Developmental vitamin D-deficiency produces autism-relevant behaviours and gut-health associated alterations in a rat model

Abstract

Developmental vitamin D (DVD)-deficiency is an epidemiologically established risk factor for autism. Emerging studies also highlight the involvement of gut microbiome/gut physiology in autism. The current study aims to examine the effect of DVD-deficiency on a broad range of autism-relevant behavioural phenotypes and gut health. Vitamin D deficient rat dams exhibited altered maternal care, DVD-deficient pups showed increased ultrasonic vocalizations and as adolescents, social behaviour impairments and increased repetitive self-grooming behaviour. There were significant impacts of DVD-deficiency on gut health demonstrated by alterations to the microbiome, decreased villi length and increased ileal propionate levels. Overall, our animal model of this epidemiologically validated risk exposure for autism shows an expanded range of autism-related behavioural phenotypes and now alterations in gut microbiome that correlate with social behavioural deficits raising the possibility that DVD-deficiency induced ASD-like behaviours are due to alterations in gut health.

Fig. 1. DVD-deficiency alters both maternal and pup behaviour.

Maternal behaviour was video-recorded and observed from P2 to P6. A DVD-deficient dams showed increased licking/grooming (LG) compared to control dams (F_1,21=2.625, p < 0.05). B There was no effect of DVD-deficiency on arched-back nursing. The circles represent individual dams. C, D Maternal retrieval of the pups was tested at two time points (P7 and P9) immediately following USV measurement. Data reported as latencies to retrieve the first, second, third, fourth and fifth pup (in seconds). Analysis was done using a mixed effects model. Overall DVD-deficient dams retrieved their pups slower than the control dams at P7 (F_1,90=6.950, p = 0.0099), whereas this was reversed at P9 (F_1,85=3.963, p = 0.049). DVD dams n = 12, CON dams n = 8 at P7. DVD dams n = 11, CON dams n = 8 at P9. *p < 0.05, **p < 0.01. Error bars show SEM. E–H Measurement of pup’s USVs at two time points (P7 and P9) in a brief isolation from their dams. No significant differences were found in the four USV parameters at P7. However, DVD-deficient pups exhibited greater number (E) and longer duration (F) of calls at P9 compared to control pups. There were no differences in the call amplitude (G) and call frequency (H) at P9 between DVD-deficient and control pups. At P9, male pups emitted longer duration of calls (F) and louder calls (G) compared to female pups. DVD males n = 89, DVD females n = 93, CON male n = 64, CON females n = 59. *p < 0.05. Black lines differences by diet, red lines difference by sex. CON Control, DVD DVD-deficient. The lower boundary of the box plot indicates 25th percentile, the middle line median and upper boundary indicates 75th percentile.

Fig. 2. DVD-deficiency decreases social play and increases stereotyped behaviour in adolescent rats.

A–C Social play behaviour was measured in P35 offspring. (A)DVD-deficient offspring showed significantly less pouncing compared to control offspring (F_{1, 94} = 7.328, p = 0.008). No differences were found in (B) total play duration or (C) latency to interact. There was no main effect of sex on frequency of pouncing (F_1,94 = 0.006, p = 0.939), total play duration (F_1,94 = 0.074, p = 0.787) and latency to interact (F_1,94 = 1.123, p = 0.292). CON males n = 24, CON females n = 24, DVD males n = 24, DVD females n = 24. D, E Self-grooming behaviour was analysed from social play recordings. D DVD-deficient rats showed increased bouts of self-grooming compared to controls (F_1,191 = 11.595, p = 0.001). E There was no difference in the duration of self-grooming between DVD-deficient and control animals (F_1,191 = 2.823, p = 0.095). CON males = 48, CON females = 48, DVD males = 48, DVD females = 47. **p < 0.01.

Fig. 3. Alpha and beta diversity measures in the DVD-deficient and control offspring.

All results are presented at genus level. A–D Alpha diversity. Observed species, Simpson’s index, Shannon index and Chao1 were not significantly different between DVD-deficient and control animals. E, F Measures of Beta diversity between DVD-deficient and control offspring. Differences in the microbial composition between DVD-deficient and control animals were determined by Bray Curtis (E) and Weighted Unifrac (F). Bray Curtis (ANOSIM, R = 0.196, P = 0.001), Weighted Unifrac (ADONIS, R² = 0.048, P = 0.001). CON n = 16, DVD n = 16.

Fig. 4. Spearman correlations between gut bacterial genera with social play behaviour (frequency of pouncing).

The top correlations for any genus with pouncing behaviour in either DVD-deficient or control offspring are depicted. There was a negative correlations between Phascolarctobacterium (R = −0.77, p = 0.015) (A) and Paracteroides (R = −0.79, p = 0.00044) (B) with the frequency of pouncing in DVD-deficient group, but not in control. The genus Parasutterella was also negatively correlated with frequency of pouncing in controls but not in DVD-deficient group (R = −0.64, p = 0.0081) (C). CON n = 16, DVD n = 16, *p < 0.05. The small circles represent individual samples. The p-values shown for each bacterium genus are nominally significant values, not adjusted for multiple correction.

Fig. 5. Measurement of three SCFAs in the ileum of the offspring.

DVD-deficiency increased propionate (A) levels but did not affect the levels of acetate (B) or butyrate (C). CON male n = 8, CON female n = 8, DVD male n = 8, DVD female n = 8. *p < 0.05.

Fig. 6. DVD-deficiency decreases gut villi length but has no effect on gut inflammatory cytokines.

A–D Histological examination of jejunal sections of P35 offspring. Jejunal tissues were collected four hours after the animals were injected with either saline or poly(I:C). A A 5 µm thick section was prepared by Swiss roll technique. B DVD-deficiency resulted in decreased villi length, but there was no effect of poly(I:C) or diet × poly(I:C) interaction on villi length. DVD-deficiency had no effect on lymphocyte (C) or goblet cell number (D). SAL = Saline, POL = Poly(I:C). CON SAL n = 8, CON POL n = 8, DVD SAL n = 6, DVD POL n = 8. Only males were used for the gut histology analysis. **p < 0.01. E–H Inflammatory cytokines were measured in the P35 offspring ileum and sera. As expected, compared to a vehicle injection, Poly(I:C) clearly elevated the levels of both IL-6 and TNF-α in ileum and sera (p < 0.05). There was no main effect of DVD-deficiency on the levels of IL-6 (E, F) and TNF-α (G, H) in ileum and sera. However, there was interaction of diet × Poly(I:C) on TNF-α levels in ileum (F_1,64 = 7.394, p = 0.009). SAL = Saline, POL = Poly(I:C). CON SAL n = 16, CON Poly(I:C) n = 16, DVD SAL n = 16, DVD PoL n = 16. *p < 0.05.