Fig. 2. Quantification of the phenotypic changes in hormone-producing cell populations following the mono- or bi-allelic loss of Arx and/or Pax4. P2-independent pancreata estimated to be of the same size were serially sectioned. n, number of pancreata analyzed for each genotype. Every tenth section was stained as indicated and the numbers of positive cells were counted and compared with the total islet cell content (estimated on adjacent sections using a mixture of antibodies raised against the different endocrine hormones). Data are shown as percentage±s.e.m. of hormone-positive cells contributing to the total endocrine population. On average, the lack of one Arx and/or Pax4 allele does not alter the endocrine cell content. Overall, the islet-cell alterations observed in Arx^- and Arx^-Pax4^+/-, or in Pax4^-/- and Pax4^-/- Arx^+/-, appear to be similar. However, the loss of a single Pax4 allele in Arx^- Pax4^+/- animals results in a significant reduction of the ß-cell content together with an increase of the -cell population when compared with Arx mutants. Note, in mice depleted in Arx and Pax4, the loss of the insulin- and glucagon-expressing cell populations, and the substantial increase in the numbers of somatostatin- or PP-producing cells. It should be underlined that the total endocrine cell content is not statistically modified in all of the genotypes analyzed. Multiple comparisons of the data obtained for each endocrine population in each genotype were processed with a single-factor ANOVA coupled to Newman-Keuls test using the wild-type genotype as a reference (^*P<0.05, ^**P<0.01). A similar study was performed comparing subtype-specific cell numbers between Arx^-Pax4^+/- mice and Arx mutants (P<0.05).