Isolation and characterization of Drosophila cAMP-dependent protein kinase genes Daniel Kalderon I and Gerald M. Rubin
Howard Hughes Medical Institute, Department of Biochemistry, University of Califomia, Berkeley, Califomia 94720 USA
We have used mammalian probes to clone genes encoding the catalytic (C) and type I regulatory (RI) components of the cAMP-dependent protein kinase in Drosophila. Both Drosophila gene products are very similar in amino acid sequence (RI, 71%; C, 82%) to their respective mammalian counterparts, implying homologous activity. A single Drosophila type I regulatory subunit gene is the source of at least three distinct transcripts originating from different promoters and spliced to a common body that would encode a full-length analog and two amino-terminally truncated variants of the mammalian RI protein. The RI locus also includes two intronic genes of unknown function. A single highly conserved catalytic subunit gene (DC0) was found that codes for a single polypeptide. It was used to isolate 11 further more distantly related apparent protein kinase genes. Two of these genes (DC1 and DC2) are sufficiently similar to DC0 in sequence (45% and 49% amino acid identity, respectively) that they could conceivably encode products of overlapping function. Two further genes are very similar in sequence to bovine cGMP-dependent protein kinase. The remaining putative geue products include amino acid sequence motifs characteristic of serine-threonine protein kinases but cannot, from the available data, be defined as homologous to specific protein kinases of other organisms.
[Key Words: Drosophila; cAMP-dependent protein kinasel protein kinase genes]
Received June 8 1988~ revised version accepted October 3, 1988.
Many extracellular stimuli elicit diverse intracellular responses through the action of the second messenger, cAMP {Sutherland 1972). The binding of extracellular li- gands to a variety of specific cell-surface receptors acti- vates membrane-associated GTP-binding proteins that either stimulate {Gs) or inhibit [Gi) adenylate cyclase, which catalyzes the synthesis of cAMP at the inner face of the plasma membrane [Gilman 1984}. In this way, in- tracellular cAMP concentration can respond to different hormones and neurotransmitters in different cell types {Nimmo and Cohen 1977~ Nestler et al. 1984). A cAMP signal is propagated by binding of cAMP to the regula- tory subunits [RJ of the inactive cAMP-dependent pro- tein kinase holoenzyme {R~C2), causing dissociation of catalytic (CI and regulatory subunits. The activated monomeric catalytic subunits phosphorylate, and thereby alter, the activity of one or more protein sub- strates appropriate for the execution of a given intracel- lular response (Krebs and Beavo 1979). Activation of the cAMP-dependent protein kinase is held to be virtually the only direct response to intracellular cAMP signals in eukaryotes [Kuo and Greengard 19691 Coffino et al. 19761 Walter et al. 19771 Nakamura and Gold 19871
tPresent address: Department of Biological Sciences, Fairchild Center, Columbia University, New York, New York 10027 USA.
Toda et al. 1987b). Therefore, cAMP-dependent protein kinase is a key enzyme in linking many agonists to cor- responding intracellular effectors and is expected to par- ticipate in cellular communication in several different physiological situations.
The specificity of the response mediated by cAMP-de- pendent protein kinase in a given cell depends upon the array of functional receptors on the cell surface and the spectrum of kinase substrates and other effector proteins present in the target cell. Specific physiological roles for cAMP-mediated signal transduction have been defined in a number of organisms. Most prominent among these are the control of glycogen and lipid metabolism in mammals by cascades of enzyme phosphorylation [Cohen 19821, modulation of activity of excitable cells, probably by direct phosphorylation of ion channel pro- teins {Kandel and Schwartz 19821 Osterreider et al. 19821 Siegelbaum et al. 1982), and less well-characterized roles in the control of meiosis [Maller and Krebs 19771 Wa- tanabe et al. 19881 and mitosis [Matsumoto et al. 19821 Shin et al. 19871 Toda et al. 19871 Tanaka et al. 19881, neuronal development [Haydon et al. 19841 Lankford et al. 19881 Rydel and Greene 19881, sensory transduction [Pace et al. 19851 Avenet et al. 1988~ Tonosaki and Funa- koski 19881, and selective transcriptional activation of genes {Lamers et al. 1982~ Hashimoto et al. 19841 Wyn-
shaw-Boris et al. 1984; Comb et al. 1986; Montminy et al. 1986; Ran et al. 1986; Deutsch et al. 1987; Montminy and Bilezikjian 1987; Nakagawa et al. 1988). These roles of cAMP signaling have been inferred largely from bio- chemical and pharmacological investigation and genetic study of yeast. It is possible that genetic investigation in a metazoan would reveal an involvement of cAMP-me- diated signal transduction in developmental decisions and specialized differentiated functions other than those defined by well-characterized hormone or neurotrans- mit ter signaling agents.
Two roles for cAMP signaling in Drosophila already have been inferred from genetic analysis. Drosophila carrying mutat ions that alter cAMP phosphodiesterase (dunce) or adenylate cyclase (rutabaga) activities are de- ficient in learning and memory tests (Byers et al. 1981; Livingstone et al. 1984). This has been interpreted to suggest a role for cAMP in the modulat ion of ion channel activity at the appropriate synapses (Quinn and Greenspan 1984). Female Drosophila carrying the dunce mutat ion are also sterile, further investigation of which suggests that cAMP signals are used in early embryo- genesis (Kiger and Salz 1985; Bellen et al. 1987). The mutat ions rutabaga and dunce affect the rates of syn- thesis and degradation of cAMP, respectively, and alter ambient cAMP levels, but retain the ability to alter cAMP levels and to respond to such alterations. Genetic inactivation of the cAMP-dependent protein kinase should prevent all cAMP signaling. Therefore, it should be possible by genetic manipula t ion of cAMP-dependent protein kinase to address the entire spectrum of cAMP- mediated signal transduction in a mult icel lular or- ganism and to pursue the molecular details of specific transduction pathways.
Here we report the isolation and characterization of Drosophila genes for the type I regulatory subunit and a catalytic subunit of cAMP-dependent protein kinase. Eleven further apparent se r ine- th reonine kinase genes have been isolated by cross-hybridization between DNAs encoding kinase domains and characterized by l imited sequencing. The complete coding sequence of
two of these genes has been determined and suggests that they may encode additional catalytic subunit variants of the cAMP-dependent protein kinases.
R e s u l t s
Isolation of Drosophila type I regulatory subunit gene
A single segment of DNA in the Drosophila genome was identified by hybridization of a bovine eDNA probe (p62C12; Lee et al. 1983)for the type I regulatory sub- uni t (RI) at low stringency to Drosophila genomic DNA blots and a Drosophila genomic library (Maniatis et al. 1978). Several overlapping genomic ~ bacteriophage were isolated, from which a short sequence recognized by the bovine probe was identified and used to isolate corresponding cDNAs. The sequence of one of these cDNAs was determined and shown to include an open reading frame (ORF) that could be translated into a pro- tein of s imilar length and sequence to m a m m a l i a n regu- latory subunits (class I RNA in Figs. 1 and 3). The trans- lation product is clearly more similar to the m a m m a l i a n type I (71% amino acid identity) than type II (32% amino acid identity) regulatory subunit sequence (Takio et al. 1982; Titani et al. 1984; G.S. McKnight, pers. comm.; see legend to Fig. 1).
Hybridization of this eDNA to polytene chromosomes showed the Drosophila RI gene to lie on the left arm of the third chromosome in polytene division 77F. This gene appears to be the only gene in the Drosophila genome capable of encoding a type I regulatory subunit, as no related sequences could be detected by low-strin- gency nucleic acid hybridization.
Diverse RNAs are generated from the RI gene by differential splicing and polyadenylation
Characterization of cDNAs, RNA blots, and S 1-nuclease protection experiments have shown that mul t iple RNA species derive from the RI gene. They result from differ- ential splicing both upstream of and wi th in the coding
Figure 1. Sequence of Drosophila RI gene. Genomic DNA sequence around four different exons, I-IV {indicated as CLASS I-CLASS IV), that may be spliced to the common region [COMMON) of the RI gene are shown. All genomic sequences surrounding splice junctions conform to the consensus {Mount 19821. Transcribed regions are represented by capital letters; putative initiator and up- stream methionine codons are highlighted for each of the forms of RNA as are polyadenylation signals and the optional splice junc- tions at nucleotides A982, A1048, and A1408 that are used to generate different forms of class I RNA. Class Ib RNA {Fig. 2) is spliced between A982 and A1048, class Ic between A982 and A1408. Heterogeneous start sites for class I RNA are between A939 and A964 and were mapped by primer extension using oligonucleotides complementary to A989-AlO16 and A1028-A1053. Thus, some class I RNAs include the methionine codon at A959 whereas others do not. Class Ic splicing removes methionine codon A1013 and Ib splicing removes methionine codons A1014 and A1064. The transcriptional start site of class II RNA was mapped by primer extension using an oligonucleotide complementary to B249-274. The 5' end of class III RNA has not been mapped; capitalized letters correspond to sequences present in the only eDNA of this RNA form. The start site of class IV RNA was mapped using three oligonucleotides complementary to D301-D331, D360-D371/E48-E63, and E48-E78, respectively. Major sites of polyadenylation evident from S1- nuclease protection experiments are around E1866 and E3286. cDNAs have been recovered that carry poly(A) sequences following nucleotides E1862-8, E2181, E3277-9, E3285-7, and E3318. The translation product of class I RNA is represented in single-letter code {Dr) underneath the DNA sequence and compared to that of the mouse Riot subunit (Mo) below (G.S. McKnight, pers. comm.I. Amino acid identity between the two sequences is 71% overall, 53% in the amino-terminal domain (1-137), and 85% and 79% in the two carboxy-terminal cAMP binding domains (138-257; 258-377). Translation of class II RNA is expected to initiate at the methionine codon immediately prior to the splice donor junction of the class II-specific exon, whereas translation of RNAs III and IV would be expected to initiate at M-81 in the common region of the RI gene.
1540 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
ATCCcACCTCCACCGTAcCACATGTCcTACATGATGGcCAAAA•G•TGGAGGAGcAG---AG••TGcGGGAGTGCGAACACTACATCCAGACG•ATGGGATCCAGCGAGTCCTC A1593 M S Y M M A K T L E E Q - S L R E C E H Y I Q T H G I Q R V L Dr30 M A S G S M A T S E E E R S L R E C E L Y V Q K H N I Q A L L Mo31
AAGGACTGCATCGTcCAGCTGTGCGTCTGCAGACCCGAGAATCCCGTGCAGTTCCTGCGACAGTACTTTCAGAAACTCGAGAGGgtgagtccaaaaccaaaattataaataata A1707 K D C I V Q L C V C R P E N P V Q F L R Q Y F Q K L E R Dr58 K D S I V Q L C T T R P E R Q M A F L R E Y F E R L E K Mo59
E Q V K L D A S R Q v I S P D D C E D L $ P DrS0 E E A R Q I Q C L Q K T G I R T D S R E D E MoSl
ATGCCACAAACAGCTGCTCCACCGGTT CGCAGGCGAGGCGGAATCTCCGCCGAACCCGTAACCGAAGAGGATGCCACCAACTATGTTAAGAAGGTGGTGCCC E216 M P Q T A A P P V R R R G G I S A E P V T E E D A T N Y V K K V V P Drll4 I S P P P P N P V V K G R R R R G A I S A E V Y T E E D A A S Y V R K V I P Moll9 AAGGACTACAAGACGATGAATG•CCTGTCCAAGGCGATTGCCAAGAACGTC•TGTTCGCTCA•TTGGA•GAGAGCGAGAGAT•CGACATATTTGATG•CATGTTCCCAGTGAAT E330 K D Y K T M N A L S K A I A K N V L F A H L D E S E R S D I F D A M F P V N Dr152 K D Y K T M A A L A K A I E K N V L F S H L D D N E R S D I F D A M F P V S Mo157 CACATCGCCGGCGAGAATATCATCCAGCAGGGCGATGAAGGcGACAACTTCTACGTGATTGATGTGGGAGAGGT•GATgtaagtatatttatttttcttgctatgacctttgaa E444 H I A G E N I I Q Q G D E G D N F Y V I D V G E V D Dr178 F I A G E T V I O O G D E G D N F Y V I F Q G E M D Mo183 ttgggtttgcatcat•atgaatatgttcacatctccgccagGTTTTCGTCAACTcCGAACTAGTGACCAC•ATCAGcGAGGGCGGCAG•TTTGGTGAACTGGCCCTCATCTATG E558
V F V N S E L V T T I S E G G S F G E L A L I Y G Dr203 V Y V N N E W A T S V G E G G S F G E L A L I Y G Mo208
GCACTC•TCGCGCCGCCA•CGTG•GCG••AAAAC•GATGTGAAGCTGTGGGGAAT•GA•CGCGA•TCCTACCGCCGCATTCTGATGGGCTCGACCATCCGCAAGCGCAAGATGT E672 T P R A A T V R A K T D V K L W G I D R D S Y R R I L M G S T I R K R K M Y Dr241 T P R A A T V K A K T N V K L W G I D R D S Y R R I L M G S T L R K R K M Y Mo246
ACGAGGAGTTCTTATCGCGCGTGTCCATTCTGGAGAGCTTGGACAAATGGGAGCGCCT•ACAGTTGCCGATTCCCTGGAGA•GTGCTC•TTCGATGACGGTGAGA•GATTGT•A E786 E E F L 5 R V S I L E S L D K W E R L T V A D S L E T C S F D D G E T I V K Dr279 E E F L S K V S I L E S L D K W E R L T V A D A L E P V Q R E D G Q K I V V Mo284
AG•AGGGAG•AGCTGGcGATGACTTCTACATCATCCTCGAGGGCTGTGCGGTGGTGCTGCAGCAG•GTTCTGAGgtgagtgaccgcgatcagacaccatgggacacctaaaaag E900 Q G A A G D D F Y I I L E G C A V V L Q Q R S E Dr303 Q G E P G D E F F I I L E G T A A V L Q R R S E Mo308
agacactctaatcaactcgtttatctaatcagCAGGGCGAGGATCCCGCCGAGGTGGGCCGTCTG•GTAGCAGCGATTACTTTGG•GAGATTGCTCTGCTTTTGGACCGAC•A• E1014 Q G E D P A E V G R L G S S D Y F G E I A L L L D R P R Dr331 N E E F V - E V G R L G P S D Y F G E I A L L M N R P R Mo335
GTGCAGCGACTGTTGTAGCTCGTGGGCCGCTCAAATGCGTCAAACTGGACCGGGCGAGgtaagtgcagtggcttgagtqctaagacaaataggcaggtgtatgttaatgacctt EI128 A A T V V A R G P L K C V K L D R A R Dr350 A A T V V A R G P L K C V K L D R P R Mo354
F E R V L G P C A D I L K Dr363 F E R V L G P C S D I L K Mo367
AAcG•AACATCACGCAATACAAcAGTTTCGTAT•GTTGT•CGTTTAAGACCGCAACAACGACAACAACCACTACAGCAGATAAACAACAAAGAGCAGCAGCAGCAAAGAAAGCA E1356 R N I T Q Y N S F V S L S V Dr377 R N I Q Q Y N S F V S L S V . Mo381
sequence of the gene and by the use of different poly- adenylation sites distal to the coding region [Fig. 2}. Dif- ferential splicing of transcripts leads to the generation of at least three putative RI polypeptide products, two of which are truncated by 57 and 80 residues, respectively, at the amino terminus relative to the largest Drosophila RI protein {Figs. 2 and 3}.
Restriction enzyme mapping and partial sequence analysis of 18 independent eDNA clones indicated that any of four different sequences can be spliced to a single site immediately prior to the codon for amino acid 59 of the largest {class I} RI product. The location of these four possible upstream exons has been determined by com- parison of genomic and eDNA sequences. The methio- nine codons indicated in Figures 1 and 2 as likely ini- tiators of translation are the furthest 5' in the respective long ORFs of class I, II, III, and IV RNAs. The theoretical primary translation product of class I RNA is of roughly the same length as the murine RI protein, the product of class II RNA is missing residues 2-58 but is otherwise identical, and the product of class III and IV RNAs lacks residues 2-81 {Figs. 1-3J.
Two minor variants of class I RNA which specify the same polypeptide can be inferred from individual
cDNAs (Ib and Ic in Fig. 2) that indicate splicing in the 5'-untranslated sequence between a single donor and two altemative acceptors (see legend to Fig. 1 for exact intron locations}. The major form of class I RNA IIa in Fig. 2) is not spliced in this region, as shown by Sl-nu- clease protection of the entire Ia exon using adult head RNA [data not shown) and by the similarity of RNA blot hybridization pattems using probes derived from class Ib intron and exon sequences [Fig. 4).
The 5' ends of class I, II, and IV RNAs have been mapped by primer extensions {see legend to Fig. 11 data not shown) and also by Sl-nuclease protection for class I and II RNAs {data not shown). Class III RNA is suffi- ciently rare that it has not been detected by these means or on RNA blots and is exemplified by only a single eDNA. Heterogeneous transcription initiation sites for class I RNAs span a region of 36 nucleotides. The appar- ently unique 5' ends of class II and IV RNAs are sepa- rated from the 5' end of type I RNA and each other by more than a kb {Fig. 2). At least three different pro- moters therefore are used by the Drosophila RI gene, each apparently specifying a distinctive RNA form of unique coding potential. The three major forms of RI transcript, I, II, and IV, appear to be regulated indepen-
Figure 2. Transcripts of the RI locus. The RI gene transcripts shown are inferred from eDNA structures and have been veri- fied by direct RNA analysis for classes I, II, and IV by using specific probes on RNA blots (Fig. 4), by primer extension, and, in addition, for classes I and II, by S 1-nuclease protection using eDNA probes spanning the variable splice junction [data not shown). Class I, II, and IV RNAs have different 5' ends. The 5' end of class III RNA has not been mapped. Each of these RNAs is spliced from a unique 5' exon to a common series of exons and may be polyadenylated at one of two major sites in the 3'-untranslated sequence. Presumed initiator methionine codons (AUG) for polypeptides derived from class I-IV RNAs are indicated. Their use would lead to the generation of a full- length homolog of mouse RI protein (class I) and truncated ver- sions lacking 57 (class II) or 80 {classes III and IV) amino-ter- minal residues. There are two or three methionine codons up- stream of the presumed initiator codon of class Ia RNA, depending on the exact start site of transcription (Fig. 1 }. Single cDNAs indicate two further variants of class I RNA that are spliced in the 5'-untranslated region, thereby removing one (class Ic) or two (class Ib) of the upstream methionine codons of class Ia RNA. The locations of sequences used as hybridization probes for the RNA blots shown in Fig. 4 are depicted under- neath the RI gene transcripts. Between the splice donor specific to class I RNAs and the 5' end of class II RNAs are two tran- scripts (designated RI intron transcripts) of opposite polarity to the RI gene, spanning 1093 bp (18c) and 1386 bp {13a} and sepa- rated by 350 bp. The site of polyadenylation of 18c is 68 nu- cleotides from the splice donor of type I RNA. A region of DNA that includes the 18c transcript and the first exon of class I RI RNA is found in related form (-90% sequence identity) in about six other loci in the genome, all clustered at polytene band 77F (see legend to Fig. 5), and is therefore indicated as repeated.
RI GENE TRANSCRIPTS
[a
Ib
Ic
I I
I I I
IV
major 3' ends
AUG
AUG ~ , ~ UAA
AUG ~ ~
AUG
AUG
HYBRIDIZATION PROBES
common-a
I ~ . . I I ,.--,--
3' common-b •
Ib exon • ,--,
Ib intron - -
Repeated I -I
18c
13a
AUG
t UAA
m
ram=
RI INTRON TRANSCRIPTS
0 2 4 6 8 10 12 14 1 6 k b I l I [ I l I l I I i I i I , I l
1542 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
Figure 3. Drosophila type I regulatory, catalytic subunit, and related polypeptides. Polypeptides theoretically encoded by the differently spliced RNA forms I-IV of the Drosophila RI gene are shown. The product of class I RNA is roughly the same size as mammalian type I regulatory subunit. By analogy with mammalian RII protein (Weldon and Taylor 1985), both trun- cated Drosophila RI proteins (RI class II and RI class III/IV) would be expected to retain domains for binding of catalytic subunit (C) and cAMP (cAMP), but only the full-length analog would retain a dimerization (2 °) domain. The Drosophila cata- lytic subunit homolog, DC0 is shown above two related pro- teins derived from the DC1 gene by alternative splicing and the DC2 gene product. Similar shading denotes sequence simi- larity. DC0, DCI-a, and DC2 include sequences similar to the whole length of mammalian catalytic subunit. DC2 contains additional unique sequence at its amino terminus and DCI-b has unique sequences immediately distal to the kinase domain.
dently during the Drosophila life cycle (Fig. 4). Only class I RNAs, the templates for full-length RI protein, are detected at all stages of development, whereas class II and IV RNAs are only reproducibly seen in adults, at which time class IV RNA levels exceed those of class I and class II RNA is comparatively rare. In adults, S 1-nu- clease protection, primer extension (data not shown), and RNA blot analyses (Fig. 4) show all forms of RI gene transcript to be enriched (about fivefold) in head as op- posed to body tissue.
Multiple possible sites of polyadenylation for RI RNAs are apparent from the isolation of cDNAs car- rying poly(A) sequences at different positions and two major sites separated by 1400 nucleotides are discernible by Sl-nuclease protection (see legend to Fig. 1, data not shown). Variable polyadenylation of class I and IV RNAs can also be inferred from RNA blots in which major bands differing by -1400 nucleotides are seen (Fig. 4). Only the larger species are seen when using probes from the extreme 3' end of the gene (Fig. 4). There is no ap-
Drosophila cAMP-dependent protein kinase genes
parent association between the site of polyadenylation and the use of a given promoter or splicing pathway.
RI polypeptides
It is not known whether stable polypeptides that are in- corporated into cAMP-dependent protein kinase holoen- zyme are produced from each of the RNA species de- scribed. RNAs synthesized in vitro from cDNA tem- plates to resemble the splicing variants I - IV can be translated in a rabbit reticulocyte lysate to yield poly- peptides of a size consistent with initiation at amino acid 1 (class I), 58 (class II), and 81 (classes ]I1 and IV) of a protein of equivalent size to murine type I regulatory subunit (data not shown). Regulatory subunit protein can be detected specifically in vivo by labeling crude ex- tracts with the radioactive photoaffinity analog, [a2P]8- azido-cAMP (Walter et al. 1977). For all stages of devel- opment from embryo to adult, two major bands are seen corresponding to full-length RI and RII proteins by the criteria of gel mobility and a characteristic shift in mo- bility when 6 M urea is included in the gel (Foster et al. 1984; and data not shown). Hence, both class I RNA and its translation product are detected throughout the Dro- sophila life cycle. Truncated versions of RI, the theoret- ical primary translation products of RNAs II-IV, are not clearly evident at any developmental stage but may be obscured by proteolytic products of full-length RI and RII protein from which they could not be definitively distinguished in this assay.
Genes within the largest RI intron
Hybridization and sequence studies have defined a sepa- rate gene {18c) within the largest intron of the RI gene that is transcribed in opposite orientation to RI at high levels in pupae and adults (Figs. 2 and 4). There is also evidence from RNA blots and the isolation of a cDNA to suggest that a second low-abundance transcript [13al de- rives from a distinct location in the same intron {Fig. 2). Because sequences from the region of the RI locus con- taining the 18c transcript and possibly also the 13a tran- script are found in similar form in about six other loca- tions in the genome, it was necessary to compare 13a and 18c cDNAs with genomic sequences form the RI in- tron to verify that they represent transcripts of the RI gene intron and not of similar cross-hybridizing se- quences from another location [Fig. 5J. The longest ORF in the 18c cDNA sequence is 215 amino acids long. This deduced protein product contains no stretches of hydro- phobic residues sufficient to span a lipid bilayer, is very basic in the carboxy-terminal half, and has not been found to be similar in sequence to any protein in the NBRF-PIR protein sequence database (December 19871. The genomic loci that cross-hybridize to the 18c gene are clustered at polytene chromosome division 77F and some, if not all, are transcribed and translated into poly- peptides similar to the 18c product as judged by hybrid- ization to RNA blots and sequence determination of two cDNAs, B4A and H15A (Fig. 5).
GENES & DEVELOPMENT 1543
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
cAMP-dependent protein kinase catalytic subunit gene
A cDNA probe for the mouse cAMP-dependent prote in kinase cata lyt ic subuni t (pMC4; Uhler et al. 1986a) hy- bridized at low s t r ingency in D N A blots of Drosophila genomic D N A p redominan t ly to one segment of D N A and detec tably to at least two others (data not shown). G e n o m i c phage corresponding to the major site of hy- br id izat ion were isolated and subsequent ly used to iden- t ify corresponding cDNAs. D N A sequences of cDNA and genomic D N A revealed a coding sequence uninter - rupted by in t rons and capable of specifying a prote in of
Figure 4. RI gene transcripts. {AI RNA blots of adult head poly{A) + RNA (5 t~g per lane) were hybridized to the indicated single-stranded probes, the locations of which are shown in Fig. 2. The probes used were specific to class I RNA (I) (I in Fig. 2; A1674-A1, written 5' to 3' in Fig. 1), from the first common exon [C) {common-a in Fig. 2; E1150-E106 in Fig. 1), specific to class II RNA ill} {II in Fig. 2; B629-B112 in Fig. 1) and the ex- treme 3' end of the RI gene (3' in Fig. 2; E3205-E2570 in Fig. 1). The relative positions and sizes of hybridizing bands were de- termined from the average of several such RNA blots with ref- erence to DNA markers and from successive probing of RNA blots with different probes. The position and sizes of bands cor- responding to class I, II, and IV RNAs are indicated on the left. Probes specific to class IV RNA hybridize weakly to RNA blots presumably because the class IV specific exon is very short ( 115 nucleotides, Fig. 11. The bands designated as class IV RNA hy- bridize to the common region probe but not to class I- or class II-specific probes and are the two most prominent bands in blot C. Using adult head RNA as template, the major extension product of a primer complementary to sequences in the first common exon corresponds in length to extension on class IV RNA {data not shown). This confirms the conclusion from the RNA blots that class IV RNA is the predominant RI transcript in adult head RNA. The sizes of the longest form of class I, II, and IV RNAs [3.8, 3.4, 3.2 kb) correspond roughly to the size of the longest cDNA representative of each RNA form II, 3.7 kb; II, 3.4 kb; IV, 3.1 kb). {B) Poly(A) + RNA (5 ~g per lanel from different developmental stages [0-24 hr embryos, E; first-larval instar, L1; one- and two-day-old pupae, Pl and P2; adult head, H; and adult body, B) was hybridized on RNA blots to the fol- lowing probes. Common-b probe [common-b in Fig. 2), a cDNA probe that contains both common and class I-specific se- quences (E106-E48, A1674-A1398, and A972-A955 in Fig. 11; class Ib exon probe {Ib exon in Figs. 2 and 4; A1674-A1398, and A972-A955 in Fig. 11; class Ib intron probe (Ib intron in Figs. 2 and 4; A1363-A1030 in Fig. 11; class II probe {II in Figs. 2 and 4; B629-Bl12 in Fig. 11; intron transcript, 18c probe {nick-trans- lated 18 cDNA). Class IV RNA is seen here in the adult head lane as hybridizing to common but not class I and II probes. In the blot, common-b, class I RNA levels are deceptively high because the probe included class I-specific as well as common region sequences. Class IV and class II RNA are reproducibly detectable only in adult RNA, whereas class I RNAs are visible throughout development. Most of the bands detected by a class Ib exon probe, are also detected by a class Ib intron probe, indi- cating that the majority of class I RNA is not spliced as the Ib RNA in the 5'-untranslated sequence. Faint bands of about 3.4 kb and 2.0 kb, detected only by the Ib exon probe, may repre- sent Ib RNAs. Class Ia and Ic RNA species have not been dis- tinguished directly on RNA blots, but Sl-nuclease protection [data not shownl and the number of corresponding cDNAs indi- cate that type Ia is the more abundant species. The apparently rare upstream splices of class Ic and Ib RNAs remove one and two methionine codons, respectively, from the RNA preceding the presumed mitiator codon for the class I RI protein (Fig. 1). RNA from late larval and pupal stages showed the same pattern of bands as early larvae and pupae but are not shown here be- cause the RNA samples appeared to be slightly degraded. Sub- sequent hybridization of these blots with an actin 5C probe [Fyrberg et al. 19831 showed that all lanes have equal amounts of intact RNA of the size of actin 5C (data not shown).
the same length and wi th 82% amino acid iden t i ty to the mouse cata lyt ic subunit , C~ {Figs. 6 and 7). A glycine is present as the second residue in the pr imary transla- t ion product of bo th m a m m a l i a n and Drosophila genes.
1544 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
a t t c t a t t t c g t g a a t a t c a g a t a c g c g t t a c t c a g c t a g t c g g a t t g c a a a c a g a a t a t t a a a a c c t t t t t c a a g a c a c a 81
E G E H R M H S E I S G C A D T F F G A G G S N S C T 141
A G C C GT T B4A CCAAGAAAGGAGAACAAGTGTTCCAAGCCCAGGGTGCGGAAGAGTTGCCCCAAACCACGGGCGAAAACCTCGAAGCAACGT 1040 P R K E N K C S K P R V R K S C P K P R A K T S K Q R 168
Figure 5. DNA sequence of the RI intronic gene, 18c. The DNA sequence of the RI locus extending from between intronic transcripts 18c and 13a to within the first exon of class Ia RNA {Fig. 2) is shown. Residues present in mature transcripts of the intronic gene, 18c (358-756; 819-1452], or the RI gene (1620-1520} are in capitals. The 18c eDNA sequence was identical to the RI intron genomic sequence in all but one position (G/C 12891, indicating that RNA is indeed transcribed from this region. The genomic loci detected by high-stringency hybridization of probes from this region were investigated by cloning, limited sequencing, and genomic DNA blots that collectively indicated the presence of at least six related loci. Sequences from these loci were compared over a limited region and found on average to be 91% identical between any pair (125 point mutations in 1457 bp) and included deletions or insertions of variable size every 90 bp on average. Some, if not all, of these related sequences appear to be transcribed, since cDNAs [B4A and H15A) similar to but distinct from 18c have been isolated. All of these related genomic loci appear to be clustered because probes from the RI gene, from the related eDNA, B4A, and from one of the related genomic phage all hybridize to the same polytene chromosome division, 77F. Below the sequence of 18c is the deduced amino acid sequence of its protein product and above are differences in DNA sequence between the 18c eDNA and cDNAs of two related transcription units, represented by cDNAs B4 and H15A. Dashes indicate deletions relative to 18c. The extent of B4A (399-11531 and H15A {1067-1308) cDNAs are indicated. The B4A sequence is contiguous with that of eDNA 18c across the splice junction of the 18c transcript. The B4A sequence includes an insertion of 12 bp, three deletions (3 bp, 1 bp, 1 bpl and 55 substitutions relative to 18c. The deduced B4A protein product can be aligned with 18c over the first 189 amino acids, of which 25 are different. Thereafter, the B4A and 18c reading frames differ and the B4A product terminates after another 15 residues and is therefore 11 residues shorter than 18c. The H15A eDNA only includes the carboxy-terminal portion of the analogous reading frame. Within this putative coding sequence there are three amino acid changes relative to 18c before a change in reading frame at amino acid 199 of 18c that leads to a termination codon 18 residues distal to that used by 18c. eDNA H15A continues further 5' but its sequence is no longer related to 18c. The AATAAA polyadenylation signal of 18c is underlined. The site of polyade- nylation (1452) of 18c is 68 nucleotides from the splice donor junction {boldface underlined} of class I RI RNA (written here in opposite polarity to transcription).
This residue is myristylated in the bovine protein (Cart et al. 1982). The Drosophila catalytic subunit gene ho- molog (DC0) has been mapped by in situ hybridization to polytene chromosome region 30C1-6. This gene has
also been isolated by Foster et al. (1988). Only a single third-base subst i tut ion dist inguishes the two reported sequences of DC0.
At least four R N A species of different size derive from
GENES & DEVELOPMENT 1545
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
M G N N A T T $ N K K V D A A E T V K E F L 22 TCGAGCAGGCCAAGGAGGAGTTCGAGGACAAGTGGCGACGCAATCCGACCAACACCGCCGCCCTCGATGACTTCGAACGGA 2268
E Q A K E E F E D K W R R N P T N T A A L D D F E R I 49 TCAAGACCCTGGGCACCGGCTCCTTTGGCCGTGTCATGATCGTCCAGCACAAGCCCACGAAAGACTATTATGCCATGAAGA 2349
K T L G T G S F G R V M I V Q H K P T K D Y Y A M K I 76 TCCTCGACAAGCAGAAGGTGGTCAAGCTGAAGCAGGTGGAGCACACGCTGAACGAGAAGCGAATCCTGCAGGCCATTCAGT 2430
L D K Q K V V K L K Q V E H T L N E K R I L Q A I Q F 103 TCCCCTTCCTCGTCTCGCTGCGCTACCACTTCAAGGACAACTCCAACCTTTACATGGTGCTGGAGTATGTTCCCGGTGGCG 2511
P F L V S L R Y H F K D N S N L Y M V L E Y V P G G E 130 AGATGTTCTCCCACCTGCGCAAGGTGGGCCGCTTCTCGGAGCCGCACTCGCGCTTCTACGCGGCGCAAATCGTGCTGGCCT 2592
M F S H L R K V G R F S E P H S R F Y A A Q I V L A F 157 TCGAGTACCTGCACTACTTGGACCTCATCTACCGTGATCTGAAGCCGGAGAATCTGCTGATTGACTCGCAGGGCTACCTCA 2673
E Y L H Y L D L I Y R D L K P E N L L I D S Q G Y L K 184 AGGTGACGGACTTCGGTTTTGCCAAGCGCGTCAAGGGCCGCACCTGGACGCTGTGCGGCACGCCCGAATACCTGGCCCCGG 2754
V T D F G F A K R V K G R T W T L C G T P E Y L A P E 211 AGATTATTCTGTCCAAGGGCTATAACAAGGCCGTCGACTGGTGGGCGTTGGGCGTACTCGTCTACGAAATGGCCGCCGGCT 2835
I I L S K G Y N K A V D W W A L G V L V Y E M A A G Y 238 ATCCGCCGTTCTTTGCGGATCAGCCGATCCAGATCTATGAGAAGATCGTCTCCGGCAAGGTGCGCTTCCCATCGCACTTTG 2916
P P F F A D Q P I Q I Y E K I V S G K V R F P S H F G 265 GCTCCGATCTGAAGGACCTACTGCGCAACCTGCTGCAGGTGGATCTGACCAAACGCTACGGCAATCTGAAGGCGGGCGTCA 2997
S D L K D L L R N L L Q V D L T K R Y G N L K A G V N 292 ATGATATTAAGAACCAGAAGTGGTTCGCCTCCACCGACTGGATTGCGATCTTCCAAAAGAAAATCGAGGCGCCGTTCATTC 3078
D I K N Q K W F A S T D W I A I F Q K K I E A P F I P 319 CGCGGTGTAAGGGTCCCGGCGACACGAGCAATTTCGATGACTACGAGGAGGAGGCGCTGCGGATCTCCAGCACCGAGAAGT 3159
R C K G P G D T S N F D D Y E E E A L R I $ S T E K C 346 GTGCCAAGGAGTTTGCTGAATTCTAGAGGAGCTCTTCGCCAGTCAGCCCAATCCCCTCGTTCTGCAG 3226
A K E F A E F 353
Figure 6. Drosophila catalytic subunit gene, DC0. The genomic DNA sequence that includes the coding region for the Drosophila catalytic subunit gene, DC0, is shown. Residues present in the mRNA form typified by cDNA DC0-cDCAT are in capitals. A second, cDNA, DC0-cDpH5, begins at nucleotide 1458 and is contiguous with genomic sequence thereafter. This cDNA, DC0-cDpH5, either represents an incompletely processed RNA or indicates that the cDCAT intron is not always removed. The presumed initiator and upstream methionine codons present in RNA represented by cDCAT are highlighted. The 5' end of RNA of the DC0-cDCAT form has been mapped to position 391 by primer extension of oligonucleotides complementary to nucleotides 424-456 and 495-524. Under- neath is the deduced translation product of the longest ORF, which is 82% identical in sequence to mouse catalytic subunit, Ca (Fig. 9).
the Drosophila catalytic subunit gene, DC0 (Fig. 7). The relative abundance of these RNA forms varies during de- velopment, the shortest species being particularly prom- inent in embryos and the longest in adults. Comparison of a representative eDNA, cDCAT with genomic se- quences indicates a single intron in the 5'-untranslated sequence. The first DC0 exon was shown by primer ex- tension to initiate one nucleotide upstream of cDCAT
{data not shown} and is shared apparently by all four DC0 RNA species (Fig. 7). Probes derived from the 3' ends of two overlapping cDNAs, cDpH5 and cDD3, that extend progressively further 3' (Fig. 8) hybridized to only the two longest (5.1 kb and 4.1 kb) and the longest (5.1 kb) DC0 transcripts, respectively, indicating that the four DC0 RNA species differ principally at their 3' ends, perhaps by the use of different sites of polyadenylation.
1546 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
Figure 7. Temporal expression of catalytic subunit, DC0, and related genes, DC1 and DC2. Single-stranded probes specific for the coding regions of (A) DC0 (3180-2951, written 5' to 3', Fig. 6), (B)DC1 (1067-260, Fig. 10), and (C)DC2 (2017-1770, Fig. 11) were hybridized to equivalent RNA blots of poly(A) + RNA (5 ~g per lane) from the indicated developmental stages (0-24 hr embryos, E; first-larval instar, L1; one-day-old pupae, P1; adult head, H; and adult body, B). As for Fig. 4, all samples con- tained the same amount of intact actin 5C RNA, and similar patterns of bands were seen in the late larval and pupal stages as for early larval and pupae, respectively (data not shown). DC0 and DC2 transcripts are present at all stages and the largest transcript in each case is particularly abundant in adult head. DC1 transcripts are not detectable in larvae, are more prevalent in adult body than head, and are of apparently altered size in embryos. (D) RNA samples [5 p.g poly(A) ÷ RNA per lane] from 0- to 15-hr embryos (E) and adults (A) have been hybridized to single-stranded probes specific for the 5' exon of the DC0 gene (5'; 14777-1471:564-392, Fig. 6}, the 3' end of cDNA, DC0- cDpH5 (M; spanning a region 3.9 to 3.6 kb downstream of the 5' end of the gene, Fig. 8), and the 3' end of eDNA, DC0-cDD3 (3', spanning a region 5.1 to 4.8 kb downstream of the 5' end of the gene, Fig. 8). The latter probe (3') also hybridized strongly to a transcript of 0.7 kb (not shown), indicating either an overlap- ping transcript or that the 3' end of cDNA DC0-cDD3 is com- posite. (E) The use of a probe biased toward DC2-cDH2-specific sequences of DC2 (DC2-5' in Fig. 8; corresponding to DC2- cDH2 nucleotides that replace nucleotides 952-649 in DC2- cDDS, Fig. 11) indicates that the 2.4-kb RNA species corre- sponds to DC2-cDH2 and the 3.0-kb species to DC2-cDDS, confirming that the two major DC2 transcripts have different 5'-untranslated sequences spliced to the same coding region.
Genes related to DCO
Both the mouse and Drosophila catalytic subunit genes recognize several other regions of DNA in the Droso- phila genome under relaxed hybridization conditions (data not shown). Successive attempts were made to iso- late these loci by screening a genomic library with Dro- sophila catalytic subunit gene probes at low stringency. In the second round, previously isolated loci were used to counterscreen at high stringency so that new loci could be discerned immediately. Eleven new distinct loci were isolated and the region responsible for hybrid- ization determined and sequenced. In all cases, portions of the DNA sequence could be translated into amino acid sequences that are conserved among serine- threonine protein kinases, indicating that each of these loci may represent a distinct protein kinase gene (Fig. 9). The chromosomal location of each of these genes is given in Table 1. Several of the genes have been shown to be expressed by hybridization to RNA blots and cDNAs have been isolated for all but one (5-23), indi- cating that they are transcribed. Two of the genes (DG1, DG2) appear to encode proteins akin to the kinase do-
main of mammalian cGMP-dependent protein kinase (Takio et al. 1984). Two others (DC1, DC2) are particu- larly similar to the cAMP-dependent protein kinase cat- alytic subunit and have been analyzed further. There are insufficient sequence data from the remaining putative protein kinase genes to determine whether they are ho- mologs of specific protein kinases characterized in other species. There is substantial similarity between the de- duced protein product of 8-6 and the Saccharomyces cerevisiae CDC28 gene product (52 identities in 94 amino acids; Lorincz and Reed 1984), and the 3-2 de- duced amino acid sequence is similar to that of protein kinase C species (50 identities in 85 residues with bo- vine PKC-et; Parker et al. 1986).
Two different forms of cDNA (DCI-cDH1 and DC1- oDD1) have been isolated for the DC1 gene that are dis- tinguished by the use of different splice acceptor se- quences located 5 bp apart in the genomic DNA (Figs. 8 and 10). The point of divergence is in coding sequence exactly at the carboxy-terminal border of the kinase do- main and therefore leads to the generation of two dif- ferent polypeptides, each expected to exhibit protein ki- nase activity (Fig. 3). One product (DCla) is of the same
GENES & DEVELOPMENT 1547
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
Figure 8. Transcripts of DC0, DCI, and DC2 genes. The DC0 gene is the source of at least four different transcripts (Fig. 7). The coding and 5'-untranslated re- gion of one RNA form, represented by cDNA cDCAT has been determined by comparison of cDNA and ge- nomic sequences. Primer extension showed the 5' end of this RNA to be one nucleotide from the 5' end of DC0-cDCAT. A second cDNA DC0-cDpH5 is contig- uous with genomic sequence around the splice acceptor of DC0-cDCAT and may represent an incompletely pro- cessed RNA. Both cDNAs DC0-cDpH5 and DC0-cDD3 extend further 3' than DC0-cDCAT, and by restriction site comparison to genomic DNA each lacks an intron of about 11 kb. The structure of these cDNAs relative to genomic DNA further 3' is not known, as indicated by the dashed lines. Both 5' and coding region probes from DC0-cDCAT hybridize to all four DC0 RNAs on RNA blots, whereas probes derived from the middle (M) and the extreme 3' end (3') of the apparent transcription unit only detect the two largest and the single largest DC0 RNA species, respectively {Fig. 7). The transcrip- tion pattern of the DC1 gene has been deduced from
comparison of genomic sequence to two cDNAs, DCI-cDH1 and DCI-cDD1. The two cDNAs indicate identical splicing patterns except for the different choice of splice acceptors for the third intron that are located 5 bp apart. The reading frames of the polypeptides encoded by the two RNAs therefore differ beyond this point, which corresponds to the carboxy-terminal border of the kinase domain (Fig. 3). The DC2 gene generates two major transcripts (Fig. 7) that are detected by the probe DC2-coding. Two forms of cDNA, represented by DC2-cDD5 and DC2-cDH2, of roughly equivalent size to the two RNA species (3.0 kb and 2.4 kb, respectively) differ 5' of the coding sequence, presumably as a result of alternative splicing. Consistent with this proposal, a probe containing largely sequences specific to DC2-cDH2 (DC2-5) highlights the shorter of the two RNA species on RNA blots (Fig. 7). The two DC2 cDNAs are represented by dashed lines because the corresponding genomic sequences are not known. For all three genes the termination codon and furthest upstream methionine coding in the long ORF are indicated.
length as the mouse catalytic subunit and of similar se- quence throughout (45% amino acid identity), whereas the other (DClb) is sl ightly longer and is not related in sequence to catalytic subuni t at its unique carboxyl ter- minus. The relative abundance of RNAs represented by these cDNAs is not known. Collectively, they are tran- scribed into a species of about 1.6 kb that is detected in embryonic, pupal, and adult RNA samples (Fig. 7). DC1 transcripts, in contrast to DC0 and DC2, are more abun- dant in adult body than adult head RNA (Fig. 7).
The second catalytic subunit related gene, DC2 is ex- pressed as two major RNA species of 3.0 kb and 2.4 kb, for which corresponding cDNAs have been isolated (Figs. 7 and 8). The complete coding sequence of DC2 has been determined from the longer of these cDNAs (DC2-cDDS, Figs. 8 and 11). Two other independent cDNAs (DC2-cDH2, DC2-cDD6), the longer of which is 2150 bp, appear to derive from the use of an alternative splice donor to pair wi th an acceptor 81 nucleotides up- stream of the initiator codon, although the genomic se- quence in this region is not known (Figs. 8 and 11). As for the DC0 gene, the relative abundance of transcripts of identical coding sequence but different untranslated sequence is regulated during development. In this case the shorter DC2 transcript is only detectable post-pu- pariation (Fig: 7). The inferred amino acid sequence of the DC2 gene product is 49% identical over the entire length of the mouse catalytic subunit but includes an additional 151 amino acids at the amino terminus (Fig.
3). The unique amino-terminal portion of the DC2 gene product is extremely rich in serine, threonine, and acidic residues (Fig. 11) and is not related in sequence in this region to any other known protein sequences (NBRF-PIR protein sequence database December 1987).
Discussion
Drosophila genes have been isolated that can encode proteins very similar in primary sequence, and hence presumably of analogous activity, to the m a m m a l i a n RI and C subunits of the cAMP-dependent protein ki- nase. The conservation in sequence between these genes (71% amino acid identi ty for RI, 82% for C) is somewhat greater than for other kinase genes that have been char- acterized in Drosophila and m a m m a l s (abl - 7 5 % , Hen- kenmeyer et al. 1988; protein kinase C - 6 4 % , Ro- senthal et al. 1987; raf - 6 5 % , Mark et al. 1987; src - 4 0 % , Simon et al. 1985). The deduced amino acid se- quence of the Drosophila RI protein is particularly sim- ilar to its m a m m a l i a n homolog in the two carboxy-ter- mina l cAMP binding domains (see Fig. 1). Clearly, it is more related to m a m m a l i a n RI than RII protein (71% versus 32% identity) and in particular, includes a serine (Set = 96) in a context analogous to that of the bovine RI serine that is phosphorylated by cGMP-dependent pro- tein kinase (Titani et al. 1984) but does not include a sequence analogous to that around the serine in bovine RII protein (RII Ser = 95; Drosophila RI Gly = 94) that
1548 G E N E S & D E V E L O P M E N T
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
~ V A ~ A I ~ K M H ~ M ~ I F ~ L ~ R ! ~ K I ~ T ~ - - - B S ~ Q L R P I 217 ~ ~ I S A L ~ Q I ~ N R D I L H L K I ~ K L R D - - - Q ~ 364 IlIGCVLQA~RGI~MI~ER~I~VKIV~ ~ I ~ V ~ ~ T ~ V ~ E A I D ~ g ~ R N I ~ N E R ~ G K L . ~ K L Q T G - R K ~ ~ V ~ ~ R ~ ~'~LSGLN~QRGIAH~M~~L~EHDNVKIS~M~T
ELLTGVDF~RI~H~Q~SS~LKIA~L~YGSEMKL~-VVV~W~P~VLIAQP~ L Q ~ E N K I ~ L ~ E ~ V K I ~ ~ G M ~ F G D ~ G ~ ~ ~ V L T E T S I
~TQ~GI~S~I~A~HVK~LC~HIQEGIV~~I~ H ~ E K N N I K I A ~
RIL~D~IL~Q~ ~LCGEPi
LH RDLKPEN D DFGF CGTP YLAPEII G GM Y E HRDL A N LV K DFG R Y G P W APE
[NKAVDWWALGVLI YEMAAGYPPFFAD-QP I Q- I YEK IVSGKVRFP S HFS SDLKDLLRNLLQVDLTK%GNLKNG 288 -- " , ~ ~ - ~ - ~ ~ G ~ ~ ~ A 1 290
I~S~F~IIV~FV~RS~AI HNRDVI LMIS~C ICDYKM~YITIQ~RS~ES~TS~L~ND~ 291 ~ - ~ ~ - - i - - ~ ] I L ~ Y D E - ~ F G - ~ L W ~ I EWER~MDP I A ~ K K ~ V ~ R ~ ~ 4 37
D ~ / I ~ g - I ~ L L ~ S ~ P - DRM~- T~q L ~ t<l~I D M I AF D ! S ; ~ I Y ~ M F ' ~ L LT
T[~ I~S AAC I ~ F ~ NRRAY S G T I ~ _ ~ _ I ~ G ~ I~-~V~E G Pl~mHm~lC F ~ M T ~ N D E- T~R GH~IACQMLTQK
D W G Y G PPF I G P L R DVWSFG L E G PY E G R P CP M CW RP F
Mouse C VND IKNHKWFATTDWIAI YQRKVEAPF IPKFKGPGDTSNFDDYEEEE I RVS IN--EKCGKEFTEF 351 D r o s o p h i l a DC0 ~ Q ~ ~ F ~ ] ~ I ~ -.--.:.~-.. " L ~ S T - - ~ A ~ A ~ 353 D ros ophi i a DC 1 S S~V~QGV~G|LNQE~YQUT I SaIAEmL~ ENFIFKDRYKIR I --NRHPEL~ANN 354 Dro s ophi i a DC2 AD~V~R~R~KHLN~NDV~SK~LKP~I L~DVHHD~K~P~KDWKPAKAVDQRDLQY~NDI 502
Figure 9. Catalytic subunit-related kinase genes in Drosophila. The complete amino acid sequences, deduced from cDNA sequence, for the mouse catalytic subunit, C~ (Uhler et al. 1986a); Drosophila catalytic subunit, DC0, and related genes DC1 and DC2 (from amino acid 151] are shown. The mature mouse C~ and DC0 proteins probably lack the amino-terminal methionine residue (Carr et al. 1982), altering residue numbering by one. Partial amino acid sequences deduced from genomic DNA sequence for nine other genes isolated by cross-hybridization to DC0 can also be aligned with each other and with amino acid sequence motifs characteristic of protein kinases, shown on the bottom two lines (Hunter and Cooper 1985). Two of these genes, DG1 and DG2, clearly encode products closely related to the cGMP-dependent protein kinase (Takio et al. 1984). The remaining deduced gene products are more similar in sequence to serine-threonine-specific than tyrosine-specific protein kinases. When compared with other known protein kinase amino acid sequences, substantial similarities between 3-2 and bovine protein kinase C-e~ (50 identities in 85 residues; Parker et al. 1986) and between 8-6 and S. cerevisiae CDC28 (52 identities in 94 residues; Lorincz and Reed 1984) are apparent. The kinase domain of mouse catalytic subunit defined by sequence similarity to other protein kinases is boxed and extends from amino acid 46 to 281.
GENES & DEVELOPMENT 1549
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
The percentage amino acid identity to mouse cAMP-dependent protein kinase catalytic subunit [C~x) [Uhler et al. 1986a) of the deduced products of the putative Drosophila protein kinase genes described in this study and of three representative mammalian serine protein kinases [bovine cGMP-dependent protein kinase, designated G kinase {Takio et al. 1984}; rabbit phosphorylase b kinase {Reimann et al. 1984); and rabbit myosin light-chain kinase (Takio et al. 1985)] is tabulated for regions within the kinase domain {amino acids 46--281 of mouse Ca) and outside {residues 1--45 and 282--351). Only limited amino acid sequence is available for some gene products. The number of amino acids over which the comparisons are made is indicated. The chromosomal location of the putative Drosophila kinase genes as determined by in situ hybridization to polytene chromosomes is also given. {AA) Amino acids.
is phosphorylated by the catalytic subunit of the cAMP- dependent protein kinase {Takio et al. 19841. The de- duced Drosophila catalytic subunit gene product in- cludes all of the conserved features of serine-threonine kinases and is substantially closer in primary sequence to the cAMP-dependent protein kinase than to the ki- nase domain of even its closest acknowledged relative, cGMP-dependent protein kinase (82% versus 42% iden- tity}.
In contrast to the mouse, where two very similar RI genes have been found {C. Clegg et al., in pressl, the Drosophila genome appears to contain only a single RI gene. However, differential splicing within the coding region of the Drosophila gene and the use of different promoters generates transcripts capable of programming the synthesis of RI proteins truncated at the amino ter- minus by 57 and 80 amino acids in addition to the au- thentic full-length RI homolog. Proteolysis studies on the mammalian RII protein suggest that the amino-ter- minal 90 residues of regulatory subunit protein are re- quired for dimerization but not for cAMP and catalytic subunit binding activities (Weldon and Taylor 1985). The two theoretical truncated polypeptide products of the Drosophila RI gene would therefore be expected to act as monomeric cAMP-dependent regulators of cata- lytic subunit activity like the Dictyostelium regulatory subunit, which is 68 amino acids shorter than mamma- lian RII protein (Mutzel et al. 1987}.
The three major transcripts of the Drosophila RI gene all derive from different promoters and therefore might
be expected to be regulated independently. Class I RNA, which codes for the largest RI protein, can be detected at roughly constant level throughout development as can full-length RI protein by labeling with a radioactive cAMP photoaffinity analog. In contrast, class II and IV RNAs have been detected only in adults where class IV RNA, which codes for the shortest RI polypeptide, is the most abundant RI species. The inferred truncated pro- teins have only been visualized directly as products of in vitro translation and, indeed, will be hard to distinguish in vivo from proteolytic products of the full-length pro- tein.
The Drosophila RI gene includes two transcription units of polarity opposite to RI within the first and largest intron. The better-characterized intronic gene (18c) appears to be expressed at high levels (10- to 20-fold higher than RI) only in pupae and adults and to be one of at least six members of a clustered gene family that en- code related polypeptides of unknown function. The po- lyadenylation site of this intronic transcript is 68 nu- cleotides distant from the first exon of the RI gene. Hence, fully processed transcripts of the 18c and RI genes do not overlap.
In mammals there are two {Showers and Maurer 1986; Uhler et al. 1986a, b) and in yeast three (Toda et al. 1987b) very similar genes that encode catalytic subunits of the cAMP-dependent protein kinase. In Drosophila, only one locus that clearly encodes a homologous pro- tein has been found and the coding region of this gene lies on a single exon, making differential RNA pro-
1550 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
M S Q H T S Q Y V F N S K E D 15 TACAACGTAATCCTAGACAACATGAGTCGCGAGTTCGAAGAGCGGTGGAATCACCAGACCCAGTCGCCCTACACCAATCTG 243 Y N V I L D N M S R E F E E R W N H Q T Q S P Y T N L 42 GAAAACTATATAACCCGGGCCGTCCTCGGCAACGGAAGTTTCGGCACTGTGgtgagttaaagggtatttcttagatgtgct 324 E N Y I T R A V L G N G S F G T V 59
acaccttcgaaatataataatcttcagATGCTGGTCAGGGAGAAGAGTGGCAAGAACTACTATGCCGCCAAGATGATGAGC 405 M L V R E K S G K N Y Y A A K M M S 77
AAGGAGGAT~TGGTGCGGCTGAAGCAGGTGGCCCACGTACACAACGAGAAGCACGTCCTGAACGCCGCCCGATTCCCGTTC 486 K E D L V R L K Q V A H V H N E K H V L N A A R F P F 104 CTCATCTACCTGGTGGACTCGACCAAGTGCTTCGACTATCTCTACCTAATCCTCCCGCTGGTCAATGGAGGAGAGCTGTTC 567 L I Y L V D S T K C F D Y L Y L I L P L V N G G E L F 131 AGTTACCATCGCAGgtgagtatccctttctggagtgctcatttcatctgatcacagactgggttccaacgggtttcccccc 648 S Y H R R 136
ccataaaaagAGTTCGCAAGTTCAACGAGAAGCACGCCAGATTCTATGCCGCCCAGGTGGCACTTGCTCTTGAGTATATGC 729 V R K F N E K H A R F Y A A Q V A L A L E Y M H 160
ACAAGATGCACCTCATGTACCGTGATCTGAAGCCGGAGAACATTCTGCTCGATCAGCGCGGCTATATCAAAATAACGGACT 810 K M H L M Y R D L K P E N I L L D Q R G Y I K I T D F 187
TCGGTTTCACTAAGgtggggaaaagacttcacataattcccattgctttcatcctcaaataatcacagCGTGTGGATGGCC 891 G F T K R V D G R 196
GCACATCAACGCTGTGTGGAACCCCGGAATACTTGGCCCCGGAGATTGTTCAACTCCGGCCCTACAACAAATCGGTGGACT 972 T S T L C G T P E Y L A P E I V Q L R P Y N K S V D W 223
GGTGGGCCTTTGGTATCCTAGTGTACGAGTTTGTGGCAGGGCGGTCTCCCTTTGCCATTCACAATCGAGATGTAATCCTGA 1053 W A F G I L V Y E F V A G R S P F A I H N R D V I L M 250
TGTACTCCAAGATCTGCATATGCGACTACAAGATGCCCTCATACTTTACGTCCCAGCTGAGGAGCCTTGTCGAGAGCCTTA 1134 Y S K I C I C D Y K M P S Y F T S Q L R S L V E S L M 277
TGCAGGTGGACACCTCAAAGCGgtgagtggagtgggagtgggtcaaggtgcttccaatgcagtgctcatgccaccccttct 1215 Q V D T S K R 284
L G N S N D G S S D V K S H P W F Q G V D W F G I L a310 K L E R R L Q R R E E S S V V P G R R L V W H S b308
TCAACCAGGAAGTCACCGCCCCCTACCAGCCCACCATTTCCGGCGCCGAAGATCTGTCGAACTTCGAGAACTTCGAGTTCA 1377 N Q E V T A P Y Q p T I S G A E D L S N F E N F E F K a337
Q P G S H R P L P A H H F R R R R S V E L R E L R V Q b335 AGGATCGGTACAAGTCCCGAATAAACCGCCATCCCGAATTGTTTGCGAATTTTTAAATGTCAATGTGT•CTTCGAAATCGT 1458
D R Y K S R I N R H P E L F A N F a354
G S V Q V P N K P P S R I V C E F L N V N V S F E I V b362 CGTTTTCTTTAGTGTTTGCTTAATTTCAAGGGGTGCGCTATCGTGATTGGATTTTGTAAAAGTAGTTT 1526
V F F S V C L I S R S A L S b376
Figure 10. Catalytic subunit-related gene, DCI. The nucleotide sequence of the DC1 gene is shown, with residues present in cDNA DCI-cDH1 in capitals and the amino acid sequence of the deduced primary translation product underneath. The limits of the DC1 transcrip~on unit have not been determined, although both cDNAs, DC 1-cDH1 and DCI-cDD 1, are of similar length (1.3 kb}, close to that estimated from RNA blots for the co~esponding RNA (1.6 kb). The two forms of cDNA indicate the use of different splice acceptors at residues 1219 and 1224. The nucleotides unique to DCI-cDD1 (form DCla) are underlined. The translation product of the mRNA using the more 3' splice acceptor, designated b, diverges beyond this point and is written underneath. The point of divergence, residue 284, corresponds almost exactly to the carboxy-terminal border of the kinase domain (see Fig. 9). All of the splice donor sites conform well to the consensus sequence (Mount 1982) as do the two alternative splice acceptor sites at 1219 and 1224. However, the remaining splice acceptor sites show only minimal expected similarity to the consensus, particularly the acceptor at nucleotide 659.
cessing an unlikely source of polypeptide diversity. However, nucleic acid hybridization has revealed 11 ad- ditional Drosophila genes with substantial sequence similarity to the cAMP-dependent protein kinase cata- lytic subunit. From the limited sequence determined for most of these genes, it seems likely that they encode serine-threonine-specific protein kinases (Fig. 9). cDNAs have been recovered for all but one of these genes, indicating that they are transcribed. Should all of these genes be functional, this observation argues for the existence of many more kinases than have, as yet, been characterized biochemically (Hunter 1987; Hanks et al. 1988}.
Two of the kinase genes isolated, DC1 and DC2, are sufficiently similar to the catalytic subunit of cAMP-de- pendent protein kinase to raise the question of whether their activity is controlled by cAMP by virtue of associa- tion with regulatory subunits. The similarity of mouse
catalytic subunit sequence to DC1 and DC2 is substan- tially lower than to DC0 (45% and 49% versus 82% amino acid identityl, but is marginally higher than to mammalian cGMP-dependent protein kinase {42% amino acid identity; Takio et al. 19841, its nearest known relative. The sequences of the S. cerevisiae TPK1-3 gene products are only 50-53% identical to bo- vine catalytic subunit, but still serve as catalytic sub- units, of cAMP-dependent protein kinase (Toda et al. 1987b). A specific determinant for regulatory subunit binding has not been defined in the catalytic subunit of cAMP-dependent protein kinase, although it is known that alteration of a single residue in the yeast catalytic subunit TPK1 (corresponding to Thr-197 in mouse C~) selectively reduces affinity for the regulatory subunit (Levin et al. 1988) and that a cysteine residue in the same region of porcine catalytic subunit (Cys-199) can be cross-linked to the type II regulatory subunits (First et
GENES & DEVELOPMENT 1551
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
Figure 11. Catalytic subunit-related gene, DC2. The nucleotide sequence of a cDNA, DC2-cDD5, to the DC2 gene is shown. A second cDNA, DC2-cDH2, diverges ~om this sequence 5' of nucleotide 814, presumably at a splice junction, and the unique se- quences of this cDNA are shown above. Neither the location of introns nor the 5' and 3' boundaries of this gene have been determined. The common theoretical translation product of both cDNAs is shown below. The first 150 amino acids are not significantly related to any known protein sequence and are highly enriched in sefine {S), threonine (T), and acidic residues (D, E; 49 S and T in amino acids 1-119, 45 D and E in amino acids 79-175). The rest of the protein is similar in sequence to the mouse catalytic subunit of cAMP-de- pendent protein kinase as presented in Fig. 9.
al. 1988). Indeed, the supposition [Witt and Roskoski 1975} that regulatory subunits inhibit kinase activity by acting as nonproductive substrates and the observed phosphorylation of mammalian and yeast regulatory subunit {Maeno et al. 1974~ Sy and Roselle 1982) R/I by the catalytic subunit argue for substantial overlap be- tween structures required for kinase activity and regula- tory subunit association. Clearly, the question of
whether DC1 and DC2 bind to the regulatory subunit or are regulated independently of cAMP concentration must be pursued biochemically by using specific anti° sera and by overexpression of the gene products.
Ahemat ive splicing of the DC1 gene generates two m R N A species capable of directing the synthesis of pro- teins that differ at their carboxyl termini immediately distal to a common kinase domain. One inferred poly-
1552 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
peptide product is of the same length as cAMP-depen- dent protein kinase catalytic subunit and of similar se- quence throughout; the other is slightly longer and unrelated at the carboxyl terminus. Both would be ex- pected to exhibit kinase activity, but it is not clear whether the proteins would differ principally in their mode of activation or substrate specificity.
In its simplest form the role of cAMP-dependent pro- tein kinase in signal transduction demands only a single constitutively expressed form of the enzyme. The speci- ficity of its action in a given cell can be generated by restricted expression of ligands, receptors, and intracel- lular effectors subject to regulation by phosphorylation. However, it is clear that several forms of cAMP-depen- dent protein kinase are expressed in yeast, mammals, Aplysia, and now Drosophila, and that the levels of these isoforms differ among cell types and during devel- opment (Eppler et al. 1982; Uhler et al. 1986a, b; Toda et al. 1987a, bl. It now appears also that several RNA forms encode each of the components of cAMP-dependent pro- tein kinase in Drosophila and that the relative abun- dance of some of these forms varies during development. The same Drosophila RI subunit homolog may be syn- thesized from RNAs that differ at their extreme 5' ends as a result of multiple clustered transcription initiation sites, in the 5'-untranslated sequence due to differential splicing and in the 3'-untranslated sequence due to dif- ferential polyadenylation. The Drosophila catalytic sub- unit homolog is transcribed into four distinguishable RNA species, even though the coding sequence lies on a single exon. These four catalytic subunit RNAs are found in different proportions in embryos and adults (Fig. 7). It is not known if the observed heterogeneity in RNA forms is functionally important and, if so, whether it contributes to regulation of RNA localization, sta- bility, or translation. An effect of upstream methionine codons on translational efficiency has been demon- strated in yeast (Mueller and Hinnebusch 1986) and in mammals (Marth et al. 1988) and may be of relevance to the Drosophila RI gene. Type I RI RNA species include between zero and three methionine codons prior to the presumed translation initiation codon, depending on the exact 5' end and which of three splicing patterns (Ia, Ib, or Ic) is adopted (Fig. 1). Neither the upstream nor the putative initiator is in a context commonly found for initiation codons (Kozak 1984a; Cavener 1987) and con- sidered to be efficiently recognized as such by ribosomes (Kozak 1984b). This is true also for the DC0, DC1, and DC2 genes, all of which include upstream methionine codons in at least some transcripts and one of which (DC2) undergoes differential splicing in the 5'- untrans- lated sequence. Thus, study of the genes for cAMP-de- pendent protein kinase in Drosophila has uncovered a number of potentially regulated RNA and protein iso- forms that point to unanticipated sophistication in the cAMP signal transduction system.
The identification of clear Drosophila homologs of RI and catalytic components of the cAMP-dependent pro- tein kinase and knowledge of their chromosomal loca- tion should allow the isolation of null mutations in
these genes and hence insight into the biological role of cAMP-dependent protein kinase in a whole organism. It should also be possible to perturb cAMP-dependent pro- tein kinase activity in Drosophila by germ-line transfor- mation of constructs designed to overexpress wild-type or specifically altered versions of those genes that lead to constitutive activity of catalytic subunit or constitutive inhibitory activity of the regulatory subunit (Clegg et al. 1987; Levin et al. 1988).
Materials and methods
Isolation of genomic DNA and cDNA clones
Genomic clones for Drosophila RI, DC0, and 11 other kinase genes were isolated from a library of sheared Drosophila Canton S genomic DNA in Charon 4A (Maniatis et al. 1978) by low-stringency hybridization (22% formamide, 5 x SSC, 10 × Denhardt's, 42°C; wash 2 x SSC, 42°C) to bovine RI, mouse C, and Drosophila DC0 probes, respectively. Corresponding cDNA clones in kgtl0 were isolated from a variety of libraries constructed from RNA of adult heads, third-instar larvae, eye imaginal discs, or total imaginal discs (A.C. Cowman and G.M. Rubin, unpubl.; H. Steller and G.M. Rubin, unpubl.) by high- stringency hybridization (10 x Denhardt's, 5 x SSC, 60°C; wash 0.5 x SSC, 65°C).
Sequence determination of genes
Complete cDNAs or coding regions of genomic DNA deter- mined by low-stringency hybridization to appropriate probes were sequenced by the dideoxy termination method {Sanger et al. 1977), largely using sonicated subfragments cloned into M13mpl0 but also by directed cloning into M13mpl8 and mpl9 vectors (Messing 1983). The DNA sequence presented here has been determined on both strands of either cDNA or genomic DNA for all transcribed regions of the RI, DC0, DC1, and DC2 genes.
RNA analysis
Poly(A) + RNA was purified by one cycle of binding to oligo{dT)-cellulose of RNA prepared from a variety of develop- mental stages, essentially by tissue disruption and phenol ex- traction {O'Hare et al. 1983). RNA blots of RNA fractionated on RNA formaldehyde gels and S 1-nuclease protection were per- formed using standard techniques (Maniatis et al. 1982) and uniformly labeled single-stranded probes derived from recombi- nant M13 single-stranded templates. Primer extension was in each case performed with at least two different oligonucleo- tides of about 30 nucleotides labeled at the 5' end by [~/-a2P]ATP and T4 polynucleotide kinase as described previously (Zuker et al. 1985).
Chromosomal location of genes
In situ hybridization to polytene chromosomes used nick- translated biotinylated probes from cDNAs or genomic DNAs including sequenced coding regions, as described previously (Zuker et al. 1985).
A c k n o w l e d g m e n t s
The initial part of this work was supported by American Cancer Society Grant ACS 41. D.K. is a senior postdoctoral fellow of the American Cancer Society, California Division and formerly
GENES & DEVELOPMENT 1553
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
was a SERC/NATO postdoctoral fellow while conducting this work. Special thanks are due to Todd Laverty for performing all in situ hybridizations to chromosomes, to Stan McKnight (Uni- versity of Washington, Seattle) for providing mammalian probes and information prior to publication, to David Baker and Mark Fortini for determination of catalytic subunit gene se- quences, to David Bowtell, Steve Hanks, Tony Hunter, Stan McKnight, Mike Simon, and Andrew Tomlinson for critical contributions to the manuscript, and to Karen Ronan for con- siderable help in preparing the manuscript.
N o t e
Sequence data described in this paper have been submitted to the EMBL/GenBank Data Libraries under accession number Y00220.
R e f e r e n c e s
Avenet, P., F. Hofmann, B. Lindemann. 1988. Transduction in taste receptor cells requires cAMP-dependent protein ki- nase. Nature 331: 351-354.
Bellen, H.J., B.K. Gregory, C.L. Olsson, and J.A. Kiger, Jr. 1987. Two Drosophila learning mutants, dunce and rutabaga, provide evidence for a maternal role for cAMP in embryo- genesis• Dev. Biol. 121: 432-444.
Byers, D., R.L. Davis, and J.A. Kiger, Jr. 1981. Defect in cAMP phosphodiesterase due to the dunce mutation of learning in Drosophila melanogaster. Nature 289: 79- 81.
Cart, S.A., K. Biemann, S. Shoji, D.C. Parmelee, and K. Titani. 1982. n-Tetradecanoyl is the NH2-terminal blocking group of the catalytic subunit of cAMP-dependent protein kinase from bovine cardiac muscle. Proc. Natl. Acad. Sci. 79: 6128-6131•
Cavener, D.R. 1987. Comparison of the consensus sequence flanking translational start sites in Drosophila and verte- brates. Nucleic Acids Res. 15: 1353-1361.
Clegg, C.H., L.A. Correll, G.G. Cadd, and G.S. McKnight. 1987. Inhibition of intracellular cAMP-dependent protein kinase using mutant genes of the regulatory type I subunit. J. Biol. Chem. 262: 13111-13119.
Coffino, P., H.R. Bourne, U. Friedrich, J. Hochman, P.A. Insel, I. Lemaire, K.L. Melmon, and G.M. Tomkins. 1976. Molecular mechanisms of cyclic AMP action: A genetic approach. Re- cent Prog. Horm. Res. 32: 669-684.
Cohen, P. 1982. The role of protein phosphorylation in neural and hormonal control of cellular activity. Nature 296: 613- 620.
Comb, M., N.C. Birnberg, A. Seasholtz, E. Herbert, and H.M. Goodman. 1986. A cyclic AMP- and phorbol ester-inducible DNA element. Nature 323: 353-356.
Deutsch, P.J., J.L. Jameson, and J.F. Habener. 1987. cAMP re- sponsiveness of human gonadotropin-c~ gene transcription is directed by a repeated 18 bp enhancer. J. Biol. Chem. 262: 12169-12174.
Eppler, C.M., J. Palazzolo, and J.H. Schwartz. 1982. Character- ization and localization of cAMP binding proteins in the nervous system of Aplysia. J. Neurosci. 2: 1692-1704.
First, E.A., J.Bubis, and S.S. Taylor. 1988. Subunit interaction sites between the regulatory and catalytic subunits of cAMP-dependent protein kinase. I. Biol. Chem. 263: 5176- 5182.
Foster, J.L., J.J. Guttman, L.M. Hall, and O.M. Rosen. 1984. Drosophila cAMP-dependent protein kinase. J. Biol. Chem. 259: 13049-13055.
Foster, J•L., G.C. Higgins, and F.R. Jackson. 1988. Cloning, se- quence and expression of the Drosophila cAMP-dependent protein kinase catalytic subunit gene. J. Biol. Chem. 263: 1676-1681.
Fyrberg, E.A., J.W. Mahaffey, B.J. Bond, and N. Davidson. 1983. Transcripts of the six Drosophila actin genes accumulate in a stage- and tissue-specific manner. Cell 33:115-123.
Gilman, A.G. 1984. G proteins and dual control of adenylate cyclase. Cell 36: 577-579.
Hanks, S.K., A.M. Quinn, and T. Hunter. 1988. The protein ki- nase family: conserved features and deduced phylogeny of the catalytic domains. Science 241:42-52.
Haydon, P.G., D.P. McCobb, and S.B. Kater. 1984. Serotonin selectively inhibits growth cone motility and synapto- genesis of specific identified neurons. Science 226: 561- 564.
Hashimoto, S., W. Schmid, and G. Schiitz. 1984. Transcrip- tional activation of the rat liver tyrosine aminotransferase gene by cAMP. Proc. Natl. Acad. Sci. 81: 6637-6641.
Henkemeyer, M.J., R.L. Bennett, F.B. Gertler, and F.M. Hoff- mann. 1988. DNA sequence, structure, and tyrosine kinase activity of the Drosophila melanogaster Abelson proto-on- cogene homolog. Mol. Cell. Biol. 8: 843-853.
Hunter, T. and J.A. Cooper. 1985. Protein-tyrosine kinases. Annu. Rev. Biochem. 54:897-930.
Hunter, T. 1987. A thousand and one protein kinases. Cell 50: 823-829.
Kandel, E.R. and J.H. Schwartz. 1982. Molecular biology of learning: modulation of transmitter release. Science 218: 433-443.
Kiger, J.A., Jr. and H.K. Salz. 1985. Cyclic nucleotide metabo- lism and physiology of the fruit fly Drosophila melana- gaster. Adv. Insect Physiol. 18: 141-179.
Kozak, M. 1984a. Compilation and analysis of sequences up- stream from the translational start site in eukaryotic mRNAs. Nucleic Acids Res. 12: 857-872.
1984b. Point mutations close to the AUG initiator codon affect the efficiency of translation of rat preproinsulin in vivo. Nature 308: 241-246.
• 1987. Effects of intercistronic length on the efficiency of reinitiation of eukaryotic ribosomes. Mol. Cell. Biol. 7: 3438-3445.
Krebs, E.G. and J.A. Beavo. 1979. Phosphorylation-dephosphor- ylation of enzymes. Annu. Rev. Biochem. 48: 923-959.
Kuo, J.F. and P. Greengard. 1969. Cyclic nucleotide-dependent protein kinases IV. Widespread occurrence of cAMP-depen- dent protein kinase in various tissues and phyla of the an- imal kingdom. Proc. Natl. Acad. Sci. 64: 1349-1355.
Lankford, K.L., F.G. DeMello, and W.L. Klein. 1988. Dl-type dopamine receptors inhibit growth cone motility in cultured retina neurons: evidence that neurotransmitters act as mor- phogenic growth regulators in the developing nervous system. Proc. Natl. Acad. Sci. 85: 2839-2843.
Lamers, W.H., R.W. Hanson, and H.M. Meisner. 1982. cAMP stimulates transcription of the gene for cytosolic phos- phoenolpyruvate carboxykinase in rat liver nuclei. Proc. Natl. Acad. Sci. 79: 5737-5741.
Lee, D.C., D.F. Carmichael, E.G. Krebs, and G.S. McKnight. 1983. Isolation of a eDNA clone for the type I regulatory subunit of bovine cAMP-dependent protein kinase. Proc. Natl. Acad. Sci. 80: 3608-3612.
Levin, L.R., J. Kuret, K.E. Johnson, S. Powers, S. Cameron, T. Michaeli, M. Wigler, and M.J. Zoller. 1988. A mutation in the catalytic subunit of cAMP-dependent protein kinase that disrupts regulation. Science 240: 68-70.
Livingstone, M.S., P.P. Sziber, and W.G. Quinn. 1984. Loss of
1554 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
calcium/calmodulin responsiveness in adenylate cyclase of rutabaga, a Drosophila leaming mutant. Cell 37: 205-215.
Lorincz, A.T. and S.I. Reed. 1984. Primary structure homology between the product of yeast cell division control gene CDC28 and vertebrate oncogenes. Nature 307: 183-185.
Maeno. H., P.L. Reyes, T. Ueda, S.A. Rudolph, and P. Green- gard. 1974. Autophosphorylation of cAMP-dependent pro- tein kinase from bovine brain. Arch. Biochem. Biophys. 164: 551-559.
Maller, ].L. and E.G. Krebs. 1977. Progesterone-stimulated meiotic cell division in Xenopus oocytes. I. Biol. Chem. 252: 1712-1718.
Maniatis, T. and R.C. Hardison, E. Lacy, I. Lauer, C. O'Connell, D. Quon, G.K. Sim, and A. Efstratiadis. 1978. The isolation of structural genes from libraries of eukaryotic DNA. Cell 15: 687-701.
Maniatis, T., E.F. Fritsch, and ]. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Labora- tory, Cold Spring Harbor, New York.
Mark, G.E., R.I. MacIntyre, M.E. Digan, L. Ambrosio, and N. Perrimon. 1987. Drosophila melanogaster homologs of the ra/oncogene. Mol. Cell. Biol. 7: 2134-2140.
Marth, I.D., R.W. Overell, K.E. Meier, E.G. Krebs, and R.M. Perlmutter. 1988. Translational activation of the lck proto- oncogene. Nature 332:171-173.
Matsumoto, K., I. Uno, Y. Oshima, and T. Ishikawa. 1982. Iso- lation and characterization of yeast mutants deficient in ad- enylate cyclase and cAMP-dependent protein kinase. Proc. Nat]. Acad. Sci. 79: 2355-2359.
Messing, I. 1983. New M13 vectors for cloning. Methods En- zymol. 101: 20-78.
Montminy, M.R., K.A. Sevarino, J.A. Wagner, G. Mandel, and R.H. Goodman. 1986. Identification of cAMP-responsive el- ement within the rat somatostatin gene. Proc. Natl. Acad. Sci. 83: 6682-6686.
Montminy, M.R. and L.M. Bilezikjian. 1987. Binding of a nu- clear protein to the cAMP response element of the somato- statin gene. Nature 328: 175-178.
Mount, S.M. 1982. A catalogue of splice junction sequences. Nucleic Acids Res. 10: 459-472.
Mueller, P.P. and A.G. Hinnebusch. 1986. Multiple upstream AUG codons mediate translational control of GCN4. Cell 45: 201-207.
Mutzel, R., M.-L. Lacombe, M.-N. Simon, I. deGunzberg, and M. Veron. 1987. Cloning and cDNA sequence of the regula- tory subunit of cAMP-dependent protein kinase from Dic- tyostelium discoideum. Proc. Natl. Acad. Sci. 84: 6-10.
Nakagawa. I., D. vonder Ahe, D. Pearson, B.A. Hemmings, S. Shibahara, and Y. Nagamine. 1988. Transcriptional regula- tion of a plasminogen activator gene by cAMP in a homolo- gous cell-free system. ~ Biol. Chem. 263: 2460-2468.
Nakamura, T. and G.H. Gold. 1987. A cyclic nucleotide-gated conductance in olfactory receptor cilia. Nature 325: 442- 444.
Nestler, E.J., S.I. Walaas, and P. Greengard. 1984. Neuronal phosphoproteins: physiological and clinical implications. Science 225: 1357-1364.
Nimmo, H.G. and P. Cohen. 1977. Hormonal control of protein phosphorylation. Adv. Cycl_ic Nucleotide Res. 8: 146-266.
O'Hare, K., R. Levis, and G.M. Rubin. 1983. Transcription of the white locus in Drosophila melanogaster. Proc. Natl. Acad. Sci. 80: 6917-6921.
Osterreider, W., G. Brum, I. Hescheler, W. Trautwein, V. Flock- erzi, and F. Hofmann. 1982. Injection of subunits of cAMP- dependent protein kinase into cardiac myocytes modulates Ca 2 current. Nature 298: 576-578.
Pace, U., E. Hanski, Y. Salomon, and D. Lancet. 1985. Odorant sensitive adenylate cyclase may mediate olfactory recep- tion. Nature 316: 255-258.
Parker, P.J., L. Coussens, N. Totty, L. Rhee, S. Young, E. Chen, S. Stabel, M.D. Waterfield, and A. Ullrich. 1986. The com- plete primary structure of protein kinase C-the major phorbol ester receptor. Science 233: 853-859.
Quinn, W.G. and R.J. Greenspan. 1984. Leaming and courtship in Drosophila: two stories with mutants. Annu. Rev. Neur- osci. 7: 141-179.
Ran, W., M. Dean, R.A. Levine, C. Henkle, and J. Campisi. 1986. Induction of c-los and c-myc mRNA by epidermal growth factor or calcium ionosphere is cAMP dependent. Proc. Natl. Acad. Sci. 83: 8216-8220.
Reimann, E.M., K.Titani, L.H. Ericsson, R.D. Wade, E.H. Fischer, and K.A. Walsh. 1984. Homology of the ~/subunit of phosphorylase b kinase with cAMP-dependent protein ki- nase. Biochemistry 23: 4185-4192.
Rosenthal, A., L. Rhee, R. Yadegari, R. Paro, A. Ullrich, and D.V. Goeddel. 1987. Structure and nucleotide sequence of a Drosophila mdanogaster protein kinase C gene. EMBO J. 6: 433-441.
Rydel, R.E. and L.A. Greene. 1988. cAMP analogs promoter sur- vival and neurite outgrowth in cultures of rat sympathetic and sensory neurons independently of nerve growth factor. Proc. Natl. Acad. Sci. 85: 1257-1261.
Sanger, F., S. Nicklen, and A.R. Coulson. 1977. DNA se- quencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. 74: 5463-5467.
Shin, D.-Y., K. Matsumoto, H. Iida, I. Uno, and T. Ishikawa. 1987. Heat-shock response of S. cerevisiae mutants altered in cAMP-dependent protein phosphorylation. Mol. Cell. Biol. 7: 244-250.
Showers, M.O. and R.A. Maurer. 1986. A cloned bovine cDNA encodes an alternate form of the catalytic subunit of cAMP- dependent protein kinase. J. Biol. Chem. 261: 16288-16291.
Siegelbaum, S.A., J.S. Camardo, and E.R. Kandel. 1982. Sero- tonin and cAMP close single K + channels in Aplysia sen- sory neurons. Nature 299: 413-417.
Simon, M.A., B. Drees, T. Kornberg, and J.M. Bishop. 1985. The nucleotide sequence and the tissue-specific expression of Drosophila c-src. Cell 42: 831-840.
Sutherland, E.W. 1972. Studies on the mechanism of hormone action. Science 177: 401-408.
Sy, J. and M. Roselle. 1982. Phosphorylation of the regulatory subunit of yeast cAMP-dependent protein kinase. Proc. Natl. Acad. Sci. 79: 2874-2877.
Takio, K., S.B. Smith, E.G. Krebs, K.A. Walsh, and K. Titani. 1982. Primary structure of the regulatory subunit of type II cAMP-dependent protein kinase from bovine cardiac muscle. Proc. Natl. Acad. Sci. 79: 2544-2548.
Takio, K., R.D. Wade, S.B. Smith, E.G. Krebs, K.A. Walsh, and K. Titani. 1984. cGMP dependent protein kinase, a chimeric protein homologous with two separate protein families. Bio- chemistry 23: 4207-4218.
Takio, K., D.K. Blumenthal, A.M. Edelman, K.A. Walsh, E.G. Krebs, and K. Titani. 1985. Amino acid sequence of an active fragment of rabbit skeletal muscle myosin light chain ki- nase. Biochemistry 24: 6028-6037.
Tanaka, K., K. Matsumoto, and A. Toh-e. 1988. Dual regulation of the expression of the polyubiquitin gene by cAMP and heat shock in yeast. EMBO J. 7: 495-502.
Titani, K., T. Sasagawa, L.H. Ericsson, S. Kumar, S.B. Smith, E.G. Krebs, and K.A. Walsh. 1984. Amino acid sequence of the regulatory subunit of bovine type I cAMP-dependent protein kinase. Biochemistry 23: 4193-4199.
GENES & DEVELOPMENT 1555
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
Toda, T., S. Cameron, P. Sass, M. Zoller, J.D. Scott, B. McMullen, M. Hurwitz, E.G. Krebs, and M. Wiglet. 1987a. Cloning and characterization of BCY1, a locus encoding a regulatory subunit of the cAMP-dependent protein kinase in Saccharomyces cerevisiae. Mol. Cell. Biol. 7: 1371-1377.
Toda, T., S. Cameron, P. Sass, M. Zoller, and M. Wigler. 1987b. Three different genes in S. cerevisiae encode the catalytic subunits of the cAMP-dependent protein kinase. Cell 50: 277-287.
Tonosaki, K. and M. Funakoski. 1988. Cyclic nucleotides may mediate taste transduction. Nature 331: 354-356.
Uhler, M.D., D.F. Carmichael, C.D. Lee, J.C. Chrivia, E.G. Krebs, and G.S. McKnight. 1986a. Isolation of cDNA clones coding for the catalytic subunit of mouse cAMP-dependent protein kinase. Proc. Natl. Acad. Sci. 83: 1300-1304.
Uhler, M.D., J.C. Chrivia, and G.S. McKnight. 1986b. Evidence for a second isoform of the catalytic subunit of cAMP-de- pendent protein kinase. J. Biol. Chem. 261: 15360-15363.
Walter, U., I. Uno, A.Y.C. Liu, and P. Greengard. 1977. Identifi- cation, characterization, and quantitative measurement of cyclic AMP receptor proteins in cytosol of various tissues using a photoaffinity ligand. J. Biol. Chem. 252: 6494- 6500.
Watanabe, Y., Y. Iino, K. Furuhata, C. Shimoda, and M. Yamo- moto. 1988. The S. pombe mei2 gene encoding a crucial molecule for commitment to meiosis is under the regula- tion of cAMP. EMBO J. 7: 761-767.
Weldon, S.L. and S.S. Taylor. 1985. Monoclonal antibodies as probes for functional domains in cAMP-dependent protein kinase II. I. Biol. Chem. 260: 4203-4209.
Witt, J.J. and R. Roskoski, Jr. 1975. Bovine brain cAMP-depen- dent protein kinase. Mechanism of regulatory subunit inhi- bition of the catalytic subunit. Biochemistry 14: 4503-4507.
Wynshaw-Boris, A. T.G. Lugo, J.M. Short, R.E.K. Foumier, and R.W. Hanson. 1984. Identification of a cAMP regulatory re- gion in the gene for rat cytosolic phosphoenolpyruvate car- boxykinase. J. Biol. Chem. 259: 12161-12169.
Zuker, C., A. Cowman, and G.M. Rubin. 1985. Isolation and structure of a rhodopsin gene from Drosophila melano- gaster. Cell 40: 851-858.
1556 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Press on February 24, 2020 - Published by genesdev.cshlp.orgDownloaded from
