The Clock Affecting 1 mutation of Neurospora is a recurrence of the frq7 mutation

Michael A. Collett1,2, Jennifer J. Loros1, and Jay C. Dunlap2. 1Department of Biochemistry and 2Department of Genetics, Dartmouth Medical School, Hanover, NH 03755.

The clock affecting-1 (cla-1) mutation of Neurospora crassa increases the period and decreases temperature compensation of the circadian rhythm, and was thought to define an uncloned gene with a possible role in the Neurospora clock. This defect, thought to be due to a translocation, was associated with a slow growth rate and a period of about 27 h at 25<C. We mapped cla-1 and found the growth rate and period defects to be due to linked independent mutations. The translocation was not the cause of the long period. The csp-1 mutation, present in the original cla-1 strain, had a period shortening effect, thus cla-1 strains lacking csp-1 had a period length similar to that of frequency7 (frq7). The cla-1 period defect mapped to the frq locus, and sequencing of frq revealed cla-1 to be a re-isolation of frq7.

______________________________________________________________________________

The cla-1 mutation was originally reported as an apparent spontaneous mutation arising from a cross between csp-1 A; bd and a; ufa (FGSC4441). The cla-1 strain had a period length of 28 h at 22<, reduced temperature compensation (i.e. as temperature increases the period length decreases, as opposed to the wild type strain where period length remains nearly constant over a range of temperatures) and was associated with a 30% slower growth rate (Brody et al. 1988 Genome 30, Suppl. 1: 299). This strain was found to possess a chromosomal rearrangement which was thought to be responsible for the cla-1 mutation (Lakin-Thomas et al. 1990 Crit. Rev. Microbiol. 17: 365-416; Dunlap 1993 Ann. Rev. Physiol. 55: 683-728). The rearrangement was reported to be a reciprocal translocation between IL and VIIR, with the breakpoint on IL reported to be near mat (3% recombination) and that on VIIR near the oli locus (2% recombination, Perkins 1997 Adv. Genet. 36: 239-398). The oli locus is 2 map units from frequency (Loros et al. 1986 Genetics 114: 1095-1110), raising the possibility that cla-1 was a new allele of frq. However, Southern analysis of a cla-1 containing strain revealed no translocation within the 8.7-kb DNA fragment containing the frq locus and flanking regions that is sufficient to rescue a null allele of frq (Johnson 1993 Ph.D. Thesis, Dartmouth College). The 28 h period length, slow growth rate and translocation outside of the known frq locus suggested that the cla-1 mutation defined a new clock affecting mutation in Neurospora.

To confirm the placement of cla-1 in relation to markers on IL, and obtain a strain of appropriate mating type for further genetic analysis, we crossed strain T(IL;VIIR) cla-1 csp-1 A; bd (strain 7035) with a clock wild type strain, a; bd (87-3, hereafter referred to as wild type) and scored the progeny for csp-1, period length and growth rate at 25<C (cross 189, see Table 1). The presence of the translocation on IL and VIIR in the cla-1 parent will cause a decrease in recombination on these linkage groups. This analysis agreed in part with previous unpublished data; i.e. the cla-1 phenotypes were linked to csp-1 and mat. However, we found the 30% reduction in growth rate associated with the cla-1 period defect was linked to, but separable from, the period length mutation (7 map units). Hence the period length, and growth rate defects were caused by independent mutations. We named the slow growth mutation slg. Additionally, progeny which were csp-1+ had a slightly longer period than those which were csp-1- as previously reported (Dharmanada and Feldman 1979 Pl. Physiol. 63: 1049-1059). Thus those progeny with a long period fell into 3 classes (Table 1): (1) Those which possessed the phenotype of the parental 7035 cla-1 strain (that is a period length +/- SEM of 27.2 +/- 0.1 h, a slow growth rate and the csp-1 spore separation defect); (2) one strain (189-33) which had a period length of 29.2 h +/- 0.3 h, a slow growth rate, and was csp-1+; and (3) two progeny (189-24 and 189-52) which had period lengths of 28.6 h +/- 0.2 h and 28.8 h +/- 0.4 h respectively, a wild type growth rate, and were csp-1+. As the growth rate defect and period length defect from the original cla-1 strain were separable, and in the absence of the csp-1 mutation the period associated with cla-1 was almost 29 h, which is close to what has previously been reported for strains carrying the frq7 mutation, we directly compared the period of strains carrying cla-1 and lacking csp-1 to a strain bearing the frq7 mutation. The period length of the long period csp-1+ progeny (i.e. cla-1; bd, slg+ or slg-) was quite similar to the period of a strain carrying the frq7 allele (strain 585-7 from the Feldman lab, UC Santa Cruz, CA) (period 29.6 h +/- 0.1 h).

We next mapped the cla-1 period length mutation in relation to VIIR markers by crossing a strain slg a; bd; cla-1 (189-33) with A; bd; oliR frq2 for (strain 922-131, cross 208, Table 1).

This mapping placed cla-1 between oli and for, and in this cross inseparable from frq. This suggested for the first time the possibility that cla-1 was an allele of frq. As the oliR mutation alters growth rate we were unable to accurately map slg in this cross.

To determine whether the cla-1 or slg mutations from the original cla-1 isolate were associated with the translocation we tested progeny from cross 189 that were recombinant for cla-1, slg and csp-1 for the presence of the translocation. This was done by crossing these progeny to tester strains fl; a or fl; A (FGSC 4347 and 4317 respectively). Unordered asci from these crosses were collected and analyzed (Table 2). This method allows one to determine the presence of a translocation in the strain being tested. A cross of translocation X normal leads to a fraction of the ascospores from the cross being unpigmented (Awhite@), in contrast to ascospores from structurally homozygous crosses, which are almost all black. Failure of pigmentation results from genetic deficiencies generated by meiotic chromosomal reassortment which results in lethality in these ascospores. Additionally, the frequency and pattern of white and black spores characterizes different kinds of translocations. For a reciprocal translocation one typically sees a 50:50 distribution of ascus classes centered around the 4:4 class, although exceptions to this are possible, in particular white ascospores are sometimes under represented. (For a detailed description see Perkins 1974 Genet. 77: 459-489.)

This analysis (Table 2) demonstrated that all strains with slg that were analyzed also possessed a chromosomal translocation; however, not all strains with the long period had the translocation, and all strains which possessed the growth rate of the wild type parent lacked a detectable translocation. The distribution of the ascus classes in the 189-33 strain was skewed from the 50:50 black:white ratio expected for the presence of a reciprocal translocation, so whether this represents the presence of a translocation is uncertain. However, 189-38, which has a wild type period length but possesses slg clearly has a translocation. Strains 189-24 and 189-52 which both have a wild-type growth rate and the cla-1 period defect lack a detected translocation.

Thus the mutation resulting in increased period length in the cla-1 strain was not due to a chromosomal translocation. Given these data it seemed likely that the chromosomal translocation was the cause of the slow growth rate; however, as our interest was focused on the clock phenotype we did not pursue this possibility further.

Since the clock-affecting mutation in cla-1 was separable from the translocation, mapped to frq on linkage group VII, and when separated from csp-1 the long period was similar to that known for the frq7 allele, we sequenced a region of frq spanning the frq7 mutation (corresponding to nucleotides 2643 to 3147 of the frq sequence in GenBank, accession number U17073) from strains 189-19, 189-24, 189-33, 7035 and wild-type. All strains with the long period (189-24, 33 and 7035) possessed the G to A point mutation that defines frq7 (Aronson et al., 1994 Proc. Natl. Acad. Sci. USA 91: 7683-7687) resulting in a Gly to Asp mutation at position 433 of the FRQ protein, while those with wild type period (189-19 and wild type) had the wild type sequence. This was the only sequence difference detected between long period and wild type strains. This sequence data, in addition to the period length similarity and genetic data thus demonstrate that the cla-1 period phenotype is caused by the frq7 mutation.

This is not the first instance of re-isolation of clock mutations. Mutations originally reported as frq2, frq4 and frq6, all of which were known to have the same period length (Feldman and Hoyle, 1976 Genetics 82: 9-17; Gardner and Feldman, 1980 Genetics 96: 877-886), were later shown by sequence analysis to be the result of identical G 6 A mutations resulting in an Ala 6 Thr mutation at position 869 in the FRQ protein; the frq7 and frq8 alleles were similarly shown to be the result of identical G 6 A mutations resulting in a Gly 6 Asp mutation at position 433 of FRQ (Aronson et al., 1994 Proc. Natl. Acad. Sci. USA 91: 7683-7687). Likewise, the per01, per02 and per03 mutations were shown to be due to identical mutations in the period gene of Drosophila (Hamblen-Coyle et al., 1989 J. Neurogenet 5: 229-256), and the Clock mutation of Drosophila was originally believed to define a novel genetic locus, but later mapped and shown to be an allele of per (Dushay et al., 1992 J. Neurogenet 8: 173-179).

In light of the molecular basis of this clock affecting allele, one possibility for the origin of cla-1 is that a frq7 conidiospore inadvertently found its way into a cross wherein a translocation occurred, resulting in the Acla-1@ strain. Alternatively, although the G 6 A mutation of frq7 is only one of a wide variety of different mutations of frq that result in a long period lengths, the formal possibility also exists that cla-1 represents a truly independent occurrence of the frq7 change.

Stuart Brody, UC, San Diego, provided us with strain 7035, T(IL; VIIR)SB332 slg A csp-1; bd; cla-1. We thank Keith Johnson for initial characterization of the cla-1 mutation at Dartmouth. This work was supported by grants from the National Institutes of Health (GM 34985 to J.C.D., MH44651 to J.C.D. and J.J.L.), the National Science Foundation (MCB-0084509 to J.J.L.) and the Norris Cotton Cancer Center core grant at Dartmouth Medical School.

 

Table 1. Mapping the cla-1 mutation.

 

 

 

 

Number of Progeny

     

Single Cross-Overs

   

Cross Number

Zygote Genotype and Percent Recombination

Parental

Region I

Region II

Region III

Double Cross-Overs

Regions I & II

Total

189

cla-1 slg A csp-1

I II III

cla-1+ slg+ a csp-1+

7 3.5 1.8

24

27

2

1

1

0

0

1

0

1

57

208

oliR frq2 for-

I II

oliS cla-1 for+

1.0 2.0

46

52

0

1

1

1

 

 

0

0

101

 

 

 

 

Table 2. The translocation associated with the long period length in the cla-1 strain is independent of the period length phenotype. The table shows the number of unorded ascus types containing various ratios of black and white ascospores arising from different strains crossed with either fl; A or fl; a. Strains without a translocation are expected to have almost exclusively asci which have black ascospores, i.e. almost all 8:0 asci, those strains which possess a reciprocal translocation are expected to have 50:50 black:white ascospores, with a symmetrical distribution of ascus classes around the 4:4 class.

 

Male Strain Number

Genotype

Ascus class (black : white)

 

 

 

 

 

8:0

6:2

4:4

2:6

0:8

189-24

slg+ a csp-1+; bd; cla-1-

32

4

2

0

0

189-33

slg- a csp-1+; bd; cla-1-

50

3

1

0

18

189-38

slg- A csp-1-; bd; cla-1+

24

1

22

2

28

189-52

slg+ a csp-1+; bd; cla-1-

25

0

0

0

0

fl; A x fl; a

14

1

0

0

0

 

 

Return to the FGN 48 Table of contents