Upload
independent
View
0
Download
0
Embed Size (px)
Citation preview
A Family of Drosophila Genes Encodingquaking-Related Maxi-KH Domains
Christine Fyrberg,1 Jodi Becker,1 Peter Barthmaier,1 James Mahaffey,2
and Eric Fyrberg1,3
Received 4 Nov. 1997Ð Final 16 Dec. 1997
We recently identi® ed a Drosophila gene, wings held out (who), that speci® es aSTAR (signal transduction and RNA activation) protein expressed within meso-derm and muscles. Genetic evidence suggests that WHO regulates muscledevelopment and function in response to steroid hormone titer. who is related tothe mouse quaking gene, essential for embryogenesis and neural myelination, andgld-1, a nematode tumor suppressor gene necessary for oocyte differentiation,both of which contain RNA binding `̀ maxi-KH’’ domains presumed to link RNAmetabolism to cell signaling. To initiate a broader study of Drosophila WHO-related proteins we used degenerate primers encoding peptides unique to maxi-KHdomains to amplify the corresponding genes. We recovered nine genes, allspecifying single maxi-KH domain proteins having tripartite regions of similaritythat extend over 200 amino acids. One is located within the 54D chromosomesubdivision, and one within 58C, while the remaining seven are within the 58Esubdivision. At least four of these STAR proteins are expressed in a generalmanner, suggesting that maxi-KH domains are employed widely in Drosophila.
KEY WORDS: Drosophila genetics; K homology domains; STAR proteins; nucleic acid-binding motifs.
INTRODUCTION
The K homology (KH) domain is present within a variety of proteins. Described
originally as a triple repeat within the hnRNP K protein primary sequence (Siomi
1 Department of Biology, The Johns Hopkins University, 3400 North Charles Street, Baltimore,
Maryland 21218.2
Department of Genetics, North Carolina State University, Raleigh, North Carolina 27695.3
To whom correspondence should be addressed. Fax: (410) 516-5213. e-mail: [email protected].
Biochemical Genetics, Vol. 36, Nos. 1/2, 1998
51
0006-2928/98/0200-00 51$15.00/0 r 1998 Plenum Publishing Corporation
et al., 1993a), single or multiple copies of the motif have been found in
polypeptides ranging from ribonuclease PNP of Escherichia coli (Regnier et al.,
1987) to human FMR1, encoded by the fragile-X mental retardation gene (Siomi
et al., 1993b). The common property of all proteins containing KH domains is
affinity for RNA or single-stranded DNA. For example, E. coli protein Nus A
binds to newly transcribed mRNA (Liu and Hanna, 1995), while the KH domain
of the prokaryotic ribosomal protein S3 is associated with ribosomally bound
mRNA (Urlaub et al., 1995). At least two single KH domain-containing proteins,
both having a preference for poly(rC), have been demonstrated to bind nucleic
acid in vitro (Ito et al., 1994; Leffers et al., 1995). Investigations of the
three-dimensional solution structure of KH domains, speci® cally of the sixth KH
domain of vigilin, an endoplasmic reticulum protein that contacts free ribosomes
(Schmidt et al., 1992; Musco et al., 1996), have established that the motif is
typically 80±100 residues in length and revealed its overall topology as b a a b b a ,
with RNA localized between the ® rst two a helices (Musco et al., 1996).
While it is reasonably well established that KH proteins typically bind RNA
or single-stranded DNA, neither the scope of their biochemical roles nor the
functions of individual proteins in particular cellular and developmental events
are known in satisfactory detail. In Drosophila melanogaster, several investiga-
tions have de® ned the functions of particular KH domain proteins in part. For
example, Drosophila protein PS1 represses the maturation of P element RNA
transcripts in somatic cells, preventing the removal of the third intron sequence by
attaching to the exon located immediately 58 to it (Siebel et al., 1995). Another ¯ y
protein, BICAUDAL C, expressed in the female germline, speci® es the anterior of
the oocyte (Mohler and Wieschaus, 1986; Schupbach and Wieschaus, 1991). KH
domain mutations of BICAUDAL C profoundly disrupt embryo polarity, causing
improper localization of follicle cells and faulty development of anterior struc-
tures (Mahone et al., 1995), although the mechanism by which disruption of RNA
binding creates these defects is not yet understood.
We and others recently discovered that a Drosophila KH domain protein
similar to those encoded by the mouse quaking and nematode gld-1 genes is
involved in muscle development (Baehrecke, 1997; Zaffran et al., 1997; Fyrberg
et al., 1997; Lo and Frasch, 1997). Transcripts of the gene, named wings held-out,
or who (aliases how, qkr93F, and struthio), accumulate in nuclei of muscles and
epithelial tendon cells, suggesting that WHO regulates gene activity in muscles as
well as in cells to which they attach. Synthesis of WHO is ecdysone dependent,
and mutations within the protein perturb muscle development and function,
suggesting that the gene coordinates myotube migration and patterning in
response to the titer of the steroid hormone ecdysone (Baehrecke, 1997). The
WHO KH domain closely resembles those of the SAM68 (Src-associated in
mitosis; 68-kD) family of proteins (Vernet and Artzt, 1997). The amino acid
52 Fyrberg, Becker, Barthmaier, Mahaffey, and Fyrberg
similarity of this protein family extends beyond the borders of the KH domain,
implying that the shared function may entail more than nucleic acid binding. The
prototype of this group, SAM68, is a phosphoprotei n target of SRC and FYN
signal transduction in mitotic cells (Taylor and Shalloway, 1994), hence the
suggestion that related proteins may also be part of signal transduction cascades
(Ebersole et al., 1996) and the recent naming of this protein family STAR (signal
transduction and activation of RNA (Vernet and Artzt, 1997).
As a step toward evaluating these ideas we have attempted to isolate
Drosophila genes encoding maxi-KH domain proteins using structural and
sequence information generated in studies of mouse quaking and Sam68 and
Drosophila who. Using degenerate PCR primers encoding peptides that are
unique to WHO and other members of the STAR protein family, we have
succeeded in isolating nine related Drosophila genes. Here we present their partial
characterization.
MATERIALS AND METHODS
Ampli® cation and Isolation of Related Genes
By comparing the primary sequences of the mouse quaking gene and several
relatives with that of the Drosophila wings held out (who), aliases held out wings
(how), and quaking-related gene at 93F (qkr93F) (Baehrecke, 1997; Zaffran et al.,
1997; Fyrberg et al., 1997), we were able to identify similar regions and construct
degenerate oligonucleotides for amplifying related Drosophila genes. The only
regions chosen for this endeavor were portions of the KH domain that are unique
to maxi-KH domain proteins. We utilized two degenerate overlapping forward
primers specifying the peptides DHPDFNFV (GGAATTCg/cAg/cc/tAc/t
CCXGAc/tTTc/tAAc/tTTc/tGT, where lowercase letters designate degenerate
nucleotides) and PDFNFVG, (GGAATTCCCXGAc/t TTc/tAAc/tTTc/tGTXGG),
respectively, and also encoding an EcoRI endonuclease site, in conjunction with
two degenerate reverse primers. The ® rst speci® es the peptide GKPNWEHL
(GGGGATCCg/aTGc/tTCCCAg/aTTXGGc/tTTXCC) and includes a BamHI site,
while the second speci® es SMRDKKK (GGGGATCCc/tTCc/tTTc/tTTc/tTTa/
gTCXCg/tCAT) and includes the same endonuclease site. We designed our
degenerate primers in accordance with codon bias characteristic of Drosophila.
After annealing primers to genomic Drosophila DNA, or reverse-transcribed
RNA at 45ÊC, we ampli® ed products for 30 cycles and selected those having the
expected electrophoretic mobilities. PCR products were subcloned into the
pUC19 vector and sequenced by the dideoxy-extension method (Sanger et al.,
1977).
quaking-Related KH Domains 53
Gene and cDNA Isolation
Fragments ampli® ed using degenerate primers in conjunction with the PCR were
used to screen either genomic libraries or cDNA populations representing various
stages of the Drosophila life cycle (Poole et al., 1985).
DNA Sequencing
All DNA sequences were generated using the dideoxy-extension method (Sanger
et al., 1977). In all instances we sequenced both strands of genes in an
overlapping manner.
Protein Sequence Comparisons
Sequence relatedness was analyzed and visualized using the TFASTA program
and aligned using the Pileup and Pretty programs of the UWGCG package
(Devereux et al., 1984). Highlighting was done using the SEQVU program
(Garvan Institute, Sydney, Australia).
Chromosomal Localization of KH Domain-Containing Genes
We prepared larval polytene chromosomes by the technique of Gall and Pardue
(1971). 3H-Labeled cRNA was synthesized using the protocol of Wensink et al.
(1974). Hybridization RNA and visualization of labeled regions were as described
by Fyrberg et al. (1980).
RNA Localization
Embryos for whole-mount in situ localization were dechorionated and ® xed
following the procedure of Tautz and Pfei¯ e (1989). In situ analysis utilized
ribonucleotide probes generated using an RNA transcription kit (Stratagene) and
DIG-11-UTP (Boehringer-Mannheim). cDNAs for each of the KH domain genes
were cloned in Bluescript vectors, and digoxygenin-labeled transcripts prepared
by incubating linearized template DNA with either T3 or T7 polymerase.
Hybridization was carried out using the method of Tautz and Pfei¯ e (1989) with
only minor modi® cations. Anti-DIG-AP (Boehringer-Mannheim) was used to
detect hybridization .
Northern Blotting
We extracted poly(A)-containing RNA from synchronously developing Dro-
sophila cultures using the SDS±phenol technique and oligo(d)T chromatography.
54 Fyrberg, Becker, Barthmaier, Mahaffey, and Fyrberg
RNA preparations were separated by electrophoresis in 20 mM sodium acetate, 1
mM EDTA. For details of RNA transfer and hybridization , refer to Fyrberg and
colleagues (1983).
RESULTS
PCR Ampli® cation of quaking -Related Drosophila Genes
Our ® rst step in this investigation was to amplify Drosophila genes encoding
proteins closely related to quaking and other members of the SAM68 family. This
task was facilitated by recent sequence comparisons of maxi-KH domain-
containing proteins (see, e.g., Jones and Schedl, 1995; Ebersole et al., 1996;
Musco et al., 1996) and, in particular, by structural analysis of the sixth KH repeat
of human vigilin (Musco et al., 1996), which revealed that its topology is
b a a b b a . The protein sequence alignments, together with the structural character-
istics, demonstrate that mouse QK, nematode GLD-1, and other maxi-KH domain
proteins share two short sequences that are absent from other proteins of this type,
the ® rst being a seven-amino acid insertion preceding helix a 1, and the second
comprising a ¯ exible loop of 16±20 amino acids located between domain b 2 and
domain b 3. Utilizing this information, we were able to design two pairs of
oligonucleotide primers that would anneal uniquely to cDNAs encoding maxi-KH
domain proteins. The two forward primers chosen encode the overlapping
peptides DHPDFNFV (F1) and PDFNFVG (F2), both of which are part of the
seven amino acid insertion. The reverse primers specify the nonoverlapping
peptides GKPNWEHL (R1) and SMRDKKK (R2), both of which are part of the
¯ exible loop. Locations of these peptides relative to primary sequences of human
vigilin, Drosophila WHO, and additional STAR proteins are illustrated in Fig. 1.
Hybridization of the respective forward and reverse primer pairs to cDNA
representing several Drosophila developmental stages, together with the polymer-
ase chain reaction, allowed us to synthesize fragments having the expected 150-
and 200-nucleotide sizes.
To ascertain how many distinct cDNA sequences the ampli® ed fragments
represented, we ligated the population of molecules into the plasmid vector
pUC19, grew up individual transformants, prepared DNA from each, and deter-
mined their nucleotide sequences. After analyzing approximately 100 sequences
in this manner, we established that our PCR ampli® cation represented, in the
main, four distinct cDNAs and one genomic clone for which we were unable to
obtain a cDNA (and which will not be further described). We sequenced
fragments for each completely and used these as probes to isolate complete
genomic and cDNA sequences. Insofar as we are aware, expressed sequence tags
are not available for any of these genes.
quaking-Related KH Domains 55
Primary Sequence Comparisons
Alignment of the conceptual protein translations derived from full length cDNAs
for the four genes using the SEQVU program, illustrated in Fig. 2, demonstrated
that all of the primary sequences share similar tripartite regions extending for
approximately 200 amino acids, a hallmark of STAR proteins (Vernet and Artzt,
1997). The most similar portion of the conserved sequence encodes the KH
domain. Forty-® ve percent of the requisite 100 amino acids are identical in all
Fig. 1. Locations of peptides unique to Drosophila WHO and maxi-KH domain proteins. The® gure illustrates computer-generated alignments of KH domains from mammalian Vigilin,
SAM68, QUAKING, Drosophila WHO (alias QKR93F), and brine shrimp GRP33. Vigilincontains multiple `̀ conventional’ ’ KH domains (that illustrated is repeat number six, which is 68residues in length), while the latter four proteins contain single `̀ maxi’ ’ -KH domains of 100 or
102 residues. The length differences are accounted for by two peptide insertions. Locations withinthem of the sequences encoded by our degenerate PCR primers, designated forward1 (F1), F2,
reverse1 (R1), and R2, are indicated.
Fig. 2. Tripartite sequences encoded by four Drosophila quaking -related cDNAs. The ® gure illus-trates computer-generated alignments of sequences encoded by four Drosophila quaking -relatedcDNAs. The most similar portions specify the KH domains, indicated in dark gray. Also signi® cant is
the similarity of the QUA1 (residues 47±111 of Qkr58E-1) and QUA2 (residues 214±306 ofQkr58E-1) domains, indicated in light gray. The remainder of the protein sequences include only
limited similarity, indicated by unshaded boxes.
56 Fyrberg, Becker, Barthmaier, Mahaffey, and Fyrberg
four proteins, and all four conform to the structural characteristic of `̀ maxi KH’ ’
domains.
Amino acid similarity was not con® ned to the RNA binding motifs, and we
noted additional regions of protein similarity located on either side of the KH
domain. Vernet and Artzt (1997) have named these ¯ anking domains QUA1 and
QUA2. These sequences, indicated in light gray (Fig. 2), display signi® cant
similarity, albeit far less than that characteristic of the KH domains. The QUA1
domains (residues 47±111 of QKR58E-1) have only 9 of 67 residues identical in
all four proteins, while we observe only 7 of 95 identical residues in the QUA2
domains (residues 214±306 of QKR58E-1). If our quali® er was three of four
identical residues, then we obtained `̀ matches’ ’ of 24 of 67 residues in QUA1 and
15 of 95 residues in QUA2. Outside of the tripartite domains we observed little
protein similarity, although two carboxy-terminal stretches collectively included
® ve positions wherein residues are identical in three of the four proteins. Finally,
we failed to note any compelling similarities of these proteins to those established
to be involved in signal transduction, for example, SAM68.
A Closely Related quaking -Related Gene Subfamily
In the course of screening genomic and cDNA libraries with qkr58E-3, we
discovered four additional maxi-KH domain genes, all of which are extremely
closely related to qkr58E-3. Figure 3 illustrates alignments of the KH domains of
all ® ve of these genes, demonstrating their similarity. Isolation and sequencing of
corresponding genomic fragments have demonstrated that these indeed represent
four genes distinct from one another and from qkr58E-3. Together with the ® ve
genes described above, this brings the total number of Drosophila maxi-KH
domain genes isolated in this study to nine.
Chromosome Locations
To facilitate genetic approaches to KH domain protein function, we determined
the chromosomal location of each of our genes. We synthesized tritium-labeled
cRNA probes representing each according to the protocol of Wensink et al. (1974)
and hybridized these to polytene chromosomes using the technique of Gall and
Pardue (1971). We found that all of the genes are located on chromosome 2. One,
qkr54B, is located within the 54B subdivision; one, qkr58C, is located within the
58C subdivision; and seven, qkr58E1-7, are distributed within the 58E subdivi-
sion. In only one instance have we delineated the spatial relationship of individual
58E subdivision genes. qkr58E-2 and qkr58E-3 are topologically organized as
shown in Fig. 4. The physical map of a third 58E region gene, qkr58E-1, is known
as well, but its relationship to the remaining six genes within the same subdivision
58 Fyrberg, Becker, Barthmaier, Mahaffey, and Fyrberg
is not. Both the physical maps and the spatial arrangements of the remaining ® ve
58E subdivision genes remain to be derived.
Physical Maps
To re® ne further the ability to identify genetic mutations within the four quaking-
related genes described in Fig. 2, we have generated physical maps for the four
originally isolated. We used the respective full-length cDNAs as probes to screen
genomic libraries for overlapping fragments that contained complete gene copies.
By hybridizing these to the cDNA probes and sequencing the hybridizing
fragments, we were able to generate maps of the locations of all exons and introns.
These are summarized in Fig. 4. Using this information it will be possible to
screen extant Drosophila strains for corresponding mutations. We noted also, in
the course of comparing sequences and exon locations (refer to Table I), that some
intron positions are conserved between genes, strongly suggesting evolution from
an ancestral gene.
Gene Expression Patterns
To investigate temporal and spatial patterns of Drosophila quaking-related gene
expression, we synthesized digoxygenin-la beled probes for qkr54B and qkr58E-1,
Fig. 3. Comparisons of KH domains of Qkr58E-3 and four closely related genes. The ® gureillustrates alignment of KH domains encoded by these ® ve very closely related single maxi-KH
domain Drosophila genes. The marked similarity of the RNA binding motifs may imply that theproteins have similar roles.
quaking-Related KH Domains 59
-2, and -3. We hybridized these to embryo whole mounts prepared as described by
Tautz and Pfeifel (1989). We expected, on the basis of prior investigation of
mouse quaking, which is expressed speci® cally in the heart and nervous system,
and Drosophila who, expressed uniquely in muscles and their attachment points,
that a majority of these genes would be expressed in a tissue-speci® c manner.
However, on the basis of both whole mounts and RNA blotting experiments, the
Fig. 4. Physical maps of four Drosophila quaking -related genes. Theschematic diagrams illustrate the structures of the four quaking -relatedDrosophila genes that encode protein sequences summarized in Fig. 2.
Filled blocks represent translated exons, while open block representsuntranslated exons. Introns are represented by lines between blocks.
Table I. Codons Contained Within Particular Drosophila quaking -Related Gene Exonsa
Exon qkr58E-1 qkr58E-2 qkr58E-3 qkr54B
1 1±85 1±99 1±52 1±91
2 86±253 100±300 53±93 92±1333 254±288 301±333 94±215 134±284
4 289±348 334±372 216±245 285±3475 349±396 246±273 348±425
6 274±317
aThe contiguous codons, numbered according to Fig. 2, comprise each exon in the four quaking -re-
lated genes.
60 Fyrberg, Becker, Barthmaier, Mahaffey, and Fyrberg
four novel genes appear to be expressed fairly generally. Relative to our control
preparations, we noted signi® cant hybridization in each case, but the intensity is
distributed fairly uniformly throughout diverse tissues and organs, rather than
being sharply localized, as in the case of Drosophila who (data not shown). This
observation was corroborated by hybridization of the same probes to RNA blots
representing embryonic, larval, pupal, and adult developmental stages. All four
probes hybridized at more or less the same intensity to all RNA preparations, a
result expected for genes expressed at low but signi® cant levels in a variety of
tissues (data not shown).
Cladogram
We ® nally examined the evolutionary relatedness of the proteins displayed in Fig.
2 to Drosophila WHO, mouse QK and SAM68, and brine shrimp GRP33
(Cruz-Alvarez and Pellicer, 1987). This work con® rms that Drosophila WHO
(alias QKR93F) is similar to mouse QUAKING in spite of their distinct develop-
mental roles (Fig. 5). The four novel Drosophila proteins described herein, while
conforming to the tripartite structure characteristic of both QUAKING and
SAM68, are more similar to one another than to SAM68, GRP33, QUAKING, or
GLD-1. This observation suggests that the number and diversity of proteins
Fig. 5. Evolutionary relatedness of predicted
proteins. The cladogram illustrated the evolution-ary distances between mouse QUAKING(QK), Drosophila WHO (QKR93F), C. el-egans GLD-1, mouse SAM68, brine shrimpGRP33, and the four novel Drosophila se-
quences reported in this paper. Note that thenew Drosophila proteins are not especiallyclosely related to any of the above listed
proteins, although all have considerable simi-larity.
quaking-Related KH Domains 61
containing single maxi-KH domains may be considerably greater than has been
documented previously.
DISCUSSION
In this paper we report the discovery of nine Drosophila genes that contain single
maxi-KH domains isolated using structural and sequence information derived
from Drosophila WHO, mouse QUAKING, and their relatives. These novel
maxi-KH domain-containing proteins, or STAR proteins (Ebersole et al., 1996;
Vernet and Artzt, 1997), warrant the interest of biologists for at least four reasons.
First, all are presumed to be capable of binding RNA and, as such, probably
modulate some aspect of temporal and spatially programmed gene expression
(Musco et al., 1996; Vernet and Artzt, 1997). Second, failures of muscles, oocytes,
and neurons to differentiate have been correlated with gene mutations in Dro-
sophila who, C. elegans gld-1, and mouse quaking, respectively (Baehrecke,
1997; Zaffran et al., 1997; Jones and Schedl, 1995; Ebersole et al., 1996). Third,
proteins speci® ed by all of the above-named genes are closely related to SAM68,
a phosphoprote in target for SRC and FYN that is believed to couple signal
transduction with cell cycle-mediated gene expression, and in that manner to
regulate mitosis. Fourth, and ® nally, in at least some cases particular STAR
protein sequences are conserved over large evolutionary distances. For example,
mammalian splicing factor1 (SF1) appears to be the homologue of S. cerevisiae
branch point bridging protein (Bbp1p), an essential splicing factor in yeast
(Vernet and Artzt, 1997).
In this work we sought to assess the size and diversity of Drosophila proteins
having single maxi-KH domains. To achieve this we designed degenerate oligo-
nucleotides that recognize gene regions that encode unique portions of maxi-KH
domains. Analyses of the resultant PCR products, and additional genomic and
cDNA screens, yielded a total of 9 new genes, which, added to who, give a total of
at least 10 Drosophila genes encoding maxi-KH domains. We have no evidence
that our biochemical screen is exhaustive, and design of additional oligonucleo-
tides may yield yet more genes. All that we are able to state de® nitively at this
time is that maxi-KH domain proteins are not rare in Drosophila.
The similar tripartite arrangements and extensive sequence conservation
suggest that all of the maxi-KH domain proteins are performing analogous
functions, although precisely what that role is for any such protein has yet to be
established. In every case where we have determined the complete sequence of
these maxi-KH domain proteins, we have found that they consist of the RNA
binding and two ¯ anking motifs. As yet, we have found no extensive similarities
between these novel Drosophila proteins and SAM68 that lead us to believe that
any are signal transduction proteins. Evidence that the functional roles may be
diverse is, however, suggested by the distinct patterns of gene expression. For
62 Fyrberg, Becker, Barthmaier, Mahaffey, and Fyrberg
example, Drosophila WHO (alias QKR93F) is unique to muscles and the cells to
which they attach (Baehrecke, 1997; Zaffran et al., 1997; Fyrberg et al., 1997),
whereas most of the proteins described in this study appear to be expressed
ubiquitously, while mouse quaking is expressed only in the mouse heart and
nervous system (Ebersole et al., 1996). Resolution of this issue will come only
after more extensive studies of these proteins and the genes that specify them.
Perhaps the greatest progress in delineating the roles of maxi-KH domain
proteins has come through analyses of mutant alleles of quaking, gld-1, and who
(Ebersole et al., 1996; Jones and Schedl, 1995; Baehrecke, 1997; Zaffran et al.,
1997). It is our hope that the information provided in this paper, together with the
ability to utilize cell biological and genetic techniques in Drosophila, will
considerably improve opportunities for further de® ning the functions of these
proteins.
ACKNOWLEDGMENTS
This work was supported by grants from the NIH and MDA to E.A.F. and by a
grant from the NSF to J.W.M. We thank Drs. Corine Vernet, Karen Artzt, and Eric
Baehrecke for discussions and for communicating data prior to their publication.
Sequences described herein have been ® led in the Genbank Database under
accession numbers AF038581, AF038582, AF038583, and AF038584.
REFERENCES
Baehrecke, E. H. (1997). who encodes a KH RNA binding protein that functions in muscledevelopment. Development 124:1323.
Cruz-Alvarez, M., and Pellicer, A. (1987). Cloning of a full length complementar y DNA for anArtemia salina glycine-rich protein. J. Biol. Chem. 262:13377.
Devereux, J., Haeberli, P., and Smithies, O. (1984). A comprehensive set of sequence analysisprograms for the VAX. Nucleic Acids Res. 12:387.
Ebersole, T. A., Chen, Q., Justice, M. J., and Artzt, K. (1996). The quaking gene product necessary inembryogenesis and myelination combines features of RNA binding and signal transductionproteins. Nature Genet. 12:260.
Fyrberg, E. A., Kindle, K. L., and Davidson, N. (1980). The actin genes of Drosophila, a dispersedmultigene family. Cell 19:365.
Fyrberg, E. A., Mahaffey, J. W., Bond, B. J., and Davidson, N. (1983). Transcripts of the sixDrosophila actin genes accumulate in a stage- and tissue-speci® c manner. Cell 33:115.
Fyrberg, C., Becker, J., Barthmaier, P., Mahaffey, J., and Fyrberg, E. (1997). A Drosophila muscle-
speci® c gene related to the mouse quaking locus. Gene 197:315.Gall, J. G., and Pardue, M. L. (1971). Nucleic acid hybridization in cytological preparations. Methods
Enzymol. 21:470.Ito, K., Sato, K., and Endo, H. (1994). Cloning and characterization a single-stranded DNA binding
protein that speci® cally recognizes deoxycytidine stretch. Nucleic Acids. Res. 22:53.Jones, A. R., and Schedl, T. (1995). Mutations in gld-1, a female germ cell-speci® c tumor suppressor
gene in Caenorhabditis elegans, affect a conserved domain also found in Src-associated protein
SAM68. Genes Dev. 9:1491.Leffers, H., Dejgaard, K., and Celis, J. E. (1995). Characterization of two major cellular poly(rC)-
quaking-Related KH Domains 63
binding human proteins each containing three K-homologous (KH) domains. Eur. J. Biochem.230:447.
Liu, K., and Hanna, M. M. (1995). NusA contacts nascent RNA in Escherichia coli transcriptioncomplexes. J. Mol. Biol. 247:547.
Lo, P. C. H., and Frasch, M. (1997). A novel KH-domain protein mediates cell adhesion processes inDrosophila. Dev. Biol. 190:241.
Mahone, M., Saffman, E. E., and Lasko, P. F. (1995). Localized Bicaudal-C RNA encodes a proteincontaining a KH domain, the RNA binding motif of FMR1. Embo J. 14:2043.
Mohler, J., and Wieschaus, E. (1986). Dominant maternal effect mutations of Drosophila melanogas-ter causing the production of double abdomen embryos. Genetics 112:803.
Musco, G., Stier, G., Joseph, C., Castiglione Morelli, M. A., Nilges, M., Gibson, T., and Pastore, A.
(1996). Three-dimensional structure and stability of the KH domain: Molecular insights into thefragile X syndrome. Cell 85:237.
Poole, S., Kauvar, L. M., Drees, B., and Kornberg, T. (1985). The engrailed locus of Drosophila:Structural analysis of an embryonic transcript. Cell 40:37.
Regnier, P., Grunberg-Manago, M., and Portier, C. (1987). Nucleotide sequence of the pnp gene of E.coli encoding polynucleotide phosphorylase. J. Biol. Chem. 262:63.
Sanger, F., Nicklen, S., and Coulson, A. R. (1977). DNA sequencing with chain-terminati ng inhibitors.
Proc. Natl. Acad. Sci. USA 74:5463.Schmidt, C., Henkel, B., Poeschl, E., Zorbas, H., Puschke, W. E., Gloe, T. R., and Mueller, P. K.
(1992). Complete cDNA sequence of chicken vigilin, a novel protein with ampli® ed and
evolutionarily conserved domains. Eur. J. Biochem. 206:625.Schupbach, T., and Wieschaus, E. (1991). Female sterile mutations on the second chromosome of
Drosophila melanogaster. II. Mutations blocking oogenesis or altering egg morphology. Genetics129:1119.
Siebel, C. W., Admon, A., and Rio, D. C. (1995). Soma-speci® c expression and cloning of PS1, a
negative regulator of P-element pre-mRNA splicing. Genes Dev. 9:269.Siomi, H., Matunis, M. J., Michael, W. M., and Dreyfuss, G. (1993a). The pre-mRNA binding K
protein contains a novel evolutionarily conserved motif. Nucleic Acids Res. 21:1193.Siomi, H., Siomi, M. C., Nussbaum, R. L., and Dreyfuss, G. (1993b). The protein product of the fragile
X gene, FMR1, has characteristics of an RNA-binding protein. Cell 74:291.Tautz, D., and Pfei¯ e, C. (1989). A non-radioactive in situ hybridization method for the localization of
speci® c RNAs in Drosophila embryos reveals translational control of the segmentation gene
hunchback. Chromosoma 98:81.Taylor, S. J., and Shalloway, D. (1994). An RNA-binding protein associated with Src through its SH2
and SH3 domains in mitosis. Nature 368:867.Urlaub, H., Kruft, V., Bischof, O., Mueller, E. C., and Wittman-Liebold , B. (1995). Protein-rRNA
binding features and their structural and functional implications in ribosomes as determined by
cross-linking studies. EMBO J. 14:4578.Vernet, C., and Artzt, K. (1997). STAR, a gene family involved in signal transduction and activation of
RNA. Trends Genet. (in press).Wensink, P. C., Finnegan, D. G., Donelson, J. E., and Hogness, D. S. (1974). A system for mapping
DNA sequences in the chromosomes of Drosophila melanogaster. Cell 3:315.Zaffran, S., Astier, M., Gratecos, D., and Semeriva, M. (1997). The held out wings(how) Drosphila
gene encodes a putative RNA-binding protein involved in the control of muscular and cardiac
activity. Development 124:2087.
64 Fyrberg, Becker, Barthmaier, Mahaffey, and Fyrberg