Chapter 3

The b-Locus and c (Albino) Series of Alleles

II. The c (Albino) Series of Alleles

Before considering the phenotypic effects of the so-called albino or c-series of alleles (chromosome 7) it should be emphasized that although albinism is epistatic to all other coat-color determinants, i.e., all mice, regardless of genotype, lack pigment in the presence of c/c ( Plate 2-A), albino mice nevertheless possess a full complement of pigment cells. Thus the inability of albino animals to produce pigment stems not from an absence of melanocytes, as is the case for white spotting, but from a deficiency and/or alteration of the structure of tyrosinase in melanocytes which are otherwise normal.

A. Evidence for the Occurrence of Amelanotic Melanocytes

Evidence that albino animals possess a nonfunctioning population of melanocytes—appropriately known as "amelanotic melanocytes"—stems from a variety of observations, most of which were made in the mouse.

When the hair bulbs of albino mice are examined histologically and compared with those originating from white-spotted areas, they are strikingly different. Whereas the hair bulbs of white-spotted areas are characterized by matrices consisting of regularly arranged cells of equal size, (Figures 3-10 and 3-11), albino hair bulbs contain, in addition, many large "clear" cells in their upper bulb region ( Chase and Rauch, 1950; Silvers, 1956) ( Figure 3-12a). Since these large cells with an apparently hyaline cytoplasm are similar in morphology and location to the pigment-containing cells found in lightly pigmented phenotypes ( Figure 3-12b), they are considered to be amelanotic melanocytes ( Silvers, 1956).

Further evidence for this conclusion stems from the observation that the experimentally depigmented melanocytes of black and yellow mice maintained on a biotin-deficient diet are indistinguishable from the clear cells of albinos, though they retain their dopa-positive character ( Quevedo, 1956). Moreover, both clear cells and melanocytes exhibit similar sensitivities to X-rays. Thus, when the skin of albino mice in the resting stage of hair growth is exposed to 1200 r of irradiation, a dose known to destroy almost completely the melanocyte population in resting hairs of pigmented animals ( Chase, 1949; Chase and Rauch, 1950), there is a marked destruction of follicular clear cells ( Quevedo, 1957). This similar radiosensitivity of clear cells and melanocytes, added to the morphological evidence noted above, indicate further that clear cells are in fact amelanotic melanocytes. This conclusion was substantiated again when it was demonstrated that clear cells, like melanocytes, are derived from the neural crest.

The neural crest originates embryologically between the junction of the neural tube and its overlying ectoderm and is initially continuous from head to tail. As development proceeds, however, its constituent cells migrate ventrolaterally on either side of the spinal cord and at the same time become segmentally clustered (see Chapter 1, note 3). In the mouse this anterior to posterior and mediolateral migration of neural crest cells, from their place of origin to their definitive positions, takes place between the eighth and twelfth day of embryonic development (the gestation in mouse is about 20 days), as demonstrated in the classic experiments of Rawles ( 1940, 1947, 1953). Thus, by transplanting tissues derived from various regions of C57BL/6 mouse embryos of different ages to the coelom of the chick embryo, Rawles was able to demonstrate that only those explants which included cells of neural crest origin produced melanocytes. She found neural crest cells to be confined to the region of the neural tube in 8.5- to 9-day-old embryos and only when this region was included in grafts of this age did melanocytes develop. By approximately 11 days of age, however, she found that cells of neural crest origin had made their way into almost all regions of the body so that skin ectoderm and adhering mesoderm removed from almost any level of the trunk (but not from the limb buds) produced pigmented hairs when transplanted to the chick coelom. Limb buds receive migrating melanoblasts between the eleventh and twelfth days of gestation and only at this time did limb-bud ectoderm and adhering mesoderm give rise to pigmented hairs.

Once this "timetable" for the migratory pathway of neural crest cells was established, and it was substantiated that the melanocytes of pigmented animals were derived from these cells, it was easy to demonstrate that they likewise differentiated into the clear cell or putative amelanotic melanocyte population of albino animals. This was accomplished by showing that hair bulbs in the skin of grafts which differentiated from albino embryo explants possessed clear cells only when the explant was known to contain cells of neural crest origin ( Silvers, 1958c) (Figure 3-13a and d). Indeed, the fact that the hair bulbs of skin known to be deprived of its neural crest component were indistinguishable from those normally originating in white-spotted areas (Figure 3-13b and c) provided the strongest evidence that white spotting resulted from an absence of melanocytes, pigmented or otherwise ( Silvers, 1958c). ¹⁰

B. c-Series Alleles

1. Extreme Dilution (c^e)

When one considers that the albino mutation was already known and maintained in mice in Greek and Roman times, and, in fact was the first mammalian trait to be analyzed following the rediscovery of Mendel's principles ( Cúenot, 1902; Castle and Allen, 1903), it is somewhat surprising that descriptions of other c-locus mutations are of much more recent vintage. The second mutation to be recorded at this locus, c^e (for extreme dilution), was fist described in 1921 by Detlefsen. ¹¹ The mutant animal was caught in a corn crib in Illinois and on a first and cursory examination gave the appearance of being a slightly stained or dirty black-eyed white. The animal darkened with age however and eventually acquired a brownish shade, "a little lighter than an ordinary pink-eyed brown with a slight dull yellowish cast" ( Detlefsen, 1921). Examination of the hair revealed no clear evidence of an agouti pattern, the base of the hair being light and the apical portion being pigmented. The ventral surface of the mutant was noticeably lighter than the dorsum and dark pigment was quite pronounced in the skin of the ears and scrotum. Detlefsen also observed that the eyes of mutant animals at birth were somewhat less heavily pigmented than the wild type and that this difference persisted for some time. The eyes of adult animals were, however, very similar in color to those of intensely pigmented mice.

Detlefsen demonstrated quite clearly that this mutation was a c-locus allele and that it expressed itself when heterozygous with albinism. c^e/c mice (on a nonagouti background) are roughly intermediate in color between c^e and c homozygotes and, although the amount of pigment in their eyes is reduced from normal, they nevertheless have enough pigment to make them appear black-eyed on superficial inspection ( Grüneberg, 1952).

The c^e allele when combined with lethal yellow ( A^y) completely removes all phaeomelanin and consequently A^y/—;c^e/c^e and A^y/—;c^e/c mice are blackeyed white. a/a;c^e/c^e mice can be described as pale brown and A/—;c^e/c^e animals, the genotype described by Detlefsen, are significantly lighter as a consequence of the fact that the yellow band of the agouti hair is completely diluted out.

2. Chinchilla (c^ch)

Almost immediately after c^e was found and described, Feldman ( 1922) recorded another c-locus allele which he designated c^r (for ruby-eyed) but which has come to be known as chinchilla ( c^ch). This mutation, which is higher on the scale than c^e, was procured from a fancier. It too has a more drastic effect on reducing the intensity of phaeomelanin than eumelanin. Thus, as described by Feldman ( 1922) "the black agouti type of the homozygous mutant possesses black pigment which is reduced to a very dark dull slate-color, while yellow is greatly reduced and appears about intermediate between white and the normal yellow of the wild type." a/a;B/B;c^ch/c^ch animals are readily distinguishable from phenotypically gray A/A;B/B;c^ch/c^ch ( Plate 2-B) and are best described as a medium shade of sepia. A^y/—;B/—;c^ch/c^ch mice are a faint cream or ivory-color.

When the black agouti type of the mutant is heterozygous for albinism ( A/—;B/—;c^ch/c), the eumelanin pigment is further reduced to a brownish shade while the yellow band is reduced almost to white. Thus a/a;B/—;c^ch/c mice are a dull brown color, a little lighter than the typical chocolate ( a/a;b/b) phenotype.

One of the most interesting features of the c^ch mutation, and one to which we will return ( Section D), is that while as noted above it drastically reduces phaeomelanin production, and significantly reduces the expression of black pigment, it does not influence the deposition of brown pigment [a dilution of brown ( b/b) pigment starts only with c^ch/c and is progressively more severe in c^e/c^e and c^e/c genotypes, respectively ( Grüneberg, 1952)]. Thus a/a;b/b;c^ch/c^ch and a/a;b/b;c^ch/c^e are virtually indistinguishable from a/a;b/b;C/C animals. Indeed the only difference is that the hairs which occur outside and around the ears of nonagouti brown mice are yellowish in a/a;b/b;C/C animals but much lighter in the corresponding c^ch/c^ch and c^ch/c^e genotypes ( Grüneberg, 1952). ¹²

3. Intense Chinchilla (cⁱ)

A fifth allele in the albino series was also described by Feldman ( 1935) and was known as intense chinchilla ( cⁱ). As far as I am aware, this mutation is no longer available. The mutation, which represented the most intense of the series below C, as first noted in a chinchilla stock. When homozygous this allele had such a very slight affect on the synthesis of black pigment that it was not possible to distinguish a/a;B/—;cⁱ/cⁱ mice from a/a;B/—;C/— animals with certainty. On the other hand, in accord with the other alleles at the locus, it had a more pronounced influence on phaeomelanin diluting the intense yellow pigment characteristic of C/C genotypes to a pale ocherous yellow color, bordering on a lemon tint ( Feldman, 1935). Because cⁱ/cⁱ had very little, if any, influence on black pigment and a less drastic effect on yellow than the lower alleles of the series, A/—;B/—;cⁱ/cⁱ animals appeared to be not very different from the wild type. Intense chinchilla like the other members of the series was phenotypically completely recessive to full color, at least on a nonagouti background. It also displayed incomplete dominance over the lower alleles of the series.

4. Summary of the Phenotypic Effects of C, cⁱ, c^ch, c^e, and c

The effects which the c-series of alleles considered above have in the presence of nonagouti ( a/a), lethal yellow ( A^y/—), and pink-eyed dilution ( p/p) are summarized in Table 3-2. Inasmuch as p/p has a drastic effect on eumelanin synthesis, but little influence on phaeomelanin production (see Chapter 4, Section II, A), while these c-series alleles reduce yellow pigment more than black, it is not surprising that A^y/—;cⁱ/cⁱ and A/A;cⁱ/cⁱ;p/p mice are significantly lighter than a/a;cⁱ/cⁱ animals and that A^y/—;c^ch/c^ch and A/A;c^ch/c^ch;p/p genotypes are lighter than a/a;c^ch/c^ch mice. Indeed, these interactions are especially conspicuous on a black-and-tan ( a^t/—) background where one can compare the effects of these genes on eumelanin and phaeomelanin production in the same host.

5. Platinum (c^p)

In 1964 a mouse slightly different from an albino was observed in an AKR ( a/a;c/c) x DBA/2J ( a/a;b/b;d/d) litter. At maturity this animal had pink eyes and a coat a shade darker than albino with a luster or sheen. Breeding tests revealed that it carried a mutant allele at the c-locus which was named "platinum" and designated c^p. a/a;B/—;c^p/c^p animals are lighter than the corresponding c^e/c^e genotype and have pink eyes. a/a;c^p/c heterozygotes have a phenotype intermediate between c^p/c^p and albinism ( Dickie, 1966a). The most interesting feature of this mutation is its expression in lethal yellow mice. Inasmuch as c^p/c^p has a more drastic effect on eumelanin than c^e/c^e and the latter produces black-eyed whites in combination with A^y/—, one might expect A^y/a;c^p/c^p mice likewise to be white but with pink-eyes, i.e., indistinguishable from true albinos. However, this is not the case for while such animals have pink-eyes they nevertheless are pigmented, resembling a/a;c^p/c^p but with a definite yellowish tinge ( Plate 2-C) (Poole and Silvers, unpublished). ¹³ This mutation is therefore enigmatic in that it has a greater influence than c^e/c^e in inhibiting eumelanin synthesis but a less drastic effect than c^e/c^e in inhibiting eumelanin synthesis but a less drastic effect than c^e/c^e on phaeomelanin production.

6. Himalayan (c^h)

Inasmuch as mutant genes which produce light or white body fur and dark extremities are common in mammals, e.g., the Himalayan rabbit, Siamese cat, "albino" guinea pig, and "partial albino" hamster, and the mutation responsible for this phenotype appears in all cases to be one of an allelic series similar to the c-series in mice, it is surprising that it has been only within the last two decades that a "himalayan" mutation has been recorded in mice. ¹⁴ The deviant animals occurred also in an AKR/J ( a/a;c/c) x DBA/2J ( a/a;b/b;d/d) litter and breeding tests confirmed that a mutation at the c-locus, appropriately designated "himalayan" (c^h), was responsible for its phenotype ( M.C. Green, 1961).

According to M.C. Green ( 1961) a/a;c^h/c^h mice ( Plate 2-D) "are indistinguishable from albinos at birth. At about one week of age they are slightly darker than albinos, particularly on the tail. The juvenile coat, when fully grown, is pale tan with little evidence of darkening of the extremities except on the tail. The body color may be uniform or may be slightly lighter across the shoulders and darker toward the tail. At the first molt the nose, ears, tail and scrotum, but not the feet, become considerably darker, and the rest of the body becomes lighter. The dark hair on the nose extends back to about the anterior border of the eyes. The body fur, both dorsal and ventral, is lightest on the anterior half of the body and darkens gradually from the middle of the body back to the tail. There is often a particularly dark ring of body fur next to the tail and the skin of the scrotum may be very dark. The ears are darkest on their anterior border. The feet never become dark as they do in the Himalayan rabbit. At subsequent molts the body may become somewhat lighter and the extremities darker, but there is considerable variation." Green also notes that "the eyes of c^h/c^h mice are not pigmented at birth but become darker with age, and are ruby colored at weaning."

Of course the most interesting feature of the himalayan mouse is that, like the Himalayan rabbit ( Schultz, 1915) and Siamese cat ( N. Iljin and V. Iljin, 1930), its ability to produce melanin is promoted by low temperatures ( M.C. Green, 1961; Coleman, 1962; Moyer, 1966).

When heterozygous with albinism, himalayan ( c^h/c) mice are indistinguishable at birth from either c/c or c^h/c^h homozygotes (they all lack pigment). Within about a week, however, they develop a very pale buff phenotype which is roughly intermediate between the color of c^h/c^h and albino mice. Moreover, this intermediate level of pigmentation persists since the extremities of c^h/c mice never become as dark as in c^h/c^h animals.

The expression of an intermediate level of pigmentation also occurs when c^h is heterozygous with c^e and c^ch. On an agouti background, c^h/c^e mice are dark eyed from birth with a pale coat which subsequently darkens at its extremities. Consequently the extremities of the adult heterozygotes are darker than the extremities of the corresponding c^e/c^e homozygote, but lighter than the extremities of c^h/c^h mice. On the other hand, the rest of the body displays a level of pigmentation lighter than A/—;c^e/c^e but darker than A/—;c^h/c^h adult mice ( M.C. Green, 1961). c^ch/c^h heterozygotes are significantly lighter than c^ch/c^ch animals. Whereas, as noted above, the a/a;c^ch/c^ch mouse is a dull black or sepia color, the a/a;c^ch/c^h adult is a light golden brown. Furthermore, although the nose and tail of this animal are darker than the rest of the body, they too are not as dark as in the c^ch/c^ch mouse ( M.C.Green, 1961). Thus the himalayan allele like the other lower alleles of the c-locus produces intermediate phenotypes when heterozygous. However, unlike the other alleles, c^h appears to occupy two positions in the c-series hierarchy. While in depth of color produced it is below c^ch in the series, it is above c^e in the color of the extremities but below this allele insofar as its affect on the color of the trunk is concerned.

7. Chinchilla-mottled (c^m)

The final c-locus allele chinchilla-mottled ( c^m) — originally known as c^22H — was found in the progeny of a neutron-irradiated male ( R.J.S. Phillips, 1966b). c^m/c^m and c^m/c^ch animals on a nonagouti, black ( a/a;B/—) background are mottled displaying patches of normal c^ch/c^ch fur and patches of a lighter fur fairly similar to c^ch/c in intensity (R.J.S. Phillips, 1966b, 1970a). The amount of each color varies considerably. c^m/c^m mice can be distinguished from c^m/c^ch heterozygotes by their whiter belly color ( R.J.S. Phillips, 1970a). While there is no evidence that this variegation is caused by a translocation, the c^m phenotype is influenced by a dominant modifying gene, probably independent of c, which has provisionally been called "modifier" and designated M(c^m). c^m/c^m;M(c^m/M(c^m) mice are almost white; they also appear to be smaller than normal ( R.J.S. Phillips, 1970a) (see Chapter 7, Section VI).

8. Other Alleles

In addition to the specific alleles noted above, L. Russell ( 1979) has reported the occurrence of 16 mosaic, or fractional, c-locus mutants which were characterized by area(s) of lighter fur or mottling. ¹⁵ Although these deviants arose in the course of radiation experiments, they probably were not radiation induced. If they were, this would indicate that the bulk of spontaneous c-locus mutations are fractionals ( L. Russell, 1979). Since in this group of mutants about one-half of the germinal tissue carried the mutation, it appears that they were derived from an overall blastomere population that was one-half mutant. As pointed out by Russell, "such a population could result from mutation in one strand of the gamete DNA; in a daughter chromosome derived from pronuclear DNA synthesis of the zygote; or in one of the first two blastomeres prior to replication."

A considerable number of lethal c-locus deviants have also been recovered from radiation experiments (see Russell et al., 1979). Many of these deviants, in addition to preventing melanin synthesis (see Rittenhouse, 1970), have other very different effects. Indeed, because they are associated with such a diverse, and seemingly unrelated, series of abnormal conditions they undoubtedly represent chromosomal deletions for genetic material other than just the albino locus (Erickson et al., 1968, 1974a; Gluecksohn-Waelsch et al., 1974; L. Russell et al., 1979; L. Russell and Raymer, 1979). ¹⁶

C. Dominance of C

Before leaving the subject of the phenotypic effects of c-locus alleles, it should be noted that there are some situations in which C is not completely dominant over the other alleles of the series. This occurs in the presence of p/p (pink-eyed dilute); the genotypes a/a;B/B;p/p;C/c and a/a;b/b;p/p;C/c are clearly lighter in color than the corresponding C/C types ( Snell, 1941). It likewise occurs in lethal yellow, black-and-tan ( a^t), and agouti mice (and presumably in all phaeomelanin containing regions of agouti-locus genotypes). In A^y (and A) mice C/c, C/c^h, and C/c^e genotypes are noticeably lighter, especially on the ventrum, than the corresponding C-homozygote. The influence in agouti animals is limited to the yellow portion of the hair and in a^t/a^t animals to the ventrum. ¹⁷ Contrary to what might be expected, C/c^e genotypes appear slightly lighter than C/c mice (Poole and Silvers, unpublished). That this effect is a consequence of a direct interaction between the c- and a-loci, and not due to a general influence of the c-locus on phaeomelanin synthesis, is provided by the fact that it does not occur in recessive yellow mice. Thus a/a;e/e;C/C and a/a;e/e;C/c animals are indistinguishable (Poole and Silvers, unpublished). ¹⁸

D. Effect on Pigment Granules

Dunn and Einsele ( 1938) compared the amount of pigment present in the hairs of both black and brown agouti animals bearing different c-locus alleles and found that, in general, the reduction in intensity of hair color was accompanied by a parallel graded reduction in the amount of melanin as measured by weight. Moreover, since they found that in combinations with black (but not brown) the chief tangible factor accompanying this decrease was the size of the pigment granule, they concluded that it was via this influence on granule size that the c-locus controlled the quantity of melanin produced in black animals. Although this effect on granule size was confirmed by E. Russell ( 1946, 1948, 1949a), her more detailed investigations revealed that in some cases changes in the level of pigmentation produced by the albino series also involved the number, shape, color intensity, and distribution of the granules in the hair.

Some estimates of the total eumelanin present in B/B and b/b animals of different c-locus genotypes, as reported in the studies of Dunn and Einsele ( 1938) and E. Russell ( 1948), are presented in Table 3-3. It is clear from these data that c-locus substitutions reduce pigment much less in brown than in black mice. Indeed as already noted, on a nonagouti brown background the genotypes C/—, c^ch/c^ch, and c^ch/c^e all possess the same amount of pigment. This is especially conspicuous in brown-and-tan ( a^t/—;b/b) mice where these c-locus genotypes all have an intensely pigmented (chocolate) dorsum but can readily be distinguished by the color of their phaeomelanin-containing ventral hairs.

More detailed data on the effect of three albino series alleles ( C, c^ch, and c^e) on a number of pigment granule attributes in black, brown, and yellow mice, as derived from the data of E. Russell ( 1946) by Grüneberg ( 1952), are given in Table 3-4. Here it can be noted that in a/a;B/B genotypes the number of pigment granules is not reduced significantly from step C/C to c^ch/c^ch (see Figure 4-6) but is very significantly reduced in c^e homozygotes. On the other hand, there is a progressive diminution of granule size from C/C to c^e/c^e accompanied by, and believed to be responsible for ( E. Russell, 1949a), changes in granule shape and color intensity. Hence, whereas the significant reduction in color intensity displayed by a/a;B/B;c^e/c^e mice results predominantly from a reduction in the number of granules, the much less conspicuous phenotypic difference between a/a;B/B;C/C and a/a;B/B;c^ch/c^ch animals is primarily a consequence of granule size (see Figure 4-6).

In brown animals the albino series does not influence granule size, though, as noted in Table 3-4, it reduces granule number. In yellow mice, however, as well as in the yellow-pigmented regions of other agouti-locus genotypes, c-locus substitutions have a slight influence on granule size although here, too, their predominant effect is on granule number.

It therefore appears that in black mice there are a maximum number of pigment granules that can be produced and since this level is attained in c^ch/c^ch animals, any further increase in the quantity of pigment must stem from an increase in granule size. In brown mice, on the other hand, the influence of c-locus alleles seems to be limited both by the number of granules and their size, i.e., there seems to be a size beyond which b/b granules cannot grow, and since both of these attributes reach a maximum in c^ch/c^ch (actually c^ch/c^e) mice, this allele has no diluting influence on the chocolate phenotype ( E. Russell, 1949b).

E. Effect on Tyrosinase Activity

Since all the evidence indicates that the action of the c-series of alleles is a general one, affecting the level of the whole pigmentation reaction rather than influencing the type of pigment produced, it should come as no surprise that these alleles appear to produce their effect by controlling the activity of tyrosinase.

Evidence for this was initially obtained in mice by L. Russell and W. Russell ( 1948) who incubated lightly fixed, 30-micron sections of skin, from 6- to 7-day-old animals, of different c-locus genotypes, in either buffered dopa or "control" solution. The change in the intensity of pigment in the hair follicles resulting from this treatment was then compared to a series of "standards" and graded. Although this method demonstrated that the intensity of the pigment formed corresponded to what would be expected from visual examination of the genotypes, i.e., C/C > c^ch/c^ch > c^e/c^e > c/c, the reactions much more closely reflected the influence of these genotypes on phaeomelanin than on eumelanin synthesis, regardless of whether the piece of skin possessed yellow pigment or not (see also W. Russell et al., 1948).

More recently Coleman ( 1962) has determined the amount of [2-¹⁴C] tyrosine that is incorporated into slices of skin from infant mice of different c-locus genotypes and although his results ( Table 3-5) are similar to those of the Russells', they nevertheless provide some new and significant information. Probably the most significant is that while a/a;B/B;C/c and a/a;B/B;C/c^h animals are indistinguishable phenotypically from the corresponding C-homozygote, they incorporate only about 50% as much tyrosine. This suggests that at this biochemical level there is no dominance of C over other c-series alleles. It may also explain why in the presence of those a-alleles which promote the synthesis of phaeomelanin, C/c heterozygotes produce less of this pigment than C-homozygotes. Coleman's results, as noted in the table, also indicate that a/a;c^ch/c^ch mice incorporate less tyrosine than either a/a;C/C or a/a;C/c animals, a result consistent with their slightly lighter phenotype, and that the level of incorporation is still further reduced in the even more lightly pigmented a/a;c^ch/c^ch;p/p and a/a;c^ch/c;p/p genotypes. The values for animals of these last two genotypes are included to emphasize that although this situation is similar to a/a;C/C and a/a;C/c in that the amount of tyrosine incorporated by the heterozygote is about half that of the homozygote, in this instance homozygotes ( c^ch/c^ch and heterozygotes ( c^ch/c) are easily distinguished visually. ¹⁹

The results with himalayan mice are particularly interesting inasmuch as, already noted, their phenotype is temperature dependent; if raised in a cold environment their pigmentation becomes significantly darker and, conversely, at warm temperatures, lighter. ²⁰ This indicates that in these animals tyrosinase is heat labile and in accord with this is the fact that when skin slices from 5-day-old C/C and c^h/c^h animals are incubated at 55 C for periods up to 1 hour, the incorporation of tyrosine is decreased by 10% in the C/C skin but by about 70% in the himalayan skin. This thermolability of tyrosinase under the influence of the c^h allele is important because it suggests that perhaps all alleles at the c-locus control the protein structure and not the quantity of the enzyme ( Coleman, 1962; see also Foster, 1967; Foster et al., 1972). ²¹

F. Influence on Forms of Tyrosine

As already noted a number of investigations have disclosed that there are multiple forms of tyrosinase in mice (Holstein et al., 1967, 1971; Burnett et al., 1969) and some of these are influenced by the c-locus ( Wolfe and Coleman, 1966). Thus an electrophoretic study of tyrosinases isolated from the skin of different c-locus genotypes has revealed not only quantitative but qualitative differences ( Wolfe and Coleman, 1966). Tyrosinase from C/C mice produce two distinct tyrosinase bands on electrophoresis whereas similar preparations from c^ch/c^ch animals produce only a single, albeit faster moving, one. Himalayan tyrosinase also displays two tyrosinases, a fast moving one which corresponds to the fastest-moving wild type form and one which moves much slower than its counterpart in wild type mice ( Wolfe and Coleman, 1966).

If the major influence of the c-series of alleles is on the structure of tyrosinase, it is difficult to explain how such a structural change in the protein could influence both the size and number of pigment granules. One possibility, however, is that the abnormal protein is not produced at a normal rate, or, more likely, that it is unable to conjugate properly with the other proteins destined to form the melanin matrix ( Coleman, 1962). On the other hand, there is some evidence which suggests that the albino locus might be more correctly regarded as a regulator rather than a structural locus for tyrosinase ( Hearing, 1973; see also Chian and Wilgram, 1967; Pomerantz and Li, 1971).

G. Ultrastructural Effects

The influence of the c-locus on the ultrastructure of the melanin granule has received considerable attention. Moyer ( 1966) examined the granules of chinchilla, himalayan, extreme dilute, and albino mice and concluded that the development of the granule in albino genotypes parallels that in nonalbinos. Indeed, the only difference he noted was that the granules of albino mice did not possess any melanin and, therefore, did not display any second-order periodicity ( Figure 3-14). In contrast all pigmented c-locus genotypes were characterized by a second-order periodicity which was determined by their b-locus constitution, an observation consistent with the fact that the c-locus affects the amount, rather than the quality, of melanin. Moyer also observed that whereas there was no obvious difference in the size, shape, and number of c^ch/c^ch, c^h/c^h and C/— retinal granules, their number, as well as their size, was reduced in c^e/c^e and c/c mice. This coincides with the effect these alleles have on the granules of the hair.

Hearing and his colleagues ( Hearing et al., 1973) also found that the premelanosomes formed in the retinal and choroid of albino mice were similar to those of pigmented animals except for the absence of melanin. they also noticed in both of these locations that the number of premelanosomes decreased after several weeks postpartum.

Rittenhouse ( 1968b), on the other hand, has reported that the fine structure of albino and pigmented granules is significantly different. As previously noted (see Section I, F) she found that there was a distinct difference in the ultrastructure of a/a;B/B and a/a;b/b hair bulb granules and consequently one might have anticipated that B/B and b/b albino granules would likewise be different. This however, was not the case. According to Rittenhouse the failure of albino granules to melanize is accompanied by changes in granule structure such that there appears to be no distinct difference between the hair bulb granules of brown-albino and black-albino mice. Animals of both genotypes contain some granules which resemble those of the "black" type (ovals with longitudinal strands, or small circles with spiral or circular patterns) and others which are more typically those of the "brown" variety (large circles with a complex disorderly internal structure). Since this absence of melanization occurs along with a shift toward the "black" pattern in b/b;c/c melanocytes, these two changes could be related. Accordingly Rittenhouse suggests that the disorganized framework of b/b granules may tend to acquire the more orderly organization of black granules if not stabilized almost immediately by at least a light melanization. Whether this is the case and/or whether at least some of the "brown" type granules represent granules which, because they are not melanized, are disintegrating ( Rittenhouse, 1968b), remains to be resolved. Regardless, Rittenhouse ( 1968b) observed that this shift toward a pattern of longitudinal strands and membranes characteristic of B/B granules also occurs when p/p and b/b are incorporated into the same genome (see Chapter 4, Section II, E).

H. Influence on Central Visual Pathways

In addition to an absence of melanin, the central visual pathways of albino mice [and of other albino mammals ( Lund, 1965; Guillery et al., 1971; Creel, 1971; Witkop et al., 1976)] are abnormal. Parts of the retina that normally give rise to uncrossed retinofugal axons send axons across the midline. Although the manner in which the albino gene acts upon the retinal cells to affect the chiasmic growth of their axons is not known, Guillery and his associates (1973) have shown that its influence is extracellular. This was demonstrated by taking advantage of Cattanach's flecked translocation (Cattanach, 1961, 1974; Ohno and Cattanach, 1962; for excellent review see Eicher, 1970b). This translocation involves the transfer of the normal allele at the albino locus to the X-chromosome where, in females, it is subject to "Lyonization" (see Chapter 8, Section I). Thus female mice which are homozygous for c and heterozygous for this translocation display the characteristic albino-variegated or flecked coat, and in the pigment layer of the retina one sees small patches of pigmented cells intermingled with patches of albino cells ( Deol and Whitten, 1972). Guillery and his colleagues reasoned that if, in the production of the abnormal chiasmatic pathway, each ganglion cell acted independently and the specification of a cell as ipsilateral or contralateral depended upon the enzymes produced within that cell, then a flecked mouse should have an abnormal pathway approximately half as large as that of an albino mouse. This seemed especially reasonable inasmuch as the effect which c has on pigmentation is intracellular. On the other hand, if the mechanism responsible for the abnormality was intercellular, and involved materials which diffused between cells, or involved mechanical interactions between cells, then flecked mice could display an abnormality equal to that of albino mice, a defect somewhere between albino and normal animals, or no abnormality at all. This latter situation apparently prevails as in contrast to the albino mouse where the ipsilateral component is, as expected, markedly smaller than in pigmented animals, in flecked mice this component is, surprisingly, somewhat larger than in pigmented mice.

The precise relationship between the influence of the albino locus on pigment formation and on the normal development of the eye remains to be determined. On the one hand the lack of tyrosine may affect the concentration of some other substance which can diffuse between cells and influence the course taken by some of the ganglion axons; on the other hand, c could affect a process that is independent of tyrosine synthesis, and this process could act upon the ganglion cells ( Guillery et al., 1973).

I. Influence on Behavior

Before concluding this section there is one investigation which should be noted because it provides good evidence that albinism (or the absence of this condition) can influence behavior. Taking advantage of a mutation from C to c in the C57BL/6 strain, Fuller ( 1967) demonstrated that the performance of mice in certain test situation was affected solely on the basis of their c-locus genotype. c/c mice escaped more slowly from water, were less active in an open field, and made more errors on a black-and-white discrimination task than C/— animals. Although the cause of these differences in performance is not known, it raises the possibility that other coat color determinants also may influence this kind of behavior.