Proceedings of the National Academy of Sciences, 42:299-304 (1956)
NEW EXPERIMENTS ON CHEMICAL PHENOCOPIES*
RICHARD B. GOLDSCHMIDT AND L. K. PITERNICK
UNIVERSITY OF CALIFORNIA, BERKELEY, CALIFORNIA
Communicated March 5, 1956
In his early work on heat-induced phenocopies Goldschmidt1 had shown that the relative frequency and the type of phenocopies produced were dependent upon the specific genetic background of the experimental material, Drosophila. Data were presented for the different responses to the same treatment of three wild-type and five mutant stocks. They showed that some reacted preferably with wing effects, some showed phenocopies otherwise not encountered, while others had a tendency to asymmetrical effect. Additional important findings were, in one case, the presence of a subthreshold (isoallelic) mutant of vestigial which was enhanced to high penetrance by the phenocopic treatment, and, further, two cases of production of a dominant effect of the recessive, which might also be described as enhancement of a subthreshold action of the heterozygous mutant locus. Although in later work by different authors the genetic element in the production of the phenocopic effect was noted and the enhancing effect upon subthreshold or low-penetrance genetic action was also encountered (e.g., Goldschmidt;2 Child, Blanc, and Plough;3 Plaine and Glass;4 Sang and McDonald5), no specific importance was attributed to the facts, which could be regarded as more or less expected in view of the known presence in all stocks of widely recombining genetic differences of the type usually described as modifiers.
Since Rapoport6 discovered the specific phenocopies produced by different chemicals, especially metal salts (later found for still other chemicals; see, e.g., Gloor,7 Bodenstein and Abdel-Malek,8 Hiaton et al.9), a more quantitative study of the phenocopic effect became possible, as is shown best in Sang and McDonald's5 study of the action of sodium metaborate producing specifically the phenocopy of the mutant eyeless and, to a lesser extent, those of the mutants aristopedia and antennaless. It was also noted in this work — though not elaborated — that different wild-type stocks had a different sensitivity, i.e., quantitative response, to the treatment and also reacted differently as to the quality of the effect.
While performing similar experiments intermittently over the last decade, we had been fascinated especially by this genetic element in the phenocopic effect. Some results appeared which raised serious questions whether the phenomenon of phenocopy might not contain an important, hitherto neglected causative agent of genetic nature which might even require a new definition of phenocopy. This suggested a series of experiments, in which the standard treatment of the Drosophila larvae with 0.06 per cent sodium tetraborate (Na2B4O7•10 H2O) was applied to different wild stocks, some mutants and hybrids, and was varied according to need, with the object of deciding whether the influence of the genetic background upon the changes in phenotype were not more specific than could be expected from the presence of general modifiers or else whether the genetic constitution of the experimental animals might be a decisive factor in, if not the actual cause of, the specific changes of phenotype after treatment.
Without going into the details of the data, already based upon very large numbers, the most important facts will be presented. The typical results obtained by all investigators (Rapoport, Sang and McDonald; see also Gersh10 and ourselves) are that Drosophila larvae receiving food containing the borate developed into flies with the phenotype of the mutant eyeless and that this occurs in up to 100 per cent of the individuals when the concentration of the salt approaches the lethal one; further, that at lower concentrations the effect decreases quantitatively (penetrance) and qualitatively (expressivity) and also, in regard to lethality, in proportion to the decrease of concentration; that the zone between 100 per cent effect with low numbers of survivors and complete lethality is a very narrow one; and that smaller numbers may also show the phenotype of aristopedia and antennaless.
The most important facts which we have to add concerning the genetic element, if not actually the genetic basis, of the phenomenon are the following:
1. There are strains with low sensitivity, strains with high sensitivity, and also strains which do not produce the eyeless effect at all, though other effects occur. Strains also occur which as a rule remain normal with the standard treatment, but here and there a single brood reacts very strongly in all respects mentioned. There are further strains in which about one-half the individual experiments with constant concentrations of borate show only a very low incidence and low expressivity of the eyeless phenotype (if any), while the other half show a high effect. There are reasons to suppose that heterozygosity for one mutant locus is involved, as direct extraction of the high and low strains was possible.
2. We could test a strain in which only the fourth chromosome (in which the copied mutant eyeless is located) is derived from Oregon stock (which has furnished the most sensitive lines), while chromosomes 1, 2, and 3 were derived from a different, more refractory, wild-type stock. This strain reacted to the treatment as intensively as the most sensitive Oregon line tested by us (see also paragraph 6 below).
3. The mutant ey2 is recessive. Our ey2 stock was 100 per cent penetrant, with high-grade expressivity. Untreated F1 (the controls) of the cross ey2 × different wild-type stocks sometimes show ey2 completely recessive, and sometimes a small percentage of lowest-grade eyeless flies appear. In the crosses with strains of high sensitivity, more dominance and also higher grades of expressivity appear in F1, and vice versa in strains of low sensitivity. But unfortunately the cross with the most sensitive strain of all gave no dominance of eyeless! Perhaps this has a special reason, namely, the great heterosis effect in this cross. We know from other experiments that heterotic F1 require a much higher borate concentration for the eyeless effect.
4. The mutant ey2 in our stock is as a rule not clearly affected by the borate treatment in regard to expressivity of the eyeless character. But the heterozygote ey2/ reacts strongly, much more than +/+ flies do (already studied quantitatively by Sang and McDonald). Again the heterozygote with the most sensitive wild line shows the highest percentage incidence of eyelessness, and, in addition, the expression of the character goes beyond that of the treated or untreated ey2 stock, producing many almost headless flies. (Though the control did not show dominance, see the end of paragraph 3.)
5. While homozygous ey2 is hardly enhanced by borate, the expression of dominant eyeless (eyD), a fourth-chromosome duplication with much lower expressivity (in our stock) than ey2, and of a different phenotype, is greatly enhanced by the treatment up to the absence of eyes. (This enhancement parallels that found by Plaine and Glass4) for erupt after L-tryptophan treatment, apart from the details involving indirect action via a suppressor.)
6. Two different derivatives of the Samarkand stock never gave the typical eyeless effect and are, in addition, rather refractory to borate treatment in all tolerated concentrations; but if they reacted, a specific effect resembling that of the mutant kidney was seen (as well as a Scutenick type; see below). If the fourth chromosome of a Samarkand line is combined with the others from a variably reactive stock (Canton) — with typical eyeless effect — a much stronger effect upon the eye appears after borate treatment with a phenotype very different from that of the eyeless effect, namely, partly the kidney type and partly a type resembling the mutant small eye, i.e., small but normally shaped eyes. Heterozygotes also show the specific effect. However, in an experiment with a stock containing chromosomes 1, 2, and 3 from a Samarkand line and chromosome 4 from another stock (Iowa), typical eyeless phenocopies were produced. The absence of the typical eyeless effect seems therefore to be based upon something in the fourth chromosome of Samarkand.
7. Another type of eye change after borate treatment was confined thus far to the stock spineless (ss). The eyeless effect has never been produced in this stock, but, instead, a condition of the eye closely resembling the mutant Lobe, especially the one studied by Zimm11 in which the surface of the eye increases and becomes crinkled and forms excrescences, while, simultaneously, the antenna on the same side becomes duplicated.
8. Another well-known borate effect is the production of the phenotype of the lower grades of the mutant aristopedia (more or less of a tarsus, instead of the arista of the antenna). According to Bodenstein and Abdel-Malek,8 this phenocopy is typically produced by mustards. In borate experiments it is independent of the eyeless effect, inasmuch as one may appear without the other, but, as a rule, the aristopedia type appears only when a strong eyeless effect is produced. No wildtype line has been found meanwhile with a specific tendency for aristopedia effect. But there are lines which never produced an aristopedia type in spite of other strong reactions. This again led to the suspicion that the aristopedia effect requires the presence of a genetic basis. It is now known that the mutant aristopedia (ssa) is an allele of spineless (ss), one of the rare cases in which two alleles seem to affect completely different organs (though some but not all lines have shorter bristles). It is possible that a pseudoallelic condition will be found, but that does not affect the present analysis. In view of the borate phenocopy of ssa the idea arose that spineless might contain a high potency for the aristopedia effect, but at a subthreshold level. Treatment of spineless and also heterozygous spineless with borate actually produced the aristopedia phenotype with higher penetrance and higher expressivity than in the experiments with wild-type stocks. It should be added that in this case no eyeless effect occurs, but rather the Lobe phenotype. This is clearly a very suggestive result.
9. A number of other specific phenocopies after borate treatment, limited to specific stocks, were found, of which the most remarkable is the Scutenick syndrome, a continuous series of changes of scutellum and thorax, beginning with one scutellar bristle missing and leading up to a characteristic type of split thorax. The lower types are identical with the fourth-chromosome dominant Scutenick, which phenotype includes also an effect upon the ocelli in both mutant and phenocopy. Two of our lines reacted specifically with this phenocopy. One, an Oregon line (our Oregon-Dempster), is a very weak line of low fertility and, in addition, very sensitive to borate treatment, as low a concentration as 0.04 per cent being almost lethal. Even with this sublethal treatment, the eyeless effect occurs only with low penetrance and expressivity. The other stock, Samarkand, is rather insensitive and is completely without the eyeless effect, as reported. These facts led to a study of the mutant Scutenick (Scn, a fourth-chromosome dominant). In our stock, Scn is balanced over eyD, a fourth-chromosome duplication with a low-grade and atypical eyeless effect, i.e., different from the mutant ey2 in our stock. But the Scn phenotype had disappeared from this stock for years-as happens in stocks-and the most careful scrutiny of the controls did not reveal a single fly. After borate treatment, a considerable percentage exhibited the Scn phenotype! It is difficult to refrain from speculating that the Samarkand and Oregon-Dempster stocks contained a subthreshold allele of Scn. But here again is a possibly contradictory fact: A line with chromosomes 1, 2, and 3 from Samarkand and chromosome 4 from Iowa produces, under strong treatment, a few Scutenick individuals. Unfortunately, Iowa stock, which might be responsible, is not available any more.
10. In the same experiments with Scn/eyD most of the flies with visible Scn characters showed also an extreme enhancement of the expression of the eyD effect, and we had met previously with such enhancement in treated hybrids of ey2 × Oregon (of a very sensitive Oregon line). Thus we had in the latter case enhancement within a normally recessive heterozygote, i.e., from a subthreshold action of the one dose of the recessive to a more extreme action than even the homozygous recessive. (Change of dominance is another way of describing the facts. This description, however, does not show the relation of this enhancement to the others.)
In the eyeless dominant case we had an enhancement in grade of a visible character; in the Scn case, the rise above the threshold of visibility of the action of a mutant known to be present but kept below the threshold action by unknown conditions in the stock. (One might also describe this as Scn having become recessive and being made dominant again by the treatment.) But it should be added to this discussion that the original Scn mutant was described (see Bridges and Brehme) as affecting the eye, at least at lower temperature. Therefore, it cannot be excluded that the described extreme effect upon the eye is not due to enhancement of eyD but, rather, is a part of the evoked Scn phenotype. The fact that all Scn types up to the most extreme can be produced in Samarkand stock without any eye effect is not in favor of such an interpretation.
11. In another set of experiments, possibly all these types of enhancement of genetic actions were involved. Podoptera is a wing mutant known to be present in most stocks with a penetrance of less than 1 per cent, which can be increased by selection, differently in different lines. In most of the lines studied the penetrance of podoptera can be doubled or tripled by the borate treatment. In the case of absence of podoptera in the controls there is no way of deciding whether an extremely low-penetrance podoptera mutant is present or not. But it is certain that in one line (Samarkand inbred) with about 0.06 per cent penetrance in the controls the treatment always enhances the penetrance up to 40 per cent, i.e., about 600 times. No other line showed such an extreme effect. But in a homozygous podoptera stock, constant for 2-3 per cent penetrance, the increase was not very considerable. Thus podoptera with smaller penetrance than the lowest podoptera stocks (i.e., in unselected stocks) is enhanced by borate treatment, and genetic conditions exist (podoptera is polygenically controlled) which make this enhancement extreme in certain lines. Since different lines have podoptera factors of different penetrance, down to a fraction of 1 per cent, which is enhanced by treatment, and since some lines do not show any podoptera at all in untreated flies within the limits of the experiment but produce them after treatment, it is impossible to say whether a border line exists between less and less penetrant podoptera factors and their complete absence.
It is clear that all these facts individually and taken as a whole are very thoughtprovoking and that one cannot help questioning whether some or all phenocopies are not due to the bringing "into the open" of already present subthreshold effects of alleles of the lowest order. Because an incontrovertible proof for such a solution of the problems of phenocopy would have tremendous consequences, and because there are many facts which require the old definition of phenocopies as induced, purely modificatory copies of mutant actions, it is preferable to wait for the outcome of further experiments before proposing an interpretation.
* To Professor Paul Buchner on his seventieth birthday.