- Email: [email protected]
Organization and expression of Drosophila tropomyosin genes
J. Mol. Biol.
(1982)
Organization
162, 231-250
and Expression of Drosophila Genes VWOKIA
Tropomyosin
L. RAI”WH. ROBERT 1’. STO~I
Department of Biological Chemistry University of Illinois Medical Center Chicago, Ill. 60612, U.S.A.
Department of Biology Massachusetts Institute of Technology Cambridge, Mass. 02139, lJ.8.A.
(Recuivd 1 April
198%)
A genomic clone,containing two tropomyosin-coding sequences has been isolated from a library of Drosophila melanogastrr DNA and identified by a positive hybridization selection and an in vitro translation procedure. In vitro translation yielded two products that comigrated aith chicken and Drosophila tropomyosins in sodium dodecyl sulfate/polyacrylamide gels and underwent the mobility shift characteristic of vertebrate tropomyosins in sodium dodecyl sulfate/urea/polyacrylamide gels. The Drosophila polypeptides also shared several proteolytic fragments with chicken tropomyosins. The cloned DNA hybridized to a single site in region 88F 2-5 on the right arm of chromosome 3 of polytene chromosomes and to the set of restriction fragments in genomic DNA predicted from the cloned sequences, indicating that similar tropomyosin-coding sequences are not located at other sites in the genome. The 18 x IO3 base-pair cloned segment contains three regions complementary to Drosophila embryo RNA separated by non-coding sequences. Two of these coding regions encode tropomyosin I and tropomyosin II ; no protein product has been identified for the third coding region. The expression of the two tropomyosin genes is developmentally regulated during embryo development and in primary myogenic cultures, with abundant transcripts occurring at the onset of muscle cell fusion. The third coding region is homologous to abundant RK’A transcripts found in earlier stages of embryo development. in primary myogenic cultures and in the Drosophila Kc cultured cell line.
1. Introduction Tropomyosin is an abundant contractile troponin, tropomyosin regulates the myosin
(Ebashi
& Endo,
t Present address: F.1t.C:.
Institut
protein of muscle calcium-sensitive
1968). In addition fuer Biologic,
cells. In association interaction of actin
to the tropomyosin
Medizinische
Hochschule
found
with and
in muscle
Luebeck, D-2400 Lurheck
1,
“31 0022-2836/82/340231-20
$03.00/O
!CI 1982 Academic Press Inc. (London) Ltd.
“32
V. 1,. HAIJT(‘H
ET Al,.
cells, non-muscle forms of tropomyosin have been isolated from several different tissues (Cohen 8: Cohen, 1972; Fine et al., 1973; Fine & Blitz, 1975). The function of non-muscle tropomyosin is not known but it, has been suggested that these tropomyosins participate in the contractile activities of the non-muscle cells (Smillie, 1979). Although tropomyosins have been isolated from several invertebrates (Kominz rt in al., 1962: Woods & Pont, 1971), these proteins have been studied more intensively vertebrates. These studies have revealed multiple forms oftropomyosin. Two major forms of vert,ebrate muscle tropomyosin subunits have been idernified: Ztropomyosin has a molecular weight of approximately 34,000, while /3-tropomyosin has a molecular weight of approximately 36,000 (Buckingham: 1977). The two tropomyosins from rabbit skeletal muscle have been seyuenced and found to differ in 39 out of 284 amino acid residues (Stone & Smillie, 1978; Mak et al., 1980). The two muscle forms of tropomyosin are highly conserved among different organisms (Cummins & Perry, 1974; Fine & Blitz, 1975). Non-muscle forms of tropomyosin have a molecular weight of approximately 30,000 with physical properties and peptide maps similar to those of the muscle subunits (Fine & Blitz, 1975). The nonmuscle forms, however, appear to differ substantially from the muscle forms at the amino and carboxy ends of the molecule (Cote et al., 1978). The biosynthesis of tropomyosin is developmentally regulated during muscle myogenesis along with the other major proteins of the contractile apparatus (Carmon et al., 1978; Moss & Schwartz, 1981). The two forms of muscle tropomyosin show different patterns of regulation. Each muscle has a characteristic ratio of CItropomyosin to /3-tropomyosin (Cummins & Perry, 1974) and this ratio may change during development (Amphlett et cd., 1976: Roy et al., 1976). Polymorphic forms of both ,I- and p-tropomyosin have been detected in vertebrates. Some of these multiple forms are produced by phosphorylation (Montarras et al., 1981). However, other forms differ in primary sequence (Mak et al., 1980) or can be identified in cell-free translation products of messenger RNA (Carmon et al., 1978), suggesting the existence of multiple tropomyosin genes. We report here the isolation and parGal characterization of two tropomyosin genes from Drosophila melanngaster contained in one recombinant DNA clone from a genomic library. These genes are separated by a few thousand base-pairs of DNA and are found at a single chromosomal locus. The proteins encoded by these genes share several of the features of the vertebrate tropomyosins described above. RNAs homologous to these genes encode different tropomyosin subunits and their is developmentally regulated during embryo development and in expression primary myogenic cultures. This is the first reported isolation of structural genes for tropomyosin.
2. Materials
and Methods
(a) DNA isolatiort The Drosophila melanogaster (Canton S) genomic DNA library was a gift from Dr Tom Maniatis. Small quantities of recombinant DNA from the phage were isolated by the PDS method of Blattner, which accompanies the Charon phage Drosophila library. Larger preparations were isolated from C&-purified phage as described by Maniatis et al. (1978).
Drosophila
TROPOMYOGIK
GENES
233
Genomic DNA from the Canton S strain of Drosophila was isolated as described by Bingham et al. (1981). Plasmid DNA was isolated from lo-ml cultures or from 1 liter cultures according to Kahn et al. (1979). (b) RNA
isolation
Frozen embryos (Oregon R) at 12 to 24 h of development were provided by Dr Anthony Mahowald. The frozen embryos were thawed directly in lysis buffer (30 mM-Tris (pH 8.3), 100 mM-h’ac], 10 mMCaCl,, 1% sodium dodecyl sulfate, to which 0.5% diethyl pyrocarbonate was added just prior to use) and total RNA was isolated by extraction with phenol/chloroform as described previously (Scott et al., 1979). Staged embryos (Oregon-R, provided by Dr Sarah Elgin) were collected at 25°C over a 2 or 4-h period and aged at 25°C to give 1 to 3, 1 to 59 to 11, or 19 to 23.h-old embryos. Drosophila muscle cell cultures were prepared as described previously (Seecof et al., 1973; Storti et al., 1978). Kc cells were maintained in spinner culture at 25°C. Cytoplasmic RNA from the embryos, myogenic cultures, and Kc cells was isolated by extraction with phenol/chloroform of Triton X-100.lysed cells as described by Storti et al. (1980). 411 RNAs were fractionated by oligo(dT)-cellulose chromatography (Spradling et al., 1977). (c) Myo$bril
protein
extraction
Myofibrils were prepared from 10 g of live second and third instar Canton 8 larvae by the procedure of Bullard et al. (1973). The final washed myofibril pellet was resuspended in rlectrophoresis buffer. (d) Hybrid-selection
translation,
assay
Hybrid-selection translation was performed according to Ricciardi et al. (1979). Approximately 1 to 5 pg of DNA was heat-denatured and spotted on one-half of a 10 mm diameter nitrocellulose filter, air-dried and baked for 2 h at 80°C in vacua. When a single restriction fragment of DNA was used the DNA-containing filter was cut from a Southern (1975) blot. Hybridization was for 2 to 4 h at 37°C in 100 ~1 of 50% deionized formamide (MCB), O-75 M-NaCl, 2 mM-EDTA, 0.1 M-Tris (pH 7.5), 0.2% SDS? and approximately 5 pg of poly(A)+ RNA from 12 to 24-h embryos. Following hybridization the filters were washed 5 times in 150 mw-NaCl, 15 mM-sodium citrate, 0.5% SDS at 60°C and twice in 10 mM-Tris (pH 7.2), 2 mM-EDTA at 60°C. The filters were boiled for 1 min in the presence of 2 mM-EDTA (pH 7.0) and 10 pg calf liver transfer RNA to dissociate the hybrid RNA. The RNA was collected by precipitation with ethanol and translated in the presence of 40&i [35S]methionine (800 to 1200 Ci/mmol, New England Nuclear.) Translation was in a microccal nuclease-treated rabbit reticulocyte lysate cell-free protein-synthesizing system as described bv Pelman $ Jackson (1976). After translation the reactions were treated with 2 pg of pancreatic RNAase for 5 min at, 37°C. (e) SDSJpolyacrylamidu gel &ctrophowsis of proteins One-dimensional electrophoresis was in SDS/l3O/d polyacrylamide slab gelsand performed as described by Laemmli (1970). Two-dimensional gel electrophoresis was according to O’Farrell (1975) with modifications as described previously (Storti at al., 1978). In some experiments the second-dimension gel was supplemented with 8 M-urea. Gels were stained with Coomassie brilliant blue and either dried for autoradiography or fluorographed (Bonner & Laskry, 1974) and exposed to Kodak XR-5 or XAR-5 film.
(f) Partial proteolytic peptide mapping [35S]methionine-labelled tropomyosin and actin were synthesized by translation in zjitro of 12 to 24-h poly(A)+ embryo RNA in reticulocyte lysates. The tropomyosin and actin were t Abbreviations used: SDS, sodium dodecgl sulfate: kb. lo3 base-pairs or bases.
234
V. L. BAIJTCH
ET AL.
separated by one-dimensional gel electrophoresis, located by autoradiography and cut from the dried gels. The labelled proteins were eluted electrophoretically into dialysis bags, dialyzed against 0.1% SDS and concentrated (Storti & Rich, 1976). Partial chymotryptic digestion was according to Cleveland et al. (1977). A total of 30 pg purified chicken skeletal muscle tropomyosin was added to each labelled protein, and 10 pg of freshly prepared chymotrypsin (Worthington) was added. Digestion was for 10 min at 37°C. The products were separated in a 15% polyacrylamide slab gel, stained with Coomassie blue. destained and fluorographed. The stained chicken polypeptides were used for comparison Drosophila tropomyosin polypeptides. with [35S]methionine-labelled (g) Radiolahelling
of nucleic acids
DNA was labelled by nick-translation according to the protocol of Maniatis et al. (1975) with some modifications. DNA (1 pg) was mixed with 60 to 80 pmol of [a-32P]dATP and/or dCTP (600 to 800 Ci/mmol, New England Nuclear) and 400 pmol of each of the other unlabelled dNTPs. The same protocol was also used with 180 pmol [3H]dATP and [3H]TTP in situ. precursors to generate 3H-labelled hDm85 DNA for hybridizations [32P]cDNA complementary to poly(A)+ RNA was synthesized by a modification of the method of Friedman & Rosbash (1977). [5’-32P]RNA was prepared in vitro by brief treatment with alkali, labelling with [y-32P]ATP (>2000 Ci/mmol, New England Nuclear) and bacteriophage T4 polynucleotide kinase (New England Biolabs) according to Maizels (1976). (h) DNA -agarose gel elrctrophmesis and j&r hybridization Horizontal agarose gels were l(& or 1.5% in Tris-acetate running buffer (40 mw-Tris (pH 8.3), 20 mw-sodium acetate, 2 mM-EDTA), with 5% glycerol. DI?II’Awas electrophoresed for 16 h at 40 mA. Restriction fragments were visualized by ethidium bromide staining and transferred to nitrocellulose filters by the method of Southern (1975). These filters were baked for 2 h at 80°C in VUCUO.Filters were prehybridized in 5 x SSC (SSC is 0.15 M-NaCl. 0.015 Msodium citrate), 10 x Denhardt’s (1966) solution (0.027; bovine serum albumin, 0.02g0 polyvinylpyrrolidone, 0.02% Ficoll), 0.1 y0 SDS, 1.25 mM-NaPP, and 50 pg denatured salmon sperm DNA/ml for 4 h at 68°C. Hybridization was in 2 to 5 ml of the prehybridization buffer with lo6 cts/min per ml radioactive probe at 68°C for 16 t,o 20 h. In some experiments hybridization was for 40 h in the same salt buffer made 50% in deionized formamide and incubated at 42°C. Filters were washed 3 times for 20 min each in 2 x SSC at room temperature and 3 times for 20 min each in 5 mM-Tris (pH 8.2), 1 mM-EDTA, O.l’$/, SDS, 1.25 mM-NaPP,, 1 x Denhardt’s solution at 60°C. Filters were blotted, wrapped in Saranwrap and exposed to Kodak XR-5 or XAR-5 film with intensifying screens at -70°C. (i) Pwijkation
of wstrictio~i
fragments
Restriction fragments were often isolated from agarose gels with low melting temperatures (Sea Plaque). The low-temperature agarose gels were l”, in SP running buffer (40 miv-Tris (pH 7.4), 5 mM-sodium acetate, 1 mM-EDTA) and were formed at 4°C. Digested DNA was electrophoresed at 40 mA overnight. The restriction fragments were visualized with ethidium bromide, excised, and the gel slice was melted at 65°C. The DNA was isolated by extraction with phenol and precipitation with ethanol. The glass-powder procedure of Vogelstein R: Gillespie (1979) was also used in some cases.
pBS85-2 was constructed by subcloning a Pstl digest of hDm85 into pBR322 linearized with PstI, as described by Kahn et al. (1979) with slight modifications. The TcR ApS colonies were selected and grown in 10 ml cultures (Kahn et al., 1979). pVB85-1 was const,ructed by ligating the purified 2.8 kb HindIII-R amH1 restriction fragment of hDm85 into pACYC184 (Chang & Cohen, 1978) digested with Hind111 and BarnHI. The ligation mixture was transformed into Escherichia coli strain MC1061 (Casadaban & Cohen, 1980) according to the
Drosophila
TKOPOMYOSIS
235
GENES
protocol of Kushner (1978). The DNA from CmR TcS colonies was isolated as described. All manipulations involving cells containing recombinant molecules were done as specified by the National Institutes of Health guidelines for research involving recombinant DNA molecules. (k) RNA glyoxal gel rlectrophoresis
and filter
hybridization
Poly(A)+ RiVAs were glyoxalated, electrophoresed on agarose gels, and transferred to nitrocellulose filters (Thomas, 1980). After baking at 80°C for 2 h the filters were prehybridized for 2 h at 60°C in 4 x SET (SET is 0.15 iv-NaCl, 0.03 M-Tris HCI (pH 8.0). 2 mM-EDTA), 10 x Denhardt’s solution, and 0.1% SDS. Overnight hybridization at 60°C was carried out under the same conditions with the nick-translated probe (2 x 106 to 4 x IO’ Cerenkov counts/ml), and 200 pg unlabelled, sheared E. coli DNA/ml was added. Filters VWY~ washed with 1 x SSC, 0.1% SDS, 4 times at 66°C. Exposure to Kodak XAR-5 filters was at, -70°C. with intensifying screens. As molecular weight markers glyoxalated single-stranded restriction fragments of bact’eriophage lambda DNA (HindID) and +X174 (HaeIII) were run on the same gel (M&faster & Carmichael, 1977) and visualized, after treatment with alkali. wit,h ethidium bromide and ultraviolet light. in situ 3H-la.belled hDm85 DNA was prepared by nick-translation and hybridized in situ to squashes of larval salivary gland polytene chromosomes according to Gall 8 Pardue (197 I ). Squashes were exposed to emulsion for 7 days or longer. (1) Hybridization
3. Results (a) Isolation
and identi$cation
of tropomyosin
genes
A library of D. melanogaster genomic DNA fragments inserted into Charon 4 lambda phage (Maniatis et al., 1978) was screened with a 32P-labelled cDNA probe enriched in muscle cell-specific sequences. (R. V. Storti, D. Mischke & M. L. Pardue, unpublished data). The clones that hybridized to the probe were characterized further by hybrid-selected translation using poly(A)+ RNA from Drosophila embryos at 12 to 24 hours of development (Ricciardi et al., 1979). One genomic clone from this screen, designated hDm85, hybrid-selected RNA that directed in vitro synthesis of a polypeptide of 35,000 molecular weight (Fig. 1 lane 3). The 35,000 molecular weight polypeptide had the same electrophoretic mobility as purified chicken skeletal muscle tropomyosin (a preparation consisting of two-thirds #I- and one third fi-tropomyosin) and also comigrated with a protein that was abundant in both a Drosophila myofibrillar extract and the translation products of poly(A)+ RNA from late embryos. The single band that we have identifed as Drosophila tropomyosin was resolved into two components of nearly equal intensity when the in vitro translation product’s were separated in a gradient of higher voltage (Fig. 1 lane 4). The faster migrating protein comigrates with chicken n-tropomyosin and has a molecular weight of approximately 34,000. We have designated this protein tropomyosin I. The slower migrating protein has a molecular weight of approximately 35,000 and has been designated tropomyosin II. The hybrid-selected translation products of hDm85 were analyzed by twodimensional gel electrophoresis with a gradient of pH 4 to pH 6 in the isoelectric
E A _P
isad TM-II .TM-I
1
2
3
Drosophila
TROPOMYORIN
GESES
“35
FIN:. 2. Migration of hDm85 translation products in SDS and SDS/urea/polya(~rylamitir gels. The 1st dimension separation was by isoelectric focusing in a pH 4 to pH 6 gradient (basic to the left). The 2nd dimension (downward) was by electrophoresis in SDS (in (a) and (c)) or urea/SDS (in (b) and (d))/polyacrylamide gels. (a) and (b) Fluorograms of [35S]methionine-labelled polgpeptides hybridselected with hDm85 and ADmA filter-bound DN$ (see Fig. 1, lanes 3 and 4). (c) and (d) A small section of Coomassie blue-stained gels showing tropomyosins (TM) and actin (A) from Drosophila myofibrils extracted from .second and third instar larvae. E, endogenous. The actin is included as a reference marker to illustrate the shift in relative mobility of TM.
focusing dimension (O’Farrell, 1975: Storti et al., 1978). The two unresolved [ 35S]methionine-labelled proteins (Fig. 2(a)) have isoelectric points of approximately pH 5 and comigrate with chicken cu-tropomyosin under these conditions. Close inspection of the elongated spot in Figure 2(a) shows that the faster migrating tropomyosin I has a p1 value slightly more basic than that of tropomyosin II. Both subunits of chicken and other vertebrate tropomyosins have been reported to show an apparent increase in molecular weight when electrophoresis is done in the presence of urea (Sender, 1971). Therefore, we analyzed the proteins translated by RNA complementary to /\Dm85 on gels containing 8 M-urea in the second dimension. tinder these conditions the two proteins are clearly separated and both show a decrease in mobility relative to actin marker (Fig. 2(b)): that is, the tropomyosins now migrate with an apparent molecular weight of approximately 50,000 to 53,000. Myofibrils isolated from second and third instar larvae yield seven major proteins
\‘.
I,. HAIT’I’VH
&‘/‘A/,.
FIN:. 3. Partial protrolytic digest of D~rosophila Jo c+tro tratrslation products and chicken tropomyosin. [ 35S]mrthionine-labelled Zn ~~itro translation products of RX.4 complementary to hDm85 and actin DNA WPPCcut from rlwtrophorrtic~ gels, added to chickt,l> trnpomyosin. and digesttbd ait,h chymotrypsin in the prewncr of SDS. The digrstion prodovt?; ww~~ wparatrd iu an SDS/l5?,, polyacrylamidv gvl, stained with C’oomassir blur snd Huorograrrhed. Lanes I and 3. (‘oomassie blue-stained gel showing the chicken tropomyosin polypeptides. Lanes 2 and 1 arr tht> firlorograms oflancs 1 and 3, and contain thr ] 35S]mc~tt~io~~inr-lat~ellrtl i/c vilro prwumptive I~rosophiln tropomyosiu and Drosophila actin pept,idw. respwtively. (‘, chymotrypsin Arrows denotth thr major staincad chivkrn (lane 1) and labelkd (lane 2) ~rosnphila tropomyosin peptidw t.hat. cemigratv.
with mobilities similar to the contractile myofibrillar proteins of vertebrates, on two-dimensional gel electrophoresis. Two of’ these Drosophila myofibrillar proteins comigrate with the hybrid selection products of ADm85 and show the same shift in mobility in the presence of 8 M-urea as does chicken tropomyosin when analyzed under these conditions (Fig. 2(c) and (d)). A comparison of the two-dimensional autoradiograms also reveals that the two Drosophila tropomyosins have shifted in mobility relative to each other. The more basic tropomyosin I now migrates more slowly than the more acidic tropomyosin IT in the urea/SDS dimension. l’ropomyosin II now comigrates with chicken +tropomyosin. This switch in
I I
I 1
-1
kb
Thr map ww constructed using single and double digests ofthe whole clonr DNA or isolated restriction fragments. The darkened areasare thaw restric%ion fragmmts with homology to poly(A)+ RNA from I:! to 24-h l)rosophi/a c~mhryos. Thr horizontal lines hclow the diagram denotr the regions ofXDmR5 nswl as h,vt)ridization probes for regions I, 2 and 3 as described in the text and the legends to Figs 8 and 9
mobility
of t,he two Drosophila tropomyosins in urea/SDS may be correlated with differences in secondary structure. To strengthen our identification of the proteins translated by RNA complementary to hDm85 as Drosophila tropomyosins, we have compared the partial proteolytic digests of these proteins to t,hose of purified chicken tropomyosin. The [35S]methionine-labelled Drosophila tropomyosin and actin that had been translated in vitro were eluted from one-dimensional SDS/polyacrylamide gels, mixed with unlabelled chicken tropomyosin and digested with chymotrypsin (Cleveland et al.. 1977). The results (Fig. 3) show that the digests ofthe Drosophila tropomyosins contain at least five major identifiable peptides that comigrate with digestion products of Coomassie blue-stained chicken tropomyosin. There are, however. other Drosophila and chicken tropomyosin peptidrs that do not comigrat,e. As a of [3sS]methionine peptides to excess control against non-specific “sticking” unlabelled chicken peptides, digests of radioactive Drosophila actin were also analyzed and compared to chicken tropomyosin digests. These non-homologous proteins have only one comigrating peptide when digested under the same conditions. From the series of analyses described we conclude that the translation products of RNA selected by /\Dm85 are Drosophila tropomyosins and that these tropomyosins are partially homologous t’o chicken tropomyosins.
(b) Restriction
et~donuclrasr
map
A restriction endonuclease ma,p of hDm85 was determined using data from single and double restriction endonuclease digests of the emire clone or of isolated restriction fragments (Fig. 4). The transcribed regions of clone hDm85 were identified by determining the restriction fragments homologous to tot,al poly(A)+ RNA from 12 t,o 24-hour embryos. This was accomplished by hybridization of ( 32P]RPI;A or of complementary ]32P]DNA to transfers of the restriction digest’ fragments bound to nitrocellulose filters (Southern, 1975). Three distinct regions of /\Dm85 homologous to the RNA probe were identified and named regions I,2 and 3 (Fig. 4). Based on the intensity of the hybridization signal to each of-the t,ranscribed regions, we estimate that RNAs homologous to regions 1 and 2 are more abundant, in our 12 to 24-hour embryo RNA probe than RNAs homologous to region 3. Although
240
V. 1,. BAL-T(“H
ET .A/,.
we have not determined the exact boundaries of the transcripts within the restriction fragments encompassing the regions: we can conclude that transcribed regions 1 and 2 are separated by at least 3.0 kb of DNA and regions 2 and 3 by approximately 1.3 kb of DNA. When restriction fragments of hDm85 are hybridized with RNA from one to fivehour embryos or Kc cells, little or no hybridization to regions 1 and 2 is detected; however, the restriction fragments of region 3 showing hybridization are identical to those hybridized with 12 to 24-hour embryo RNA. Thus, as discussed below, the RNA transcripts of region 3 from one to five-hour embryos differ from those seen in late embryos, yet we detect’ no difference in the DNA fragment transcribed.
(c) Identijcation
of tropomyosirr
genes on ADm85
We identified the regions of ADm85 encoding the tropomyosin genes by determining the hybrid-selected translation products of regions 1 and 2. A Pstl restriction fragment of hDm85 was subcloned into pBR322 (Bolivar et al., 1977). The resulting chimera plasmid, designated pBS852, showed homology to the 32Plabelled RNA probe. pBS852 was labelled by nick-translation and hybridized to Southern transfers of restriction enzyme-digested hDm85. In this manner the subcloned fragment was localized and mapped within region 2 of hDm85 (Fig. 4). The hybrid-selected translation product of RNA complementary to this subclone was identified as tropomysin II by two-dimensional gel electrophoresis (Fig. 5(a)). The second dimension contained 8 M-urea and unlabelled chicken tropomyosin was included as an identity marker. The trace amounts of actin and tropomysin T attributed to non-specific sticking of RNA to the filter or of partial homology of the RNA were also used as reference markers. The DNA of region 1 was isolated using a different strategy. A restriction enzyme digest of hDm85 was transferred onto a nitrocellulose filter. The piece of filter containing region 1 DNA was cut from the filter and used in a hybrid-selection translation assay. The translation product of RN,4 selected by region 1 was identified as tropomyosin I by two-dimensional gel electrophoresis as described above (Fig. 5(b)). Th e interpretation of these results is that the genes encoding tropomyosin I and tropomyosin II are located within regions 1 and 2, respectively. No polypeptide product was detected when region 3 DNA was bound to a nitrocellulose filter and used in a hybrid-selection translation assay.
(d) Chromosomal
location
oj tropomyosi,n
genes
The chromosomal location of hDm85 was determined by hybridization salivary gland polytene chromosomes of Drosophila melanogaster (Gall 1971). hDm85, nick-translated with 3H-labelled deoxynucleotides, was to salivary gland polytene chromosomes and located by autoradiography. band of hybridization was detected in region 88F2-5 on chromosome arm after relatively short exposures (1 week). No additional hybridization seen even after longer exposure times (7 to 8 weeks), although small
in situ to & Pardue, hybridized A single 3R (Fig. 6) bands were regions of
Drosophila
TROPOMYOAIN
GENES
241
FK:. 5. Hybrid-selected translation products of isolated region I and region 2 DSA (:els are pH 4 to 6 (basic to the left) in the isoelectric focusing dimension and SDS/l3”,, polyacrylamide with X ~-urea in the molecular weight dimension. Isolated region 1 or region 2 DNA (see the text) was used in a hybrid-selection translation assay with 5~8 poly(A)+ RNA from 12 to 24-h embryos. [35S]methionine-labelled products were electrophoresed with unlabelled chicken tropomyosin and the gels were stained and prepared for fluorography. (a) The product of pBS85-2 (region 2) DNA. (h) The product of the 6.0 kb PcoRI-HamHI restriction fragment of hDm85 (region 1) DSA. E, endogenous; A. actin: TM, tropomyosin. The arrows indicate the slight amount of TM I in (a) and the sticking of [%]methionine to the Coomassie blue-stained chicken \-TM in (h).
homology outside 88F 2-5 might not have been detected because of the relatively high sequence complexity of the probe. We would, however, expect to detect other copies of the tropomyosin genes if the additional genes had a degree of homology equal to that of members of the r-tubulin gene family (Mischke & Pardue, 1982). We extended our search for genomic DNA sequences homologous to hDm85 by an analysis of whole-genome hybridizations. An EcoRI restriction digest of D. melanogaster (Canton S) genomic DNA and a similar genomic digest containing the equivalent of one genomic copy of ADm85 DNA were transferred onto nitrocellulose filters and hybridized with 32P-labelled nick translated hDm85. Because hDm85 was constructed by creating artificial EcoRI restriction sites, an EcoRI digest of genomic DNA should generate at least two sets of restriction fragments homologous to /\Dm85: a set of fragments with internal restriction sites that comigrate with the 7.2 kb and 24 kb EcoRI restriction fragments of hDm85. and two fragments larger than the 7.5 kb and 0.9 kb end fragments of the Drosophila DNA insert in hDm85 that extend to the next genomic EcoRI site on either side. All of the hybridization bands seen in the EcoRI digest of genomic DNA could be accounted for by these two sets of fragments (Fig. 7). The @9 kb and
FIG. 6. Hybridization in situ of hDm85. ADm8.5 was labelled with I’H]deoxynucleodites by nick-translation and hybridized to Drosophila salivary gland polyt,enr chromosomes. Hybridization is seen only to region 88F 2-5 on chromosome arm 3R. Hybridization was in 0.3 r+NaCI, 02 rv-Tris (pH 6.8). at, 65°C. Exposure was for 7 days.
7.5 kb EcoRI restriction fragments located at the ends of the Drosophila DNA insert in hDm85 are replaced by two hybridization bands 3.9 kb and 115 kb in size in the genomic digest. From these results we conclude that all genomic sequences homologous to hDm85 are located within 22 kb of contiguous DNA. Longer exposures do not show hybridization of the tropomyosin-coding sequences to other restriction fragments, as might be expected if Drosophila had additional copies of these genes elsewhere in the genome.
(e) Expression
of transcripts
homologous to /\Dm85 during development
embryonic
and muscle
The developmental regulation of RNAs homologous to hDm85 has been investigated by analyzing the RNAs of staged Drosophila embryos and RNAs of Kc cells, a Drosophila cell line of embryonic origin (Echalier & Ohanessian, 1970). In order to characterize the transcripts and identify their corresponding genes, the RNAs were subjected to glyoxal/agarose gel electrophoresis, transferred to nitrocellulose filters, and hybridized to 32P-labelled nick translated probes (Thomas, 1980). The RNA blots were probed with DNA segments corresponding to each of the transcribed regions in XDm85. The results of these hybridizations (Fig. 8) show that several transcripts with different patterns of expression are homologous to these regions. DNA containing coding region 1 and encoding tropomyosin I (the 6-O kb EcoRIBamHI fragment) shows no hybridization to RNA from one-to-three-hour or oneto-five-hour embryos. In 9 to 11-hour embryos, however, there is an RNA band of 1.8 kb and a more diffuse band with a mean size of 145 kb that show homology to the region 1 probe (visible on longer exposures). Hybridization to both bands is much stronger in RNA from 19 to 23-hour embryos (Fig. 8 lanes 1). The 1.45 kb
11rosophiln
TROPOMYOYIS
GESES
FIG. 7. Hybridization of hDmX5 to genomic D?r‘A. DSA was digested with EcoRI, elcctrophoresed in a 1” 0 agarose gel and transferred to nitrocellulose. 32P-labelled hDm85 (2 x 10’ cts/min per pg) was hybridized to the filter. After washing, the filter was exposed to Kodak XAR-5 film with intensifying scwcn at -70°C for 10 h. Lane I, 1 ng hDmR:i DSA+lOpg total Drosophila (Canton S) DSA (approx. I genome equivalent): lane 2, 1Opg total Drosophila (Canton 8) DNA. On the right the sizes of the hybridizing genomic bands arc given, in kb. The arrows indicate the vector arms in hDm85. The molecular weight standards were a Hind111 digest of h DSA.
“41
V. 1,. B.AlrT(‘H
ET .-1L.
band is resolved into two bands of 1.4 and 15 kb in shorter exposures (data not shown). DNA from coding region 2 (pBSS5S), encoding tropomyosin II. also does not hybridize to any transcripts in early embryos but shows some hybridization to a 1.5 kb transcript that appears in 9 to 1 l-hour RNA (visible on longer exposures) and increases in abundance in 19 to 23-hour RNA (Fig. 8 lanes 2). The probes for regions 1 and 2 do not show any hybridization to Kc cell RNA (data not shown). Thus the temporal pattern of appearance of homologous RXA is similar fol region 1 and 2. DNA containing coding region 3 was subcloned by inserting the purified 2% kb HindIlLBumHI restriction fragment into the plasmid PACYC184 (Chang & Cohen. 1978). This construction, designated pVB85-1, hybridizes to an abundant 2.3 kb transcript in RNA from one- to three- and one- to five-hour embryos. This RNA, however, is not detected in older embryos. The coding region 3 probe also hybridizes to a 2.8 kb transcript that is present in rather low amounts in one- to three- and oneto five-hour embryos, becomes very abundant in 9 to I l-hour RNA. and then decreases by 19 to 23-hours of development. In addition, the region 3 probe also hybridizes weakly to a transcript of 1.5 kb in the RNA from older embryos (Fig. 8 lanes 3). The coding region 3 probe hybridizes to a single transcript of 2.8 kb in Kc cell RSA (Fig. 8 lane 4). The data show that the major transcripts homologous t,o region 1. containing the tropomyosin I coding region, and to region 2, containing the tropomyosin II coding
.2.8 .P3
ab
cd 1
a
b
c 2
d
a
b
c 3
d 4
FIG;. 8. Hybridization oftranscribed regions ofhDm85 DSA to Dro,so~hila embryonic and Kc call RSA. Poly(A)+ RNA (3 pg) from staged embryos or Kc cells was subjected to glyoxal/agarose gel electrophoresis and transferred to nitrocellulose filters. The filters were hybridized with 32P-labelled DNA corresponding to each of the 3 transcribed regions of ADm85. After washing, the filters were exposed to Kodak XAR-5 film with intensifying screen at -70°C for 15 h. Lane a of filters 1 to 3 is 1 to 3-h embryo RNA : lane b is 1 to 5-h embryo RNA ; lane c is 9 to 11-h embryo RNA ; and lane d is 19 to 23-h embryo RXA. Filter 4 contains Kc cell RNA. The probes were labelled to a specific activity of 10s to 109 Cerenkov cts/pg and arc as follows: filter 1. 6.0 kb EcoRI-RamHI restriction fragment (region 1) of hDm85 DNA: filter 2, pBR85-2 (region 2) DlVA: filters 3 and 4. pVB%-1 (region 3) DPU’A. In the margins the sizes of the hybridizing bands are given. in kb. The molecular weight standards were a Hind111 digest of h DNA and an HaeIII digest of +X174 DNA.
Drosophila
TROPOMYOHIN
GENES
245
region, first become detectable in 9 to 11-hour embryos and increase in abundance in later stages. This pattern of gene expression correlates with the observation that muscle cell fusion is first detectable at about 9 to 11 hours of development (Poulson, 1950). It is also consistent with analyses of embryo poly(A)+ RNA by in vitro translation. Tropomyosins I and 11 are very abundant translation products directed by poly(A)+ RNA from 18 to 20.hour embryos, but these proteins are not present in appreciable quantities in the translation products of poly(A)+ RNA from bwo- to three-hour embryos (data not shown). The pattern ofexpression oftranscripts homologous to XDm85 in muscle cells was also studied in primary myogenic cultures of Drosophila cells (Storti et al., 1978). Total poly(A)+ RNA was isolated from myogenic cells collected 4, 12 and 18 hours after plating of cell suspensions made from two- to three-hour-old embryos. These times correspond to <3% fusion, 50% fusion and SOY/,fusion (Storti et al.. 1978). RNAs from the cell cultures were separated by electrophoresis on glyoxal/agarosr gels, t,ransferred to nitrocellulose filters and hybridized with the same DNA probes that had been used to analyze the embryo RNA. DNA containing coding region 1 hybridizes to a 1.8 kb transcript, and a broad band of hybridization at 1.4 to 1.5 kb (Fig. 9 lanes 1). These transcripts are detected only in cultures with significant myotube formation (12 h and 18 h post-plating). The region 2 probe hybridizes to a single transcript of I.5 kb that is present in approximately equal quantities in 12 and 1%hour RNA (Fig. 9 lanes 2). Neither the region 1 probe nor the region 2 show hybridization with myoblast (4 h post-plating) RNA. As is the case in the intact embryos, the appearance of RNA homologous to the regions of the tropomyosin genes is correlated with the differentiation of myotubes. These observations are supported by experiments that show that the tropomyosins are greatly enriched in the in vitro translation two Drosophila products of RNA from cultures containing a high proportion of mpot’ubes (data not, shown). The probe containing region 3 hybridizes to a 2.8 kb transcript that is present in unfused myoblasts and increases in abundance as fusion progresses (Fig. 9 lanes 3). This probe also hybridizes weakly to a transcript of 1.5 kb that is detectable only in cultures enriched in myotubes (12 h and 18 h post-plating). We consider it likely that the 2.8 kb transcript in the myogenic cell RNA is the same as the 2.3 kb transcript seen in the embryo Rh’As and in Kc cell RXA. In cultured myogenic cells this transcript increases in abundance after the time of fusion and does not show the decreased abundance seen in 19 to 23.hour embryos. Not transcript corresponding to the 2.3 kb RX’A seen in early embryos is detected in myogenic cells or in the Kc cells. It is possible that the myogenic cells and the embryo cells from which the Kc cell line was derived have passed the appropriate developmental time for expression of the 2.3 kb transcript. 4. Discussion We have isolated two Drosophila tropomyosin genes on a single genomic clone (ADmE%). These genes were identified by characterizing the hybrid-selected translation products of RNA homologous to the clone. /\Dm85 hybrid-selected RSA
\‘.
a
b
1
c
L. BAITTC’H
a
b
2
ET AL
c
a
b
c
3
lQ(:. 9. Hybridization of transcribed regions of hDtn8.i DSA to RN.4 from 11roaophi/a myogrnic cells. Poly(A)’ RNA (3 pg) from myogenic cells collected at 1. I:! or 18 h post-plating was subject,ed to glyoxal/agarose gel electrophoresis and transferred to nitrorellulose filters. The filters xcrc hybridized with 32P-jabelled DSA corresponding to the transcribed regions of hDm85. After washing, thtl filters were exposed to Kodak XAR-5 film with intensifying screen at - 70°C for 1 days. Lane a offilters 1 to 3 is RNA collected 4 h post-plating: lane b is RNA collected 12 h post-plating; and lane c is K?rTA collected 18 h post-plating. Probes were labelled to a specific activity of 108 to lo9 rerenkov cts/pg and are as follows: filter 1, 6.0 kb EcoRI-BamHI restriction fragment (region 1) of hDm85 DNA: filter 2. pBS85-2 (region 2) DNA ; filter 3, pVB85-I (region 3) DSA. In the margins the sizes of the hybridizing bands are given. in kb. The molecular weight standards were a Hind111 digest of X DNA and a HafIII digest of ~5x174 DS.4.
encoding 35,000 molecular weight and 34,000 molecular weight polypeptides. These polypeptides comigrate with purified chicken r-tropomyosin and with two polypeptides present in llrosophila myofibrillar protein preparations on one- and two-dimensional gels. The translation products also show the apparent increase in molecular weight when electrophoresed in the presence of 8 M-urea that is characteristic of vertebrate tropomyosins. The peptides generated by partial proteolytic digestion of the proteins hybrid-selected by hDm85 show partial homology to t,hose of purified chicken tropomyosin. Thus the Drosophila tropomyosins have a number of properties similar to those that characterize the vertebrate and invertebrate tropomyosins that have been investigated. We have designated the 34,000 M, polypeptide tropomyosin I and 35,000 M, polypeptide tropomyosin II. The 18 kb of Drosophila DNA in hDm85 contains three distinct transcribed regions. Transcribed regions 1 and 2 encode tropomyosins I and II, respectively. Region 1 is approximately 49 kb and region 2 is approximately 2.1 kb in size, and they are separated by about 3 kb of DNA. The gene region encoding tropomyosin II is homologous to a single transcript of
I~rocsophi/u
TROPOMYOSIK
GEKES
211
1.5 kb. This RNA is approximately twice the size necessary to encode a protein of about 280 amino acids, which would be expected from the size of vertebrate tropomyosins. This discrepancy suggests a significant amount of untranslated RNA. The gene region encoding tropomyosin I hybridizes to a transcript of 1.8 kb and a broad band of RNA that is composed of two transcripts of 1.5 and I.4 kb. Because of the abundance of the 1.8 kb transcript it seems unlikely that it is a precursor to the smaller transcripts. Moreover, preliminary results show that a 1.3 kb probe from the left end of coding region 1 hybridizes to all three transcripts. These RNAs might, therefore, be the result ofalternative processing ofa single gene transcript (Hagenbuchle et al., 1981; Marie et aE., 1981) rather than the transcripts arising from two adjacent genes. We have not determined which of these transcripts encodes tropomyosin I. The mRNAs for tropomyosin I and tropomyosin TI show little, if any, cross hybridization either in the hybrid-selection assays or in Southern blot hybridization. Thus, our data are consistent with the finding that a cDNA clone of chicken I-tropomyosin does not cross-hybridize to chicken fi-tropomyosin RNA under similar conditions (MacLeod, 1981). The expression of /\Dm85 regions 1 and 2 (containing the tropomyosin 1 and I I genes. respectively) is developmentally regulated. Embryonic transcripts homologous to each of these coding regions are first detected in RNA from embryos at, 9 to 1 I hours of development. The expression of the two tropomyosin genes may not be exactly synchronous, however, since transcripts of region 1 are considerably more abundant than those of region 2 at 9 to 11 hours. The developmental time period during which the first transcripts of these genes can be observed correlates very well with the onset of muscle cell fusion in the embryo (Poulson, 1950). In addition t,o the two tropomyosin-coding regions, ADrn8.5 contains a third transcribed region that is homologous to several transcripts. R’egion 3 is approximately 2.8 kb in size and is separated from the tropomyosin II-coding region by about, 1.3 kb of DNA. We have not identified a protein encoded by this region. The first major transcript homologous to this region, a 2.3 kb transcript,, is present from the onset of development but disappears at about, the time of’ gastrulation (3 to 5 h) and, therefore. is most likely a mat)ernal transcript. The other major transcript is 2.8 kb in size and appears at, or shortly after gastrulation and in mvogenic cultures both before andafter myotube format,ion. The 2.8 kb transcript is also found in Drosophila Kc cells. Region 3 is not large enough to contain two separate transcribed regions of2.3 and 2.8 kb. Thus it seems probable that these tw o transcripts arise by differential transcription or by differential processing ofa. single primary transcript (Hagenbuchle et al., 1981 : Marie et al.. 1981), although other possibilities such as a rearrangement of the gene (Early et al.. 1980: Hicks rf al., 1979) cannot be completely ruled out. The coding region 3 probe also hybridizes weakly to a 1.5 kb t)ranscript in late embryos. The fact that transcripts of approximately the same size are detected with the tropomyosins I and II coding region probes suggests a possible low level of homology among the coding regions. We are currently constructing subclones of all the coding regions in order to determine their homologies at the sequence level. F’reliminary hybridization experiments indicate that t,here are weak homologies
24x
T. L. BAITT(“H
El’
AI,.
between coding region 3 and the tropomyosin coding regions. It is, therefore, tempting to speculate that coding region 3 may encode a non-muscle tropomyosin, although it has yet to be shown that Drosophila has tropomyosin in non-muscle cells and we have not yet identified a protein product of the rather abundant transcripts homologous to coding region 3. We have localized the Drosophila tropomyosin genes to the chromosomal region 88F 2-5 by hybridization in situ. One of the six Drosophila actin genes has also been mapped to region 88F (Tobin et al., 1980: Fyrberg et al., 1980). We do not know whether there is a functional basis for this relative proximity. Particularly, it. should be noted that it is not known whether the 88F actin gene encodes a muscle-specific isoform and that the chromosomal region may well span more than 100 kb (Bridges, 1935 : Rudkin, 1972). The hybridization experiments in situ give no evidence that sequences homologous to hDm85 are located elsewhere in t,he genome. This conclusion is strengthened by hybridization of hDm85 to restriction enzyme-cleaved genomic DNA. These results indicate that no portion of hDm85 is homologous to sequences outside the 22 kb of contiguous DSA surrounding the tropomyosin genes. Thus the tropomyosin genes of Drosophila do not appear to belong to a family of partially homologous genes scattered over the genome, as do the Drosophila actin (Tobin et al., 1980: Fyrberg et al., 1980) and tubulin genes (Sanchez et al., 1980: Kalfayan gL Wensink, 1981; Mischke & Pardue, 1982). The correlation between the appearance of abundant transcripts homologous to the tropomyosin genes in hDm85 and the differentiation of myotubes is strong evidence that both tropomyosin genes are expressed in muscle tissue. The data do not’ exclude the possibility that these same genes code for tropomyosin in nonmuscle. cells. However, if these genes are expressed in non-muscle cells, their level of expression is substantially lower than that of muscle cells. The possibility that hDm85 also contains a structural gene for non-muscle tropomyosin is currently under invest,igation. Fortunately the single locus and probable single-copy nature of the cloned tropomyosin genes will facilitate genetic approaches to this question. The authors thank Dr Michael Barany for providing us with chicken tropomyosin. We are also indebted to Dr Anthony Mahowald and Dr Sarah Elgin for a generous supply of embryos and to Dr Pieter Wrnsink for providing us with the genomic actin clone XDmAl. R.V.S. and V.L.B. also acknowledge the expert technical assistance of MS Alice Szwast. This from the National Tnstitutes of Health and work was supported hy a grant Campus Research Board and Biomedical Research Support grants from Cnirersity of Illinois (to R.V.S.). and a grant from the NIH (to M.L.P.). R.V.S. is a recipient of a Research Career Development award from the National Institutes of Health. V.L.B. was supported by a University of Illinois Graduate College fellowship. D.M. was supported by a fellowship from the Deutschr Forschungsgemeinschaft.
REFERENCES Amphlrtt, (;. W., Syska. H. Br Perry. S. V. (1976). PE’RS Letters, 63, 22%. Bingham, P. M., Levis, R. 8: Rubin, (4. M. (1981). (‘~41, 25, 693-704. Bolivar, F., Rodriguez, K. L., Greene. P. J., Betlach. M. C.. Heyneker, H. L., Boyer, Cross, J. H. & Falkow, S. (1977). (Gene, 2, 95-l 13.
H. W..
Drosophiln
TROPOMYOSIX
(GENES
219
Bonner, W. M. & Laskey, R. A. (1974). Eur. J. Biochem. 46, 83-88. Bridges, C. B. (1935). J. Hered. 26, 60-64. Buckingham, M. E. (1977). In Biochemistry of Cell Diff erentiatio,n, vol. 13 (Paul,
“50
V. I,. BAlJTCH
E’T AL.
Rudkin, G. T. (1972). In DevelopmerLtal Studies on Giant Chromosomes (Beermann, W., ed.), pp. 58-85, Springer, New York. Sanchez, F., Natzle, J. E., Cleveland, D. W., Kirschner, M. W. & McCarthy, B. J. (1980). (le/l,
22, 845-854. Scott. M. P., Storti, R. V., Pardue, M. L. & Rich, A. ( 1979). Biochemistry, 18, 1588-1598. Seecoff, R. L., Gerson, I., Donady, J. J. & Teplitz, R. 1,. (1973). Develop. Biol. 35. 250-261. Sender, F’. M. (1971). FEBS Letters, 17, 106-l 10. Smillie, L. B. (1979). Trans. Znt. Biochem. Sot. 4. 151.-155. Southern, E. M. (1975). J. Mol. Biol. 98, 503- 518. Spradling, A. C., Pardue, M. L. & Penman, S. (1977). J. MoZ. Hiol. 199> 5599571. Stone, D. & Smillie, L. B. (1978). J. Biol. Chem. 253, 113771148. Storti, R. V. & Rich, A. (1976). Proc. Nat. Acad. Sri., I’.S.d. 73, 2346-2350. Storti, R. V., Horovitch, 8. J., Scott, M. I’., Rich, A. Pr I’ardue, M. L. (1978). Cell, 13,589-598. Storti. R. V., Scott, M. P., Rich, A. & Pardue, M. L. (1980). CelZ, 22, 825-834. Thomas, I’. 8. (1980). Proc. Nat. dcad. Sci.. C's.d. 77. 5201-5205. Tobin, S. L., Zulauf, E., Sanchez, F., Craig, E. A. Br McCarthy, B. J. (1980). CelZ, 19, 121-131. Vogelstein, B. & Gillespie, D. (1979). Proc. Nat. Acad. Sci., li.S.il. 76, 615419. Woods. E. F. & Pont, M. J. (1971). Riochemistry, 10, 270-276.
Edited
by S. Brwnrr
(1982)
Organization
162, 231-250
and Expression of Drosophila Genes VWOKIA
Tropomyosin
L. RAI”WH. ROBERT 1’. STO~I
Department of Biological Chemistry University of Illinois Medical Center Chicago, Ill. 60612, U.S.A.
Department of Biology Massachusetts Institute of Technology Cambridge, Mass. 02139, lJ.8.A.
(Recuivd 1 April
198%)
A genomic clone,containing two tropomyosin-coding sequences has been isolated from a library of Drosophila melanogastrr DNA and identified by a positive hybridization selection and an in vitro translation procedure. In vitro translation yielded two products that comigrated aith chicken and Drosophila tropomyosins in sodium dodecyl sulfate/polyacrylamide gels and underwent the mobility shift characteristic of vertebrate tropomyosins in sodium dodecyl sulfate/urea/polyacrylamide gels. The Drosophila polypeptides also shared several proteolytic fragments with chicken tropomyosins. The cloned DNA hybridized to a single site in region 88F 2-5 on the right arm of chromosome 3 of polytene chromosomes and to the set of restriction fragments in genomic DNA predicted from the cloned sequences, indicating that similar tropomyosin-coding sequences are not located at other sites in the genome. The 18 x IO3 base-pair cloned segment contains three regions complementary to Drosophila embryo RNA separated by non-coding sequences. Two of these coding regions encode tropomyosin I and tropomyosin II ; no protein product has been identified for the third coding region. The expression of the two tropomyosin genes is developmentally regulated during embryo development and in primary myogenic cultures, with abundant transcripts occurring at the onset of muscle cell fusion. The third coding region is homologous to abundant RK’A transcripts found in earlier stages of embryo development. in primary myogenic cultures and in the Drosophila Kc cultured cell line.
1. Introduction Tropomyosin is an abundant contractile troponin, tropomyosin regulates the myosin
(Ebashi
& Endo,
t Present address: F.1t.C:.
Institut
protein of muscle calcium-sensitive
1968). In addition fuer Biologic,
cells. In association interaction of actin
to the tropomyosin
Medizinische
Hochschule
found
with and
in muscle
Luebeck, D-2400 Lurheck
1,
“31 0022-2836/82/340231-20
$03.00/O
!CI 1982 Academic Press Inc. (London) Ltd.
“32
V. 1,. HAIJT(‘H
ET Al,.
cells, non-muscle forms of tropomyosin have been isolated from several different tissues (Cohen 8: Cohen, 1972; Fine et al., 1973; Fine & Blitz, 1975). The function of non-muscle tropomyosin is not known but it, has been suggested that these tropomyosins participate in the contractile activities of the non-muscle cells (Smillie, 1979). Although tropomyosins have been isolated from several invertebrates (Kominz rt in al., 1962: Woods & Pont, 1971), these proteins have been studied more intensively vertebrates. These studies have revealed multiple forms oftropomyosin. Two major forms of vert,ebrate muscle tropomyosin subunits have been idernified: Ztropomyosin has a molecular weight of approximately 34,000, while /3-tropomyosin has a molecular weight of approximately 36,000 (Buckingham: 1977). The two tropomyosins from rabbit skeletal muscle have been seyuenced and found to differ in 39 out of 284 amino acid residues (Stone & Smillie, 1978; Mak et al., 1980). The two muscle forms of tropomyosin are highly conserved among different organisms (Cummins & Perry, 1974; Fine & Blitz, 1975). Non-muscle forms of tropomyosin have a molecular weight of approximately 30,000 with physical properties and peptide maps similar to those of the muscle subunits (Fine & Blitz, 1975). The nonmuscle forms, however, appear to differ substantially from the muscle forms at the amino and carboxy ends of the molecule (Cote et al., 1978). The biosynthesis of tropomyosin is developmentally regulated during muscle myogenesis along with the other major proteins of the contractile apparatus (Carmon et al., 1978; Moss & Schwartz, 1981). The two forms of muscle tropomyosin show different patterns of regulation. Each muscle has a characteristic ratio of CItropomyosin to /3-tropomyosin (Cummins & Perry, 1974) and this ratio may change during development (Amphlett et cd., 1976: Roy et al., 1976). Polymorphic forms of both ,I- and p-tropomyosin have been detected in vertebrates. Some of these multiple forms are produced by phosphorylation (Montarras et al., 1981). However, other forms differ in primary sequence (Mak et al., 1980) or can be identified in cell-free translation products of messenger RNA (Carmon et al., 1978), suggesting the existence of multiple tropomyosin genes. We report here the isolation and parGal characterization of two tropomyosin genes from Drosophila melanngaster contained in one recombinant DNA clone from a genomic library. These genes are separated by a few thousand base-pairs of DNA and are found at a single chromosomal locus. The proteins encoded by these genes share several of the features of the vertebrate tropomyosins described above. RNAs homologous to these genes encode different tropomyosin subunits and their is developmentally regulated during embryo development and in expression primary myogenic cultures. This is the first reported isolation of structural genes for tropomyosin.
2. Materials
and Methods
(a) DNA isolatiort The Drosophila melanogaster (Canton S) genomic DNA library was a gift from Dr Tom Maniatis. Small quantities of recombinant DNA from the phage were isolated by the PDS method of Blattner, which accompanies the Charon phage Drosophila library. Larger preparations were isolated from C&-purified phage as described by Maniatis et al. (1978).
Drosophila
TROPOMYOGIK
GENES
233
Genomic DNA from the Canton S strain of Drosophila was isolated as described by Bingham et al. (1981). Plasmid DNA was isolated from lo-ml cultures or from 1 liter cultures according to Kahn et al. (1979). (b) RNA
isolation
Frozen embryos (Oregon R) at 12 to 24 h of development were provided by Dr Anthony Mahowald. The frozen embryos were thawed directly in lysis buffer (30 mM-Tris (pH 8.3), 100 mM-h’ac], 10 mMCaCl,, 1% sodium dodecyl sulfate, to which 0.5% diethyl pyrocarbonate was added just prior to use) and total RNA was isolated by extraction with phenol/chloroform as described previously (Scott et al., 1979). Staged embryos (Oregon-R, provided by Dr Sarah Elgin) were collected at 25°C over a 2 or 4-h period and aged at 25°C to give 1 to 3, 1 to 59 to 11, or 19 to 23.h-old embryos. Drosophila muscle cell cultures were prepared as described previously (Seecof et al., 1973; Storti et al., 1978). Kc cells were maintained in spinner culture at 25°C. Cytoplasmic RNA from the embryos, myogenic cultures, and Kc cells was isolated by extraction with phenol/chloroform of Triton X-100.lysed cells as described by Storti et al. (1980). 411 RNAs were fractionated by oligo(dT)-cellulose chromatography (Spradling et al., 1977). (c) Myo$bril
protein
extraction
Myofibrils were prepared from 10 g of live second and third instar Canton 8 larvae by the procedure of Bullard et al. (1973). The final washed myofibril pellet was resuspended in rlectrophoresis buffer. (d) Hybrid-selection
translation,
assay
Hybrid-selection translation was performed according to Ricciardi et al. (1979). Approximately 1 to 5 pg of DNA was heat-denatured and spotted on one-half of a 10 mm diameter nitrocellulose filter, air-dried and baked for 2 h at 80°C in vacua. When a single restriction fragment of DNA was used the DNA-containing filter was cut from a Southern (1975) blot. Hybridization was for 2 to 4 h at 37°C in 100 ~1 of 50% deionized formamide (MCB), O-75 M-NaCl, 2 mM-EDTA, 0.1 M-Tris (pH 7.5), 0.2% SDS? and approximately 5 pg of poly(A)+ RNA from 12 to 24-h embryos. Following hybridization the filters were washed 5 times in 150 mw-NaCl, 15 mM-sodium citrate, 0.5% SDS at 60°C and twice in 10 mM-Tris (pH 7.2), 2 mM-EDTA at 60°C. The filters were boiled for 1 min in the presence of 2 mM-EDTA (pH 7.0) and 10 pg calf liver transfer RNA to dissociate the hybrid RNA. The RNA was collected by precipitation with ethanol and translated in the presence of 40&i [35S]methionine (800 to 1200 Ci/mmol, New England Nuclear.) Translation was in a microccal nuclease-treated rabbit reticulocyte lysate cell-free protein-synthesizing system as described bv Pelman $ Jackson (1976). After translation the reactions were treated with 2 pg of pancreatic RNAase for 5 min at, 37°C. (e) SDSJpolyacrylamidu gel &ctrophowsis of proteins One-dimensional electrophoresis was in SDS/l3O/d polyacrylamide slab gelsand performed as described by Laemmli (1970). Two-dimensional gel electrophoresis was according to O’Farrell (1975) with modifications as described previously (Storti at al., 1978). In some experiments the second-dimension gel was supplemented with 8 M-urea. Gels were stained with Coomassie brilliant blue and either dried for autoradiography or fluorographed (Bonner & Laskry, 1974) and exposed to Kodak XR-5 or XAR-5 film.
(f) Partial proteolytic peptide mapping [35S]methionine-labelled tropomyosin and actin were synthesized by translation in zjitro of 12 to 24-h poly(A)+ embryo RNA in reticulocyte lysates. The tropomyosin and actin were t Abbreviations used: SDS, sodium dodecgl sulfate: kb. lo3 base-pairs or bases.
234
V. L. BAIJTCH
ET AL.
separated by one-dimensional gel electrophoresis, located by autoradiography and cut from the dried gels. The labelled proteins were eluted electrophoretically into dialysis bags, dialyzed against 0.1% SDS and concentrated (Storti & Rich, 1976). Partial chymotryptic digestion was according to Cleveland et al. (1977). A total of 30 pg purified chicken skeletal muscle tropomyosin was added to each labelled protein, and 10 pg of freshly prepared chymotrypsin (Worthington) was added. Digestion was for 10 min at 37°C. The products were separated in a 15% polyacrylamide slab gel, stained with Coomassie blue. destained and fluorographed. The stained chicken polypeptides were used for comparison Drosophila tropomyosin polypeptides. with [35S]methionine-labelled (g) Radiolahelling
of nucleic acids
DNA was labelled by nick-translation according to the protocol of Maniatis et al. (1975) with some modifications. DNA (1 pg) was mixed with 60 to 80 pmol of [a-32P]dATP and/or dCTP (600 to 800 Ci/mmol, New England Nuclear) and 400 pmol of each of the other unlabelled dNTPs. The same protocol was also used with 180 pmol [3H]dATP and [3H]TTP in situ. precursors to generate 3H-labelled hDm85 DNA for hybridizations [32P]cDNA complementary to poly(A)+ RNA was synthesized by a modification of the method of Friedman & Rosbash (1977). [5’-32P]RNA was prepared in vitro by brief treatment with alkali, labelling with [y-32P]ATP (>2000 Ci/mmol, New England Nuclear) and bacteriophage T4 polynucleotide kinase (New England Biolabs) according to Maizels (1976). (h) DNA -agarose gel elrctrophmesis and j&r hybridization Horizontal agarose gels were l(& or 1.5% in Tris-acetate running buffer (40 mw-Tris (pH 8.3), 20 mw-sodium acetate, 2 mM-EDTA), with 5% glycerol. DI?II’Awas electrophoresed for 16 h at 40 mA. Restriction fragments were visualized by ethidium bromide staining and transferred to nitrocellulose filters by the method of Southern (1975). These filters were baked for 2 h at 80°C in VUCUO.Filters were prehybridized in 5 x SSC (SSC is 0.15 M-NaCl. 0.015 Msodium citrate), 10 x Denhardt’s (1966) solution (0.027; bovine serum albumin, 0.02g0 polyvinylpyrrolidone, 0.02% Ficoll), 0.1 y0 SDS, 1.25 mM-NaPP, and 50 pg denatured salmon sperm DNA/ml for 4 h at 68°C. Hybridization was in 2 to 5 ml of the prehybridization buffer with lo6 cts/min per ml radioactive probe at 68°C for 16 t,o 20 h. In some experiments hybridization was for 40 h in the same salt buffer made 50% in deionized formamide and incubated at 42°C. Filters were washed 3 times for 20 min each in 2 x SSC at room temperature and 3 times for 20 min each in 5 mM-Tris (pH 8.2), 1 mM-EDTA, O.l’$/, SDS, 1.25 mM-NaPP,, 1 x Denhardt’s solution at 60°C. Filters were blotted, wrapped in Saranwrap and exposed to Kodak XR-5 or XAR-5 film with intensifying screens at -70°C. (i) Pwijkation
of wstrictio~i
fragments
Restriction fragments were often isolated from agarose gels with low melting temperatures (Sea Plaque). The low-temperature agarose gels were l”, in SP running buffer (40 miv-Tris (pH 7.4), 5 mM-sodium acetate, 1 mM-EDTA) and were formed at 4°C. Digested DNA was electrophoresed at 40 mA overnight. The restriction fragments were visualized with ethidium bromide, excised, and the gel slice was melted at 65°C. The DNA was isolated by extraction with phenol and precipitation with ethanol. The glass-powder procedure of Vogelstein R: Gillespie (1979) was also used in some cases.
pBS85-2 was constructed by subcloning a Pstl digest of hDm85 into pBR322 linearized with PstI, as described by Kahn et al. (1979) with slight modifications. The TcR ApS colonies were selected and grown in 10 ml cultures (Kahn et al., 1979). pVB85-1 was const,ructed by ligating the purified 2.8 kb HindIII-R amH1 restriction fragment of hDm85 into pACYC184 (Chang & Cohen, 1978) digested with Hind111 and BarnHI. The ligation mixture was transformed into Escherichia coli strain MC1061 (Casadaban & Cohen, 1980) according to the
Drosophila
TKOPOMYOSIS
235
GENES
protocol of Kushner (1978). The DNA from CmR TcS colonies was isolated as described. All manipulations involving cells containing recombinant molecules were done as specified by the National Institutes of Health guidelines for research involving recombinant DNA molecules. (k) RNA glyoxal gel rlectrophoresis
and filter
hybridization
Poly(A)+ RiVAs were glyoxalated, electrophoresed on agarose gels, and transferred to nitrocellulose filters (Thomas, 1980). After baking at 80°C for 2 h the filters were prehybridized for 2 h at 60°C in 4 x SET (SET is 0.15 iv-NaCl, 0.03 M-Tris HCI (pH 8.0). 2 mM-EDTA), 10 x Denhardt’s solution, and 0.1% SDS. Overnight hybridization at 60°C was carried out under the same conditions with the nick-translated probe (2 x 106 to 4 x IO’ Cerenkov counts/ml), and 200 pg unlabelled, sheared E. coli DNA/ml was added. Filters VWY~ washed with 1 x SSC, 0.1% SDS, 4 times at 66°C. Exposure to Kodak XAR-5 filters was at, -70°C. with intensifying screens. As molecular weight markers glyoxalated single-stranded restriction fragments of bact’eriophage lambda DNA (HindID) and +X174 (HaeIII) were run on the same gel (M&faster & Carmichael, 1977) and visualized, after treatment with alkali. wit,h ethidium bromide and ultraviolet light. in situ 3H-la.belled hDm85 DNA was prepared by nick-translation and hybridized in situ to squashes of larval salivary gland polytene chromosomes according to Gall 8 Pardue (197 I ). Squashes were exposed to emulsion for 7 days or longer. (1) Hybridization
3. Results (a) Isolation
and identi$cation
of tropomyosin
genes
A library of D. melanogaster genomic DNA fragments inserted into Charon 4 lambda phage (Maniatis et al., 1978) was screened with a 32P-labelled cDNA probe enriched in muscle cell-specific sequences. (R. V. Storti, D. Mischke & M. L. Pardue, unpublished data). The clones that hybridized to the probe were characterized further by hybrid-selected translation using poly(A)+ RNA from Drosophila embryos at 12 to 24 hours of development (Ricciardi et al., 1979). One genomic clone from this screen, designated hDm85, hybrid-selected RNA that directed in vitro synthesis of a polypeptide of 35,000 molecular weight (Fig. 1 lane 3). The 35,000 molecular weight polypeptide had the same electrophoretic mobility as purified chicken skeletal muscle tropomyosin (a preparation consisting of two-thirds #I- and one third fi-tropomyosin) and also comigrated with a protein that was abundant in both a Drosophila myofibrillar extract and the translation products of poly(A)+ RNA from late embryos. The single band that we have identifed as Drosophila tropomyosin was resolved into two components of nearly equal intensity when the in vitro translation product’s were separated in a gradient of higher voltage (Fig. 1 lane 4). The faster migrating protein comigrates with chicken n-tropomyosin and has a molecular weight of approximately 34,000. We have designated this protein tropomyosin I. The slower migrating protein has a molecular weight of approximately 35,000 and has been designated tropomyosin II. The hybrid-selected translation products of hDm85 were analyzed by twodimensional gel electrophoresis with a gradient of pH 4 to pH 6 in the isoelectric
E A _P
isad TM-II .TM-I
1
2
3
Drosophila
TROPOMYORIN
GESES
“35
FIN:. 2. Migration of hDm85 translation products in SDS and SDS/urea/polya(~rylamitir gels. The 1st dimension separation was by isoelectric focusing in a pH 4 to pH 6 gradient (basic to the left). The 2nd dimension (downward) was by electrophoresis in SDS (in (a) and (c)) or urea/SDS (in (b) and (d))/polyacrylamide gels. (a) and (b) Fluorograms of [35S]methionine-labelled polgpeptides hybridselected with hDm85 and ADmA filter-bound DN$ (see Fig. 1, lanes 3 and 4). (c) and (d) A small section of Coomassie blue-stained gels showing tropomyosins (TM) and actin (A) from Drosophila myofibrils extracted from .second and third instar larvae. E, endogenous. The actin is included as a reference marker to illustrate the shift in relative mobility of TM.
focusing dimension (O’Farrell, 1975: Storti et al., 1978). The two unresolved [ 35S]methionine-labelled proteins (Fig. 2(a)) have isoelectric points of approximately pH 5 and comigrate with chicken cu-tropomyosin under these conditions. Close inspection of the elongated spot in Figure 2(a) shows that the faster migrating tropomyosin I has a p1 value slightly more basic than that of tropomyosin II. Both subunits of chicken and other vertebrate tropomyosins have been reported to show an apparent increase in molecular weight when electrophoresis is done in the presence of urea (Sender, 1971). Therefore, we analyzed the proteins translated by RNA complementary to /\Dm85 on gels containing 8 M-urea in the second dimension. tinder these conditions the two proteins are clearly separated and both show a decrease in mobility relative to actin marker (Fig. 2(b)): that is, the tropomyosins now migrate with an apparent molecular weight of approximately 50,000 to 53,000. Myofibrils isolated from second and third instar larvae yield seven major proteins
\‘.
I,. HAIT’I’VH
&‘/‘A/,.
FIN:. 3. Partial protrolytic digest of D~rosophila Jo c+tro tratrslation products and chicken tropomyosin. [ 35S]mrthionine-labelled Zn ~~itro translation products of RX.4 complementary to hDm85 and actin DNA WPPCcut from rlwtrophorrtic~ gels, added to chickt,l> trnpomyosin. and digesttbd ait,h chymotrypsin in the prewncr of SDS. The digrstion prodovt?; ww~~ wparatrd iu an SDS/l5?,, polyacrylamidv gvl, stained with C’oomassir blur snd Huorograrrhed. Lanes I and 3. (‘oomassie blue-stained gel showing the chicken tropomyosin polypeptides. Lanes 2 and 1 arr tht> firlorograms oflancs 1 and 3, and contain thr ] 35S]mc~tt~io~~inr-lat~ellrtl i/c vilro prwumptive I~rosophiln tropomyosiu and Drosophila actin pept,idw. respwtively. (‘, chymotrypsin Arrows denotth thr major staincad chivkrn (lane 1) and labelkd (lane 2) ~rosnphila tropomyosin peptidw t.hat. cemigratv.
with mobilities similar to the contractile myofibrillar proteins of vertebrates, on two-dimensional gel electrophoresis. Two of’ these Drosophila myofibrillar proteins comigrate with the hybrid selection products of ADm85 and show the same shift in mobility in the presence of 8 M-urea as does chicken tropomyosin when analyzed under these conditions (Fig. 2(c) and (d)). A comparison of the two-dimensional autoradiograms also reveals that the two Drosophila tropomyosins have shifted in mobility relative to each other. The more basic tropomyosin I now migrates more slowly than the more acidic tropomyosin IT in the urea/SDS dimension. l’ropomyosin II now comigrates with chicken +tropomyosin. This switch in
I I
I 1
-1
kb
Thr map ww constructed using single and double digests ofthe whole clonr DNA or isolated restriction fragments. The darkened areasare thaw restric%ion fragmmts with homology to poly(A)+ RNA from I:! to 24-h l)rosophi/a c~mhryos. Thr horizontal lines hclow the diagram denotr the regions ofXDmR5 nswl as h,vt)ridization probes for regions I, 2 and 3 as described in the text and the legends to Figs 8 and 9
mobility
of t,he two Drosophila tropomyosins in urea/SDS may be correlated with differences in secondary structure. To strengthen our identification of the proteins translated by RNA complementary to hDm85 as Drosophila tropomyosins, we have compared the partial proteolytic digests of these proteins to t,hose of purified chicken tropomyosin. The [35S]methionine-labelled Drosophila tropomyosin and actin that had been translated in vitro were eluted from one-dimensional SDS/polyacrylamide gels, mixed with unlabelled chicken tropomyosin and digested with chymotrypsin (Cleveland et al.. 1977). The results (Fig. 3) show that the digests ofthe Drosophila tropomyosins contain at least five major identifiable peptides that comigrate with digestion products of Coomassie blue-stained chicken tropomyosin. There are, however. other Drosophila and chicken tropomyosin peptidrs that do not comigrat,e. As a of [3sS]methionine peptides to excess control against non-specific “sticking” unlabelled chicken peptides, digests of radioactive Drosophila actin were also analyzed and compared to chicken tropomyosin digests. These non-homologous proteins have only one comigrating peptide when digested under the same conditions. From the series of analyses described we conclude that the translation products of RNA selected by /\Dm85 are Drosophila tropomyosins and that these tropomyosins are partially homologous t’o chicken tropomyosins.
(b) Restriction
et~donuclrasr
map
A restriction endonuclease ma,p of hDm85 was determined using data from single and double restriction endonuclease digests of the emire clone or of isolated restriction fragments (Fig. 4). The transcribed regions of clone hDm85 were identified by determining the restriction fragments homologous to tot,al poly(A)+ RNA from 12 t,o 24-hour embryos. This was accomplished by hybridization of ( 32P]RPI;A or of complementary ]32P]DNA to transfers of the restriction digest’ fragments bound to nitrocellulose filters (Southern, 1975). Three distinct regions of /\Dm85 homologous to the RNA probe were identified and named regions I,2 and 3 (Fig. 4). Based on the intensity of the hybridization signal to each of-the t,ranscribed regions, we estimate that RNAs homologous to regions 1 and 2 are more abundant, in our 12 to 24-hour embryo RNA probe than RNAs homologous to region 3. Although
240
V. 1,. BAL-T(“H
ET .A/,.
we have not determined the exact boundaries of the transcripts within the restriction fragments encompassing the regions: we can conclude that transcribed regions 1 and 2 are separated by at least 3.0 kb of DNA and regions 2 and 3 by approximately 1.3 kb of DNA. When restriction fragments of hDm85 are hybridized with RNA from one to fivehour embryos or Kc cells, little or no hybridization to regions 1 and 2 is detected; however, the restriction fragments of region 3 showing hybridization are identical to those hybridized with 12 to 24-hour embryo RNA. Thus, as discussed below, the RNA transcripts of region 3 from one to five-hour embryos differ from those seen in late embryos, yet we detect’ no difference in the DNA fragment transcribed.
(c) Identijcation
of tropomyosirr
genes on ADm85
We identified the regions of ADm85 encoding the tropomyosin genes by determining the hybrid-selected translation products of regions 1 and 2. A Pstl restriction fragment of hDm85 was subcloned into pBR322 (Bolivar et al., 1977). The resulting chimera plasmid, designated pBS852, showed homology to the 32Plabelled RNA probe. pBS852 was labelled by nick-translation and hybridized to Southern transfers of restriction enzyme-digested hDm85. In this manner the subcloned fragment was localized and mapped within region 2 of hDm85 (Fig. 4). The hybrid-selected translation product of RNA complementary to this subclone was identified as tropomysin II by two-dimensional gel electrophoresis (Fig. 5(a)). The second dimension contained 8 M-urea and unlabelled chicken tropomyosin was included as an identity marker. The trace amounts of actin and tropomysin T attributed to non-specific sticking of RNA to the filter or of partial homology of the RNA were also used as reference markers. The DNA of region 1 was isolated using a different strategy. A restriction enzyme digest of hDm85 was transferred onto a nitrocellulose filter. The piece of filter containing region 1 DNA was cut from the filter and used in a hybrid-selection translation assay. The translation product of RN,4 selected by region 1 was identified as tropomyosin I by two-dimensional gel electrophoresis as described above (Fig. 5(b)). Th e interpretation of these results is that the genes encoding tropomyosin I and tropomyosin II are located within regions 1 and 2, respectively. No polypeptide product was detected when region 3 DNA was bound to a nitrocellulose filter and used in a hybrid-selection translation assay.
(d) Chromosomal
location
oj tropomyosi,n
genes
The chromosomal location of hDm85 was determined by hybridization salivary gland polytene chromosomes of Drosophila melanogaster (Gall 1971). hDm85, nick-translated with 3H-labelled deoxynucleotides, was to salivary gland polytene chromosomes and located by autoradiography. band of hybridization was detected in region 88F2-5 on chromosome arm after relatively short exposures (1 week). No additional hybridization seen even after longer exposure times (7 to 8 weeks), although small
in situ to & Pardue, hybridized A single 3R (Fig. 6) bands were regions of
Drosophila
TROPOMYOAIN
GENES
241
FK:. 5. Hybrid-selected translation products of isolated region I and region 2 DSA (:els are pH 4 to 6 (basic to the left) in the isoelectric focusing dimension and SDS/l3”,, polyacrylamide with X ~-urea in the molecular weight dimension. Isolated region 1 or region 2 DNA (see the text) was used in a hybrid-selection translation assay with 5~8 poly(A)+ RNA from 12 to 24-h embryos. [35S]methionine-labelled products were electrophoresed with unlabelled chicken tropomyosin and the gels were stained and prepared for fluorography. (a) The product of pBS85-2 (region 2) DNA. (h) The product of the 6.0 kb PcoRI-HamHI restriction fragment of hDm85 (region 1) DSA. E, endogenous; A. actin: TM, tropomyosin. The arrows indicate the slight amount of TM I in (a) and the sticking of [%]methionine to the Coomassie blue-stained chicken \-TM in (h).
homology outside 88F 2-5 might not have been detected because of the relatively high sequence complexity of the probe. We would, however, expect to detect other copies of the tropomyosin genes if the additional genes had a degree of homology equal to that of members of the r-tubulin gene family (Mischke & Pardue, 1982). We extended our search for genomic DNA sequences homologous to hDm85 by an analysis of whole-genome hybridizations. An EcoRI restriction digest of D. melanogaster (Canton S) genomic DNA and a similar genomic digest containing the equivalent of one genomic copy of ADm85 DNA were transferred onto nitrocellulose filters and hybridized with 32P-labelled nick translated hDm85. Because hDm85 was constructed by creating artificial EcoRI restriction sites, an EcoRI digest of genomic DNA should generate at least two sets of restriction fragments homologous to /\Dm85: a set of fragments with internal restriction sites that comigrate with the 7.2 kb and 24 kb EcoRI restriction fragments of hDm85. and two fragments larger than the 7.5 kb and 0.9 kb end fragments of the Drosophila DNA insert in hDm85 that extend to the next genomic EcoRI site on either side. All of the hybridization bands seen in the EcoRI digest of genomic DNA could be accounted for by these two sets of fragments (Fig. 7). The @9 kb and
FIG. 6. Hybridization in situ of hDm85. ADm8.5 was labelled with I’H]deoxynucleodites by nick-translation and hybridized to Drosophila salivary gland polyt,enr chromosomes. Hybridization is seen only to region 88F 2-5 on chromosome arm 3R. Hybridization was in 0.3 r+NaCI, 02 rv-Tris (pH 6.8). at, 65°C. Exposure was for 7 days.
7.5 kb EcoRI restriction fragments located at the ends of the Drosophila DNA insert in hDm85 are replaced by two hybridization bands 3.9 kb and 115 kb in size in the genomic digest. From these results we conclude that all genomic sequences homologous to hDm85 are located within 22 kb of contiguous DNA. Longer exposures do not show hybridization of the tropomyosin-coding sequences to other restriction fragments, as might be expected if Drosophila had additional copies of these genes elsewhere in the genome.
(e) Expression
of transcripts
homologous to /\Dm85 during development
embryonic
and muscle
The developmental regulation of RNAs homologous to hDm85 has been investigated by analyzing the RNAs of staged Drosophila embryos and RNAs of Kc cells, a Drosophila cell line of embryonic origin (Echalier & Ohanessian, 1970). In order to characterize the transcripts and identify their corresponding genes, the RNAs were subjected to glyoxal/agarose gel electrophoresis, transferred to nitrocellulose filters, and hybridized to 32P-labelled nick translated probes (Thomas, 1980). The RNA blots were probed with DNA segments corresponding to each of the transcribed regions in XDm85. The results of these hybridizations (Fig. 8) show that several transcripts with different patterns of expression are homologous to these regions. DNA containing coding region 1 and encoding tropomyosin I (the 6-O kb EcoRIBamHI fragment) shows no hybridization to RNA from one-to-three-hour or oneto-five-hour embryos. In 9 to 11-hour embryos, however, there is an RNA band of 1.8 kb and a more diffuse band with a mean size of 145 kb that show homology to the region 1 probe (visible on longer exposures). Hybridization to both bands is much stronger in RNA from 19 to 23-hour embryos (Fig. 8 lanes 1). The 1.45 kb
11rosophiln
TROPOMYOYIS
GESES
FIG. 7. Hybridization of hDmX5 to genomic D?r‘A. DSA was digested with EcoRI, elcctrophoresed in a 1” 0 agarose gel and transferred to nitrocellulose. 32P-labelled hDm85 (2 x 10’ cts/min per pg) was hybridized to the filter. After washing, the filter was exposed to Kodak XAR-5 film with intensifying scwcn at -70°C for 10 h. Lane I, 1 ng hDmR:i DSA+lOpg total Drosophila (Canton S) DSA (approx. I genome equivalent): lane 2, 1Opg total Drosophila (Canton 8) DNA. On the right the sizes of the hybridizing genomic bands arc given, in kb. The arrows indicate the vector arms in hDm85. The molecular weight standards were a Hind111 digest of h DSA.
“41
V. 1,. B.AlrT(‘H
ET .-1L.
band is resolved into two bands of 1.4 and 15 kb in shorter exposures (data not shown). DNA from coding region 2 (pBSS5S), encoding tropomyosin II. also does not hybridize to any transcripts in early embryos but shows some hybridization to a 1.5 kb transcript that appears in 9 to 1 l-hour RNA (visible on longer exposures) and increases in abundance in 19 to 23-hour RNA (Fig. 8 lanes 2). The probes for regions 1 and 2 do not show any hybridization to Kc cell RNA (data not shown). Thus the temporal pattern of appearance of homologous RXA is similar fol region 1 and 2. DNA containing coding region 3 was subcloned by inserting the purified 2% kb HindIlLBumHI restriction fragment into the plasmid PACYC184 (Chang & Cohen. 1978). This construction, designated pVB85-1, hybridizes to an abundant 2.3 kb transcript in RNA from one- to three- and one- to five-hour embryos. This RNA, however, is not detected in older embryos. The coding region 3 probe also hybridizes to a 2.8 kb transcript that is present in rather low amounts in one- to three- and oneto five-hour embryos, becomes very abundant in 9 to I l-hour RNA. and then decreases by 19 to 23-hours of development. In addition, the region 3 probe also hybridizes weakly to a transcript of 1.5 kb in the RNA from older embryos (Fig. 8 lanes 3). The coding region 3 probe hybridizes to a single transcript of 2.8 kb in Kc cell RSA (Fig. 8 lane 4). The data show that the major transcripts homologous t,o region 1. containing the tropomyosin I coding region, and to region 2, containing the tropomyosin II coding
.2.8 .P3
ab
cd 1
a
b
c 2
d
a
b
c 3
d 4
FIG;. 8. Hybridization oftranscribed regions ofhDm85 DSA to Dro,so~hila embryonic and Kc call RSA. Poly(A)+ RNA (3 pg) from staged embryos or Kc cells was subjected to glyoxal/agarose gel electrophoresis and transferred to nitrocellulose filters. The filters were hybridized with 32P-labelled DNA corresponding to each of the 3 transcribed regions of ADm85. After washing, the filters were exposed to Kodak XAR-5 film with intensifying screen at -70°C for 15 h. Lane a of filters 1 to 3 is 1 to 3-h embryo RNA : lane b is 1 to 5-h embryo RNA ; lane c is 9 to 11-h embryo RNA ; and lane d is 19 to 23-h embryo RXA. Filter 4 contains Kc cell RNA. The probes were labelled to a specific activity of 10s to 109 Cerenkov cts/pg and arc as follows: filter 1. 6.0 kb EcoRI-RamHI restriction fragment (region 1) of hDm85 DNA: filter 2, pBR85-2 (region 2) DlVA: filters 3 and 4. pVB%-1 (region 3) DPU’A. In the margins the sizes of the hybridizing bands are given. in kb. The molecular weight standards were a Hind111 digest of h DNA and an HaeIII digest of +X174 DNA.
Drosophila
TROPOMYOHIN
GENES
245
region, first become detectable in 9 to 11-hour embryos and increase in abundance in later stages. This pattern of gene expression correlates with the observation that muscle cell fusion is first detectable at about 9 to 11 hours of development (Poulson, 1950). It is also consistent with analyses of embryo poly(A)+ RNA by in vitro translation. Tropomyosins I and 11 are very abundant translation products directed by poly(A)+ RNA from 18 to 20.hour embryos, but these proteins are not present in appreciable quantities in the translation products of poly(A)+ RNA from bwo- to three-hour embryos (data not shown). The pattern ofexpression oftranscripts homologous to XDm85 in muscle cells was also studied in primary myogenic cultures of Drosophila cells (Storti et al., 1978). Total poly(A)+ RNA was isolated from myogenic cells collected 4, 12 and 18 hours after plating of cell suspensions made from two- to three-hour-old embryos. These times correspond to <3% fusion, 50% fusion and SOY/,fusion (Storti et al.. 1978). RNAs from the cell cultures were separated by electrophoresis on glyoxal/agarosr gels, t,ransferred to nitrocellulose filters and hybridized with the same DNA probes that had been used to analyze the embryo RNA. DNA containing coding region 1 hybridizes to a 1.8 kb transcript, and a broad band of hybridization at 1.4 to 1.5 kb (Fig. 9 lanes 1). These transcripts are detected only in cultures with significant myotube formation (12 h and 18 h post-plating). The region 2 probe hybridizes to a single transcript of I.5 kb that is present in approximately equal quantities in 12 and 1%hour RNA (Fig. 9 lanes 2). Neither the region 1 probe nor the region 2 show hybridization with myoblast (4 h post-plating) RNA. As is the case in the intact embryos, the appearance of RNA homologous to the regions of the tropomyosin genes is correlated with the differentiation of myotubes. These observations are supported by experiments that show that the tropomyosins are greatly enriched in the in vitro translation two Drosophila products of RNA from cultures containing a high proportion of mpot’ubes (data not, shown). The probe containing region 3 hybridizes to a 2.8 kb transcript that is present in unfused myoblasts and increases in abundance as fusion progresses (Fig. 9 lanes 3). This probe also hybridizes weakly to a transcript of 1.5 kb that is detectable only in cultures enriched in myotubes (12 h and 18 h post-plating). We consider it likely that the 2.8 kb transcript in the myogenic cell RNA is the same as the 2.3 kb transcript seen in the embryo Rh’As and in Kc cell RXA. In cultured myogenic cells this transcript increases in abundance after the time of fusion and does not show the decreased abundance seen in 19 to 23.hour embryos. Not transcript corresponding to the 2.3 kb RX’A seen in early embryos is detected in myogenic cells or in the Kc cells. It is possible that the myogenic cells and the embryo cells from which the Kc cell line was derived have passed the appropriate developmental time for expression of the 2.3 kb transcript. 4. Discussion We have isolated two Drosophila tropomyosin genes on a single genomic clone (ADmE%). These genes were identified by characterizing the hybrid-selected translation products of RNA homologous to the clone. /\Dm85 hybrid-selected RSA
\‘.
a
b
1
c
L. BAITTC’H
a
b
2
ET AL
c
a
b
c
3
lQ(:. 9. Hybridization of transcribed regions of hDtn8.i DSA to RN.4 from 11roaophi/a myogrnic cells. Poly(A)’ RNA (3 pg) from myogenic cells collected at 1. I:! or 18 h post-plating was subject,ed to glyoxal/agarose gel electrophoresis and transferred to nitrorellulose filters. The filters xcrc hybridized with 32P-jabelled DSA corresponding to the transcribed regions of hDm85. After washing, thtl filters were exposed to Kodak XAR-5 film with intensifying screen at - 70°C for 1 days. Lane a offilters 1 to 3 is RNA collected 4 h post-plating: lane b is RNA collected 12 h post-plating; and lane c is K?rTA collected 18 h post-plating. Probes were labelled to a specific activity of 108 to lo9 rerenkov cts/pg and are as follows: filter 1, 6.0 kb EcoRI-BamHI restriction fragment (region 1) of hDm85 DNA: filter 2. pBS85-2 (region 2) DNA ; filter 3, pVB85-I (region 3) DSA. In the margins the sizes of the hybridizing bands are given. in kb. The molecular weight standards were a Hind111 digest of X DNA and a HafIII digest of ~5x174 DS.4.
encoding 35,000 molecular weight and 34,000 molecular weight polypeptides. These polypeptides comigrate with purified chicken r-tropomyosin and with two polypeptides present in llrosophila myofibrillar protein preparations on one- and two-dimensional gels. The translation products also show the apparent increase in molecular weight when electrophoresed in the presence of 8 M-urea that is characteristic of vertebrate tropomyosins. The peptides generated by partial proteolytic digestion of the proteins hybrid-selected by hDm85 show partial homology to t,hose of purified chicken tropomyosin. Thus the Drosophila tropomyosins have a number of properties similar to those that characterize the vertebrate and invertebrate tropomyosins that have been investigated. We have designated the 34,000 M, polypeptide tropomyosin I and 35,000 M, polypeptide tropomyosin II. The 18 kb of Drosophila DNA in hDm85 contains three distinct transcribed regions. Transcribed regions 1 and 2 encode tropomyosins I and II, respectively. Region 1 is approximately 49 kb and region 2 is approximately 2.1 kb in size, and they are separated by about 3 kb of DNA. The gene region encoding tropomyosin II is homologous to a single transcript of
I~rocsophi/u
TROPOMYOSIK
GEKES
211
1.5 kb. This RNA is approximately twice the size necessary to encode a protein of about 280 amino acids, which would be expected from the size of vertebrate tropomyosins. This discrepancy suggests a significant amount of untranslated RNA. The gene region encoding tropomyosin I hybridizes to a transcript of 1.8 kb and a broad band of RNA that is composed of two transcripts of 1.5 and I.4 kb. Because of the abundance of the 1.8 kb transcript it seems unlikely that it is a precursor to the smaller transcripts. Moreover, preliminary results show that a 1.3 kb probe from the left end of coding region 1 hybridizes to all three transcripts. These RNAs might, therefore, be the result ofalternative processing ofa single gene transcript (Hagenbuchle et al., 1981; Marie et aE., 1981) rather than the transcripts arising from two adjacent genes. We have not determined which of these transcripts encodes tropomyosin I. The mRNAs for tropomyosin I and tropomyosin TI show little, if any, cross hybridization either in the hybrid-selection assays or in Southern blot hybridization. Thus, our data are consistent with the finding that a cDNA clone of chicken I-tropomyosin does not cross-hybridize to chicken fi-tropomyosin RNA under similar conditions (MacLeod, 1981). The expression of /\Dm85 regions 1 and 2 (containing the tropomyosin 1 and I I genes. respectively) is developmentally regulated. Embryonic transcripts homologous to each of these coding regions are first detected in RNA from embryos at, 9 to 1 I hours of development. The expression of the two tropomyosin genes may not be exactly synchronous, however, since transcripts of region 1 are considerably more abundant than those of region 2 at 9 to 11 hours. The developmental time period during which the first transcripts of these genes can be observed correlates very well with the onset of muscle cell fusion in the embryo (Poulson, 1950). In addition t,o the two tropomyosin-coding regions, ADrn8.5 contains a third transcribed region that is homologous to several transcripts. R’egion 3 is approximately 2.8 kb in size and is separated from the tropomyosin II-coding region by about, 1.3 kb of DNA. We have not identified a protein encoded by this region. The first major transcript homologous to this region, a 2.3 kb transcript,, is present from the onset of development but disappears at about, the time of’ gastrulation (3 to 5 h) and, therefore. is most likely a mat)ernal transcript. The other major transcript is 2.8 kb in size and appears at, or shortly after gastrulation and in mvogenic cultures both before andafter myotube format,ion. The 2.8 kb transcript is also found in Drosophila Kc cells. Region 3 is not large enough to contain two separate transcribed regions of2.3 and 2.8 kb. Thus it seems probable that these tw o transcripts arise by differential transcription or by differential processing ofa. single primary transcript (Hagenbuchle et al., 1981 : Marie et al.. 1981), although other possibilities such as a rearrangement of the gene (Early et al.. 1980: Hicks rf al., 1979) cannot be completely ruled out. The coding region 3 probe also hybridizes weakly to a 1.5 kb t)ranscript in late embryos. The fact that transcripts of approximately the same size are detected with the tropomyosins I and II coding region probes suggests a possible low level of homology among the coding regions. We are currently constructing subclones of all the coding regions in order to determine their homologies at the sequence level. F’reliminary hybridization experiments indicate that t,here are weak homologies
24x
T. L. BAITT(“H
El’
AI,.
between coding region 3 and the tropomyosin coding regions. It is, therefore, tempting to speculate that coding region 3 may encode a non-muscle tropomyosin, although it has yet to be shown that Drosophila has tropomyosin in non-muscle cells and we have not yet identified a protein product of the rather abundant transcripts homologous to coding region 3. We have localized the Drosophila tropomyosin genes to the chromosomal region 88F 2-5 by hybridization in situ. One of the six Drosophila actin genes has also been mapped to region 88F (Tobin et al., 1980: Fyrberg et al., 1980). We do not know whether there is a functional basis for this relative proximity. Particularly, it. should be noted that it is not known whether the 88F actin gene encodes a muscle-specific isoform and that the chromosomal region may well span more than 100 kb (Bridges, 1935 : Rudkin, 1972). The hybridization experiments in situ give no evidence that sequences homologous to hDm85 are located elsewhere in t,he genome. This conclusion is strengthened by hybridization of hDm85 to restriction enzyme-cleaved genomic DNA. These results indicate that no portion of hDm85 is homologous to sequences outside the 22 kb of contiguous DSA surrounding the tropomyosin genes. Thus the tropomyosin genes of Drosophila do not appear to belong to a family of partially homologous genes scattered over the genome, as do the Drosophila actin (Tobin et al., 1980: Fyrberg et al., 1980) and tubulin genes (Sanchez et al., 1980: Kalfayan gL Wensink, 1981; Mischke & Pardue, 1982). The correlation between the appearance of abundant transcripts homologous to the tropomyosin genes in hDm85 and the differentiation of myotubes is strong evidence that both tropomyosin genes are expressed in muscle tissue. The data do not’ exclude the possibility that these same genes code for tropomyosin in nonmuscle. cells. However, if these genes are expressed in non-muscle cells, their level of expression is substantially lower than that of muscle cells. The possibility that hDm85 also contains a structural gene for non-muscle tropomyosin is currently under invest,igation. Fortunately the single locus and probable single-copy nature of the cloned tropomyosin genes will facilitate genetic approaches to this question. The authors thank Dr Michael Barany for providing us with chicken tropomyosin. We are also indebted to Dr Anthony Mahowald and Dr Sarah Elgin for a generous supply of embryos and to Dr Pieter Wrnsink for providing us with the genomic actin clone XDmAl. R.V.S. and V.L.B. also acknowledge the expert technical assistance of MS Alice Szwast. This from the National Tnstitutes of Health and work was supported hy a grant Campus Research Board and Biomedical Research Support grants from Cnirersity of Illinois (to R.V.S.). and a grant from the NIH (to M.L.P.). R.V.S. is a recipient of a Research Career Development award from the National Institutes of Health. V.L.B. was supported by a University of Illinois Graduate College fellowship. D.M. was supported by a fellowship from the Deutschr Forschungsgemeinschaft.
REFERENCES Amphlrtt, (;. W., Syska. H. Br Perry. S. V. (1976). PE’RS Letters, 63, 22%. Bingham, P. M., Levis, R. 8: Rubin, (4. M. (1981). (‘~41, 25, 693-704. Bolivar, F., Rodriguez, K. L., Greene. P. J., Betlach. M. C.. Heyneker, H. L., Boyer, Cross, J. H. & Falkow, S. (1977). (Gene, 2, 95-l 13.
H. W..
Drosophiln
TROPOMYOSIX
(GENES
219
Bonner, W. M. & Laskey, R. A. (1974). Eur. J. Biochem. 46, 83-88. Bridges, C. B. (1935). J. Hered. 26, 60-64. Buckingham, M. E. (1977). In Biochemistry of Cell Diff erentiatio,n, vol. 13 (Paul,
“50
V. I,. BAlJTCH
E’T AL.
Rudkin, G. T. (1972). In DevelopmerLtal Studies on Giant Chromosomes (Beermann, W., ed.), pp. 58-85, Springer, New York. Sanchez, F., Natzle, J. E., Cleveland, D. W., Kirschner, M. W. & McCarthy, B. J. (1980). (le/l,
22, 845-854. Scott. M. P., Storti, R. V., Pardue, M. L. & Rich, A. ( 1979). Biochemistry, 18, 1588-1598. Seecoff, R. L., Gerson, I., Donady, J. J. & Teplitz, R. 1,. (1973). Develop. Biol. 35. 250-261. Sender, F’. M. (1971). FEBS Letters, 17, 106-l 10. Smillie, L. B. (1979). Trans. Znt. Biochem. Sot. 4. 151.-155. Southern, E. M. (1975). J. Mol. Biol. 98, 503- 518. Spradling, A. C., Pardue, M. L. & Penman, S. (1977). J. MoZ. Hiol. 199> 5599571. Stone, D. & Smillie, L. B. (1978). J. Biol. Chem. 253, 113771148. Storti, R. V. & Rich, A. (1976). Proc. Nat. Acad. Sri., I’.S.d. 73, 2346-2350. Storti, R. V., Horovitch, 8. J., Scott, M. I’., Rich, A. Pr I’ardue, M. L. (1978). Cell, 13,589-598. Storti. R. V., Scott, M. P., Rich, A. & Pardue, M. L. (1980). CelZ, 22, 825-834. Thomas, I’. 8. (1980). Proc. Nat. dcad. Sci.. C's.d. 77. 5201-5205. Tobin, S. L., Zulauf, E., Sanchez, F., Craig, E. A. Br McCarthy, B. J. (1980). CelZ, 19, 121-131. Vogelstein, B. & Gillespie, D. (1979). Proc. Nat. Acad. Sci., li.S.il. 76, 615419. Woods. E. F. & Pont, M. J. (1971). Riochemistry, 10, 270-276.
Edited
by S. Brwnrr