Osteoactivin (OA) is more highly expressed in the bones of osteopetrotic mutant rats (op/op) than in those of their normal littermates and is the homologue of human nmb, a cDNA more highly expressed in melanoma-derived cell lines of low metastatic potential, and of mouse DC-HIL, which has been implicated in endothelial cell adhesion. The human OA gene is found on chromosome 7p15.1 and consists of 11 exons spanning 28.3 kb. Murine OA is encoded by a highly similar gene of 11 exons spanning 20.2 kb on mouse chromosome 6. Human OA uses the same transcriptional initiation site in both bone and kidney as was reported for melanoma cells. OA is expressed in primary human and mouse osteoblast cultures at all stages of differentiation, with increased levels observed concurrently with the expression of osteoblast phenotype markers. OA is also expressed in a wide variety of human and mouse tissues as determined by RT-PCR analysis. Immunohistochemical investigation of OA expression in late mouse embryonic development showed very high, cell-specific expression in the nervous system, basal layer of the skin, germinal cells of hair follicles, and in the forming nephrons of the kidney. Continuing investigation of the cell-specific expression of OA in bone as well as in other tissues will lead to a better understanding of its function in the development of these cell types.

Mouse strains congenic for individual quantitative trait loci (QTLs) conferring hypnotic sensitivity to ethanol were constructed by backcrossing genotypically selected ILS x ISS N2 individuals to either inbred Long Sleep (ILS) or inbred Short Sleep (ISS) mice. We used a novel "speed congenic" approach in which N2 mice were genotyped for markers flanking each of the five originally identified QTLs. Genotypic selection for ISS regions at four of the five QTLs, and for ILS/ISS at the fifth QTL, allowed rapid fixation of the genetic background. We call this strategy "QTL-Marker-Assisted Counter Selection" or QMACS. By the N4 generation, phenotypic assessments showed that in some sublines the QTL had not been captured; these sublines were discarded and positive lines split to create new replicate sublines. One QTL, on Chromosome (Chr) 8, was not confirmed. At the N8, virtually all sublines on the remaining QTLs retained the phenotypic difference between heterozygotes and ISS homozygotes. Small numbers of interim congenics were produced at the N6 and later generations in which the ILS QTL was made homozygous on the ISS background; as expected, these congenic mice showed an increased sleep time. For later backcrosses (after the N4), the parents were selected on the basis of phenotype as well as genotype. The parent-offspring correlation over all QTLs was significant, supporting the use of phenotypic selection in congenic construction.