In previous studies, we established that a gene trap allele of Mbtps1
was lethal both in the homozygous state and in trans
(Brandl et al. 2009
). However, although hypopigmentation was the phenotype used to map the Mbtps1wrt
mutation, no firm conclusion could be drawn concerning cause and effect. Using transgenesis, we have now adduced strong evidence that the previously reported coat color anomaly of the wrt
strain does, in fact, result from the hypomorphic wrt
mutation of Mbtps1
. In a compound heterozygous cross, Mbtps1wrt/wrt
mice that were also transgenic for the wild-type Mbtps1
locus never displayed hypopigmentation and all mice that did display hypopigmentation were of the Mbtps1wrt/wrt
genotype, and lacked the transgene.
More than 300 gene products are known to affect mouse pigmentation (Montoliu et al. 2011
), including proteins that regulate melanocyte proliferation and development (e.g.
, Mitf, Kit, Edn1), melanosome biogenesis (e.g.
, Oca2, Slc45a2, BLOC-1 complex), melanosome transport (e.g.
, Rab27a, melanophilin, myosin Va), and melanogenesis (e.g.
, Tyr, Dct, Mc1r). Most classical coat color genes encode factors produced in the skin and/or hair follicle that signal locally to melanocytes or within them (Hirobe 2011
; Slominski et al. 2004
). In general, less is known of factors that circulate through the bloodstream and exert systemic control over pigmentation. One well-studied example is α-melanocyte−signaling hormone, which is produced in humans by melanocytes and keratinocytes (Chakraborty et al. 1996
; Rousseau et al. 2007
; Wakamatsu et al. 1997
) and in mice and humans by the pituitary gland (Hirobe et al. 2004
; Krude et al. 1998
; Pears et al. 1992
), through cleavage of the precursor protein proopiomelanocortin. α-melanocyte−signaling hormone from either tissue source in humans, and from the pituitary gland via the bloodstream in mice, promotes the production of eumelanin by melanocytes. Other systemic factors influencing pigmentation include steroid hormones (e.g.
, estrogen, progesterone, androgen), fatty acids (e.g.
, linoleic acid, palmitic acid), and iron (Hirobe 2011
Using reciprocal immunologically compatible skin grafts, we showed that normal pigmentation of the fur depends upon two processes, one cell-autonomous or paracrine and one systemic, both of which are disrupted by homozygosity for the wrt mutation. The effects of Mbtps1 activity on cells or tissues at remote locations may be interpreted in several ways. First, it has been noted that both an ER/Golgi membrane-anchored and a shed, soluble form of S1P exist, and the soluble form might conceivably exert a systemic effect on pigmentation. Second, it may be imagined that specific metabolites, dependent upon the enzymatic activity of S1P cleavage products, might be needed for this process. Although we showed that serum cholesterol levels were reduced in Mbtps1wrt/wrt mice, cholesterol does not seem to be the crucial metabolite needed for normal pigmentation, since a high cholesterol diet did not normalize the coat pigmentation of Mbtps1wrt/wrt mice.
In breeding experiments, maternal-zygotic Mbtps1wrt
mutant offspring (homozygotes derived from homozygous mutant mothers) displayed fully penetrant embryonic lethality, whereas zygotic mutant offspring (homozygotes derived from heterozygous mothers) displayed partial (~40%) embryonic lethality, and heterozygous mutant offspring of homozygous mothers were fully viable. These findings demonstrate a maternal-zygotic effect of Mbtps1
, to our knowledge the second mammalian gene to which such an effect has been ascribed. Mouse Zfp57
was the first identified mammalian maternal-zygotic effect gene and was found to participate in the maintenance of genomic DNA methylation imprints without which mouse embryos died in midgestation (Li et al. 2008
). ZFP57, together with its cofactor KAP1, recruits DNA methyltransferases to a methylated hexanucleotide within numerous imprinting control regions (Quenneville et al. 2011
; Zuo et al. 2011
). A role for S1P in genomic imprinting remains to be tested.
Experiments in which Drosophila
was used support an evolutionarily conserved requirement for the function of the S1P/S2P module during embryonic development. Analogously to wrt
mutant mice, dS2P
homozygous mutant fly embryos from heterozygous mothers emerged at a normal frequency, whereas less than 50% of the expected number of homozygous offspring derived from homozygous mothers survived (Matthews et al. 2009
). The survival rate of heterozygous offspring of homozygous female flies was not reported in this study, leaving open the possibility that dS2P
may function as a maternal-zygotic effect gene. However, in contrast to the nonredundant function of S1P in mice, the caspase drICE can partially compensate for dS2P deficiency in Drosophila
, thus permitting a significant proportion of homozygous fly embryos derived from homozygous mothers to survive to adulthood (Amarneh et al. 2009
). Notably, maternal effect embryonic lethality of homozygous mutant flies was completely rescued by supplementation of the embryo culture medium with fatty acids (Matthews et al. 2009
). Fatty acids and/or cholesterol may likewise be critically lacking during development of Mbtps1wrt/wrt
concepti in utero
We found that cholesterol and lipoproteins were preferentially reduced relative to triglycerides in serum from Mbtps1wrt/wrt mice, an effect that we attribute to a more severe impairment of SREBP2 processing as compared with SREBP1 processing. Y496, the residue mutated in woodrat mice, may provide a contact critical for interaction with SREBP2.