Selection and mapping of mutations affecting cysteine synthesis in Aspergillus nidulans was carried out. A new locus, cysE, is described, the mutants of which are deficient in in vivo conversion of O-acetylserine to cysteine, a step mediated by cysteine synthase. Three loci (cysB, C and E) were thus found to control this step in vivo, apparently without affecting the enzyme activity in vitro. By scoring for propargylglycine sensitivity of cys mutants, chromosomal map positions were obtained for all five cysteine loci (A, B, C, D and E).

The Rec-1 strain of Caenorhabditis elegans increases recombination frequency three-fold. In this paper, we have investigated the effect of Rec-1 on the intragenic recombination phenomena of crossing-over and gene conversion. These events were increased two- to three-fold as was X-chromosome non-disjunction. All of the recovered recombinants were independent events, indicating that Rec-1 does not act pre-meiotically. The pattern of recombination in the Rec-1 strain resembles a meiotic pattern more than a radiation expansion. We conclude from this result that the Rec-1 enhancement of recombination is not the result of an increased number of DNA lesions randomly distributed along the chromosome. The increased recombination frequency of Rec-1 was not accompanied by any detrimental effects on growth, progeny number or spontaneous mutation rate. In this regard, the results may have implications for models which propose either selective advantage or disadvantage accompanying increased recombination.

The quaking (qk) locus on mouse chromosome 17 has been defined by a single viable quaking allele. Three new alleles of quaking were selected after ENU mutagenesis by their failure to complement the quaking phenotype. The qkk2 allele was induced on wild-type chromatin and the qkkt1 and qkkt4 alleles were induced on t-chromatin. Each is a recessive embryonic lethal mutation. They fail to complement each other and are not complemented by the deletion, TtOrl. Homozygotes and hemizygotes die at 8–9·5 days gestation, but not at a single precise time. Because the classical quaking mutation complements the lethality of these new alleles, but they fail to complement its quaking phenotype (myelination defect), we conclude that the quaking+ function is required for embryonic survival as well as for myelination.

In marsupials, X chromosome inactivation is paternal and incomplete. The tissue-specific pattern of inactivation of X-linked loci (G6PD, PGK, GLA) has been attributed to a piecemeal inactivation of different regions of the X. We here propose an alternative hypothesis, in which inactivation of the marsupial X is a chromosome-wide event, but is differentially regulated in different tissues. This hypothesis was suggested by the relationship between the positions and activity of genes on the kangaroo paternal X. In the absence of an HPRT polymorphism, we have used somatic cell hybridization to assess the activity of the paternal HPRT allele in lymphocytes and fibroblasts. The absence of the paternal X, and of the paternal forms of G6PD or PGK, from 33 cell hybrids made by fusing HPRT-deficient rodent cells with lymphocytes or fibroblasts of heterozygous females, suggests that the HPRT gene on the paternal X is inactive in both tissues and therefore not selectable. Since HPRT is located medially on the Xq near GLA, which shares the same characteristics of activity, we suggest that the locus-specific and tissue-specific patterns of activity result from a differential spread of inactivation from a single control locus, located near HPRT and GLA, outwards in both directions to G6PD and PGK. The nucleolus organizer region on the short arm does not seem to be part of the inactivated unit.

Three new t-haplotypes, tp 4, tp 12 and tp 14, were isolated from M. musculus male mice captured in Central and East Bohemia, Czechoslovakia, about 400 km from the zone of hybridization between M. musculus and M. domesticus species. Complementation tests have shown that all three new t-haplotypes belong to tw 73 group. When compared with 5 t-haplotypes from M. domesticus they displayed the same pattern of BamHI restriction fragments with H-2 class I genes, and they also shared the t-specific 5·2 kb TaqI fragment of the alpha globin pseudogene. However, they differed from M. domesticus t-haplotypes at the D17Leh443 locus since they all displayed a 10·5 kb MspI fragment, labelled by the Tu443 probe, not found in wild type-chromosomes or in M. domesticus t-haplotypes. A hypothesis is proposed that t-haplotypes in M. domesticus originated by a single successful introgression from a parental species during speciation.

It has been proposed that isoenzymes functioning within cell organelles (chloroplasts, mitochondria) are genetically less variable than their cytoplasmic counterparts, as a result either of constraints imposed by the need to cross organelle membranes or from the different and specialized nature of organelle metabolism. However, some recent findings concerning chloroplast and cytoplasmic isozyme variability are not consistent with this thesis. We have analyzed a number of surveys of electrophoretically detectable enzyme variation in vertebrates, and show that for each of the four tested enzymes (malate dehydrogenase, isocitrate dehydrogenase, malic enzyme, and aspartate aminotransferase), the mitochondrial isozymes are less variable than their corresponding cytosplasmic forms. The mean heterozygosities across the four enzymes are 0·083 and 0·038 for the cytoplasmic and mitochondrial forms respectively. We conclude that mitochondrial isozymes are indeed subject to greater constraints than cytoplasmic forms and have fewer sites able to accept neutral or slightly deleterious mutations. It is also noted that of the enzymes analyzed, that with the smallest subunit molecular weight (MDH) has the least variable cytoplasmic and mitochondrial isozymes, whereas the enzyme with the largest subunits (ME) has the most variable isozymes.

The shapes of T1 and T2 vertebrae from unselected Q strain mice and from strains selected for large and small body size were studied by Fourier analysis in order to ascertain whether shape change was produced by size selection. The vertebrae of large, small and control strains were easily distinguishable, but between replicate groups shape differences were less marked. The main component of shape change was size related, but mice unselected for size also showed a non-size-related shape change.

By application of the neutral model of phenotypic evolution, quantitative estimates of the rate of input of genetic variance by polygenic mutation can be extracted from divergence experiments as well as from the response of an inbred base population to selection. The analytical methods are illustrated through a survey of data on a diversity of organisms including Drosophila, Tribolium, mice, and several crop species. The mutational rate of introduction of genetic variance (Vm) scaled by the environmental variance (VE) is shown to vary between populations, species, and characters with a range of approximately 10−4 to 5 × 10−2. Vm/VE for Drosophila viability is somewhat below this range, while hybrid dysgenesis may temporarily inflate Vm/VE beyond 10−1. Potential sources of bias and error in the estimation of Vm are discussed, as are the practical implications of the observed limits to Vm/VE for projecting the long-term response to selection and for testing adaptational hypotheses.

We have analysed data on weight and tarsus length collected during a long-term study of natural populations of Great Tits to evaluate the relative importance of genetic variation in body size. Some of our data were collected over a 25-year period, and therefore include a relatively large sample of naturally occurring environmental conditions. An overall heritability estimate calculated from the uncorrected mean weights of breeding birds amounts to 0·5. This estimate is unlikely to be influenced by resemblance in environmental conditions between relatives. Heritability estimates based on the size of fledglings vary between zero and the value for adults, depending on the environmental conditions during growth. If the feeding conditions for the nestlings are poor, no resemblance between parents and offspring is observed. Selection against small nestlings acts strongly on the environmental variance. This is concluded from the higher heritability estimates in the same cohorts after survival for at least three months after fledging, compared to measurements on nestlings. Such selection acting differentially on the genetic and environmental components of the phenotypic variance has important consequences for our ability to make predictions of phenotypic change from measured natural selection. Nevertheless, the amount of genetic variation would allow rapid response should selection on adult size occur.

Genet. Res., Camb. (1988) 51. Facing page 1.

The photograph of Douglas S. Falconer should have been dated December 1987 and credited to

ANTONIA REEVE PHOTOGRAPHY