Denatured total proteins of two maize lines and their reciprocal F1 hybrids were analysed by two-dimensional electrophoresis. Quantitative (spot more or less intense) and qualitative (presence or absence of spots) differences exist between the lines, and correspond to 11% of the total reproducible spots. Non-additive effects on spot intensities were found in the hybrids, which display spots similar to the more intense parental spot for 11% of varying spots. This may correspond to dominance for regulatory systems controlling the protein amounts. Such interactions contrast with additivity classically described for enzymes or DNA.

Mice were selected for growth from 3 to 9 weeks of age on a normal protein diet (N) containing 19·3% protein and a reduced protein diet (R) containing 5·1% protein. On each diet there were 3 high (H), 3 low (L) and 3 unselected control (C) lines. After 6 generations of selection, half of the mice in each line were tested on each diet. Responses were obtained when selecting for both increased and decreased growth on both diets. The realized heritabilities from within-family selection were 33 and 26% for the divergences on the normal and reduced protein diets, respectively. Consistent genotype-environment interactions were found when all lines were tested on both diets in generation 7. Performance on each protein level was best improved by selection on that protein level. Further, the correlated response was significantly less than the direct response when selecting on both diets. The estimates of the genetic correlation between growth on the two protein levels were low, rN = 0·16 from selection on the normal protein diet and rR = 0·51 from selection on the reduced protein diet. Selection resulted in a change in environmental sensitivity in the lines, dependent on the diet and direction of selection. The average of the divergences on the two diets was not dependent on the selection environment.

The protein spots representing the haptoglobin α1F, α1S and α2 chains in two-dimensional gels of human plasma samples representative of the six common haptoglobin phenotypes were identified by comparing their position with those of purified haptoglobin and distinguished from other spots in the vicinity by comparison with plasma with undetectably low levels of haptoglobin. Silver staining indicated that the α1F chain was represented by one spot in subtypes 1F-1F, 1F-1S and 1F-2, the α1S chain by three spots in subtypes 1S-1S, 1F-1S and 1S-2 and the α2 chain by six spots in 1F-2, 1S-2 and 2–2

Expression of X-linked genes for G6PD and GALA in interspecific hybrids between Microtus arvalis, M. subarvalis and M. kirgisorum voles was studied. Quantitative predominance of the enzyme activities of M. arvalis over G6PD activity of M. subarvalis and the GALA activity of M. kirgisorum in the female hybrids was shown. The definitive patterns of these enzyme activities was found on day 6·5 of embryonic development. Non-random inactivation of X-chromosomes derived from M. subarvalis and M. kirgisorum in the interspecific hybrids with M. arvalis is supposed to be the cause of the phenomenon observed. A hypothesis is proposed that there is a connection between the presence of large heterochromatin regions in the X-chromosomes derived from M. subarvalis or M. kirgisorum and the preferential inactivation of these in female hybrids with M. arvalis.

The transmission ratio distortion seen in males heterozygous for a mouse t-complex has been explained on the basis of trans-acting distorter genes, having a harmful effect on a responder gene. The t-complex form of the responder is relatively resistant to these harmful effects and hence is preferentially transmitted. Animals homozygous for the t-complex responder would be expected to show equal transmission of the two homologous chromosomes, but this is not always so. Studies described in this paper have shown differences among complete t's in their transmission when opposite a constant responder carrying partial t-haplotype. In addition, the proximal partial haplotypes th49 and tw18, both derived from tw5 but of different lengths, behave differently when opposite a responder. The three central partial haplotypes, tlowH, tlow2H and tlow3H, also differ, in that tlow3H shows lower transmission than tlowH or tloW2H when opposite either wild-type, or another responder, or distorter genes. These results can be explained either on the basis of differences in the responder region of various haplotypes, including the possibility of varying numbers of copies of the relevant sequences, or on the basis of differences in cis-acting (as opposed to trans-acting) distorter genes.

A method is developed for calculating the probability of establishment of an allele which is favoured in some places, but not others, in a large subdivided population. This method is quite general, and could be used to calculate the chance that any system which is linear near an absorbing boundary will move away from that boundary. The results are applied to a population distributed along one dimension. Only mutants which arise within a distance ∼ σ/ √2s of the region in which they are favoured stand an appreciable chance of establishment. The net chance of establishment of mutations distributed randomly across the habitat will be decreased by gene flow if selection against them is sufficiently strong. However, if the mutations are only weakly deleterious outside some limited region, gene flow may increase the net chance of establishment.

A model of the evolutionary accumulation of highly repeated DNA (HRDNA) is proposed. The accumulation of HRDNA sequences, which are organized largely in tandem arrays and whose functional significance is obscure, is explained here as a consequence of the action of the forces of amplification (promoting increase in copy numbers) and unequal crossing over, random drift and natural selection (controlling copy numbers). This model provides a general framework (i) to study the chromosomal location of satellite DNAs present in the genomes of all higher eukaryotes, and (ii) to explain the significant variation in the amounts of satellites which is frequently found among closely related species, but only rarely within a species. A review of the relevant data is included and open questions are identified.

A model of the balance between mutations and stabilizing selection affecting a quantitative character is developed and analysed. This model is essentially a discretized version of the continuum-of-alleles models analysed previously by Kimura, Lande, Turelli and others, and is formally similar to the stepwise mutation models used to interpret electrophoretic data. The complete model cannot be solved even for a haploid species, but there are useful approximations for most parameter values of interest. The ‘house-of-cards’ approximation can be used when selection is strong relative to mutation, and a normal approximation can be used when selection is relatively weak. For intermediate levels of selection a new ‘five-allele’ approximation provides accurate results over a wide range of parameter values. The house-of-cards and five-allele approximations applied to recessive alleles in a diploid population show that, for a given mutation rate, a somewhat larger genetic variance is maintained at equilibrium than in a comparable model of additive alleles. Under directional selection, the increase in genetic variance is largest for alleles of large effect and is much smaller for alleles of intermediate or small effect. At an equilibrium under stabilizing selection, homozygotes would tend to have a higher average fitness than heterozygotes when each mutation has a relatively large effect (the house-of-cards approximation), with the reverse if each mutation has a small effect (the normal approximation).

The covariance of an ancestor and the average of its descendants in generation t(CPt) is formulated for a breeding system which is a mixture of selfing and outcrossing. This covariance is partitioned into least squares additive (σA. At) and non-additive (σD. Dt) components summed over individual loci, and the combined inbreeding effects of all loci (H′), such that CPt = σA. At + (s/2)t[σD. Do + H′]. For the mth locus the covariance mσA. At = 2(1 + F)Σipiα(0)iα(t)i, in which pi is the frequency of the ith allele whose additive effect (α(t)) depends on the generation for which it is defined. For distant descendants α(∞) is equal to half of the derivative of the population mean with respect to the frequency of the allele. The covariance CP∞ = σA. A∞ thus relates directly to permanent selection response measured in the equilibrium population, any additional responses observed in earlier generations being due to temporary disturbances in population genotypic structure. It is only for these distant descendants that the least squares additive component has any direct interpretation in terms of selection response. The definitions of α(0) and α(∞), lead to two distinct definitions of the average effect of an allele substitution for a model with two alleles (Fisher, 1941), and to a clarification of their significance for this breeding system.

Mice are the most widely used experimental mammals, and many inbred strains are available. However, except for the relatively recent strains derived from known wild populations, the relationships between wild and laboratory mice are not well understood. Based on the Y-chromosomal restriction fragmentlength polymorphism, seventeen inbred strains were classified into two groups: strains with the Mus musculus musculus type Y chromosome and those with the M. m. domesticus type Y chromosome. We extended the survey to an additional twenty-two inbred strains. The M. m. musculus type Y chromosome was found in AEJ/GnLe, AAU/SsJ, BDP/J, BXSB/MpJ, DA/HuSn, HTG/GoSfSn, I/LnJ, LP/J, NZW/LacJ, RIIIS/J, SB/Le, SEA/GnJ, SF/CamEi, SK/CamEi, SM/J, WB/ReJ, WC/ReJ and YBR/Ei, while the M. m. domesticus type Y chromosome was present in BUB/BnJ, MA/MyJ, PL/J and ST/bJ.

Twelve isogenic X chromosome lines from a single natural population of Drosophila melanogaster were tested for their potential to induce gonadal dysgenesis and singed-weak mutability in P-M hybrid dysgenesis. The correlation between sterility and mutability was significantly positive for Cross A, confirming the results reported by Engels (1984) and Kocur, Drier & Simmons (1986). In Cross A* cytotype tests, however, two of the lines gave strikingly different results when measured by the gonadal dysgenesis test as compared to the singed-weak test. Positive correlations between traits within a given line were generally not observed. The results suggest that the relationship between gonadal dysgenesis production and the mobilization of P elements in singed-weak mutability is more complicated than that proposed by Engels (1984). The two phenomena may be separable under certain conditions. Neither test can be taken as an adequate characterization of the hybrid dysgenesis ‘profile’ of a line of flies.

Genet. Res., Camb. (1986), 48, pp. 175–178

A chromosomal segment conserved since divergence of lineages leading to man and mouse: the gene order of aminoacylase-1, transferrin and beta-galactosidase on mouse chromosome 9

In the summary on page 175 and in the last paragraph on page 176, the second aminoacylase locus, Acy-2, was inadvertently described as requiring MgC12. The correct compound is MnC12.