<0<h<000 <0P0h<<<<0 0<0<0<000h <h<0<<<0<0 00<<<<00<0 <0<0h0h<h< 00<0<0<0hh hh<hhh0h00 «can
hh<000h00< 0h000<<h00 0<<h0h0<00 <00000<<h0 090<00<00h h<00<0<<00 09<0¢00<hh 0<0<<<03hh 0h00h0hh00 0<00<0h<0< 00hh0<0<05 0thPPOh<< aonu
¢h00h<0hhh H<<<000<<0 h00<<0h<0< h00<<0h<<h 0h<00<h0b< 0h<<<0<<00 0<0¢0<<00H <hh<h<00<< <0000<0<0h h00<0<0h0< <0h00h0<0h h0<50hh0<0 «vow

 

<00<<<hhhh 0h0h00090% 0<<00<<hkh hh0h0h<h<< hh<<0h0hh0 0h0h0h0b0< 00¢H<<H<0< hh<<0h00<< <<hh0h<fihh 080<h0h0hh hh0hhh0000 h0h<0h0<00 «flaw
h<0hh<hh0h h9900h0<<< 0<Hhh9000h <<<h000<0k 00h<h0h<00 h0<<<h0<00 000<0<<0<h 0h<0<<000< 0<0<<0<<00 0hh00h0000 <000hh<<0h <009hh0hhh mocu
4H0000<<H0 hh0<<00h80 0<hhb0<<<h 000<0h0<0< 0<0<0h0000 <0<0h0h000 0h<00h0<00 <<<<<0<<<< 0<<h<50<0< <0Ph00h0<0 hk<h<<00yh <0hP00<<<h uawu
0h000h0<0< 0<<F080<0h 0090000<<0 h050h000<h (0000(00<0 <000000<00 <0<H<h<<<h <400<0h00h 0hh00<00<k 0hh0h0<<0h <0hh00h0hh 0hhhh00hhh ~o~u
hhh<0hhh00 h<0h0<h0h0 00<0<0<<0h <0<0P00<<< 00hhho0h<0 80<0hh0h0h 0h0b<0hUh< <0<<0hh0h0 <0<00h0h00 h0h000h00h 080h0h0h0< hh00h0b000 «can
b00hh0<00h 00<080000< 009<<h0<0< 0h<0<0<0hh 0<0h<<h<0h Hh<0h0<h0h 0h00<<0<h< 000¢<<0<hh 0<<00<0h<0 h0h09<<<hh 0h<<00h<h0 h<0<00<h0h amm—
<<0hh0h<hh 0<<0<<0<00 <00h0<h<<< 0<0<<00<h0 <00<0h0h0< 0h0<h00h00 0<0<0<<0<0 H00h<<hk<0 hh<00000h0 hhh000<000 0h0hh<0h0< <0<00h00h0 «can
0<0000<h<0 0<<<000H0< 0000(000<0 0<0<00<000 <0hh000<00 0000000<00 0000<000<0 04000<0<00 <h4<00h000 h00hh0h0<< 00h<00h000 00000000h0 moo—
0<<00<<<00 0000<0<<0h 01¢00<0000 0<00<<00<< 000<000<00 <0<<¢0<00< <0¢00<00<0 0h00000<<0 h<0<<00<0< 0h<h0¢09<0 900hk0000< <<00h0<00< «emu
<0000<<0<< 00<hh<000h <00h00h<00 0h0<0h<0<0 0<00<00hhh 0000h0hh<< 00<<0<HOH0 hhh00<00hh h0hhho<00h <0hh0000h0 0<0Ph00<<H 00Uh0<<0h< «ec—
00000<<0<< <0<0<00h<0 h<000hh<00 0h<<0<0<<0 0H00b000<0 000<00H00< 0h00<<0<<0 00h0hb<<0h 00H00hh008 00h09000b0 04400<0<00 h00<000hh0 «an—
00<<00<0h< 00J<0h0h0h (050900000 <0h<0h<0h< 0000h<0h0h h<<<00<<00 0<0<0<0h<0 0h00h<0h<0 <h0<000<00 h0<00<00<0 h0<0<00<h< 0<0<<H0h0< «on—
0h<00h00h< 0<<<<00hb0 h<00<00<<< <0<<0hhh<h 0hh000hh00 <0<0<h00hh <0h<000hh< 0400h<0h<0 <00h<<00h< 00<0h<00<0 00<000h<<0 h00<<0¢0hh «oc—
h0h<h<<<h0 00<0<0h<0< 0<0hh0h0<0 <<<<h00hhh 0<00Hh0<00 <0<0<<<00h 90h<00h00h 000h<00h00 <<000080<0 00<<0<000h <<<0h<0<00 <h090h<<0< «om

<h0<000<0< 0<00<0hh<0 <00<000h0h P00000000< 0h00h<0000 HH0000<00H 00<0<0<0hh 0<H<000<<0 H00<<<0000 <0h00h00hh 00h0H90h<0 0<0<00000h new

00<<0hh000 0400000090 <<0<00<h00 <0h<00<000 0<H0000<<H 0<0H0005h< 90h<ohh000 0H00900000 <00<0h00hh <0<0Hhh<<< h<00h00h0< <<0h00<0h< -~

h0¢h0h<0<0 hh<<00h09< <<0h00900< 00H0<<000< 0h00000h00 <000<0hh0< 00hh<<<h<< hh<<<0h<h< <<0<0<00h< 0<000h00<0 <00<<00h00 h<0<0h000h ~00

0<<0<<0<<0 0<000h<<00 0h0000<00< d000b<h050 0<0<0<00<< <<0hh0h<<0 <0<00H<<<0 <<0h<<0<0h 0h0<<0h00< 00<h<<<080 0h<0<0<<00 hhhh<h<00< «cc

<00<0<<<<0 hh0<<<0<<0 h<0hh0<0h< hh0<0<0h00 H<<<hh090H 0H00h0000h 00Hh0<000< 0hhh0h0<00 h<0h000<0< 08¢00h<<<0 h<0hhh000< <4000h<0<< ~on

1000h<<0<h hhh00<0000 H0000hh0hh <<<00<00<< 0<<0<<0h0h h<<000h<h0 hhh00h0h<< <<<hh<0h<0 0h<00h0h0< 0h<00h0h00 0<0<<<0<00 hh<00hhhhh gem

h0<00¢0<0h h00000h<<0 h<0h<0<<00 hhh<0h0000 H<<<<0000h h0h<hh0<00 00h0h0000< 090<hhh000 0h0<00<0<0 0H00h00h<< 0000H<0h0< <<0h<0<0<< «N—

0144000H<0 0h00h<0<h0 <0hh<0h00< <0<hhhh0¢< 0h<<<000<< 008<0000h0 <00<000h00 h000<h000< 00H000<000 00h<0h00h0 <0000hUh0h <0hhhhhh<< d

 

 

 

“c-am ago—u

 

 

 

117

Table A1: Nucleotide Differences Observed in cDNA Clones. The top half
of the table shows changes in comparison of the composite sequence of
the a-subunit (Figure 6A) to a published sequence (29). Only one
deletion was found, which is in the 3’ untranslated region. Shown at the
bottom are the differences observed during sequence analysis of the
cDNAs. All of the differences are noted relative to that shown in Figure
6A. The amino acid changes found in the published a—subunit sequence
were at the following positions: 194, valine to alanine, 424, threonine to
alanine, and 675, phenylalanine to cysteine.

 

118

 

 

 

 

 

 

 

mtutfildm . ADV Hmms mu at. oara h .00 mmm

m Ado HmvN o adv «mam 0 “UV meow

m .00 «How a .00 wood a .ev game

m .00 mafia m .00 pmaa m .00 can

cacao ouamno sowuwoom osoao ouamno dowuwoom cacao oucuno sowuwnom
.GOAImu sommmmlmwlmmmmmmmmm dzoo cw mummmmmwwmm

.o. momm .o. Hora

Acoauoaoo 0. even .40 anvw Adv mmau .00 meow

Ace avou no. ado“ .00 moms no. vosfl

Aw. vuma A49 Hmuu A00 hmaa A00 mmm

madman cowuwmom ousmno sowuwoom oocmno nowuamom ousmno sowuwnom

 

 

 

. mu coco on nonmwanom.oa UmmmmmmmiawmmmmM!mwnug vacansmnw

 

 

 

.mocoHo «200 GM vo>uomno moocouomuwa wowuomaonz

”a4_OHQMB

 

I ' ‘1 I.“ l I

-

MICHIGAN STATE UNIV. LIBRARIES
1|1111111|”WINll”HW1111111111
31293005853191