Negative feedback defining a circadian clock: Autoregulation of the clock gene frequency

Science

Volumn 263, Issue 5153, 1994, Pages 1578-1584

  Aronson, Benjamin D ^a   Johnson, Keith A ^b   Loros, Jennifer J ^c   Dunlap, Jay C ^c  

^a STONY BROOK UNIVERSITY   (United States)

^b RICE UNIVERSITY   (United States)

^c DARTMOUTH MEDICAL SCHOOL   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

MESSENGER RNA;

ARTICLE; AUTOREGULATION; CIRCADIAN RHYTHM; FUNGUS MUTANT; NEUROSPORA CRASSA; NONHUMAN; PRIORITY JOURNAL; TRANSCRIPTION REGULATION;

EID: 0028258994     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.8128244     Document Type: Article

References (66)

1. L. E. Edmunds Jr., Cellular and Molecular Bases of Biological Clocks (Springer-Verlag, New York, 1988).
2. J. C. Dunlap, Annu. Rev. Physiol. 55, 683 (1993); Trends Genet. 6, 159 (1990); P. Lakin-Thomas, G. Cote, S. Brody, CRC Crit. Rev. Microbiol. 17, 365 (1990).
3. J. C. Dunlap, Annu. Rev. Physiol. 55, 683 (1993); Trends Genet. 6, 159 (1990); P. Lakin-Thomas, G. Cote, S. Brody, CRC Crit. Rev. Microbiol. 17, 365 (1990).
4. J. C. Dunlap, Annu. Rev. Physiol. 55, 683 (1993); Trends Genet. 6, 159 (1990); P. Lakin-Thomas, G. Cote, S. Brody, CRC Crit. Rev. Microbiol. 17, 365 (1990).
5. M. C. Moore-Ede, F. M. Sulzman, C. A. Fuller, The Clocks That Time Us (Harvard Univ. Press, Cambridge, MA, 1982).
6. J. W. Hastings, Annu. Rev. Microbiol. 13, 297 (1959); S. Michel, M. E. Geusz, J. J. Zaritsky, G. D. Block, Science 259, 239 (1993).
7. J. W. Hastings, Annu. Rev. Microbiol. 13, 297 (1959); S. Michel, M. E. Geusz, J. J. Zaritsky, G. D. Block, Science 259, 239 (1993).
8. J. Takahashi, Curr. Opin. Genet. Dev. 3, 301 (1993).
9. J. F. Feldman and M. Hoyle, Genetics 75, 605 (1973).
10. C. R. McClung, B. A. Fox, J. C. Dunlap, Nature 339, 558 (1989).
11. M. Merrow and J. C. Dunlap, EMBO J., in press.
12. B. D. Aronson, K. A. Johnson, J. C. Dunlap, Proc. Natl. Acad. Sci. U.S.A., in press (1994).
13. At least two nonoverlapping divergent transcripts map to the frq locus. A smaller transcript (1.5 kb) has no extended open reading frame and no clock-relevant functions for this transcript have been uncovered (K. A. Johnson and J. C. Dunlap, unpublished observations). The central role of the larger transcripts (∼4.5 kb) in the circadian clock has been established (9). The transcriptional product of frq referred to in our work is the ∼4.5-kb clock-relevant transcript.
14. T. Pavlidis and W. Kauzman, Arch. Biochem. Biophys. 132, 338 (1969); B. C. Goodwin, Temporal Organization in Cells (Academic Press, London, 1963).
15. T. Pavlidis and W. Kauzman, Arch. Biochem. Biophys. 132, 338 (1969); B. C. Goodwin, Temporal Organization in Cells (Academic Press, London, 1963).
16. M. Zatz, Discoveries Neurosci. 8, 67 (1992); D. J. Chadwick and K. Ackrill, Eds., Circadian Clocks and Their Adjustment (CIBA Foundation Symposium 183, Wiley, Chichester, UK, in press).
17. M. Zatz, Discoveries Neurosci. 8, 67 (1992); D. J. Chadwick and K. Ackrill, Eds., Circadian Clocks and Their Adjustment (CIBA Foundation Symposium 183, Wiley, Chichester, UK, in press).
18. R. J. Konopka and S. Benzer, Proc. Natl. Acad. Sci. U.S.A. 68, 2112 (1971).
19. K. K. Siwicki, C. Eastman, G. Petersen, M. Rosbash, J. C. Hall, Neuron 1, 141 (1988); D. M. Zerr, J. C. Hall, M. Rosbash, K. K. Siwicki, J. Neurosci. 10, 2749 (1990).
20. K. K. Siwicki, C. Eastman, G. Petersen, M. Rosbash, J. C. Hall, Neuron 1, 141 (1988); D. M. Zerr, J. C. Hall, M. Rosbash, K. K. Siwicki, J. Neurosci. 10, 2749 (1990).
21. P. E. Hardin, J. C. Hall, M. Rosbash, Nature 343, 536 (1990).
22. The standard clock wild-type genetic background used in all experiments includes the band (bd) mutation that permits clear observation of the conidial banding phenotype. This mutation has no effect on the clock itself.
23. The concept of circadian time (abbreviated CT) has been formulated to allow circadian pacemakers with different endogenous periodicities to be compared. Biological time is thus normalized by dividing the circadian cycle into 24 equal parts, each a circadian hour. By convention, CT 0 is subjective dawn and CT 12 corresponds to the end of the light period in a 12:12 light:dark cycle or to the time of transfer from constant light to constant darkness. CT 6 is thus always mid-circadian day and CT 18 is midnight.
24. J. J. Loros, S. A. Denome, J. C. Dunlap, Science 243, 385 (1989).
25. J. J. Loros and J. F. Feldman, J. Biol. Rhythms 1, 187 (1986); J. J. Loros, A. Richman, J. F. Feldman, Genetics 114, 1095 (1986).
26. J. J. Loros and J. F. Feldman, J. Biol. Rhythms 1, 187 (1986); J. J. Loros, A. Richman, J. F. Feldman, Genetics 114, 1095 (1986).
27. I. Edery, J. E. Rutila, M. Rosbash, Science 263, 237 (1994).
28. N. H. Giles et al., Microbiol. Rev. 49, 338 (1985).
29. - (Strata-gene) and further subcloned into a vector carrying the entire frq open reading frame. This clone, designated pBA40, has the entire frq open reading frame under qa-2 promoter control. The pBA40 insert was further subcloned into vector pDE3 (23), compatible with targeted transformations at the his-3 locus, and designated pBA50. All regions of pBA40 and pBA50 that had been subjected to PCR were sequenced. A single base pair alteration (C to T at coordinate 2249) in the qa-2 promoter was detected; this coincided with the most distal and weakest of the four qa-1F binding sites. This alteration is predicted to have no effect on quinic acid induction because in the wild-type case this position is nonconsensus. The inducer, quinic acid, is a well characterized utilizable carbon source for wild-type N. crassa. In the presence of glucose, quinic acid can be considered nutritionally gratuitous and at the concentrations used here its only effects should be submaximal induction of the elements of the quinic acid metabolic pathway as well as the qa-2pFRQ. The qa-2 promoter can be induced up to 500 times by growth in the presence of quinic acid (21); however, because the promoter is catabolite-repressed, induction of only 25 to 50 times basal amounts are achieved when both glucose and quinic acid are present (21). All experiments were done in the presence of glucose unless otherwise noted.
30. 9(qa-2pFRQhis-3) respectively. Where appropriate, individual transformant strains are identified by an isolate number that follows the name.
31. 9(qa-2pFRQhis-3) respectively. Where appropriate, individual transformant strains are identified by an isolate number that follows the name.
32. 9 strains used in attempts to rescue rhythmicity through constitutive expression of a FRQ-encoding RNA were shown to have a properly targeted qa-2pFRQ construct, and of the three examined by Northern analysis, all produced high amounts of inducible qa-2pFRQ mRNA.
33. J. Chaleff, J. Gen. Microbiol. 81, 357 (1974).
34. H. Nakashima, Plant Cell Physiol. 22, 231 (1981).
35. J. C. Dunlap and J. F. Feldman, Proc. Natl. Acad. Sci. U.S.A. 85, 1096 (1988).
36. S. Dharmananda and J. F. Feldman, Plant Physiol. 63, 1049 (1979).
37. J. J. R. Tyson, in The Molecular Basis of Circadian Rhythms, J. W. Hastings and H.-G. Schweiger, Eds. (Abakon Verlagsgesellschaft, Berlin, 1976) pp. 85-108.
38. J. Loros and J. C. Dunlap, Mol. Cell. Biol. 11, 558 (1991).
39. D. Bell-Pedersen, J. C. Dunlap, J. J. Loros, Genes Dev. 6, 2382 (1992).
40. _, unpublished observations.
41. F.-R. Lauter and C. Yanofsky, Proc. Natl. Acad. Sci. U.S.A. 90, 8249 (1993); F. Lauter, V. Russo, C. Yanofsky, Genes Dev. 6, 2373 (1992).
42. F.-R. Lauter and C. Yanofsky, Proc. Natl. Acad. Sci. U.S.A. 90, 8249 (1993); F. Lauter, V. Russo, C. Yanofsky, Genes Dev. 6, 2373 (1992).
43. J. C. Dunlap, S. Crosthwaite, J. J. Loros, J. F. Feldman, in preparation.
44. H. Nakashima, J. Perlman, J. F. Feldman, Am. J. Physiol. 241, R31 (1981).
45. J. W. Jacklet, Science 198, 69 (1977); S. J. Yeung and A. Eskin, J. Biol. Rhythms 3, 225 (1988); S. B. S. Khalsa and G. D. Block, Proc. Natl. Acad. Sci. U.S.A. 89, 10862 (1992); J. C. Dunlap, W. R. Taylor, J. W. Hastings, J. Comp. Physiol. 138, 1 (1980).
46. J. W. Jacklet, Science 198, 69 (1977); S. J. Yeung and A. Eskin, J. Biol. Rhythms 3, 225 (1988); S. B. S. Khalsa and G. D. Block, Proc. Natl. Acad. Sci. U.S.A. 89, 10862 (1992); J. C. Dunlap, W. R. Taylor, J. W. Hastings, J. Comp. Physiol. 138, 1 (1980).
47. J. W. Jacklet, Science 198, 69 (1977); S. J. Yeung and A. Eskin, J. Biol. Rhythms 3, 225 (1988); S. B. S. Khalsa and G. D. Block, Proc. Natl. Acad. Sci. U.S.A. 89, 10862 (1992); J. C. Dunlap, W. R. Taylor, J. W. Hastings, J. Comp. Physiol. 138, 1 (1980).
48. J. W. Jacklet, Science 198, 69 (1977); S. J. Yeung and A. Eskin, J. Biol. Rhythms 3, 225 (1988); S. B. S. Khalsa and G. D. Block, Proc. Natl. Acad. Sci. U.S.A. 89, 10862 (1992); J. C. Dunlap, W. R. Taylor, J. W. Hastings, J. Comp. Physiol. 138, 1 (1980).
49. X. Liu et al., Neurone 6, 753 (1991); R. F. Smith and R. J. Konopka, Mol. Gen. Genet. 185, 30 (1982); M. K. Baylies, T. A. Bargiello, F. R. Jackson, M. W. Young, Nature 326, 390 (1987).
50. X. Liu et al., Neurone 6, 753 (1991); R. F. Smith and R. J. Konopka, Mol. Gen. Genet. 185, 30 (1982); M. K. Baylies, T. A. Bargiello, F. R. Jackson, M. W. Young, Nature 326, 390 (1987).
51. X. Liu et al., Neurone 6, 753 (1991); R. F. Smith and R. J. Konopka, Mol. Gen. Genet. 185, 30 (1982); M. K. Baylies, T. A. Bargiello, F. R. Jackson, M. W. Young, Nature 326, 390 (1987).
52. K. Lindgren, J. C. Dunlap, J. J. Loros, unpublished observations.
53. Although in these experiments the clock is reset by high amounts of FRQ-encoding transcript to the point predicted if frq and FRQ were the only state variables, these data do not support a contention that feedback between frq and FRQ exclusively comprises the clock. Rather we expect that there are additional state variables yet to be discovered, but this raises the issue of whether the inducer release-driven phase was predictable a priori, because the final phase of the clock after release would also depend on the values of the other state variables at the time the oscillation was reinitiated. Despite this caveat, the fact that the phase was predictable is noted, and we suggest the following interpretation for our data. The extremely strong clock resetting stimulus represented by the high concentration of FRQ may artificially drive not only frq and FRQ, but the other state variables as well, to unique phase points they might assume when frq is fixed at its low point in the normal cycle. The analogy can be drawn to resetting by light. A short pulse of light minutes in length will reset (advance or delay) the clock toward the daytime part of the cycle, no matter where in the cycle the clock is when light is perceived. However, the clock is driven not to any one point in the day, but rather to phases lying broadly within the day - a span of some hours - the final phase depending on where in the cycle the clock was before the pulse. A long pulse of light (over 12 hours), though, will appear to drive the clock to a unique phase point, so that upon transfer to darkness, the clock will begin from subjective dusk (a span of about 1 hour), regardless of where the clock was when the long pulse of light began. In the present case, the "pulse" of FRQ was more than 12 hours long so it may be that FRQ is driving all of the state variables rather than just frq, in a manner analogous to the action of light.
54. A. Millarand S. Kay, Plant Cell 3, 541 (1991); G. Arpaia, J. J. Loros, J. C. Dunlap, G. Morelli, G. Macino, Plant Physiol. 102, 1299 (1993).
55. A. Millarand S. Kay, Plant Cell 3, 541 (1991); G. Arpaia, J. J. Loros, J. C. Dunlap, G. Morelli, G. Macino, Plant Physiol. 102, 1299 (1993).
56. P. E. Hardin, J. C. Hall, M. Rosbash, Proc. Nat. Acad. Sci. U.S.A. 89, 11711 (1992).
57. J. Ewer, M. Rosbash, J. C. Hall, Nature 333, 82 (1988).
58. B. Frisch, P. E. Hardin, M. J. Hamblen-Coyle, M. Rosbash, J. Hall, Neuron, in press (1994).
59. I. Edery, L. Zweibel, M. E. Dembinska, M. Rosbash, Proc. Nat. Acad. Sci. U.S.A., in press.
60. 32P-labeled frq-specific RNA. The RNA probe was generated by in vitro transcription of a frq template corresponding to a 1.8-kb fragment from the frq open reading frame [coordinates 1389 to 3203, in which 1 corresponds to the A of the initiating ATG of the FRQ protein (9)]. Hybridization was done at 60°C under standard conditions and blots were washed at high stringency (68°C, 0.1 x SSPE, 0.1 percent SDS). Northern analysis with ribosomal RNA and sod-1 [P. Chary, R. A. Hallewell, D. O. Natvig, J. Biol. Chem. 265, 18961 (1990)] for normalization of RNA was done in the same manner, except that a DNA probe was used and hybridization was done at 42°C and washing at 50°C. For all experiments, multiple autoradiographic exposures were collected so that for any lane or band, several films could be identified in which the grain density was within the linear range of autoradiographic density. For comparisons between gels, autoradiographic exposures of identical duration were collected for blots hybridized with the same probe, and in different experiments, the normalization data with the sod-1 transcript and ribosomal RNA were always collected from the same gel used to quantify frq. The DNA fragment was labeled with random hexamers [A. P. Feinberg and B. Vogelstein, Anal. Biochem. 137, 266 (1983)].
61. 32P-labeled frq-specific RNA. The RNA probe was generated by in vitro transcription of a frq template corresponding to a 1.8-kb fragment from the frq open reading frame [coordinates 1389 to 3203, in which 1 corresponds to the A of the initiating ATG of the FRQ protein (9)]. Hybridization was done at 60°C under standard conditions and blots were washed at high stringency (68°C, 0.1 x SSPE, 0.1 percent SDS). Northern analysis with ribosomal RNA and sod-1 [P. Chary, R. A. Hallewell, D. O. Natvig, J. Biol. Chem. 265, 18961 (1990)] for normalization of RNA was done in the same manner, except that a DNA probe was used and hybridization was done at 42°C and washing at 50°C. For all experiments, multiple autoradiographic exposures were collected so that for any lane or band, several films could be identified in which the grain density was within the linear range of autoradiographic density. For comparisons between gels, autoradiographic exposures of identical duration were collected for blots hybridized with the same probe, and in different experiments, the normalization data with the sod-1 transcript and ribosomal RNA were always collected from the same gel used to quantify frq. The DNA fragment was labeled with random hexamers [A. P. Feinberg and B. Vogelstein, Anal. Biochem. 137, 266 (1983)].
62. B. D. Aronson, K. M. Lindgren, J. C. Dunlap, J. J. Loros, Mol. Gen. Genet., in press.
63. -2 M.
64. Mycelial cultures from appropriate strains were grown in 2.0% glucose (except where noted), 1 x Vogel's salts, plus various concentrations of inducer. After 15 hours of induction (where appropriate) and 12 hours in the dark (subjective morning, CT 1), the cultures were harvested, RNA was isolated and Northern (RNA) analysis was done as in Fig. 1. Because of catabolite repression or inducer exclusion, when cultures were grown in 2.0% glucose, induction of the ga-2pFRQ transcript was limited (21, 22). One of two different riboprobes was used for hybridizations. The riboprobe that detects endogenous frq mRNA, but not ga-2pFRQ mRNA, was derived from an Eco RI-BgI II fragment (coordinates -18 to -1371) that lies entirely within the 5′ untranslated region and that was therefore absent from the qa-2pFRQ construct. The riboprobe that detected both endogenous frq transcript and the ga-2pFRQ transcript was the same probe used in Fig. 1 (43).
65. + strain = 23.0].
66. Supported by NIH grant GM 34985 and the National Institute of Mental Health (J.C.D.), the Norris Cotton Cancer Center Core Grant, and NIH fellowship GM14465 (B.D.A.). We thank D. Natvig for sod-1 clones and DNA and G. Block and J. Hall for critical discussion of the data.