Members of the SoxB transcription factor family play critical roles in the regulation of neurogenesis. The SoxB1 proteins are required for the induction and maintenance of a proliferating neural progenitor population in numerous vertebrates, however the role of the SoxB2 protein, Sox21, is less clear due to conflicting studies. To clarify the role of Sox21 in neurogenesis, we examined its function in the Xenopus neural plate. Here we report that misexpression of Sox21 expands the neural progenitor domain, and represses neuron formation by binding to Neurogenin (Ngn2) and blocking its function. Conversely, we found that Sox21 is also required for neuron formation, as cells lacking Sox21 undergo cell death and thus are unable to differentiate. Together our data indicate that Sox21 plays more than one role in neurogenesis, where a threshold level is required for cell viability and normal differentiation of neurons, but a higher concentration of Sox21 inhibits neuron formation and instead promotes progenitor maintenance.

Dual-photon absorptiometry (DPA) was performed on both hips of 40 patients to determine if the calculated bone-mineral density (BMD) of one hip could be used to predict the BMD of the opposite hip. For the Ward triangle, femoral neck, and greater trochanter the correlation coefficients between the BMD of the two hips was .920, .917, and .843, and the standard errors (SE) of the estimate for the linear regression of the left hip on to the right were 0.067, 0.063, and 0.077 g cm-2. The absolute error of predicting one hip from the other was not a function of BMD and thus the relative error increases with lower BMD values. The relative errors were 17%, 8%, and 7% for BMDs of 0.4, 0.8, and 1.0 g cm-2, respectively. The interobserver variability was small, with an r value of .96 and an SE of the estimate value of 0.036 g cm-2. The relative error in the mild-to-moderate osteoporosis categories was 2.5 times the precision of the instrument, indicating that the asymmetry of BMD is due to real differences between hips. Therefore the BMD of one hip cannot be used to predict that of the other with sufficient accuracy to discriminate clinically relevant trends in BMD.