Maternal AsX and DHA supplementation modulate the body weight of pups at birth and average litter size
There was no significant effect of pre, post, and perinatal undernourishment on average number of pups born per dam. However, pups’ weight at birth was significantly reduced by maternal undernutrition, and this significant body weight depletion remained throughout the study. The body weight of the pups from all the experimental groups were measured at birth and throughout the lactation and adolescent period i.e., from PD-1 to PD-56. The same trend was observed in brain weight, measured at the end of the experiment (supplementary table 1). The birth weight of the pups born to pre and perinatally undernourished dams showed significant decrease compared to pups born to the dams fed with control diet. AsX and DHA supplementation to pre and perinatally undernourished dams significantly increased the body weight of the pups at birth (Fig. 2A, C; C vs PreUN: p < 0.0001; C vs PeriUN: p < 0.0001; PreUN vs PreUN-AD: p < 0.0001; PeriUN vs PeriUN-AD: p < 0.0001, one-way ANOVA with Tukey’s multiple comparison). Even though, pups born to the undernourished dams supplemented with AsX and DHA showed lowered body weight at birth compared to controls (C vs PreUN-AD: p = 0.0018; C vs PeriUN-AD: p = 0.0177), when compared to undernourished groups, AsX and DHA supplemented groups showed significantly increased birth weight (PreUN vs PreUN-AD: p < 0.0001; PeriUN vs PeriUN-AD: p < 0.0001). Compared to control group pre, post, and perinatally undernourished offsprings showed significantly reduced body weight during lactation (PD-1 to PD-21) and postweaning period (PD-21 TO PD-57) indicating the effect of undernourishment on the body weight. AsX and DHA supplementation in all the three groups promoted the body weight. The two-way ANOVA with repeated measures revealed the significant effects for the treatment, time and time × treatment interaction. Tukey’s multiple comparison showed significant difference between the groups from PD-1 to PD-57(Fig. 1D; Time effect: F [ 2.814, 84.41] = 2288; p < 0.0001; Treatment effect: F [ 5, 30] = 54.52; p < 0.0001; Interaction effect: F [ 35, 210] = 10.84; P < 0.0001, E; Time effect: F [ 3.322, 99.67] = 2238; p < 0.0001; Treatment effect: F [ 5, 30] = 90.39; p < 0.0001; Interaction effect: F [ 35, 210] = 13.28; p < 0.0001, F; Time effect: F [ 2.871, 86.13] = 2420; p < 0.0001; Treatment effect: F [ 5, 30] = 92.50; p < 0.0001; Interaction effect: F [ 35, 210] = 18.87; p < 0.0001). Further, we also observed that the brain weight of the offspring’s were significantly reduced in pre and perinatally undernourished groups. Maternal AsX and DHA supplementation enhanced brain weight (Supplementary Figure S1).
Maternal AsX and DHA supplementation modulates cognitive decline following acute stress caused by maternal undernutrition
AsX and DHA supplementation enhances recognition memory
To evaluate the effect of pre, post and perinatal undernutrition on recognition memory, and to asses weather maternal AsX and DHA supplementation ameliorates the recognition memory, we performed NOR test. It included 3 sessions’ habituation, training and test (Fig. 2A). Discrimination Index was calculated as a measure of discrimination between familiar and novel objects 31. NOR test utilizes the innate preference of rodents to explore novel objects over the familiar ones in the environment. Repeated exposure to an object decreases exploration time as the object becomes familiar, and rodents will spend more time exploring the novel. Several previous studies have reported that protein malnutrition during the gestation and lactation period impairs recognition memory 32 33 34.
Total exploration time in the training session did not significantly differ between the objects and experimental groups (Supplementary Figure S2). Pre, post, and perinatal dietary supplementation of AsX and DHA enhanced recognition memory which was impaired by maternal undernutrition. In this study, we observed significantly decreased DI in undernourished groups compared to drug control (Fig. 2B; DC vs. PreUN: p = 0.0006 0.0038, Fig. 3C; DC vs. PostUN: p = 0.0006, Fig. 3D; DC vs. PeriUN: p = 0.0020). With maternal undernourishment, dietary supplementation of AsX and DHA showed enhanced DI compared to the undernourished group. Perinatal undernourishment and AsX and DHA supplemented group showed significantly increased DI (Fig. 2D; PeriUN vs. PeriUN-AD: p = 0.0243) when compared to the perinatally undernourished group indicating, AsX and DHA supplementation improved recognition memory in adult rats which were undernourished during their pre, post and perinatal period.
Maternal AsX and DHA supplementation enhance learning in undernutrition-induced learning impairment in partially baited radial arm maze tasks
All the experimental rats were given 12 days of training before retention. Food rewards were placed at the end of the 1st and 2nd arms as an adjacent choice and at the end of the 5th and 7th arms as a choice. Two-way ANOVA with repeated measures revealed the significant effect of treatment and interaction on experimental groups (Fig. 3B; Interaction effect: F (25, 150) = 1.827, p = 0.0147; Treatment effect: F (5, 30) = 6.201, p = 0.0005; Fig. 3C Interaction effect: F (25, 150) = 2.747, P < 0.0001; Treatment effect: F (5, 30) = 5.455, p = 0.0011; Fig. 3D Interaction effect: F (25, 150) = 2.212, p = 0.0018; Treatment effect: F (5, 30) = 5.931, p = 0.0006). The percentage of correct choice (%CC) increased, and all the experimental groups showed a standard learning curve during acquisition. Maternal undernourishment resulted in a significant decrease of %CC in B-5 and B-6 (B-5; C vs. PreUN: p = 0.015, C vs. PostUN: ns, C vs. PeriUN: p = 0.028; B-6; C vs. PreUN: P = 0.008, C vs PostUN: p = 0.009, C vs PeriUN: p = ns). AsX-DHA supplementation significantly declined learning deficit only in B-6 in PostUN-AD rats (B-6; PostUN vs PostUN-AD: p = 0.011). These results suggested differential learning only in Blocks 5 and 6. Throughout the acquisition, AsX-DHA-supplemented groups showed better learning than undernourished groups (Fig. B-D).
Control rats attained 80% correct choice criteria after 11 days of training. At 12 days of training, Control rats reached 83.92 ± 9.18, Drug control rats achieved 82.67 ± 6.56, and Vehicle control rats reached 80.08 ± 5.69% correct choice, whereas undernourished rats failed to reach the criteria of learning (80% correct choice) at 12 days of training (Fig. 3E-G). Undernourished groups showed a significantly decreased %CC on day 12th compared to control groups (C vs. PreUN: p = 0.030; C vs. PostUN: p = 0.001 and C vs. PeriUN: p < 0.0001). Pre and Post AsX-DHA supplemented groups did not show significantly increased %CC compared to the undernourished group on day 12th, indicating, AsX-DHA supplementation did not boost the learning in pre and postnatally undernourished rats. But perinatal supplementation of AsX-DHA improved learning by showing significantly enhanced %CC compared to PeriUN rats (PeriUN vs PeriUN-AD: p = 0.002).
Effect of maternal undernourishment and ASX-DHA supplementation on retention
To assess the effect of maternal undernutrition and AsX-DHA on the retention of learned information, we tested retention after ten days’ post-acquisition. During retention, subjects demonstrated a significantly decreased %CC, a significantly increased number of reference memory errors (RMEs), and the number of working memory errors (WMEs) in PreUN and PeriUN groups.
Figure 4: A) Representative track reports of experimental groups. The effect of undernourishment and AsX-DHA supplementation on %CC (B-D), number of reference memory errors (E-G), and number of working memory errors (H-J) in the retention test of partially baited radial arm maze task. In all the cases One-Way ANOVA with post hoc Tukey’s test was used. Values were represented as mean ± SEM (n = 6). Asterisks were used to represent significant differences, *p < 0.05, **p < 0.01.
Similar to the acquisition, the retention test also observed a significant decline in %CC in VC-UN, Pre-UN, and PeriUN groups [Fig. 4B; F (5, 30) = 8.883, p < 0.0001, Fig. 4D; F (5, 30) = 8.635, p < 0.0001]. Compared to controls, PreUN and PeriUN groups showed decreased %CC (C vs. Pre-UN: p = 0.0008, C vs Peri-UN: p = 0.001). AsX-DHA-supplemented groups did not show any impairment in the retention test. The positive effect of AsX-DHA supplementation on %CC in the retention test was statistically significant in the PeriUN-AD group (PeriUN vs. PeriUN-AD: p = 0.006). Though we observed lower %CC in PostUN subjects on the 12th day of training, this trend has not continued the retention.
Significant increase in the number of RMEs have been observed in undernourished groups [Fig. 4E; F (5, 30) = 7.097: p = 0.0002, C vs. PreUN: p = 0.004; Fig. 4G, F (5, 30) = 5.977: p = 0.0006, C vs. Peri-UN: p = 0.02]. Supplementation ameliorated impairment in PreUN and PeriUN groups as we have observed no significant increase in the number of RMEs in the ASX-DHA-supplemented group compared to controls (Fig. 4E-G). Working memory impairment in undernourished rats was reinforced by AsX-DHA supplementation (Fig. 4H-J). Postnatal undernourishment and AsX-DHA supplementation did not affect the number of WMEs. But Pre and Perinatal undernourishment negatively affected WMEs which was restored by AsX-DHA supplementation (C vs. PreUN: p = 0.006; C vs. Peri-UN: p = 0.006; Peri-UN vs. PeriUN-AD: p = 0.006).
Effect of maternal undernourishment and AsX-DHA supplementation on BDNF, CREB, NT-3 and UCP-2 gene expression
It has been shown that expression of BDNF, NT-3, CREB and UCP-2 genes modulate the synapse development, transmission and plasticity. We wanted to investigate whether pre, post and perinatal undernourishment alters the long-term expression of these genes in the hippocampus. Transcriptomic changes of these factors were studied by quantifying mRNA levels using real time RT-PCR.
Figure 5 displays the hippocampus's relative expression of BDNF, NT-3, CREB, and UCP-2. In pre, post, and perinatally undernourished rats, BDNF mRNA levels decreased significantly when compared to their controls (Fig. 5A-C; C vs. PreUN: p = 0.0003; C vs. PostUN: p < 0.0001; C vs. PostUN: p = 0.009). AsX-DHA supplemented groups showed significantly higher levels of BDNF compared to their undernourished counterparts except for PostUN-AD group (Fig. 5A-C; PreUN vs. PreUN-AD: p = 0.001; PeriUN vs. PeriUN-AD: p = 0.005). Only PeriUN rats showed significantly lower levels of NT-3 compared to controls and PeriUN-AD rats, but in the case of pre and postnatal conditions, AsX-DHA supplementation increased the NT-3 expression, which was not significant (Fig. 5D-F; C vs PeriUN: P = 0.028; PeriUN vs. PeriUN-AD: p = 0.001). Undernourishment resulted in significantly lower levels of CREB and UCP-2 mRNA, and in PreUN-AD and PeriUN-AD conditions significant increase in CREB and UCP-2 mRNA levels was observed (Fig. 5G-I; C vs PreUN: p < 0.0001; C vs. PostUN: p < 0.0001; C vs. PeriUN: p < 0.0001; PreUN vs. PreUN-AD: p < 0.0001; PeriUN vs. PeriUN-AD: p = 0.004, Fig. 5J-L; C vs PreUN: p = 0.0004; C vs PostUN: p = 0.0003; C vs PeriUN: p = 0.0001; PreUN vs. PreUN-AD: p = 0.003).
Using the immunohistochemistry technique, we studied the expression of synapsin-1 (Fig. 6A) and PSD-95 (Fig. 7A) in the hippocampus's CA1, CA2, CA3, and DG regions. The results showed that the percentage of positive areas significantly differed between the experimental groups. The expressions of synapsin-1 and PSD-95 decreased in undernourished groups compared to control and AsX-DHA treated groups at CA1, CA2, CA3, and DG regions, respectively. CA1-Synapsin 1 [Fig. 6B. Control vs PreUN: p = 0.0003, PreUN vs PreUN-AD: p = 0.022; Control vs PostUN: p = 0.0002, PostUN vs PostUN-AD: p = 0.039; Control vs PeriUN: p < 0.0001, PeriUN vs PeriUN-AD: p < 0.0001], CA2-Synapsin-1 [Fig. 6B. Control vs PreUN: p < 0.0001, PreUN vs PreUN-AD: p = 0.0001; Control vs PostUN: p < 0.0001, PostUN vs PostUN-AD: p = 0.0014; Control vs PeriUN: p < 0.0001, PeriUN vs PeriUN-AD: p = 0.0006], CA3-Synapsin-1 [Fig. 6B. Control vs PreUN: p < 0.0001, PreUN vs PreUN-AD: p = 0.0007; Control vs PostUN: p < 0.0001: Control vs PeriUN: p < 0.0001, PeriUN vs PeriUN-AD: p < 0.0001], DG-Synapsin-1 [Fig. 6B. Control vs PreUN: p < 0.0001, PreUN vs PreUN-AD: p = 0.0002; Control vs PostUN: p < 0.0001; Control vs PeriUN: p < 0.0001, PeriUN vs PeriUN-AD: p = 0.032].
CA1-PSD-95 [Fig. 6B. Control vs PreUN: p < 0.0001, PreUN vs PreUN-AD: p < 0.0001; Control vs PostUN: p < 0.0001, PostUN vs PostUN-AD: p = 0.015; Control vs PeriUN: p < 0.0001, PeriUN vs PeriUN-AD: p = 0.0001], CA2-PSD-95[Fig. 6B. Control vs PreUN: p < 0.0001, PreUN vs PreUN-AD: p < 0.0001; Control vs PostUN: p < 0.0001; Control vs PeriUN: p < 0.0001, PeriUN vs PeriUN-AD: p = 0.0003] CA3-PSD-95 [Fig. 6B. Control vs PreUN: P < 0.0001, PreUN vs PreUN-AD: p < 0.0001; Control vs PostUN: p < 0.0001, PostUN vs PostUN-AD: p < 0.0001; Control vs PeriUN: p < 0.0001, PeriUN vs PeriUN-AD: p < 0.0001], DG-PSD-95 [Fig. 6B. Control vs PreUN: p = 0.024; Control vs PostUN: p < 0.0001, PostUN vs PostUN-AD: p < 0.0001; Control vs PeriUN: p < 0.0001, PeriUN vs PeriUN-AD: p = 0.007].
However, our study showed that AsX-DHA-treated animals did not show a significantly increased percentage of positive area at CA3 and DG compared to the PostUN group. Similarly, PSD95 in AsX-DHA treated animals also did not show a significantly increased portion of positive area at CA2 in the PostUN group and DG in the PreUN group.
Correlation OF behaviour with hippocampal protein levels
The Pearson correlation between behavioural parameters (% CC, RME, WME, and DI) with the expression of synapsin-1 and PSD-95 in hippocampal subfields, CA1, CA2, CA3, and DG revealed a significant positive correlation between %CC retention and DI with synapsin-1 and PSD-95 expression and a significant negative correlation between RME and WME with synapsin-1 and PSD-95 (Table no. 1). On the other hand, the expression of PSD-95 in DG did not show a significant correlation with behavioural parameters (PSD-95: %CC retention: r = 0.504, P = 0.137; RME: r=-0.564, P = 0.318; WME: r=-0.367, P = 0.296; DI: r = 0.627, P = 0.052).