We have examined the methylation status of the zebrafish genome during early embryogenesis and we find evidence that methylation fluxes do occur in that organism. The parental genetic contributions to the zygote are, initially, differently methylated with the genome of the sperm being hypermethylated relative to the genome of the oocyte. Post-fertilization there is an immediate decrease in methylation of the embryonic genome but the methylation begins to increase rapidly and is re-established by the gastrulation stage. These results are consistent with the results of Santos et al. (Dev Biol 241:172-182, 2002), who examined the methylation of early mouse embryos, and this conservation argues that demethylation/re-methylation is an important part of vertebrate development.

Fetuses were examined to produce a developmental profile of renin expression in the kidneys and adrenal glands in single renin gene and two renin gene strains of mice. Sites of renin expression were detected by in situ hybridization using an 35S-labeled antisense RNA probe complimentary to the renin cDNA. Accumulation of renin transcripts in the adrenal gland reached a maximum at 15.5 days post coitum for all strains examined, but declined to undetectable levels by birth in one gene strains, while in two gene strains, the levels of renin transcripts lessened and by birth became limited to the developing inner cortex. Kidney renin transcripts were first detected at 14.5 days post coitum in the newly developing arteries in fetuses of both genotypes of mice. As the renal arterial tree developed, renin mRNA containing cells were progressively localized to more distal blood vessels and finally to the specialized cells of the afferent arteriole (juxtaglomerular cells). These results were confirmed by examining the localization of immunoreactive T antigen in transgenic fetuses. These mice carried a transgene which placed the SV40 T antigen structural gene under control of renin regulatory elements. Expression of T antigen occurred at the same sites in the kidneys and adrenal glands as renin mRNA. Furthermore, in strains with two renin genes, primer extension analysis indicated transcripts from both genes were present in equal proportion in combined kidney and adrenal gland extracts of total RNA. These transcripts were full length in size. The transient localization of renin mRNA in cells of the fetal intrarenal arteries is consistent with the notion that renin may be a useful marker for the developing renal vasculature.

The Ren-1 locus of mice encodes the protease renin, which with converting enzyme processes angiotensinogen to the potent vasopressor angiotensin II. Some strains of mice appear to carry a duplication of the renin structural gene (Ren-2) near the Ren-1 locus. Strains with the gene duplication can exhibit as much as 100-fold higher levels of submaxillary gland renin compared to strains with a single gene copy. In contrast, kidney renin levels appear to be unaffected by the gene duplication. Sequence analysis of a 319 bp renin cDNA recombinant isolated from a kidney library from the two-gene strain DBA/2Ha corresponds to a transcript of the Ren-1 gene. Moreover, a single base substitution of A for G at residue #996 in the kidney renin mRNA creates a potential glycosylation recognition site that may, in part, account for the differential glycosylation of kidney and submaxillary gland renins. In addition, our tissue surveys indicate that mature mRNAs from the Ren loci are detectable in adrenal gland and testes, as well as sublingual and parotid salivary glands, and reveal length variation for the renin transcripts in at least the submaxillary gland.

Proposed models for the inheritance of locus-specific methylation phenotypes in somatic cells include those in which there is stable inheritance of a methylation pattern such that all cells contain a similarly methylated locus, as well as models in which the inheritance of methylation can be variable. We investigated these possibilities by examining the methylation and expression of hemizygous loci in the mouse. Our results demonstrate that differences in both methylation and expression can exist between apparently identical cells and that such mosaicism is genetically controlled.