Identification of nonsense mutations in Neurospora: Application to the complex arg-6 locus

The arg-6 locus of Neurospora crassa encodes two enzymes of arginine synthesis, acetylglutamate kinase and acetylglutamyl phosphate reductase. Mutants lacking one or the other enzyme fall into two different complementation groups; a large non-complementing group lacks both enzymes. We wished to survey over 50 alleles for suppressibility by a nonsense suppressor. We compared two methods of assessing suppressibility. One, based on trans-action of a nonsense suppressor, was simple, but not efficient in detecting all suppressible alleles. The other, based on crosses involving a marker linked to the locus surveyed was very efficient in methodology and is suited to all cases, such as the arg-6 locus, in which allelic crosses are sterile. The data indicate that the arg-6 locus encodes a bifunctional protein.


Introduction
The arg-6 locus of Neurospora crassa appears to be a complex genetic region which controls two early steps in arginine biosynthesis.On the basis of complementation analysis, a large group of mutants recently isolated (Davis 1979) can be grouped into three categories.Two groups, A and B, fail to complement within themselves, but complement readily with each other.A third, non-complementing (NC) group does not complement within itself, nor with groups A or B. Enzymatic tests (Davis 1979;Wolf and Weiss 1980, and unpublished observations) reveal that the N C group lacks activity for acetylglutamate kinase (ATP: N-acetylglutamate 5-phosphotransferase, EC 2.7.2.8) and acetylglutamylphosphate reductase (N-acetyl-L-glutamyl-5-semialdehyde: N A D P + oxidoreductase [phosphorylating], EC 1.2.1.38)(Wolf and Weiss 1980).Group A mutants lack the kinase; group B mutants lack the reductase.The locus is very similar in its functions and complementation behavior to the homologous °'argBC'' region of Saccharomyces cerevisiae (Minet et al. 1979;Jacobs et al. 1980).The questions posed by observations on the two fungal systems are whether the two singly-deficient groups of mutants represent two genes transcribed into separate m R N A ' s ; whether the entire region is transcribed into one m R N A and, if so, whether one or two proteins finally emerge from the translation of this m R N A .
Offprint requests to: R.H. Davis The action of a genetic locus is most clearly demonstrated with null mutants of the nonsense or frameshift type, particularly if mutations reveal translational polarity with respect to component functions (Manney 1964;Radford 1969Radford , 1970;;Case and Giles 1968).The identification of nonsense mutations usually requires crosses of mutants with a strain carrying an informational suppressor gene.Present methods for such analysis are not well developed in Neurospora, in part because of sterility in crosses called for by certain direct rationales, and in part because of the lack of a vegetative diploid phase.Accordingly, in this paper, two other tests of nonsense suppressor action were made on an array of arg-6 mutants.One was based on a heterokaryon test for suppressor action, in which transdominance of the suppressor mutation is required.The other was based on a crossing regime in which a visible marker linked to arg-6 is used to facilitate assessment of the progeny genotypes of a sexual cross.The arg-6 nonsense mutations identified are used to test hypotheses of arg-6 gene structure and action.

Materials and Methods
Strains.A list of tester strains, with constituent loci and alleles, is given in Table 1.All are related by origin or backcrossing to the standard wild-type strain 74A of Neurospora crassa, and were constructed by appropriate crosses.The mutations in these testers are as follows.(a) Y319-44 (" ssu-1 ?"), isolated by Case and Giles (1974)is a nonsense suppressor on linkage group VIIR, similar in its linkage relations and action spectrum to the classic ssu-1 allele, W R N 33 (Scale 1968(Scale , 1972(Scale , 1976)).Y319-44 has not been  1970) were made by combining drops of conidial suspensions in 1 ml of minimal medium N of Vogel (1964) in 13 mm x 100 mm culture tubes with loose aluminum caps.
Growth was scored each day for 10 days.Genetic analysis involving crosses were done by standard methods (Davis and de Serres 1970), and rationales are given in the Results.
Intercrosses among arg-6 mutants were sterile, which precluded fine-structure analysis of the locus.Sterility also precluded a straightforward test of suppressor action; namely, mating a tester strain carrying a non-suppressible arg-6 allele (e.g., CD-63) and ssu-1 to all other arg-6 strains, and plating each cross on minimal medium.While this rationale works well in other systems (e.g.Kinsey and Hung 1981), it was impossible to apply here.

Complementation Analysis" of Mutants
All but two auxotrophic arg-6 mutants, as well as several that displayed extremely poor growth on minimal medium, could be classified into the A (4 mutants), B (28 mutants) and NC (12 mutants) groups by complementation tests.The two exceptions were CD-189, which complemented with all A mutants and a few B mutants, and CD-219, an A mutant which complemented with only some B mutants.
(Neither, as later tests showed, was suppressible.)There were 9 leaky mutants which were impossible to classify by complementation.

Crosses to Detect Nonsense Mutants
In this method, crosses are made to reveal whether arg-6, ssu-1 double mutants will grow without an arginine supplement.The efficiency of the technique was greatly enhanced by use of the al-2 (albino) marker, which is approximately 0.5 map units from arg-6.In practice, the arg-6 mutants to be tested were mated individually to either IC10 or IC11 (ssu-1, al-2).Ascospores were plated on arginine-containing plating medium, and 50 colonies were transferred the next day to 10 mmx 75 mm slants of minimal medium.Because of tight linkage of arg-6 and al-2, only four genotypic categories were expected (Table 2).In essence, all strains carry- The crosses detected twelve suppressible alleles.Eleven were group B, and one was non-complementing (Table 3).Several points regarding the crossing tests should be made.(a) "Leaky" alleles distinguished themselves from weakly suppressible alleles by the absence of complete auxotrophs among progeny in the minimal tubes.No leaky alleles were suppressible (i.e., less leaky with ssu-1).(b) Weakly suppressible alleles which failed to produce conidia could be transferred from the minimal tubes to arginine-supplemented media, where expectations about conidial color could be confirmed.(c)A slight orange tinge of ssu-1, al-2 double mutants indicated that the al-2 allele used was slightly suppressible.This did not confuse classificatin.(d)Trials in which ascospores were plated directly on minimal plating Argo This isolate is presumed to be misclassified for ssu-1 medium, followed by observations of colony color on plates were not sufficiently rigorous in detecting suppressible arg-6 alleles.(e) All arg-6 alleles tested were found to show linkage to al-2 in the crosses; at most, only one of the Ssu ÷ prototrophs was orange.Thus, the method confirmed the genetic location of the mutations in the arg-6 region.(f) The progeny bearing ssu-1 were selected against to some extent (Table 4); this would compromise tests for suppressible alleles on the basis of Arg ÷ :Arg-ratios alone.

Heterokaryon Tests for Nonsense Suppression
This method depends upon forming heterokaryons of arg-6, ssu-1 strains, IC12 or IC13, in which the arg-6 allele is non-complementing and non-suppressible, with an array of arg-6 single mutants to be tested.Growth of the heterokaryon indicates that the dominant ssu-1 mutation in the tester nucleus suppresses the arg-6 mutation in the other nucleus.The heterokaryon test identified five mutations as suppressible:   (Table 2).The first four were suppressible by the crossing test.The last mutant was sterile in the cross, and the heterokaryon test is the only evidence of its suppressibility.Three controls accompanied the heterokaryon tests.(a) The arg-6 mutants were tested alone to detect leakiness, a factor which must be excluded as a reason for growth in the test condition.In fact, some mutants were leaky and are listed as such in Table 2. (b) The arg-6 mutants were tested against the IC14 or IC15 strain to confirm non-complementarity with CD-63.(c)IC12 and IC13, the tester strains, were themselves tested against bona fide group A and group B mutants to prove that ssu-1 has not converted CD-63 to a complementing form.(This control is inherent in the survey of arg-6 strains.)The second and third controls showed, as expected, that CD-63 was non-complementing, both in the presence and in the absence of ssu-1.

Test for Suppression of Translational Polarity Effects
Nonsense mutations are polar in their effect on translation.In a bifunctional locus, a nonsense mutation in a transla-tional-proximal region may eliminate both proximal and distal functions.A nonsense suppressor may relieve the polarity effect without relieving the deficiency for the proximal function.If so, some non-complementers (doubly deficient) may be converted to complementers (singly deficient) by ssu-1, rather than to prototrophs.This phenomenon occurs in the argBC locus of yeast (Minet et al. 1979).
To test this possibility, all non-complementing (NC) alleles, alone or in combination with the ssu-1 mutation, were tested for complementation by forming heterokaryons with CD-29 (A), CD-34 (B), and CD-63 (NC).(None of these testers is suppressible.)No heterokaryon displayed growth, indicating that ssu-1 did not relieve polarity of the mutations sufficiently for growth to occur.The test is not rigorous, however, given the proven inefficiency of suppressor action in heterokaryons.

Discussion
The major goal of this work was to classify arg-6 mutants and, in the process, to compare two methods of identifying nonsense mutations.The heterokaryon test, formally equivalent to diploid formation in yeast (e.g., Minet et al. 1979), identified positively only five of the 13 suppressible alleles.This probably reflects the inefficiency of forming heterokaryons involving a suppressor gene, or of achieving a nuclear ratio allowing the dominance of the ssu-I allele to be expressed at the level of growth.In yeast, diploid formation assures a 1:1 ratio of ssu alleles and the dominance of ssu mutations is generally perceptible in growth phenotypes (e.g., Minet et al. /979;Chattoo et al. 1979).Despite its inefficiency, the Neurospora heterokaryon test is very simple and would be useful in quickly identifying at least one nonsense mutant in a large array of mutants.
In a previous study of the dominance of ssu mutations at three loci, Griffiths (1976) concluded that none of those tested (Y319-44, the one used here, Y319-37, and Y319-45) could exert its effect in trans.It was therefore suggested that these ssu mutations act only in cellular domains controlled by their own nuclei.The reason for the disparity in Griffiths' results and our own is not certain.It is possible that our test for suppression -even weak growth in a very restricted culture volume -is more likely to yield positive evidence than that of Griffiths, who asked for continued growth of heterokaryons, established on supplemented medium, upon transfer to minimal "race tubes".Moreover, the locus studied by Griffiths, ad-3B, is one whose suppressed mutants (i.e., ad-ssu-homokaryons) grow with a stop-start pattern on minimal medium (Griffiths 1975).This is suggestive of weak restoration of function at best, a problem which would be exacerbated by superimposing heterokaryosis for both the suppressor and the suppressible mutation.Nevertheless, in a qualitative sense, Griffiths' results agree with ours in showing that heterokaryon tests for super-suppressor action are inefficient and are certainly not definitive.
The crossing method, in which arg-6 ssu-1 recombinants were tested directly for growth, took advantage of the morphology of ssu-1 strains, and of the linked al-2 gene.We could select progeny first for the Ssu-phenotype, and absence of auxotrophs among them signified suppression.Moreover, the color marker (in this case, the A1 + phenotype) identified the progeny carrying the arg-6 mutation, This rationale can be adapted to testing alleles of any locus for which a nearby unsuppressible marker -visible or nutritional -is available, assuming an unlinked suppressor gene is being used.The major advantages of the rationale are that tetrads need not be isolated, and that non-ideal segregation ratios can be tolerated.In the most common method of surveying for suppressible alleles in Neurospora, reliance is placed upon departure from a 1 : 1 ratio of prototrophs to auxotrophs, an excess of prototrophs being presumptive evidence of positive suppressor action (Scale 1968).Poor germination of strains carrying arg-6-alleles, which was not a serious problem here, would have given false-positive evidence for suppression.Poor germination of ssu-1 alleles, as was seen here, reduced the impact of positive suppressor action on the overall prototroph:auxotroph ratio.In the actual crosses, the proportion of auxotrophs in crosses of suppressible mutants often exceeded 0.4, where only 0.25 was expected.
The "linked marker" rationale has not been widely used in Neurospora, despite Scale's description of such a use in his initial work (Seale 1968).This undoubletedly reflects the lower density of markers available in the past, and because in some cases (e.g.Kinsey and Hung 1981 ) it is possible to use a non-suppressible allele of the locus involved in place of the linked marker.Because a locus of interest is now likely to lie within 10 map units of a good visible or auxotrophic marker, we suggest the following general rationale, taking advantage of the selection one can impose against an auxotrophic marker.Mutants of locus m (to be tested) are mated to a single strain carrying an unlinked suppressor mutation, ssu-, and a diagnostic auxotrophic mutation, aux , linked to m (m-, aux + ; ssu + x m +, aux-; ssu ).The progeny are plated on minimal medium.Only suppressed mutants (m-, aux +, ssu-) which arise by independent assortment, and prototrophic crossover progeny (m +, aux +) will grow (Table 5).The latter category will be minimized by tight linkage of m and aux, and its magnitude will be a standard percentage seen in all crosses involving m-alleles.A control plating on a medium singly supplemented with the requirement ofm strains will aid in assessing the significance of colonies appearing on minimal medium.In practice, even a 10 map-unit distance between m and aux will yield more than a 5-fold difference in minimal plate counts of crosses involving suppressible and unsuppressible alleles (Table 5).This is a far better distinction than the 75% vs 50% difference in random spore platings of rnssu ÷ x m ÷ ssu crosses usually used to assess suppressibility.Moreover, the linked marker is useful in confirming the genetic location of each allele of the complementation group being surveyed for suppression.
The structure of the arg-6 locus can be inferred in part from the complementation and suppression data.First, noncomplementing mutants are found which lack both enzyme activities (Wolf and Weiss, in preparation).One is a nonsense mutation, and this implies that it lacks two enzyme activities, rather than one, owing to translational polarity.The group of four A mutants lacks the kinase; none is a suppressible mutation.The B group, lacking the reductase, has a high proportion (12/28) of nonsense mutants.The simplest interpretation of these data is that the arg-6 locus encodes a bifunctional protein, the reductase being translation-distal.Processing of this protein may then yield the separate enzymes found in isolated mitochondrion (Wolf and Weiss, unpublished observations), and as found in yeast (Minet et al. 1979).

Table 1 .
List of tester strains

Table 2 .
should grow into orange cultures if they are suppressible, and should not grow at all if they are not suppressible.Thus, any set of spores yielding orange cultures signifies that the arg-6 allele in question is suppressible.The test was strengthened by initial recognition of Ssu-(slow growing) strains (see Materials and Methods).When orange progeny were found, they always appeared in the Ssu class.

Table 4 .
Complete analysis of arg-6 crosses involving the nonsuppressible CD-63 and four suppressible alleles: arg-6, al+; ssu+x arg-6 +, al-1; ssu-1 in pooled data: 153:88.Classification of ssu alleles was done first on the basis of retarded growth.Color and auxotrophy were determined later b Not tested for arginine requirement; A1 + are presumed to be