Chapter 11

Steel, Flexed-Tail, Splotch, and Varitint-Waddler

I. Steel ( Sl)

The alleles at the Sl locus (chromosome 10) ¹ are of special interest because, like those of the W-series, they too adversely affect melanogenesis, erythropoiesis, and germ cell development. The first mutation at this locus, Sl, arose spontaneously in a C3H line at Oak Ridge ( Sarvella and L. Russell, 1956).

A. Heterozygotes

Steel receives its name from the fact that when heterozygous it produces an overall dilution of hair pigment ( Plate 3-F) which is more severe on the ventrum. On a C3H background Sl/+ mice also have a white snout tip and almost invariably, a small white spot in the middle of the forehead as well as one on the belly (see note 27). Very occasionally Sl/+ mice may also have a white blaze between the eyes or a few white hairs on the back. They also display an almost complete lack of pigmentation in the skin of the feet and frequently an unpigmented tail tip ( D. Bennett, 1956).

On some genetic backgrounds the fur dilution of Sl/+ heterozygotes is very slight, and is not invariably accompanied by the pattern of small white spots present on the C3H background, the head spot being absent most often, the belly spot the least ( Sarvella and L. Russell, 1956).

Although Sl/+ mice have a moderate macrocytic anemia ( D. Bennett, 1956), their life span is normal. At 14 days gestation they possess about 75% of the normal number of erythrocytes and from 2 weeks after birth onward about 80%. ² When heterozygous Sl also produces a reduction in gonad size which, according to Bennett, but not Mintz ( 1957b, 1960), ³ may be associated with a slight deficiency of primordial germ cells.

The macrocytosis associated with Sl is barely detectable in the heterozygote. At birth the mean cell diameter of Sl/+ vs +/+ erythrocytes is on the borderline of statistical significance but the maximum diameter in the heterozygote is considerably greater than the maximum in the +/+ group (and, conversely, the minimum diameter in wild type is much smaller than in the heterozygotes). At 15 days gestation the distribution of cell diameters of the former group may still be slightly larger. No differences are discernable in 14-day-old embryos ( D. Bennett, 1956).

Although the average weight of the testis of Sl/+ males is about 75% that of normals (+/+) it looks normal histologically and the fertility of heterozygotes is not diminished.

B. Homozygotes

In contrast to the relatively slight effects which Sl has when heterozygous, it is lethal when homozygous, Sl/Sl mice rarely surviving to birth. The anemia already is severe at 14 days gestation when Sl/Sl embryos have only 28% of the normal color of erythrocytes and are readily distinguished by their pale color. At 15 days gestation the anemia is even more pronounced with Sl homozygotes having less than 16% of the normal number of red cells. The very severe anemia at this stage of development marks the beginning of the lethal period for these embryos, which die on or after the fifteenth day ( Sarvella and L. Russell, 1956). Very occasionally some Sl/Sl homozygotes survive to birth with an erythrocyte count about 26% of normal, a value that is undoubtedly higher than it really is since such homozygotes never nurse and are probably dehydrated when assayed ( D. Bennett, 1956).

Paralleling the more severe anemia suffered by Sl homozygotes over heterozygotes is a more marked macrocytosis. Although this macrocytosis is not evident in 14-day-old embryos it is very obvious at 15 days. Moreover, the percentage of nucleated cells, i.e., cells of the primitive generation, is much higher in 14- and 15-day-old Sl homozygotes than in +/+ mice. Since at this stage of development most of the blood cells are being produced by the liver this organ is normally (and in Sl/+ mice) densely packed with hematopoietic foci of intermediate cells. In Sl/Sl embryos, however, large areas of the liver are totally devoid of erythropoietic activity, and the foci which are present are smaller and much less dense ( D. Bennett, 1956).

Sl homozygotes are also completely deficient in germ cells. As previously noted ( Chapter 10, Section I, H, 1) in normal mice germ cells can first be detected in the yolk sac splanchnopleure, the allantoic mesoderm, and the caudal primitive streak area of 8-day-old embryos, and at 9 days they can be found in increasing numbers migrating anteriorly to take up their definitive positions in the genital ridge ( Chiquoine, 1954). While Sl/Sl embryos cannot be distinguished from their +/+ or Sl/+ littermates at 8 days, a paucity of germ cells is evident at 9 days and by 14 days gestation they are completely deficient in these cells. Bennett suggests that this progressively increasing deficiency of germ cells in Sl homozygotes may be due to the fact that the cells which are formed fail to divide or to migrate to their proper location, and so degenerate.

Because Sl is lethal when homozygous it was not possible to evaluate directly how two does of the mutation affected pigmentation. Histological examination of the few homozygotes which survived till birth revealed a total absence of pigment granules in the follicles of the vibrissae, but this could hardly be taken as evidence that the entire body lacked pigment. That such was the case, however, was indicated from the results of inserting skin from the mid-dorsal area at the level of the forelimbs of 14- and 15-day-old embryos and from newborn animals, all of which had been classified as Sl/Sl on the basis of blood counts, under the skin of the head of histocompatible adult hosts. All of these grafts, when maintained in these hosts for from 10 to 20 days, produced normal but unpigmented hairs, whereas similar transplants from Sl and +/+ littermates always gave rise to pigmented hairs ( D. Bennett, 1956). It thus appears that Sl/Sl, like W/W and W/W^v (as well as like other W and Mi-series ( Chapter 12, Section I) genotypes, and like Sp/Sp ( Section III), results in a complete absence of neural crest-derived melanocytes. This was confirmed when other, more viable, Sl allelic combinations became available.

C. Other Alleles

The Sl locus seems to be highly mutable. In populations exposed to X-irradition at Oak Ridge and at Harwell more than 30 mutations have been encountered which behave as Sl alleles. Moreover, in nonirradiated populations of C57BL/6J and DBA/2J mice maintained at the Jackson Laboratory, at least five Sl mutations were reported over a 3-year period ( E. Russell and Bernstein, 1966). Induced and spontaneous mutations include the following:

1. Steel-Dickie ( Sl^d)

Next to Sl, this allele has been the most widely studied. Indeed, most of the experimental efforts on the effects of steel have involved the Sl/Sl^d heterozygote since these heterozygotes display all the pleiotropic effects of Sl/Sl, i.e., they are anemic, sterile, and, except for their retinas, lack pigment, but are viable and may live for as long as a year.

The Sl^d mutation appeared spontaneously in the DBA/2J strain ( Bernstein, 1960). Sl^d/+ mice have a similar phenotype as Sl/+ animals; they have a slight overall dilution, which is more pronounced on the belly, an occasional belly spot, and a mild macrocytic anemia ( Deol, 1970b). Although Sl^d/+ mice have no inner ear abnormalities, such abnormalities are produced when Sl^d/+ is combined with recessive spotting ( rs/rs) (see Chapter 10, note 32) another mutation which, by itself, is not associated with inner ear lesions. Thus Deol ( 1970b) has reported that Sl^d/+;rs/rs genotypes are completely white and suffer from moderate to severe abnormalities of the inner ear with every ear affected. Moreover, although such genotypes usually (in 91% of the cases) have some pigment in their inner ear, they never have a full complement of melanocytes.

Sl^d homozygotes, like Sl/Sl^d heterozygotes, are black-eyed whites, sterile, and severely anemic. These genotypes also have similar life spans which vary with their genetic background. For example, on a C57BL/6J background 20% of Sl^d/Sl^d mice survive until they are a month old, and their mean survival is 79 days, whereas when incorporated into a (C57BL/6J x DBA/2J)F₁ background these values increase 25% and 113 days, respectively ( E. Russell and Bernstein, 1966).

Erythropoiesis has been studied extensively in Sl/Sl^d fetuses and all the observations are consistent with the fact that Sl genes do not affect the primitive erythroid cell lineage derived from the yolk sac blood islands, but interfere seriously with the development of the definitive erythroid cell lineage of fetal liver origin ( Chui and Loyer, 1975a). ⁴ Thus whereas the number and sizes of yolk sac-derived nucleated red blood cells are similar in Sl/Sl^d an +/+ fetuses from days 13 - 17 of gestation, Sl/Sl^d fetuses have significantly fewer and much larger fetal liver derived nonnucleated cells ( Chui and E. Russell, 1974; Chui and Loyer, 1975a). Because the number of hemoglobin-containing mature erythroblasts in Sl/Sl^d fetal livers is markedly reduced from that in +/+ livers, while the number of immature erythroid precursors per unit area of fetal liver is not, it has been suggested that "the mutant Sl gene product(s) interferes with or fails to support the differentiation of immature erythroid precursors into hemoglobin synthesizing cells" ( Chui and E. Russell, 1974).

Adult Sl/Sl^d mice display a reduction in the absolute number of nucleated marrow cells ( Bernstein et al., 1968; Travassoli et al., 1973; Wilson and O'Grady, 1976), have normal numbers of blood platelets, reduced numbers of megakaryocytes (Ebbe et al., 1973a, 1973a, 1977) and neutrophils ( Ruscetti et al., 1976), and their granulocytopoiesis is not totally normal ( Sutherland et al., 1970; Knospe et al., 1976). ⁵ Furthermore, it has been shown that although these animals, like W/W^v anemics, are very unresponsive to erythropoietin ( Bernstein et al., 1968; Harrison and E. Russell, 1972; E. Russell and Keighley, 1972), ⁶ they have much more difficulty than the W-anemics in adapting to sudden exposure to constant hypoxia ( Bernstein et al., 1968). ⁷

In contrast to the situation in Sl/Sl mice where no primordial germ cells reach the genital ridges, in Sl/Sl^d embryos the germ cell defect does not seem to be as severe and some cells reach these ridges. In fact McCoshen and McCallion ( 1975) found that while there is a great paucity of primordial germ cells in Sl/Sl^d mice, these cells appear in the same locations on about the same days as in normal genotypes. In normal mice they found that less than 2% of the germ cells had reached the gonadal ridges by day 10 and more than 70% after day 11. On the other hand, in Sl/Sl^d embryos, 23% of the germ cells reached the ridges by day 11 and an additional 12% were found in the adjacent mesenteric root and coelomic angles. Indeed, these investigators point out that considering the fact that the mutant's germ cells do not proliferate (or, if they do, have a high death rate), their rate of migration seems to be comparable to those of normal animals. ⁸

2. Cloud-Gray ( Sl^cg)

This mutation occurred in an F₁ of a neutron-irradiated male. Sl^cg/+ mice are barely lighter than wild type whereas Sl^cg homozygotes are very light grey with white blotches ( Owens, 1972). When heterozygous with Sl, Sl^cg produces an off-white coat, black eyes, and dark ears ( Kelly, 1974). Sl^cg/Sl^cg mice are not very fertile ( Owens, 1972) and about one-third of the females have an imperforate vagina ( Kelly, 1974).

3. Contrasted ( Sl^con)

This mutation occurred in a neutron irradiation experiment ( Searle, 1968c). Sl^con/+ mice can be classified at or soon after birth by the presence of dark pigmentation on the genital papilla. The adult coat, however, tends to be a little lighter than normal. Animals homozygous for this allele also have dark external genitalia but a markedly diluted coat. ⁹ Both eumelanin and phaeomelanin are affected. The underfur is more severely diluted than the overfur, especially in nonagouti ( a/a) mice. While Sl^con homozygous males are fertile ( Beechy and Searle, 1971, many Sl^con/Sl^con females are sterile and none has borne more than one litter ( Searle and Beechy, 1974). Vaginal smears indicate that they generally do not come into oestrus ( Searle, 1968c), and histological studies show that there is a gradual degeneration of oocytes in Graafian follicles so that practically all have gone by 2 months ( Beechy and Searle, 1971). Sl^con/Sl^d heterozygotes display markedly increased white-spotting and dilution ( Beechy and Searle, 1972). Moreover, these heterozygous females are completely sterile with very small ovaries. This allele also has hematological consequences. Thus Sl^con homozygotes, and Sl^con/Sl^d and Sl^con/+ heterozygotes, have fewer erythrocytes than Sl^con/+, Sl^d/+, and +/+ mice, respectively. There is also evidence that this anemic condition involves a macrocytosis ( Searle and Beechy, 1974).

4. Grizzle-belly ( Sl^gb)

This Sl allele arose spontaneously in a white belted female ( Schaible, 1960). Sl^gb/+ heterozygotes have light bellies while Sl^gb homozygotes are anemic and die during the first week of life ( Schaible, 1961). Sl^gb/Sl^d and Sl^gb/Sl genotypes are black-eyed white anemics ( Nash, 1963); Schaible, 1963c). Sl^gb interacts with Mi^wh/+ to give a white coronal area, as well as a grizzled belly ( Schaible, 1960). It also interacts with W^a/+ to produce a roan phenotype with white coronal area ( Schaible, 1961).

5. Sooty ( Sl^so)

Sooty arose in the C57BL/6 strain ( D.S. Miller, 1963). Sl^so/+ heterozygotes have a dilute coat and light tail whereas Sl^so homozygotes are white with black eyes and anemic ( Hollander, 1964). On a C57BL background Sl^so homozygotes usually die before maturity. On other backgrounds, however, they may survive longer but are sterile (Hollander, personal communication).

6. Steel-Miller ( Sl^m)

This mutation occurred in the same colony as Sl^so ( D.S. Miller, 1963). The coat of Sl^m/+ mice is diluted but displays no white spotting. Sl^m homozygotes are black-eyed whites and somewhat anemic but viable. All Sl^m/Sl^m males are sterile but a few females have borne one litter ( D.S. Miller, 1963; Hollander, 1964).

The anemia of C57BL/6-Sl^m/Sl^m mice has been studied in some detail by Kales and his associates ( 1966). It is a macrocytic condition characterized by an increased number of reticulocytes which they attribute to a gastrointestinal bleeding defect of unknown etiology. These investigators also report that Sl^m/Sl^m homozygotes, like Sl/Sl^d heterozygotes, do not respond normally to erythropoietin or to hypoxia. Although they also found this erythropoietic or to hypoxia. Although they also found this erythropoietic defect in Sl^m/+ heterozygotes, in these animals it evidently remains latent as they are not anemic.

7. Dusty ( sl^du)

Dusty, which appears to act as a recessive, occurred as a spontaneous mutation in the C3H/HeJ strain and has been described briefly by M.C. Green and H. Sweet ( 1973). Homozygotes have a slightly diluted coat and small white patches which vary in number from none to extensive white speckling. They are not noticeably anemic. occasionally sl^d/+ heterozygotes can be recognized by the fact that they possess whitish toes and the ventral side of their lower jaw is lighter than normal. Very infrequently sl^du homozygotes cannot be distinguished from wild type as Green and Sweet found that matings of Sl/sl^du x sl^du/sl^du produced six apparently wild type offspring (out of a total of 334), four of which (the other two died) proved to be sl^du/sl^du. Sl/sl^du heterozygotes are moderately anemic at birth and, as a consequence, paler than normal. These heterozygotes also usually have more white spotting than Sl/+ mice.

The effects of the Sl-locus genotypes considered above are summarized in Table 11-1.

D. Influence on Pigmentation

Steel's influence on pigmentation has been studied by Mayer ( 1970, 1973a) and by Mayer and M.C. Green ( 1968) who on the basis of their findings concluded that Sl acts solely via the developing skin to block the survival or differentiation of melanoblasts. In support of this contention they offer the following observations:

1. When grafts of neural tubes, including neural crest cells, from 9-day-old embryos from matings segregating for Sl and Sl^d, were combined with 11-day-old neural crest-free embryonic +/+ skin, and grafted to the chick coelom, melanocytes always were found in either the hairs of recovered grafts or in the tissues of the host in the operated region. However, when +/+ neural tubes were combined with embryonic skin derived from Sl/+ x Sl^d/+ matings pigment was absent in 33% of the cases even though all grafts produced normal hairs. Moreover, these pigment-free grafts could be subdivided into two groups with respect to pigment development in the tissues of the host embryos. In one group the grafts were surrounded by host tissues possessing melanocytes of graft origin, whereas in the other group the coelomic lining and skin of the host chick, like the graft itself, was completely devoid of melanocytes. From these observations it was concluded that the 33% of the cases in which pigment failed to be produced represented combinations which (by chance) the skin was Sl/Sl^d, and that the presence of this skin adversely affected the differentiation of +/+ melanoblasts which migrated out into the surrounding tissues of the host ( Mayer and M.C. Green, 1968).


2. When the occurrence of melanocytes was surveyed in a number of tissues of +/+, Sl/+, Sl^d/+, and Sl/Sl^d 5-day-old mice it was found that Sl/Sl^d mice were devoid of any pigment in all locations save the retinal layer of the eyes (see Table 11-2). From this it was concluded either that Sl/Sl^d skin affected the differentiation of nonepidermal melanoblasts or that other tissues too were affected by the Sl/Sl^d genotype ( Mayer and M.C. Green, 1968).


3. In a later study Mayer ( 1970) extended his earlier efforts by combining neural tubes (including neural crest) from 9-day-old +/+ embryos with skin from 13- to 18-day-old Sl^d/Sl^d embryos. By employing skin of this age Mayer was able to assess its genotype accurately. ¹⁰ The results were consistent with his and Green's earlier observations, i.e., the large majority of the Sl^d/Sl^d skin grafts failed to display pigmented hairs when combined with +/+ neural tubes even though such neural tubes always produced pigmented hairs when combined with W^v/W skin or with 11-day-old (neural crest-free) albino skin. ¹¹ Nevertheless, this absence of pigment in Sl^d/Sl^d hair follicles was not absolute since in a few grafts (3 of 63) some pigmented hairs were found along with those lacking pigment. In these experiments, too, the Sl^d/Sl^d skin appeared to inhibit the differentiation of melanoblasts which migrated into the surrounding tissues of the chick host for in only five cases were such melanocytes found, the area of the host immediately adjacent to the remaining 58 grafts being completely melanocyte free.


4. Finally, in an attempt to analyze more critically this block of pigmentation by the skin, small pieces of embryonic skin from 13-day-old +/+ and Sl/Sl^d embryos were separated into their dermal and epidermal components, recombined, and transplanted to the chick coelom ( Mayer, 1973a). ¹² The results indicated that both components of steel skin adversely affected the development of pigment. Thus, when +/+ epidermis was combined with Sl/Sl^d dermis, all of the grafts failed to display pigment in the dermis as well as in the host tissues surrounding the graft. Nevertheless, in all but 2 of 24 grafts pigmented hairs were produced. Conversely, when Sl/Sl^d epidermis was combined with +/+ dermis none of the recombinants had pigmented hairs (and in only 2 of the 18 cases were melanocytes found in the host tissues surrounding the graft) even though in 10 of 18 cases melanocytes were observed between the hair follicles in the dermis. On the basis of these results, and those involving recombinations of W/W^v and +/+ dermis and epidermis (see Chapter 10, Section I, D), Mayer concluded that both the dermis and epidermis of steel skin adversely affects the development of pigment, and that there is no observable or consistent evidence of a transfer of an inhibitory effect from a Sl/Sl^d skin component to its associated normal component. ¹³

The results of these experiments certainly indicate that Sl acts solely via the environment to block the survival or differentiation of melanoblasts. ¹⁴ The most convincing experiment in this regard is the one in which neural tubes from 9-day-old embryos produced from Sl/+ x Sl^d/+ matings were implanted with normal, 11-day-old neural crest-free skin. The fact that all 33 such combinations in which donor skin grafts were recovered were characterized by pigmented hairs, along with the observation that the 12 others which did not form skin nevertheless produced large numbers of melanocytes (which occurred in the chick tissues extending out from the operated region), implies that the expected 10 or so Sl/Sl^d neural tube-containing composites in this population of transplants must have possessed viable melanocyte clones. The only other possibility is that the putatively neural crest-free +/+ skin was not actually free of melanoblasts. This, however, does not seem likely since skin of the same genotype, age, and origin usually did not produce pigment when combined with neural tubes derived from W^r/+ matings (see Chapter 10, Section I, D). It is therefore hard to escape the conclusion that Sl has no direct effect on melanoblast survival, ¹⁵ a conclusion which, at least to the author, is a little surprising not because it implies that there may be more than one mechanism involved for producing white spotting, but because it is somewhat out of line with Mintz's attractive model. ¹⁶ Thus, if Mayer and Green are correct, and Mintz is also, not only might one expect Sl/+ mice to look differently, but one would also anticipate Sl/Sl <--> allophenics to display pigment patterns different from W/W <--> +/+ animals. Indeed, on the basis of Mayer and Green's observations one might expect such allophenics to resemble more closely the pigment patterns produced by hair follicle phenoclones [such as perhaps the phenotypes displayed by silver ( si/si) mice] than those produced by viable and inviable melanoblast clones. Such allophenics are eagerly awaited.

1. Behavior of +/+ Melanocytes in Adult Sl/Sl^d Skin

While the results noted above indicate that embryonic steel skin adversely affects the survival and/or differentiation of +/+ melanoblasts, this may not be the case with adult Sl skin. Thus when small histocompatible black ( a/a;B/B) ear skin grafts are placed in the center of well-established adult Sl/Sl^d trunk skin grafts, some of the ( Sl/Sl^d) hairs immediately surrounding these implants become pigmented (Poole and Silvers, unpublished). It therefore appears either that the adverse influence Sl skin has on pigment cells is a transitory one limited to embryogenesis and early life, or, alternatively, that the melanocyte is susceptible to the "steel environment" only during a stage in its differentiation, i.e., pigment cells which have passed this "critical" stage can continue to survive and function in Sl/Sl^d hair follicles. ¹⁷

E. Influence on Erythropoiesis

The influence which steel has on melanogenesis appears to parallel its effect on erythropoiesis as there are now a considerable number of studies which indicate that the anemia of Sl/Sl^d mice results from a derangement in the hematopoietic microenvironment in which erythropoiesis takes place ( McCulloch et al., 1965; Bernstein, 1970; M. Bennett, et al., 1968b; Altus et al., 1971; Fried et al., 1973; McCusky and Meineke, 1973; Wolf, 1974). Thus when marrow cells from Sl/Sl^d or Sl^d/Sl^d animals were implanted into heavily irradiated normal (+/+) mice, they were capable of forming macroscopic spleen colonies with approximately the same frequency as cells from +/+ mice. Indeed, Sl/Sl^d marrow cells were found to be as capable as +/+ cells in curing the anemia of histocompatible W/W^v mice. On the other hand, when +/+ marrow cells were transplanted into irradiated Sl/Sl^d animals they failed to proliferate and differentiate normally ( McCulloch et al., 1965; see also Bernstein et al., 1968). ¹⁸

Further evidence that steel adversely affects the hematopoietic microenvironment is provided by the fact that the anemic condition of Sl/Sl^d mice can be ameliorated if they are grafted with an intact spleen from a genetically normal (and histocompatible) donor. Moreover, the anemic condition of these mice also improves following receipt of a W/W^v histocompatible spleen ( Bernstein, 1970). ¹⁹