Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (205K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (18)
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Myers, M. P.
Right arrow Articles by Young, M. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Myers, M. P.
Right arrow Articles by Young, M. W.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© 1997 Oxford University Press 4710-4714

Comparison of chromosomal DNA composing timeless in Drosophila melanogaster and D.virilis suggests a new conserved structure for the TIMELESS protein

Comparison of chromosomal DNA composing timeless in Drosophila melanogaster and D.virilis suggests a new conserved structure for the TIMELESS protein Michael P. Myers+, Adrian Rothenfluh, Michael Chang and Michael W. Young*

National Science Foundation Science and Technology Center for Biological Timing and Laboratory of Genetics, The Rockefeller University, 1230 York Avenue, New York, NY 10021, USA

Received September 8, 1997; Revised and Accepted October 9, 1997

ABSTRACT

Two proteins, TIM and PER, physically interact to control circadian cycles of tim and per transcription in Drosophila melanogaster. In the present study the structure of TIM protein expressed by D.virilis was determined by isolation and sequence analysis of genomic DNA (gDNA) corresponding to the D.virilistim locus (vtim). Comparison of vtim and mtim gDNA revealed high conservation of the TIM protein. This contrasts with poor sequence conservation previously observed for the TIM partner protein PER in these species. Inspection of the vtim sequence suggests an alternative structure for most TIM proteins. Sequences forming an intron in a previously characterized D.melanogastertim cDNA appear to be most often translated to produce a longer TIM protein in both species. The N-terminal sequence of vTIM and sequence analysis of genomic DNA from several strains of D.melanogaster suggest that only one of two possible translation initiation sites found in tim mRNA is sufficient to generate circadian rhythms in D.melanogaster. TIM translation may be affected by multiple AUG codons that appear to have been conserved in sequences composing the 5'-untranslated tim mRNA leader.

INTRODUCTION

Drosophila circadian behavioral rhythms are regulated by the interaction of two proteins PERIOD (PER) and TIMELESS (TIM) (1 -5 ). The proteins promote molecular cycles of per and tim transcription. Behavioral and molecular oscillations require the activities of both proteins, which are believed to function in transcriptional control upon translocation to the nucleus (6 -7 ). Heterodimerization of the proteins regulates timing of nuclear localization of PER and TIM and involves suppression of protein domains (CLDs) that confer cytoplasmic localization of monomeric forms of PER and TIM (4 ). Drosophila circadian rhythms are entrained to environmental cycles through a pathway mediated by TIM, which is degraded when flies are exposed to daylight (5 -9 ). PER is related to a family of proteins that possess a protein interaction motif termed PAS. This domain functions in homotypic and heterotypic protein associations (reviewed in 6 -7 ). PAS has been identified in proteins regulating circadian rhythmicity in Drosophila, Neurospora and the mouse (10 -12 ) and some PAS family proteins are essential for blue light phototransduction in Neurospora, algae and certain bacteria (reviewed in 13 ). These observations suggest a common evolutionary origin for a group of proteins underlying photoreception and circadian regulation.

The PAS domain of PER contains binding sites that direct association with TIM (1 ,4 ). PAS is also closely associated with sequences forming the PER CLD (4 ). However, sequences composing the TIM binding site for PER are unrelated to PAS and there is no sequence similarity revealed by comparing the PER and TIM CLDs (14 ). Since no functional properties for TIM could be inferred from the remaining sequence, we sought to isolate the tim homolog from another fly species, Drosophila virilis. A prior comparison of vPER and mPER revealed high conservation of PAS and CLD, but poor conservation of much of the remaining sequence (15 ). Thus analysis of a tim homolog might similarly reveal sequence elements critical for TIM function and assist future attempts to isolate tim-related genes from other organisms.

We report here the cloning and genomic DNA sequence of D.virilis tim (vtim), as well as the genomic DNA sequence of D.melanogaster tim (mtim). We find that, in contrast to PER, TIM is highly conserved in these two fly species. The strong homology includes regions of TIM required for nuclear localization and heterodimerization with PER. In contrast, the region of vTIM composing the mTIM CLD shows greater divergence and may identify a subset of sequences that promotes cytoplasmic localization of monomeric TIM proteins. Our analysis of the vtim sequence also indicates an alternative structure for the mtim transcription unit (14 ). First, an additional exon for mtim (and vtim) has been identified, coding for 32 amino acids. Secondly, of two possible sites for initiation of translation in D.melanogaster only one is available for use in vtim. These changes bring the D.melanogaster TIM protein to 1398 amino acids in length, while D.virilis TIM consists of 1343 amino acids.

RESULTS AND DISCUSSION

We used low stringency hybridization to clone vtim from a bacteriophage [lambda] library containing D.virilis genomic DNA (see the legend to Fig. 1 for probe and hybridization conditions). The insert (~15 kb) was subcloned from the [lambda] isolate as four separate SalI fragments. Two of the larger SalI fragments were further subcloned to accelerate the pace of sequencing. The intron/exon junctions for vtim were deduced by comparison with the mtim cDNA and genomic (gDNA) sequences. The assignments were assisted by computer generated alignments (BestFit program of the Wisconsin Package, Genetics Computer Group Inc., Madison, WI) between translated vtim gDNA and the mTIM protein sequence. With the possible exception of an intron in the mtim 5'-untranslated region (UTR) (no splicing sites were found in the predicted 5'-UTR of vtim), the architecture of the two genes is nearly identical (see discussion below and Fig. 2 B); both the number and sizes of the introns are well conserved. Like the majority of Drosophila introns (16 ), eight of 12 shared introns in mtim and vtim are <100 bp in length. The largest introns in mtim and vtim are ~3 kb in length and have not been completely sequenced. Partial sequence for this region in each gene is available upon request.


Figure 1. Conceptual translation of TIM proteins from D.melanogaster (mTIM) and D.virilis (vTIM). Protein sequences for mTIM were determined by comparison of tim cDNA (14) and mtim genomic DNA and by RNase protection assays of D.melanogaster tim mRNA (see text). The vTIM sequence was predicted by comparing the vtim genomic DNA sequence with the D.melanogaster sequences above. All DNA sequences have been deposited in GenBank. vtim DNA was cloned from an EMBL3 genomic D.virilis library (provided by Ron K.Blackman, University of Illinois at Urbana-Champaign). Full-length D.melanogaster tim cDNA (pSKtim2; 14, M.Myers, unpublished results) was hybridized to the library using standard conditions (24) except for formamide concentration (reduced to 25%) and temperature (37°C). Final washes were in 0.5* SSC at room temperature. [brvbar], Identical amino acids in the vTIM sequence; *, gaps in the vTIM or mTIM sequence.


Figure 2. (A) The long, but not the short, form of mtim RNA is found in wild-type head extracts. RNase protection assays were performed as described (2). The probe (a) spans the 5'-boundary of the putative intron/exon [including nt 829-1019 of Myers et al. (14) plus 59 nt of the presumed intron]. Protected fragments produced in head RNA were compared with protected fragments generated from in vitro transcribed long RNA (b) (249 nt) and short RNA (c) (190 nt). The long, but not the short, RNA form was detected in the fly head extract. As indicated, the protected long RNA fragment from the head extract oscillated with a daily rhythm. (B) Map depicting the tim exon/intron structure for the D.melanogaster (and D. virilis) locus. The restriction map is for D.melanogaster. tim exon/intron structure of the two species is equivalent at the scale depicted except introns 1 and are absent in vtim (see text). Complete genomic DNA sequences were obtained for both species except for a portion of the largest tim intron, indicated by a dashed line in the restriction map (see text). GenBank provides both genomic sequences and exon/intron boundaries.


Figure 3. Sequence of the mtim 5'-leader sequence. Four AUGs are indicated, together with their respective predicted products. Initiation at AUG-3 leads to the longest predicted TIM sequence (14; bottom reading frame). Sequence derived from the tim loci carried by two genetically marked chromosomes, cn bw and Cy Roi respectively, show two point mutations in the indicated sequence. Both chromosomes support wild-type behavioral rhythms (text). For both chromosomes a single base is deleted (vertical arrow), leading to a predicted peptide of only 19 amino acids upon initiation at AUG-3 (boxed sequence). A second change, A -> C transversion, is found three bases 3' of the deleted base pair in cn bw and Cy Roi (not shown). A third, independent wild-type strain (Oregon R) shows complete agreement with the displayed sequence (from 14).

We noted that an apparent intron for the mtim locus (this segment was spliced out of the mtim cDNA previously sequenced by Myers et al.; 14 ) was highly conserved in the vtim sequence. When present, the 96 bp segment preserves the existing reading frame and codes for an additional 32 amino acids that differ at only one position between the two species. Moreover, DNA sequencing has revealed that a mutation in D.melanogaster tim which affects period length maps to this possible coding segment (A.Rothenfluh, unpublished data).

Since this sequence is flanked by canonical splicing signals and might be subject to regulated splicing, we used RNase protection analysis to determine the relative abundance of the spliced and unspliced forms in D.melanogaster. For the purposes of this discussion we refer to transcripts containing the added coding sequence as the `long' form and those without it as the `short' form. RNA was purified from heads of D.melanogaster that were entrained to a cycle of 12 h light/12 h dark. Drosophila heads were harvested at 4 h intervals. The probe was designed to permit detection of the two transcript forms, the spliced (short) form yielding a shorter fragment after cleavage of the probe by RNase (Fig. 2 , legend). Surprisingly, we found that only the long transcript was detected in fly head RNA (Fig. 2 A). There was no significant expression of the shorter (spliced) form at any time in the clock cycle. We conclude that the short form originally isolated from the cDNA library represents either a rare splice product or contamination of the head cDNA library by another tissue which expresses the short isoform. A schematic map depicting the predominant exon/intron structure of the D.melanogaster and D.virilis genomic loci is shown in Figure 2 B.

In vitro heterodimerization studies have demonstrated that both long and short versions of the mTIM protein will physically associate with PER (4 ; L.Saez and M.Young, unpublished observation) and that all sites promoting PER-TIM interaction are contained in the short form of TIM (L.Saez and M.Young, unpublished observation). No differences have been detected in the behavior of the long and short forms of TIM upon expression in cultured Drosophila cells: both forms of TIM accumulate in the cytoplasm when expressed alone, but translocate to nuclei when co-expressed with PER (4 ; L.Saez and M.Young, unpublished observation).

Assignment of the initiation codons

A previously sequenced mtim cDNA (14 ) contained four spaced AUGs within a 107 bp segment of the predicted mRNA 5'-leader. These potential translation start sites are depicted in Figure 3 and two, AUG-1 and AUG-2, would terminate after translation of only 12 and 9 residues respectively. The initiation codon for mtim was initially presumed to be the 5'-most AUG within the 1389 amino acid reading frame of the cDNA (AUG-3, Fig. 3 ). However, comparison of the mtim and vtim sequences suggest that a downstream (and in-frame) AUG is the more likely translation start site for mtim (AUG-4, Fig. 3 ). Although this would give a slightly shorter mTIM protein, identical N-terminal sequences (MDWLLATPQLL...; see Figs 1 and 3 ) would then be predicted for both mTIM and vTIM. A longer vTIM protein is not possible in this reading frame due to an ochre codon 15 nt upstream of the predicted translation start site in vtim gDNA. The strong homology of the predicted vTIM sequence to mTIM and the absence of any splicing signals that would fuse the lone vtim AUG to an upstream exon suggested that initiation at the sequence corresponding to mtim AUG-4 produces the true N-terminus for both mTIM and vTIM.

Other evidence for this new AUG assignment comes from our discovery of nucleotide polymorphisms in the 5'-leader region of tim in different D.melanogaster fly stocks. We discovered that tim gDNA derived from two second chromosomes (cn bw and Cy Roi), when compared with the sequence of our mtim cDNA, carried a single nucleotide deletion between AUG-3 and AUG-4 that places AUG-3 out of register with the tim open reading frame (ORF) (Fig. 3 , the G marked with a vertical arrow). cn bw and Cy Roi/Df(2 )tim02 flies (the latter carry tim only on the Cy Roi chromosome) both produce wild-type rhythms (A.Rothenfluh-Hilfiker, unpublished observation). Thus translation initiated at mtim AUG-4 is not only sufficient to generate functional TIM protein, but in some strains appears to be the only form of the protein that can be produced. Similar results were recently reported following an independent sequence analysis of the D.melanogaster sibling species D.simulans and D.yakuba and certain unrelated D.melanogaster stocks (17 ).

Although a vtim cDNA sequence is not available, AUG codons are found upstream of the predicted site of translation initiation in vtim gDNA, with spacing comparable with that seen for AUG-2 and AUG-3 of the mtim mRNA leader. AUG-1, -2 and -3 were also fully conserved in the predicted 5'-UTR of cn bw and Cy Roi tim mRNAs (Fig. 3 , legend). The occurrence of such upstream AUGs is uncommon. A recent survey of vertebrate mRNAs suggested that <10% have an AUG codon within their untranslated leaders (18 ). For several mRNAs ORFs within the untranslated leader have been shown to suppress translation from the downstream reading frame and it has been suggested that such upstream AUG codons may provide a cis-regulatory function (reviewed in 18 ,19 ). It may also be significant that upstream AUG codons have been recognized for frq, a Neurospora clock gene that is subject to translational regulation (20 -21 ). tim RNA accumulates several hours in advance of TIM protein (5 ,8 -9 ). Possibly alternative sites for initiation of translation within the mtim leader contribute to delays in TIM accumulation.

Amino acid sequence comparisons

An alignment of the deduced amino acid sequences for mtim and vtim using BestFit shows that the TIM protein is highly conserved between D.melanogaster and D.virilis; the overall identity is 77% without accounting for gaps (Fig. 1 ). If gaps are ignored, extensive portions of the sequences are nearly 100% identical. The overall level of conservation is significantly higher for TIM than for PER in these comparisons of the D.melanogaster and D.virilis coding sequences. The basis for these differences lies in the poor conservation of large tracts of sequence in PER, particularly sequences composing the C-terminal half of the protein (15 ; Fig. 4 ). TIM and PER regions of high conservation include the TIM-interacting region of PER (PAS and CLD, amino acids 233-512 of mPER) and the TIM PER binding regions (amino acids 514-587 and 724-923 of mTIM) (Fig. 4 ; 4 ). A functional nuclear localization sequence has been mapped near the N-terminus of mPER (amino acids 66-80) and between amino acids 550 and 562 in mTIM (Fig. 4 ; 4 ,22 -23 ). Both nuclear localization sequences are well conserved in D.virilis (Figs 1 and 4 ; 15 ). It was surprising to see overall differences in sequence divergence rates for TIM and PER, as the two proteins function together as partners in a heterodimeric complex. PER has also been found to physically associate with other D.melanogaster proteins in yeast (1 ), so differences in patterns of sequence divergence for PER and TIM might reflect protein activities in addition to those previously linked to the PER-TIM complex.


Figure 4. Similarity plots comparing deduced protein sequences for mTIM and vTIM (left) and mPER and vPER (right). The horizontal line in each plot indicates level of similarity calculated by comparison of complete protein sequences of the two species. Default local scoring matrix BLOSUM62 was used with a running average window size of 10 amino acids (PlotSimilarity program of the Wisconsin Package, Genetics Computer Group Inc., Madison, WI). Protein sequence for D.melanogaster PER was deduced from Canton S cDNA (22). The D.virilis PER sequence was described by Colot et al. (15). Acidic, region of acidic amino acid residues in mTIM which are not conserved in vTIM (see text); NLS, nuclear localization sequence; PB1 and PB2, TIM regions binding PAS and CLD domains of PER; CLD, cytoplasmic localization domains.

The similarity plot (Fig. 4 ) shows that the C-terminus of TIM, which includes sequences composing the TIM CLD, is not as well conserved; the strong similarity tapers off after residue 1219 of mTIM. The presence of several small gaps and non-conservative substitutions throughout the region suggests that the active CLD may be located within conserved regions spanning TIM amino acids 1283-1300 or amino acids 1345-1388 (Figs 1 and 4 ). These sequence comparisons should help to specify the TIM CLD.

The region in vTIM corresponding to amino acids 323-397 of mTIM (74 residues) has been replaced by an unrelated sequence of only 19 residues in vTIM. It was previously noted that for mTIM this region contains a very high concentration of the acidic residues Asp and Glu, a feature of some transcriptional activators (14 ). The lack of a similar domain within vTIM would seem to diminish the likelihood of such a role for the sequence in mTIM. There are no significant portions of vTIM that are absent in mTIM. Thus it should be instructive to determine whether such a truncated D.virilis TIM protein can provide full function as a component of the D.melanogaster clock.

NOTE

The genomic DNA sequences for both vtim and mtim have been deposited in GenBank (complete except for the large 3' intron in each gene). We have constructed the predicted, full-length mtim cDNA and will provide the clone upon request.

ACKNOWLEDGEMENT

This work was supported by a grant from the National Institutes of Health to M.W.Y. (GM54339).

REFERENCES

1 Gekakis, N., Saez,L., Delahaye-Brown,A.-M., Myers,M.P., Sehgal,A., Young,M.W. and Weitz,C.J. (1995) Science, 270, 811-815. MEDLINE Abstract

2 Sehgal,A., Rothenfluh-Hilfiker,A., Hunter-Ensor,M., Chen,Y., Myers,M.P. and Young,M.W. (1995) Science, 270, 808-810. MEDLINE Abstract

3 Lee,C., Parikh,V., Itsukaichi,T., Bae,K. and Edery,I. (1996) Science, 271, 1740-1744. MEDLINE Abstract

4 Saez,L. and Young,M.W. (1996) Neuron, 17, 911-920. MEDLINE Abstract

5 Zeng,H., Qian,Z., Myers,M.P. and Rosbash,M.(1996) Nature, 380, 129-135. MEDLINE Abstract

6 Rosbash,M., Allada,R., Dembinska,M., Guo,W.Q, Le,M., Marrus,S., Qian,Z., Rutila,J., Yaglom,J. and Zeng,H.(1996) Cold Spring Harbor Symp. Quant. Biol., 61, 265-278. MEDLINE Abstract

7 Young,M.W., Wager-Smith,K., Vosshall,L.B., Saez,L. and Myers,M.P. (1996) Cold Spring Harbor Symp. Quant. Biol., 61, 279-284. MEDLINE Abstract

8 Huntor-Ensor,M., Ousley,A. and Sehgal,A.(1996) Cell, 84, 677-685.

9 Myers,M.P., Wager-Smith,K., Rothenfluh-Hilfiker,A. and Young,M.W. (1996) Science, 271, 1736-1740. MEDLINE Abstract

10 Crews,S.T., Thomas,J.B. and Goodman,C.S. (1988) Cell, 52, 143-151. MEDLINE Abstract

11 Crosthwaite,S.K., Dunlap,J.C. and Loros,J.J. (1997) Science, 276, 763-770. MEDLINE Abstract

12 King,D.P., Zhao,Y., Sangoram,A.M., Wilsbacher,L.D., Tanaka,M., Antoch,M.P., Steeves,T.D.L., Vitaterna,M.H., Kornhauser,J.M., Lowrey,P.L., Turek,F.W. and Takahashi,J.S. (1997) Cell, 89, 641-653. MEDLINE Abstract

13 Kay,S. (1997) Science, 276, 753-755. MEDLINE Abstract

14 Myers,M.P., Wager-Smith,K., Wesley,C.S., Young,M.W. and Sehgal,A. (1995) Science, 270, 805-808. MEDLINE Abstract

15 Colot,H.V., Hall,J.C. and Rosbash,M. (1988) EMBO J., 7, 3929-3937. MEDLINE Abstract

16 Mount,S.M., Burks,C., Hertz,G., Stormo,G.D., White,O. and Fields,C. (1992) Nucleic Acids Res., 20, 4255-4262. MEDLINE Abstract

17 Rosato,E., Trevisan,A., Sandrelli,F., Zordan,M., Kyriacou,C.P. and Costa,R. (1997) Nucleic Acids Res., 25, 455-457. MEDLINE Abstract

18 Geballe,A,P. and Morris,D.R. (1994) Trends Biochem. Sci., 19, 159-164. MEDLINE Abstract

19 Geballe,A.P. (1996) In Hershey,J.W.B., Mathews,M.B. and Sonenberg,N. (eds), Translational Control. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 173-197.

20 Garceau,N.Y., Liu,Y., Loros,J.J. and Dunlap,J.C. (1997) Cell, 89, 469-476. MEDLINE Abstract

21 Liu,Y., Garceau,N.Y., Loros,J.J and Dunlap,J.C. (1997) Cell, 89, 477-486. MEDLINE Abstract

22 Baylies,M.K., Weiner,L., Vosshall,L.B., Saez,L. and Young,M.W. (1993) In Young,M.W. (ed.), Molecular Genetics of Biological Rhythms. Marcel Dekker, New York, NY, pp. 123-153.

23 Vosshall,L.B., Price,J.L., Sehgal,A., Saez,L. and Young,M.W. (1994) Science, 263, 1606-1609. MEDLINE Abstract

24 Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A Laboratory Manual, 2nd Edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

*To whom correspondence should be addressed. Tel: +1 212 327 8645; Fax: +1 212 327 8695; Email: young@rockvax.rockefeller.edu
+Present address: Preclinical Research and Development, Hoffman LaRoche, 340 Kingsland Street, Nutley, NJ 07110, USA
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 J Biol RhythmsHome page
J. Stehlik, R. Zavodska, K. Shimada, I. Sauman, and V. Kostal
Photoperiodic Induction of Diapause Requires Regulated Transcription of timeless in the Larval Brain of Chymomyza costata
J Biol Rhythms, April 1, 2008; 23(2): 129 - 139.
[Abstract] [PDF]

Home page
 GeneticsHome page
B. F. McAllister
Sequence Differentiation Associated With an Inversion on the Neo-X Chromosome of Drosophila americana
Genetics, November 1, 2003; 165(3): 1317 - 1328.
[Abstract] [Full Text] [PDF]

Home page
 Mol. Cell. Biol.Home page
A. Matsumoto, K. Tomioka, Y. Chiba, and T. Tanimura
timrit Lengthens Circadian Period in a Temperature-Dependent Manner through Suppression of PERIOD Protein Cycling and Nuclear Localization
Mol. Cell. Biol., June 1, 1999; 19(6): 4343 - 4354.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (205K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (18)
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Myers, M. P.
Right arrow Articles by Young, M. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Myers, M. P.
Right arrow Articles by Young, M. W.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?