Loss of maternal calcitriol reversibly alters early offspring growth and skeletal development in mice

Abstract

Ablation of Cyp27b1 eliminates calcitriol but does not disturb fetal mineral homeostasis or skeletal development. However, independent of fetal genotypes, maternal loss of Cyp27b1 altered fetal mineral and hormonal levels compared to offspring of WT dams. We hypothesized that these maternal influences would alter postnatal skeletal development. Cyp27b1 null and WT females were mated to bear only Cyp27b1+/- offspring. Forty-eight hours after birth, pups were cross-fostered to dams of the same or opposite genotype that bore them. Maternal and offspring samples were collected on days 21 (weaning) and 42. Offspring measurements included minerals and hormones, BMC by DXA, ash weight and mineral content, gene expression, 3-point bending tests, and microCT. Maternal lactational behavior was evaluated. Milk was analyzed for nutritional content. At day 21, offspring fostered by nulls, independent of birth dam, had ~20% lower weight, BMC, ash weight, and ash calcium than pups fostered by WT dams. Adjustment for body weight accounted for the lower BMC but not the lower ash weight and ash calcium. Hormones and serum/urine minerals did not differ across offspring groups. Offspring fostered by nulls had shorter femurs and lower cortical thickness, mean polar moment of inertia, cortical area, trabecular bone volume, and trabecular number. Dam lactational behaviors and milk nutritional content did not differ between groups. At day 42, body weight, ash weight, lengths, BMC, and tibial bone strength were no longer different between pups fostered by null vs WT dams. In summary, pups fostered by Cyp27b1 nulls, regardless of birth dam, have proportionately smaller skeletons at 21 d, impaired microstructure, but normal mineral homeostasis. The skeletal effects are largely recovered by day 42 (3 wk after weaning). In conclusion, maternal loss of calcitriol impairs early postnatal cortical bone growth and trabecular bone mass, but affected offspring catch up after weaning.

Keywords: 1α-hydroxylase; Cyp27b1; calcitriol; calcium; mineralization; neonate; parathyroid hormone; phosphorus; skeleton.

Figure 1

Mating and cross-fostering schemes. (A) WT dams were mated to Cyp27b1 null males, while in B, Cyp27b1 null dams were mated to WT males. All dams bore only Cyp27b1^+/− pups. (C) On day 2 after birth, pups born of WT dams were cross-fostered to Cyp27b1 null dams and vice versa. The differing colors of dams and pups reflect that pups have been cross-fostered to dams of the opposite genotype from the birth dams. (D and E) Controls included cross-fostering pups of WT dams to WT dams, and pups of Cyp27b1 null dams to Cyp27b1 null dams. The similar colors of dams and pups reflect that the pups have been cross-fostered to dams of the same genotype as the birth dams. Images created with BioRender.com.

Figure 2

Skeletal parameters of offspring at day 21. Pups fostered by Cyp27b1 null dams had lower body weight (A), length (B), and ash weight (C); the difference in ash weight was attenuated but remained significant when corrected for body weight (D). Ash calcium and magnesium were reduced in pups fostered by Cyp27b1 nulls, while ash phosphorus content was not (EFG); however, ash calcium and phosphorus content were both lower in null→WT as compared to WT → WT pups. Whole body BMC was reduced in pups fostered by Cyp27b1 null dams (H), but the difference was eliminated after adjustment for body weight (I). Significant subgroup differences included that null→null pups had significantly lower body weight, body length, ash weight, ash calcium, ash magnesium, and BMC, than WT → WT pups. Data are shown as individual values, with mean and standard deviation. P-values are from 2-way ANOVA (bold) with Šidák post hoc tests.

Figure 3

MicroCT analysis of offspring femora at day 21. Femora of pups fostered by Cyp27b1 null dams had lower femur lengths (A), cortical thickness (B), cortical area (C), and mean polar moment of inertia (D). Trabecular number (E) and trabecular bone volume (F) were reduced, while there were no significant differences in trabecular thickness (G), separation (H), or periosteal perimeter (I). Data are shown as individual values, with mean and standard deviation. P-values are from 2-way ANOVA (bold) with Šidák post hoc tests.

Figure 4

Biomechanical testing of offspring tibias at day 21. Three weeks after birth, a 3-point bend test indicated a higher ultimate load in WT→WT tibias (A), higher displacement in null→WT tibias and in tibias of pups fostered by WT vs fostered by Cyp27b1 null dams (B), increased maximum strain in WT → WT tibias (C), and no difference in stiffness across groups (D). The tibias broke easily and many bent rather than breaking, as described in the Results section. Data are shown as individual values, with mean and standard deviation. P-values are from 2-way ANOVA (bold) with Šidák post hoc tests.

Figure 5

Serum calcifediol and IGF-I in pups at day 21. Serum 25(OH)D₃ was modestly but significantly higher in pups fostered by Cyp27b1 null dams as compared to pups fostered by WT dams (A). Serum IGF-I was higher in pups fostered by null dams vs fostered by WT dams; there was also a birth dam effect with lower IGF-I in WT→WT pups as compared to null→null pups (B). Data are shown as individual values, with mean and standard deviation. P-values are from 2-way ANOVA (bold) with Šidák post hoc tests.

Figure 6

Gene expression in pup duodena and kidneys at day 21. Offspring duodena showed no significant changes in the mRNA levels of Atpb2 (A), S100g (B), Trpv6 (C), Slc8a1 (D), and Kcnma1 (E). Offspring kidneys similarly showed no significant changes in the mRNA levels of Atpb2 (F), S100g (G), Trpv6 (H), whereas small but statistically significant changes in Slc8a1 (I) and Kcnma1 (J) mRNA levels were likely not physiologically important given their very low levels of expression. Data are shown as individual fold values relative to the housekeeping gene, with mean fold and standard deviation. P-values are from 2-way ANOVA (bold) with Šidák post hoc tests.

Figure 7

Maternal parameters at day 21. Consistent with our early study, Cyp27b1 null females were hypocalcemic (A) with low serum phosphorus (B), increased PTH (C), lower FGF23 (D), low calcitriol (E), and higher 25(OH)D₃ (F). IGF-I was variable with no significant differences among the groups (G). Data are shown as individual values, with mean and standard deviation. P-values are from t-tests.

Figure 8

Milk composition at day 10. Milk nutritional content was no different between Cyp27b1 null and WT, including calcium (A), protein (B), triglycerides (C), moisture (D), creamatocrit (E), fat (F), and energy (G). Data are shown as individual values, with mean and standard deviation. P-values are from t-tests.

Figure 9

Maternal lactational behavior. During a total of 6 d of continuous video surveillance for each dam with its litter, Cyp27b1 null and WT dams showed no differences in the percentage of time spent off the litter (A) or on the litter while active (B) or inactive (C). The inactive state, with the dam resting on the litter, is when the pups are most likely to obtain the most milk. Data are shown as individual values, with mean and standard deviation. P-values are from t-tests.

Figure 10

Resolution of offspring phenotype by day 42. Three weeks after weaning, and while consuming normal chow, there were no longer any between-group differences in body weight (A), body length (B), ash weight (C), and whole body BMC (D). A 3-point bend test on the tibias indicated no effects of the birth or foster dam on ultimate strength, displacement, maximum strain, and stiffness (E–H). Data are shown as individual values, with mean and standard deviation. P-values are from 2-way ANOVA (bold) with Šidák post hoc tests.