By T. Yamamoto
In: "MEDAKA(killifish) : Biology and Strains"
Yamamoto, T. (ed.) , Keigaku Pub. Co., Tokyo, 1975, pp. 243-250.

In the medaka, where XX is normally female and XY is male it is possible to obtain estrogen-induced fertile XY females at will (Yamamoto, 1953, 1955, 1958, 1959), thus providing sufficient materials to approach the problem of viability of YY zygotes.

The original incentive of experiments described in the preceding chapter was to explain the rarity of viable YRYR zygotes which could be obtained by mating estrogen-induced XrYR females with normal XrYR males. The rarity of surviving YRYR zygotes might suggest "weakness" of YY zygotes with essentially similar genetic constitution that is a selection effect. As a matter of fact, however, in spite of their rarity, the only two YRYR males ever to have been detected were full of vigor. Furthermore, nearly all YRYr and YrYr males have been found to be viable. Hence, the rarity of viable YRYR zygotes cannot be ascribed to the general weakness of YY as such. As an outcome of several lines of experiments, we arrived at the conclusion that a few surviving YRYR zygotes have a particular genetic make-up which is radically different from that of nonviable ones.

The bulk of evidence reported in the preceding chapter leaves little doubt that there is an inert section (-) in the regular YR chromosome and a corresponding "viability" section (+) in the ordinary Xr and XR chromosomes. By viability segment we mean a segment which contains major genes essential for viability. The inert section (-) is considered to be a segment in which major genes are deteriorated or absent. The regular YR chromosome may be denoted by YR,- and ordinary Xr and XR chromosomes by Xr,+ and XR,+. The majority of YRYR zygotes are interpreted as nonviable because of their constitution YR,-YR,-, i. e. duplex in the (-) segment. A few viable YRYR zygotes are considered as crossovers with the constitution YR,-YcR,+, where YcR,+ stands for a recombinant YR whose (+) was derived from the Xr,+ by crossing over.

Various lines of evidence supporting the hypothesis were presented (Yamamoto, 1964a). The frequency of crossing over between the differential segment and the r locus has been estimated previously (Yamamoto 1961; see also 1964a). The present paper deals with the estimation of the crossover unit delimited by the r locus and the inert section (-).

In the argument and estimation of recombination fractions, occurrence of double crossing over does not enter into our consideration because the homologous section is very short, i. e. the r locus is only 0.2 ± 0.01 crossover unit from the differential segment in normal XrYR males, and 1.0 ± 0.4 unit in induced XrYR females.

The well-known three-point test cannot be applied to decide the linear order of the loci concerned, since the (-) section manifests itself no visible traits. Accordingly, we reasoned by reductio ad absurdum. The linear order of the segments and the r locus was decided to be (x)-r-(+) in the ordinary Xr,+ and (y)-R-(-) in the regular YR,- where (x) and (y) stand for the differential segments in the X and Y. The supposition of the other two linear orders leads to absurdity. Thus, if the (x)-(+)-r and (y)-(-)-R were the case, there would be no viable YRYR zygotes. And, if the order (+)-(x)-r and (-)-(y)-R were assumed, no curious genetic behavior of crossover XrXcR females would be expected, with the consequence that XcRYR zygotes would be viable, which they were not (see. Yamamoto, 1964a). Having obtained the linear order of the loci, our next step is to estimate crossover units between the interval of the r locus and the (-) segment.

Although the viable YRYR male was detected without difficulty in the experiments reported previously (Yamamoto, 1955, 1959), the actual proportion of XrYR to YRYR genotypes among offspring from estrogen-induced XrYR females by normal XrYR males deviated seriously from the expectation for a 3 : 1 ratio. Of 57 F1 orange-red sons (XrYR and YRYR) singly tested by white females XrXr, only two proved to be YRYR, that is, only one in 28.5 sons instead of the theoretical one in three. This ratio is used to estimate the value under consideration. One of fascinating results obtained previously (Yamamoto, 1961) is that recombination frequency between Xr and YR is quite different in males than it is in induced females. The recombination value between the differential segment and the r locus in estrone-induced XrYR females is five times higher than that in normal XrYR males. From this fact it is assumed that the recombination between the r locus and the (-) segment in induced XrYR females also might be five times as frequent as that in normal XrYR males. This assumption may be taken for granted since both intervals are within a short homologous region.

Let pl be the recombination fraction between the differential segment and the r-locus, and let p2 be that between the r-locus and the inert section (-) in the normal XrYR male. Let p'l be the recombination fraction between (y) and r-locus, and let p'2 be that between r-locus and (-) in the estrogen-induced XrYR female (see. Table 20-1). The gametic output is shown in Table 20-2, in which the subscript c is used to indicate crossover sex chromosomes. Neither XcR,+ nor Ycr,- is included in this table, because they could be produced only by double crossing over, which is practically nil.

The sum of all fractions relevant to viable YRYR males is taken as the numerator and that of all fractions of viable genotypes mentioned above as the denominator. Then, the numerator becomes (1/4) (l-p'l-p'2)p2 + (1/4) (1-pl-p2)p'2 + (1/4)p2p'2 and the denominator (2/4) (1-p'l-p'2) (1-pl-p2) + (2/4) (1-p'l-p'2)p2 + (2/4) (1-pl-p2)p'2 + (1/4) (1-p'l-p'2)pl + (1/4)(1-pl-p2)p'l + (3/4)p2p'2 + (2/4)p'2pl + (2/4)p2p'l + (2/4)plp'l. This reduces to and is equal to the observed value 2/57. Since p'l = 5pl (Yamamoto, l961), the equation becomes Of the three fractions, p1 is known to be 0.002. Substituting this value and assuming that p'2 = 5p2 as observed p'l = 5pl, the equation reduces to the quadratic

In normal XrYR males, the established linkage relationship in the regular YR is (y) 0.2R1.2(-) and in estrogen-induced XrYR females it is (y) 1.0R6.0(-), where numerals represent crossover units (percent). The linkage map is illustrated in Figure 20-1. In this map, the ends of both differential and (+), (-) segments are represented as saw-toothed edges because their units are as yet unknown.

Second, the problem is raised as to the effect of one versus two doses of the (+) segment. Although the (-) segment in the Y of a normal male is located only 1.4(0.2 + 1.2) crossover units from the differential segment, a few X+Yc+ males might be produced by crossing over. If two doses of the (+) had twice the effect of one dose, the X+Yc+ male might have been more vigorous than the X+Y- male. Had natural selection affected the X+Yc+ and X+Y- genotypes differentially in favor of the former type, the former would be expected to predominate among present-day descendants. Actually, however, the majority of normal XrYR males when tested either by estrogen-induced XrYR females (presumably Xr,+YR,-) or by crossover females XrXcR (Xr,+XcR,-) proved to be the genotype Xr,+YR,-. Our findings seem to indicate that a single (+) dose is sufficient for development and viability as well as vigor, as far as the gene products of the (+) section is concerned. Basic similarity of development and vitality or ordinary X+X+, crossover X+Xc- females, and regular X+Y- males, also, supports the concept.

Third, medaka-mouse-man comparisons reveal a basic similarity in respect to the male determining property of the Y chromosome. Appearance of a fertile XXY male (Yamamoto, 1963) can be taken as an evidence that the Y carries a superior male-gene or a group of male-genes epistatic in effect. (Distribution of autosomal sex-genes or modifiers in this fish was discussed in the chapter 15.) Russell, Russell and Gower (1959), Welshons and Russell (1959), Russell (1961) and Russell and Chu (1961) in the mouse, and Jacobs and Strong (1959) and Ford, Polani, Almeida, and Briggs (1959) in man, discovered that sex determination in mammals depends primarily on the Y chromosome. It is conceded, that the Y chromosome of the medaka is much more primitive or generalized in view of evolution of this chromosome. Besides the differential segment, it has a homologous section and the (-) segment. In the human, the famous case of a completely Y-linked inheritance, the "porcupine man" has been shown to be false (Stern, 1960). The presence of partially Y-Linked genes in man, also, has become dubious on a statistical basis (Morton, 1957). In the mouse and man the supposed inert section, that might have been present in their early evolutional stage, and perhaps also most of homologous section of the Y chromosome must have been lost during the course of its retrogressive evolution, leaving mainly the male determining differential segment. It is hardly necessary to add that mammals, including man, must have evolved from a fish-like organism which had the XX-XY type of sex-chromosomes, in which the Y had the male determining property.