Altered Developmental Trajectory in Male and Female Rats in a Prenatal Valproic Acid Exposure Model of Autism Spectrum Disorder

Early motor and sensory developmental delays precede Autism Spectrum Disorder (ASD) diagnosis and may serve as early indicators of ASD. The literature on sensorimotor development in animal models is sparse, male centered, and has mixed findings. We characterized early development in a prenatal valproic acid (VPA) model of ASD and found sex-specific developmental delays in VPA rats. We created a developmental composite score combining 15 test readouts, yielding a reliable gestalt measure spanning physical, sensory, and motor development, that effectively discriminated between VPA and control groups. Considering the heterogeneity in ASD phenotype, the developmental composite offers a robust metric that can enable comparison across different animal models of ASD and can serve as an outcome measure for early intervention studies.

Keywords: Autism; Developmental milestones; Sex differences; VPA; Animal models; Composite score

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s10803-022-05684-y.

The brain undergoes a series of complex, critical developmental stages over a long period of time starting from the embryonic period, through puberty, into young adulthood. The initial developmental period (birth to 5 years) is especially crucial as the child starts to interact with the environment and rapidly develops sensory, motor, social, linguistic, and cognitive skills (Moodie et al., [81]). The temporal sequence in which these skills appear is predictable, making it possible to identify normative milestones in the early developmental journey of a child's growth. Timely acquisition of these milestones indicates healthy development of a child, and deviation from this normal trajectory signals potential for neurodevelopmental disorders such as autism spectrum disorder (ASD), and other adverse behavioral consequences (Davidovitch et al., [28]; R. J. Landa et al., [66]; Sheldrick et al., [106]).

Retrospective studies in children with ASD using parent reports and old home-videos have reported deviations from normal developmental trajectories (Grace T Baranek, [8]; Barbaro & Dissanayake, [10]; Osterling & Dawson, [86]). Similarly, prospective studies in children at high risk for ASD (younger siblings of ASD diagnosed children) have demonstrated delayed skill acquisition in sensory, motor, and social domains (Davidovitch et al., [28]; Landa & Garrett-Mayer, [65]; Ozonoff et al., [87]; Zwaigenbaum et al., [131]). Both gross and fine motor skills have been found to be delayed in children later diagnosed with ASD (Bolton et al., [17]; Lemcke et al., [68]). During early development, atypical responsiveness to sensory inputs across auditory, visual, and somatosensory modalities have been observed in children with ASD or at high risk for ASD (Grace T Baranek, [8]; Leekam et al., [67]; McCormick et al., [77]; Wolff et al., [128]; Zwaigenbaum et al., [131]). Furthermore, delayed acquisition of early motor and sensory milestones have been shown to predict subsequent cognitive and behavioral outcomes in children with ASD (G. T. Baranek et al., [9]; Bedford et al., [13]; Gernsbacher et al., [43]; Thye et al., [118]; Uljarević et al., [120]). Importantly, intervention programs for children with ASD are likely to be more effective when started early in the developmental period (Dawson, [29]; Rebecca J. Landa, [64]). In summary, there is mounting evidence for the need to study early developmental trajectory in ASD, i.e., to understand how ASD unfolds from birth towards the appearance of core symptoms (Jones et al., [55]; Shen & Piven, [107]).

ASD has high clinical and biological heterogeneity that arises, at least in part, due to the large number of genetic, environmental, and epigenetic factors and their interactions that may contribute to ASD etiology (Banerjee et al., [7]; Bölte et al., [16]; Cheroni et al., [25]; Kim & Leventhal, [59]). Animal models are typically based on experimental manipulation of one of the known genetic and environmental risk factors and thus provide opportunities to delineate specific pathways in ASD pathogenesis (Ergaz et al., [36]; Halladay et al., [47]). For example, mutations of genes related to synaptic scaffolding and neuronal cell adhesion such as SHANK3, CNTNAP2, and NLGN3 genes affect synaptic development and impact excitatory-inhibitory neurotransmission. Mutations in PTEN, and TSC1/ TSC2 inhibit activation of the mechanistic target of rapamycin complex and thus impact neuronal protein synthesis and growth regulation(Chen et al., [24]; Möhrle et al., [80]). Among environmental risk factors, maternal immune activation due to polyinosine:cytosine (poly(I:C), viral mimic) and lipopolysaccharide (LPS, bacterial mimic), triggers microglial activation, elevations in pro-inflammatory cytokines and transcriptome dysregulation (Knuesel et al., [60]; Lombardo et al., [71]; Patterson, [88]). Prenatal exposure to the drug valproate (VPA, a GABA agonist that also acts as a histone deacetylase inhibitor) acts via epigenetic mechanisms to influence the expression profile of different genes involved in cell proliferation and differentiation (Favre et al., [38]; Kataoka et al., [56]; Nicolini & Fahnestock, [84]). Behavioral phenotyping of these animal models using a battery of assays can link these causal mechanisms to specific domains of dysfunction found in ASD and facilitate the development of targeted therapeutic strategies (Möhrle et al., [80]; Michela Servadio et al., [105]; Silverman et al., [109]). Further, characterization of early developmental trajectories in these animal models can provide clues towards a mechanistic understanding of the emergence of ASD phenotype.

In the present study, we assessed early developmental trajectory in the VPA model. The VPA model mimics ASD at multiple levels, starting from the behavioral alterations seen in ASD, but also ranging from the molecular and cellular levels to anatomical and circuit level atypicalities. The VPA model has been well-validated for core ASD features—reduced social interaction, increased stereotyped behavior—as well as additional features such as presence of hypersensitivity, enhanced anxiety, inattention, hyperactivity, and sleep disturbances (Chaliha et al., [22]; Cusmano & Mong, [26]; K. Markram, Rinaldi, Mendola, Sandi, & Markram, [76]; Schneider & Przewłocki, [101]). At the molecular level, prenatal VPA exposure leads to reduced expression of neuroligin (NLGN, a key ASD risk gene) mRNA in somatosensory cortex and hippocampus (Kolozsi et al., [61]). Neuroanatomical hallmarks of VPA exposure in rodents include lower Purkinje cell counts in the cerebellum (widely observed in postmortem ASD brains), as well as changes in neuronal density and neuronal cell count in prefrontal, somatosensory, and motor cortices, regions impacted in ASD (Ingram et al., [54]; Nicolini & Fahnestock, [84]; Roullet et al., [97]; R. Zhang et al., [129]). There is robust evidence implicating excitatory-inhibitory (E/I) imbalance in ASD pathogenesis and have been documented in the VPA model in terms of both increased glutamatergic as well as decreased GABAergic signaling (Gogolla et al., [44]; Rinaldi et al., [93]; Rubenstein & Merzenich, [99]). Finally, at the circuit level, altered connectivity and excitability in cortical microcircuits have been reported in the VPA model (Rinaldi et al., [94], [95]).

Animal models can never recapitulate the full spectrum of human condition, let alone span the heterogeneity found in ASD. Nevertheless, the VPA model mimics ASD characteristics remarkably well across many different levels, making it an excellent model for studying ASD pathophysiology (Mabunga et al., [73]; H. Markram, Rinaldi, & Markram, [75]; Roullet et al., [97]).

In the early development literature however, very few studies have carried out a detailed examination of early development in the VPA model (Hou et al., [52]) as most studies have checked a few isolated developmental milestones as part of a larger body of work (for example, Al Sagheer et al., [1]; Dobrovolsky et al., [33]; Scheggi et al., [100]). Another limitation is that most of these studies have focused exclusively on development in VPA males. ASD is more prevalent among males and there are studies suggestive of sex-specific differences in the acquisition of early developmental milestones (Carter et al., [20]; Harrop et al., [49]; Messinger et al., [78]).

We carried out a detailed evaluation of the ontogeny of developmental milestones in both male and female VPA rats. We systematically characterized the early developmental trajectory (from postnatal day 4 till weaning) of VPA and control rats using a comprehensive test battery spanning physical, sensory, and motor developmental milestones. Finally, considering the heterogeneity of the ASD phenotype, we developed a rodent developmental composite score by incorporating all test readouts into a single score and evaluated the efficacy of this overall developmental profile in discriminating between VPA and control rats.

Experimental protocols were carried out with approval from the Institutional Animal Ethics Committee, at the National Institute of Mental Health and Neurosciences (NIMHANS), Bengaluru. Adult Sprague Dawley female and male rats (2–3 months old) were procured from the Central Animal Research Facility (CARF), NIMHANS. These rats were housed in standard home cages with corncob bedding and were maintained on 12:12 h light/dark cycle with lights on at 7:00 AM in the morning. Standard chow (Special Diets Services) and water were available ad libitum. Every effort was made to minimize the number of animals used, and to minimize their suffering.

Female rats were examined for their estrous phase by visual inspection of external genitalia prior to mating (Byers et al., [19]). Female rats, found to be ready or receptive for the mating (in the estrous or proestrous phase), were kept for overnight mating (1:2 male to female ratio), and pregnancy was determined by the presence of a vaginal plug on the following morning (designated as embryonic day E1). Weight gain in dams was checked throughout the gestation period to confirm the pregnancy. On E12.5, prenatal valproate (VPA) exposure was provided to treated dams (n = 8) using a single intraperitoneal (IP) injection of VPA (450 mg/kg sodium salt of valproic acid (NaVPA, Sigma), dissolved in 0.9% saline solution at a concentration of 100 mg/ml) while control dams (n = 8) received a single IP injection of saline solution (SAL). We have previously used the above protocol (VPA dose and treatment time window) and reported core ASD features such as reduced social interaction as well as additional features such as impaired sensorimotor gating and attentional atypicalities, in VPA rats (Anshu et al., [5]). Four female dams were housed together in a standard cage until E18 and subsequently, dams were housed individually in standard home cages and left undisturbed to raise their own litters until weaning. No additional enrichment was provided. Sex of rat pups was determined by examining the anogenital distance, which is larger in males as compared to females.

Experiments were carried out on male and female offspring of the VPA treated and control dams described above. We found complete fetal reabsorption (zero pups born) in 50% (4/8) VPA injected dams, and 0% (0/8) SAL injected dams. Complete fetal reabsorption has been reported to occur in the VPA model (Favre et al., [38]). All offspring from the four VPA injected dams that gave birth were used in the study. Accordingly, offspring from four SAL injected dams, with similar litter sizes, were used as controls. There was no difference in average litter size for the VPA and SAL groups used in the present study (t6 = 0, p = 0.873; SAL (n = 4): 6.5 ± 0.866; VPA (n = 4): 6.5 ± 0.957). See supplementary information (SI section 1.2 and Figure S1) for additional details of pregnancy outcomes in SAL and VPA rats. Overall, eight litters were used in the study, with a total of 52 offspring (12 SAL male, 14 SAL female, 18 VPA male, and 8 VPA female). Day of birth was considered postnatal day 1 (P1). Developmental assays were performed in SAL and VPA pups between postnatal day 4 (P4) and postnatal day 21 (P21, weaning) or until all pups achieved developmental milestones. To assess the overall growth of rat pups, body weight was measured daily during this period. Finally, locomotor activity was assessed in these pups using an open field task (Smirnov & Sitnikova, [111]; Subhadeep et al., [113]) on P21. See Fig. 1 for the experimental timeline.

Graph: Fig. 1Experimental Design. A schematic presentation to show the timeline of experiments performed on rat pups prenatally exposed to VPA. This is for illustration purposes and the distance between the two points on the developmental age line does not represent the actual age gap in days. E = Embryonic day, P = Postnatal day. Developmental milestones were tested between P4-P20. Body weights of pups were recorded from P4-P21. On P21, pups were tested in open field to assess their locomotor activity

Physical maturity and the ontogeny of sensory and motor milestones were assessed (Fig. 2) using an extensive rodent developmental test battery. All tests were conducted in the lights-on period and all behavioral assessments were performed by the first author. Each day, at the time of testing, rat pups from the same litter were taken out from their home-cage and transferred to a heated holding chamber. To avoid any order effect, for each test, rat pups were randomly removed one by one from the holding chamber, assessed for developmental milestones, and then returned to the holding chamber. Once all pups from a litter were tested, they were returned to their home-cage at the same time. There is a natural progression in terms of the age (postnatal day) at which different developmental milestones are achieved. To estimate the approximate time of onset, prior to actual data acquisition, developmental testing was performed on 3 control rat litters. To avoid unnecessary testing and to minimize the time of separation of pups from their mother, testing for each milestones began two days prior to the day of expected milestone achievement.

Graph: Fig. 2Images of development milestones testing. Physical milestones (A–E), Vibrissa placing (F), Cliff avoidance (G), Negative geotaxis turn (H), Limb Placing (I), Surface righting reflex (J), Grasp reflex (K), Bar holding test (L), Vertical screen test (M), Horizontal screen test (N), Negative geotaxis climb (O)

Rat pups are immature at birth, i.e., their eyes and ears are closed, and they are devoid of their fur. Appearance of these physical milestones reflects the physical maturity of pups. Development of the following physical characteristics was checked: opening of both eyes (eye opening), detachment of both pinnae from the cranium (pinnae detachment), opening of the ear canal (ear opening), protrusion of the upper incisors (incisor eruption), and appearance of body hair (fur development). The day of first appearance was noted as achievement of each milestone.

Sensory and motor reflexes were tested daily based on earlier protocols (Altman & Sudarshan, [3]; Heyser, [51]; Nguyen, Armstrong, & Yager, [83]) until the first positive response was observed and that day was considered as day of onset for that milestone. These reflexes appear at different stages of development, resulting in variable end dates for the tests.

• Righting reflex—surface righting: The pup was gently held and placed on a padded horizontal surface in supine position for 5 s and then released to check if the pup was able to turn over onto its belly, i.e., the prone position. If the pup took longer than 60 s to turn, the test was stopped, and the pup was turned to its normal prone position.
• Negative geotaxis—turn: The pup was placed head facing downwards, on the center of a 35 cm long inclined board (30 degrees from horizontal plane). The inclined board had a rough surface (i.e., enough friction) to prevent sliding and to support the pups when they tried to turn around. A pile of cotton was kept below the board to protect pups from injury in case of a fall during unsuccessful trials. The ability of the pup to successfully turn around 180° (to change its orientation from face downwards to face upwards on the inclined plane) within 60 s was recorded. This successful postural reaction is considered a negative geotaxis-turn.
• Negative geotaxis—climb: After turning is achieved on the negative geotaxis test, pups usually begin climbing up the inclined plane. If the pup successfully reached the top of the board within 60 s after the turn, it was considered a negative geotaxis-climb.
• Cliff avoidance: The 'cliff' was a cardboard platform set at a height of 30 cm above a table. A pile of cotton was kept at the bottom of the cliff as a precautionary safety measure. The pup was placed at the edge of the cliff such that its snout and forepaws were placed beyond the edge. The cliff avoidance response depends on the vestibular system and does not depend on visual information and thus can be present before the eyes are open. It was recorded if the pup successfully avoided the cliff by crawling away from the edge of the platform by moving backwards or by turning towards either side (maximum time 30 s).
• Hind limb placing response: The pup was held gently around the trunk and suspended in air. A thin wooden stick was used to gently touch the dorsum of the hind paw and it was checked if the pup withdrew that paw, raised it, and then placed it on the wooden stick. If both hind limbs showed the placing response, the milestone was recorded as achieved.
• Vibrissae placing response: The pup was held gently around the trunk and suspended in air. Vibrissae were slowly stroked once using the blunt end of a wooden toothpick. It was noted if the pup moved its head away and extended its forelimbs to grasp the stick.
• Grasp reflex: The pup was gently held by the nape of its neck making the pup immobile and relaxed. Each forelimb paw was gently stroked once with the blunt end of a wooden toothpick and the presence or absence of grasping response, i.e., whether the pup was able to grasp the toothpick, was recorded.
• Bar holding: Each pup was gently lifted by the nape of its neck and brought close to a thin wooden bar (3 mm diameter) suspended 30 cm above a table surface. The pup was positioned to allow it to grab the wooden bar with its forepaws. Once the pup held the bar tightly, the pup was released such that it was hanging from the bar. If the pup was able to hang onto the bar by holding it with both forepaws for 10 s, the test was marked as successful. If the pup fell immediately, two more trials were given. To prevent injury to the pups during fall, piles of cotton were kept on the table surface.
• Horizontal screen test (also called level screen test): The pup was placed on a horizontal metal wire screen (16 mesh) and its tail was gently pulled to drag it backwards. The ability of the pup to hold on to the horizontal screen and not get dragged behind was noted.
• Vertical screen test: Each pup was placed on the horizontal screen (16 mesh) and allowed to hold the mesh properly. The screen was then slowly rotated by 90° to make it vertical such that the pup's head was facing up due to the rotation. The climbing response of the pup on the vertical screen was checked for 60 s. To prevent pups from injury during fall, piles of cotton were kept on the table surface.

To determine the effect of prenatal VPA exposure on locomotor activity, SAL and VPA pups were tested in the open field test on P21 (Smirnov & Sitnikova, [111]). Each pup was gently placed in the center of an evenly illuminated (30 lx) rectangular open field apparatus (80 cm × 40 cm × 40 cm) and allowed to freely explore the apparatus for 5 min. At the end of the test, the pup was immediately removed from the apparatus and returned to its home-cage. Open field apparatus was cleaned with 70% alcohol and allowed to dry after each experiment. All sessions were recorded with a top mounted video camera and videos were tracked and analyzed offline for locomotor activity by using "animal tracker" software (Image J plugin). Total distance covered, and mean velocity were calculated to evaluate locomotor functions. Additionally, center and periphery regions of the apparatus were defined during offline tracking and time spent in periphery and time spent in center were calculated. If the pup covered less than 2 cm in one second, it was considered immobile. Total immobility time was also recorded.

We combined the results of all 15 tests spanning physical, sensory, and motor domains in our developmental test battery into a single developmental composite score. First, each test readout was converted into a Z-score to standardize it to a common scale with mean 0 and standard deviation of 1 (El-Kordi et al., [35]). To facilitate comparisons, standardization was done across all rats (both sexes and groups) by subtracting the mean and dividing by the standard deviation for that test readout. Then, the Z-scores for all tests were averaged to arrive at the developmental composite score for each rat. Additionally, Z-scores from the tests belonging to physical, sensory, and motor domains were averaged separately to arrive at domain specific composite scores— see SI Sect. 1.1 for details. A higher value on the developmental composite score denotes greater developmental delay. Thus, a developmental composite score of 0 indicates that the development milestones for that rat was representative of the overall group. A rat with a score of -1 would indicate that the rat had earlier development than the group mean, and a rat with a score of 1 indicates that the milestones for this rat were one standard deviation later than the mean for the overall group. Cronbach's alpha was used to assess the reliability (internal consistency) of the developmental composite score.

Binary logistic regression was used (El-Kordi et al., [35]) to determine if the developmental composite score could reliably discriminate between SAL and VPA rats. The data for each sex were separated and further split into a training set (80%) and a validation (or 'hold out') set (20%) using the caret package in R (R Core Team, [115]). The training set was used to fit a logistic regression model and fivefold cross-validation was carried out with five repeats to arrive at the cross-validated accuracy scores. Finally, the model performance was assessed using the validation sets for both males and females.

Shapiro-Wilks test was used to assess normality in the data. As milestones data did not pass the normality test, we used the nonparametric Mann–Whitney U test to assess any significant differences in developmental milestones between SAL and VPA treatment groups for males and females separately. Locomotor activity, and developmental composite score data met the normality assumption and accordingly, two-way ANOVA followed by Tukey's post hoc tests were carried out. A linear mixed model (LMM) was used for body weight-gain comparisons across the developmental time course. Subject was taken as a random effect parameter and Sex, Treatment and Postnatal Day were taken as fixed effects. We did not control for litter effects in our analyses. Data were checked for outliers (1.5 times the inter-quartile range or IQR) and extreme outliers (3 times the IQR). All reported findings were consistent with or without removal of outliers (if any), therefore, all results reported in the main text are based on the complete dataset. See SI Sect. 1.7 for additional details. R software version 3.6.0 (R Core Team, [115]) and the packages tidyverse, psych, ggpubr, caret, rstatix, nlme, phia, emmeans, Hmisc, kableExtra, corrplot were used for statistical analyses. Descriptive results were represented as mean ± SEM and statistical significance was set at p < 0.05. p < 0.1 was considered to be an indicator of a trend (Thiese et al., [116]).

All rats progressively gained weight (Fig. 3) throughout the early developmental period (from age P4-P22). VPA rats did not show any difference in body weight during the entire early developmental period as compared to control rats. However, there was a clear sex difference in body weight gain as females from both VPA and control groups were lighter than males from P16 onwards. Please SI Sect. 1.3 for statistical details.

Graph: Fig. 3Body weight across early postnatal developmental period. After first two postnatal weeks, females were lighter than males in both groups. There was no effect of prenatal VPA treatment on the body weight of pups. F: Female, M: Male. Data expressed as mean ± SEM. n = 18 VPA male pups, 8 VPA female pups, 12 SAL male pups and 14 SAL female pups. Sex differences within each treatment group (Male vs. Female): ### p <.001, # # p<.01, # p<.05, ^p <.1

Major physical landmarks of rodent development such as eye and ear canal openings, pinnae detachment, fur appearance, and incisor eruption were evaluated in the rat pups (Fig. 4).

Graph: Fig. 4Development of Physical Milestones. Both male and female VPA pups were late in pinnae detachment and incisors eruption as compared to sex matched SAL pups. F: Female, M: Male. Data expressed as box and whisker plots. n = 18 VPA male pups, 8 VPA female pups, 12 SAL male pups and 14 SAL female pups. Differences between treatment groups (VPA vs. SAL): ***p <.001, **p <.01, *p <.05

Comparisons between treatment groups revealed that as compared to controls, incisor eruption was delayed in VPA males (9.0 ± 0 days versus 8.41 ± 0.15 days, U = 45, p < 0.001) as well as females (9.0 ± 0 days versus 8.35 ± 0.13 days, U = 20, p = 0.004). Similarly, pinnae detachment was also delayed in both VPA males (6.11 ± 0.30 days versus 5.0 ± 0 days, U = 48, p = 0.002) and females (6.63 ± 0.53 days versus 5.0 ± 0 days, U = 21, p = 0.001). There were no differences between SAL and VPA rats in the onset of other physical milestones such as eye opening (males: U = 129, p = 0.315; females: U = 59.5, p = 0.813), ear canal opening (males: U = 105, p = 0.901; females: U = 50, p = 0.642) and fur appearance (males: U = 105, p = 0.901; females: U = 50, p = 0.642).

VPA pups showed significant developmental delays when tested for sensory and motor reflexes crucial for overall growth and development (Fig. 5 and Fig. 6). Among sensory milestones, as compared to control males, VPA males showed delayed ontogeny of negative geotaxis-turn (8.16 ± 0.27 days versus 6.41 ± 0.14 days, U = 15, p < 0.001), and hind limb placing response (6.5 ± 0.33 days versus 5.0 ± 0 days, U = 36, p < 0.001) but not vibrissa placing response (8.44 ± 0.39 days versus 7.83 ± 0.27 days, U = 106, p = 0.946). On the other hand, female VPA pups showed delayed onset of all the above reflexes (negative geotaxis-turn: 8.75 ± 0.49 days versus 7.35 ± 0.19 days, U = 24, p < 0.001; hind limb placing response: 6.62 ± 0.53 days versus 5.0 ± 0 days, U = 14, p < 0.001; vibrissa placing response: 9.37 ± 0.62 days versus 6.78 ± 0.39 days, U = 18, p = 0.008) in comparison to control females. Within treatment groups, comparisons showed a sex dependent difference in the ontogeny of negative geotaxis-turn as control females took more days to develop this reflex as compared to control males (7.35 ± 0.199 versus 6.41 ± 0.149 days, U = 140, p = 0.002) whereas VPA females and males did not show this difference (8.75 ± 0.491 versus 8.16 ± 0.271 days, U = 89, p = 0.334). Additionally, control females showed a trend towards earlier onset of vibrissa-placing as compared to control males (6.79 ± 0.395 versus 7.83 ± 0.271 days, U = 51, p = 0.074) but VPA females and males did not show this difference (9.38 ± 0.625 versus 8.44 ± 0.398 days, U = 95.5, p = 0.175).

Graph: Fig. 5Development of Sensory Milestones. Both male and female VPA pups showed delayed ontogeny of negative geotaxis and hind limb placing response as compared to sex matched SAL pups. Female VPA pups also showed delayed onset of vibrissa placing response as compared to female control pups. Additionally, SAL females were slower than SAL males in acquiring the negative geotaxis reflex and had a trend towards an earlier onset of the vibrissa placing response as compared to SAL males. It is noteworthy that VPA rats did not show these sex-differences. F: Female, M: Male. Data expressed as box and whisker plots. n = 18 VPA male pups, 8 VPA female pups, 12 SAL male pups and 14 SAL female pups. Differences between treatment groups (VPA vs. SAL): ***p <.001, **p <.01, *p <.05. Sex differences within each treatment group (Male vs. Female): ##p <.01, ^p <.1

Graph: Fig. 6Development of Motor Milestones. Female VPA pups showed delayed development of grasp reflex and horizontal screen test in comparison with female control pups. In contrast, male VPA pups showed delayed development in grasp reflex test, bar holding test and horizontal screen test as compared to male SAL pups, showing widespread motor delays in male VPA pups. There was also a sex difference as female control rats took more time to develop vertical screen test response than control males and this difference was not seen in VPA rats. F: Female, M: Male. Data expressed as box and whisker plots. n = 18 VPA male pups, 8 VPA female pups, 12 SAL male pups and 14 SAL female pups. Differences between treatment groups (VPA vs. SAL): ***p <.001, **p <.01, *p <.05. Sex differences within each treatment group (Male vs. Female): #p <.05

Among motor milestones, VPA males showed delayed onset of grasp reflex (9.27 ± 0.28 days versus 8.0 ± 0 days, U = 36, p < 0.001), bar holding performance (13.33 ± 0.36 days versus 11.33 ± 0.51 days, U = 43.5, p = 0.005), and horizontal screen (13.22 ± 0.12 days versus 12.08 ± 0.41 days, U = 59.5, p = 0.026) as compared to control males. VPA females had delayed onset of grasp reflex (9.75 ± 0.49 days versus 8.0 ± 0 days, U = 14, p < 0.001) and horizontal screen (13.25 ± 0.49 days versus 12.64 ± 0.19 days, U = 28, p = 0.048) but not bar holding performance as compared to female controls. Thus, there was a sex-specific difference in terms of bar-holding performance where VPA males showed a developmental delay, but VPA females did not. Other motor milestones such as onset of surface righting reflex, vertical screen and negative geotaxis-climb were unaffected in male and female VPA pups. Additionally, within treatment groups, comparisons showed a sex dependent ontogeny of vertical screen performance, as control males took less time to develop this motor milestone than control females (12.83 ± 0.423 days versus 14.00 ± 0.182 days, U = 123, p = 0.03) but this differential development was not seen in VPA male and female pups (13.77 ± 0.173 days versus 13.50 ± 0.189 days, U = 58, p = 0.409). In summary, both male and female VPA rats demonstrated a delayed developmental trajectory in terms of physical growth and the ontogeny of sensory and motor milestones.

We consolidated z-scores domain-wise (by averaging z-scores from individual milestones within each domain) to be able to examine differential patterns, if any, in developmental delay. VPA females had significantly higher physical and sensory domain composite scores (Fig. 7A and B) whereas VPA males had higher sensory and motor domain composite scores (Fig. 7B and C) as compared to their respective controls (see SI Sect. 1.4 for statistical details). Thus, there were clear sex-specific differences in developmental trajectory in male and female VPA rats.

Graph: Fig. 7Domain-specific composite scores. Composite scores were computed for A physical milestones, B sensory milestones and C motor milestones, by combining z-transformed raw data from corresponding individual milestone tests. Data presented as box and whisker plots. Higher values of a domain-specific composite score represent greater delay in that domain. Differences between treatment groups (VPA vs. SAL): ***p <.001, **p <.01, *p <.05

Having observed the delay in acquisition of motor milestones in VPA rats, we used open field test (OFT) to explore if their general locomotor activity was also affected.

The locomotor activity was evaluated in terms of total distance covered and the average velocity during the open field activity (Fig. 8A and Fig. 8B). We did not find any effect of VPA treatment in total distance covered as well as in mean velocity in open field. Two-way ANOVA showed no significant main effects of Treatment (F(1, 46) = 0.98, p = 0.33), or Sex (F(1, 46) = 0.02, p = 0.88) or interaction effect of Treatment X Sex (F(1, 46) = 0.68, p = 0.41) on total distance covered. Similarly, there were no main effects of Treatment (F(1, 46) = 1.32, p = 0.26), or Sex (F(1, 46) = 0.01, p = 0.95), or interaction effect of Treatment X Sex (F(1, 46) = 0.55, p = 0.46) for mean velocity. Additionally, we assessed time spent in center and periphery of the open field chamber, as well as total immobility time in the chamber, and did not find any significant differences between SAL and VPA (SI Sect. 1.5, Figure S2A-C). In summary, VPA rats did not show differences on these measures of open field test that were assessed in our study. These results are in accordance with previous findings of unchanged locomotor activity in VPA rats at weaning age (Olexova et al., [85]; Dobrovolsky et al., [33]).

Graph: Fig. 8Locomotor Activity in Open Field Test. VPA pups did not show differences in total distance covered (A) and mean velocity (B) in the open field test. Data expressed as box and whisker plots. n = 16 VPA male pups, 8 VPA female pups, 12 SAL male pups and 14 SAL female pups

We explored correlations among onset of developmental milestones and the open field test measures. The number of significant correlations between pairs of individual developmental milestones differed widely across conditions (SAL males: 2; SAL females: 6; VPA females: 8 and VPA males: 18). Among VPA males and females (but not SAL), onset of several sensory and motor milestones were highly correlated with each other. Among VPA males, onset of horizontal screen milestone was positively correlated (r = 0.76, p < 0.001) with velocity in open field chamber. See SI Sect. 1.6 for statistical details and Figures S3A-D and S4A-D.

We formulated a developmental composite score that combined results of all fifteen tests, encompassing overall physical and sensory-motor behavioral development of rat pups. This approach provides an effective gestalt measure, and thereby, a robust metric of developmental differences between control and VPA rats. The composite score showed high reliability (standardized Cronbach's alpha of 0.769) indicating high internal consistency in integrating all test readouts into a single composite.

Two-way ANOVA on the developmental composite score revealed significant main effects of Treatment (F(1, 48) = 38.84, p < 0.001), and Sex (F(1, 48) = 4.45, p = 0.04) but no interaction effect of Treatment X Sex (F(1, 48) = 0.06, p = 0.80). Post hoc comparisons showed developmental composite scores were significantly higher in both VPA males (p < 0.001) as well as VPA females (p < 0.001) as compared to their sex-matched controls (Fig. 9A), showing that prenatal VPA exposure altered the overall developmental trajectory in these rats.

Graph: Fig. 9Developmental Composite Score. A developmental composite score (A) was computed by combining the z-transformed raw data from all 15 tests used in our developmental test battery. Data presented as box and whisker plots. Higher values of developmental composite score represent greater delay in overall development of VPA rats. Plots for frequency distribution of developmental composite score (B for females and C for males) show less overlap in male rats than female rats. Developmental composite scores of individual rats (D for females and E for males) distinguish between control and VPA rats. Logistic regression analysis revealed high discrimination between VPA and control rats in both males (96% accuracy) and females (76.3%). n = 18 VPA male pups, 8 VPA female pups, 12 SAL male pups and 14 SAL female pups. Differences between treatment groups (VPA vs. SAL): ***p <.001

The frequency distributions of the developmental composite score showed less overlap in male rats than female rats (Fig. 9C and Fig. 9B) indicating that the distinction between VPA and control populations was better in males than in females. This is further revealed by the plot of individual developmental composite scores that provides a qualitative demarcation between groups among males (Fig. 9E) and females (Fig. 9D). We verified this demarcation quantitatively using binary logistic regression analyses. The models showed high discriminative ability in correctly predicting the group of a given rat (VPA or SAL) on the basis of the developmental composite score in the cross-validated training sets for both males (accuracy 96%) and females (76.3%). The models also performed well on the validation sets (prediction accuracy: males 100%, females 66.7%).

The present study aimed to characterize early development in the prenatal VPA rat model of autism. Our findings demonstrate significantly deviated developmental trajectories in both male and female VPA rats in comparison to typically developing control rats.

We checked body weight in VPA and control rats in their early developmental period until weaning (P4-P22) and did not find any group differences, in line with a few earlier studies (Favre et al., [38]; Servadio et al., [104]) using a similar VPA dosage (500 mg/kg or less). Studies where offspring were prenatally exposed to higher VPA dosage (600 mg/kg or more) have found lower body weight in the VPA group (Du et al., [34]; Zhang et al., [129]). Thus, our findings of developmental delays in VPA rats were not affected by differences in body weight.

In rodents, pups are born with eyes and ears closed, with no fur on the body, underdeveloped sensory sensitivity, and random, uncoordinated movements. These processes continue to develop in a temporally programmed manner during the first 21 postnatal days. Assessment of developmental milestones, such as physical landmarks and tests of sensory and motor abilities, during this initial critical period can indicate effects of prenatal insult on the development of brain functions and characterize early markers for later behavior (Heyser, [51]). Ontogeny of specific sensory and motor developmental milestones are known to be an index of maturation in specific brain regions or neural circuits in animals, including humans (Rice & Barone, [92]).

In the present study, early postnatal development of VPA rats was continuously monitored and compared with the typical development of sex-matched control rats. VPA male and female rats were slow to acquire some physical milestones (incisor eruption and pinnae detachment), indicating a slightly delayed physical maturation in VPA rats. Ontogeny of physical maturity milestones are not directly relevant for clinical populations. However, we included these tests in our developmental test battery as onset of somatic milestones is a crucial component of rodent neurodevelopmental trajectory (Semple et al., [103]). For example, delays in physical milestones such as eye and ear development may be indicative of altered sensory development as well as subsequent visual and auditory input-oriented behavior respectively (Boitnott et al., [15]; Smirnov & Sitnikova, [111]). Although we did not find delays in eye opening, VPA pups showed delayed ear development.

Additionally, both male and female VPA rats had delayed development of sensory systems indicated by late appearance of negative geotactic and limb placing responses. VPA females also had a delayed vibrissa placing response. These tests mainly assess dynamic postural adjustments (induced by sensory stimuli) involving the vestibular system for negative geotactic reaction, and the combined vestibular and exteroceptive systems for limb placing and vibrissa placing responses (Altman & Sudarshan, [3]; Schneider & Przewłocki, [101]). Central regulation by medullary, cerebellar and sensory-motor cortical systems also plays a major role (Altman & Sudarshan, [3]; Wagner et al., [123]; Y. Zhang, Li, Yang, Zhang, & Yang, [130]).

Finally, we also saw evidence of delayed motor system development due to prenatal VPA exposure. The ontogeny of forelimb grasp reflex was affected in both males and females. Although the grasping response is a spinal reflex, it is regulated by the action of higher brain centers comprising non-primary motor areas such as premotor cortex and supplementary motor cortex, on the spinal interneurons (Hashimoto & Tanaka, [50]; Nguyen et al., [83]). Moreover, bar holding performance, and horizontal screen performance that were used to assess muscle tone, limb strength, as well as sensorimotor coordination (Deacon, [30]; Heyser, [51]; Slamberova et al., [110]), appeared late in VPA males indicating delayed motor system development in the brain (Feather-Schussler & Ferguson, [39]; van de Wijer et al., [121]). VPA females also showed delayed horizontal screen performance indicating that motor system development was delayed in both sexes. However, VPA rats did not show any impairment in open field test at weaning, indicating that once motor milestones were acquired, locomotor functions were normal.

Overall, our findings of delayed ontogeny of critical sensory and motor developmental landmarks in first three weeks after birth are indicative of prenatal VPA induced alterations in the early development of the brain regions regulating these sensorimotor functions, mainly somatosensory cortex, motor cortex, cerebellum, and brainstem. This is supported by previous reports of lower cell density and smaller neuronal cell sizes (Al Sagheer et al., [1]; Ingram et al., [54]; Kataoka et al., [56]; Lukose et al., [72]; Spisak et al., [112]; Varghese et al., [122]) as well as altered excitatory-inhibitory (E/I) balance, i.e., increased glutamatergic and decreased GABAergic signaling, in these brain regions (Chau et al., [23]; Hou et al., [52]; Rinaldi et al., [93]; Tartaglione et al., [114]). It is important to note, however, that most of these studies were carried out during adolescence and adulthood, and not during early development. Future studies during the neonatal period can help elucidate the neuroanatomical and neurochemical underpinnings of the time course of developmental trajectory alterations in the VPA rat model of autism.

Alterations in E/I dynamics is known to impact the timing of critical periods in brain development, and eventually, alterations in the formation of functional networks (Gogolla et al., [44]; D. D. Wang & Kriegstein, [124]). For example, Wang & Kriegstein, [124], disrupted E/I balance in typically developing mice pups during the perinatal period and reported developmental delays in onset of sensory and motor milestones such as negative geotaxis and bar holding abilities. Further, Nagode and colleagues demonstrated that prenatal VPA exposure in mice altered the development of the earliest cortical circuits involved in sensory processing via altering E/I balance (Nagode et al., [82]). Indeed, hyper-connected microcircuits in the somatosensory cortex as well as altered synaptic plasticity in the somatosensory cortex and cerebellum have been found in the VPA rat model (Iijima et al., [53]; Silva et al., [108]; R. Wang et al., [125]). Future studies in the VPA model could try to restore E/I balance in these regions during the early developmental period and evaluate rescue of behavioral milestones to test the possibility of a causal role of E/I imbalance in the developmental delays seen in the VPA model.

Previous studies that explored early development in prenatal VPA rodent model of autism provided mixed findings for different milestones (Dobrovolsky et al., [33]; Li et al., [69]; Schneider & Przewłocki, [101]; Wagner et al., [123]; R. Zhang et al., [129]). There are two possibilities for this discrepancy in the literature—first, the methodology used to assess developmental milestones; and second, the data analysis steps prior to statistical testing.

Most studies investigating early development in VPA rats incorporated only few tests as part of a larger study with a different goal. Moreover, many of these studies assessed early development within a short postnatal time window or on a specific postnatal day (Dobrovolsky et al., [33]; Schneider & Przewłocki, [101]; Wagner et al., [123]). The literature on early development in rodents shows that milestones occur in a temporal sequence in a pup's early life (Heyser, [51]). For example, milestones such as incisor eruption, fur development, surface righting, cliff avoidance, horizontal screen, negative geotaxis appear early, whereas eye opening, ear canal opening, vertical screen and bar holding are late appearing milestones. Importantly, the age of onset of different developmental milestones differs between species, between strains of the same species, and between individual pups of a given species and strain. Additionally, as we observed in our data, VPA rats had delayed appearance of several milestones, but they did not miss achieving any milestone entirely. Therefore, testing for a milestone too early (when that specific milestone might not have appeared) or too late (when even VPA pups have acquired that milestone) might miss finding group differences that might be present.

Another important consideration is that many studies, including our own, used a binary assessment i.e., whether milestones were present or not, tested on each postnatal day, until the acquisition of the developmental landmarks (Li et al., [69]; R. Wang et al., [125]). Other studies have used a graded scoring system (no response to maximum response) for the acquisition of each developmental milestone (Al Sagheer et al., [1]). Graded scoring can be more informative about the process of ontogeny of developmental landmarks but is more heavily influenced by subjective bias than the binary scoring system. We recommend that future studies use the binary measure for all milestones to facilitate comparison across studies, and additionally adopt the graded system where appropriate as per study goals.

Another methodological difference in the literature involves quantification. Several studies quantified milestones such as latency to reorient in negative geotaxis and surface righting (Gandal et al., [42]; R. Zhang et al., [129]). Some studies have compared proportions–i.e., how many pups in each group achieved a milestone on a postnatal day (Reynolds et al., [91]). Considering these methodological differences in assessment and analysis, the mixed evidence in the literature is not surprising. Other contributing factors include differences in VPA dose and concentration, timing, and mode of VPA administration, species, and strains of model animals.

In summary, there is mounting evidence that the VPA model engenders developmental delays and that there is a clear need to use a unified approach to quantifying these delays in different models that vary in terms of species, strains, VPA dose etc.

Neurodevelopmental trajectories have been examined in other environmental insult-based models of ASDs, including the maternal immune activation (MIA), a prominent environmental model of ASD (Guma et al., [45]). Studies using LPS induced MIA model did not find any major difference in the onset of neonatal reflexes and physical milestones (Baharnoori et al., [6]; Fernandez de Cossio et al., [40]; Foley et al., [41]). In contrast, sensorimotor milestones such as righting reflex, gait, and negative geotaxis were delayed in another study using the LPS model (Rousset et al., [98]). Similarly, Malkova et al., [74] did not find delays in development of negative geotaxis, grasp reflex and righting reflex in a Poly I:C induced MIA model. On the other hand, Haida et al., [46] assessed onset of eye opening, negative geotaxis, and righting reflex using the Poly I:C model and found delays in these milestones.

Delays in righting reflex and negative geotaxis were found in a prenatal Vitamin D deficiency model of ASD (Ali et al., [2]) whereas physical milestones were not affected. Similarly, delayed sensorimotor development in milestones such as righting reflex, cliff avoidance and negative geotaxis were found in prenatal chlorpyrifos (CPF, an insecticide) model (Lan et al., [63]), however, physical milestones were not examined. In a pharmacological model, perinatal fluoxetine treated rats showed delayed onset of righting reflex, negative geotaxis, vibrissa placing and bar holding abilities, but no delay in eye opening, the only physical milestone that was tested (Kroeze et al., [62]).

In summary, early postnatal development in other environmental models show mixed findings for different milestones that could be due to several factors. Differences in protocols to develop the models, differential impact of these different environmental risk factors on brain development, differences in the degree of their construct and phenotypic validity for the ASD condition, as well as methodological differences in assessment of developmental milestones, could be contributing factors for these mixed findings. Detailed comprehensive studies spanning multiple physical, sensory, and motor milestones are needed to facilitate comparisons between different environmental models of ASD.

There is limited literature on early development in females in animal models of ASD. Most studies have focused on males (Kazdoba et al., [57]; Wohr et al., [127]) and this is the case in the VPA model too (Hou et al., [52]; Wang et al., [125]). Al Sagheer et al., [1] found significant delays in eye opening and surface righting reflex in male as well as female VPA mice. Similarly, Wagner et al., [123] observed delayed appearance of surface righting as well as negative geotaxis in both male and female VPA mice. Another study found delayed auditory startle reflex in female VPA mice, but no males were studied, precluding assessment of sex differences (Kazlauskas et al., [58]).

In the present study, we found overall delayed development in both male and female VPA rats (developmental composite score data). However, there were differences in the pattern of developmental delays in VPA males and females. In the motor domain, VPA exposure impacted males more than females since only males showed delayed bar holding performance and the noteworthy sex-dependent advantage shown by controls males for the vertical screen was absent among VPA males. In the sensory domain, females were disproportionately impacted by VPA exposure as the vibrissa placing response was delayed only in VPA females who also did not show the sex-dependent advantage shown by control females for this test. However, in a different sensory test (negative geotaxis), although both males and females were affected, the male advantage seen among controls was missing among VPA showing that VPA males were more affected by the prenatal VPA exposure. These differences may be attributed to the task-specific nuances in the sensory domain as vibrissa placing is a tactile-stimulus dependent test whereas the negative geotaxis test involves the vestibular system as well as sensory-motor coordination for turning the body by 180° on an inclined plane.

We constructed domain specific composite scores to smooth over test specific differences and provide a consolidated representation for the physical, sensory, and motor developmental domains. Indeed, these composites provided greater clarity into the sex-specific patterns of developmental delays in VPA. Specifically, although both VPA males and females had delays in the sensory domain composite, only VPA males had delays in the motor domain composite and further, only VPA females had delays in the physical domain composite.

Sex specific differences in developmental profiles of ASD children are not clear. There are reports showing female advantage in the acquisition of language, visual reception, and fine motor milestones among ASD children (Messinger et al., [78]). Some studies have shown better language and motor (gross and fine) skills in autistic boys and better visual reception in girls (Carter et al., [20]) but other studies have also shown no sex specific differences in early sensory and motor development of ASD children (Bedford et al., [12]; Reinhardt et al., [90]). It is possible that heterogeneity in the autism phenotype as well as methodological differences may have contributed to the inconsistencies in the literature.

Taken together, our findings suggest that prenatal VPA exposure impacted both sexes but there were important sex-dependent differences. The test-specific nuances seen in our study suggests that methodological considerations could shed light on better understanding the differential neurobiology of ASD in boys and girls.

We carried out exploratory correlation analyses to examine associations among early developmental milestones and locomotor function assessed at P22 using open field test measures. These analyses were conducted with individual developmental milestones as well as with domain specific composites. Our findings indicate that for VPA rats, onset of several milestones were highly correlated with each other. This pattern was not observed among SAL rats. Further, VPA males showed a positive correlation between velocity in open field and onset of a motor milestone (horizontal screen) possibly suggestive of a compensatory increase in motor activity with delay in motor development. Overall, these exploratory analyses support the possibility that prenatal VPA exposure provides a common underlying driver (disrupting connectivity between brain regions during development) leading to the observed correlated delays in milestones along the developmental trajectory.

Individuals on the autism spectrum show a huge diversity in the severity of symptom presentation. This variability in symptoms is also found in animal models of autism, presenting a problem in effectively capturing symptom severity and comparing across models. The problem exists even in animals containing the same autism-related mutation (Dere et al., [31]). A useful approach is to compute a composite score spanning multiple test readouts rather than comparing groups for each individual test readout (El-Kordi et al., [35]).

Our focus in this study was to develop a developmental composite score consolidating all 15 milestones comprising physical, sensory, and motor developmental domains, into a single number snapshot, a gestalt measure, of the developmental trajectory of each rat from birth to weaning. The high reliability of the developmental composite score and overall ability to accurately discriminate between VPA and control rats demonstrates its value for the comprehensive assessment of neurodevelopment in rat pups and might have translational value for studying developmental disorders during early life and in early intervention studies in animal models.

Our findings of delayed developmental milestones in VPA rats are well in line with a large body of clinical evidence indicating that early motor and sensory developmental delays precede ASD diagnosis and may serve as early indicators of ASD (Bhat et al., [14]; Dadalko & Travers, [27]; Estes et al., [37]; Harris, [48]; Robertson & Baron-Cohen, [96]).

Several studies have reported delayed development of early gross motor functions (e.g., sitting and standing without support and independent walking) in children with, and at high risk for, ASD (Davidovitch et al., [28]; R. Landa & Garrett-Mayer, [65]; Lemcke et al., [68]). Delays in fine motor functions such as fine motor coordination, reaching, and grasping have also been reported in children with ASD (Bolton et al., [17]; R. Landa & Garrett-Mayer, [65]; Libertus et al., [70]). In our study, VPA rats showed delays in bar holding and horizontal screen tests which mainly assess gross aspects of motor functions such as stamina and limb strength (Brooks & Dunnett, [18]; Feather-Schussler & Ferguson, [39]). Early fine motor skills such as involuntary grasp response as seen in the grasp reflex test and active grasp response with proximal stability as seen in bar holding test (Anekar & Bordoni, [4]; Diener & Bregman, [32]; Whishaw & Kolb, [126]) were also delayed in VPA rats. These early fine motor skills are pre-requisites for voluntary reaching and grasping—fine motor skills that develop at maturity (at least 4 weeks of age in rats).

The literature on early sensory processing problems in autistic children is mixed. However, several studies in children with ASD, or those at high risk for ASD, have reported atypical responsiveness to sensory inputs across auditory, visual, and somatosensory modalities during early development (Grace T Baranek, [8]; G. T. Baranek et al., [9]; Baum et al., [11]; Robertson & Baron-Cohen, [96]; Wolff et al., [128]; Zwaigenbaum et al., [131]). Thus, our findings of delayed somatosensory development in VPA rats, reflected by delays in negative geotaxis, vibrissa placing and limb placing tests, are in concordance with the findings in children with ASD.

It is noteworthy that the somatosensory system plays a central role in early childhood, and that somatosensory inputs are critical not only for the development of both gross and fine motor skills, but also for healthy social and communication development (Cascio, [21]; Metcalfe et al., [79]; Thompson et al., [117]). Neural computations underlying basic sensory and motor processing are well conserved between humans and rodents. Future studies on the structural and functional aspects of early sensorimotor development in VPA model would be instrumental in shedding light on the time course of ASD neurobiology.

Our study provides comprehensive evidence of developmental delay in the ontogeny of physical, sensory, and motor milestones in the prenatal VPA exposure rat model of autism. However, there are a few limitations that warrant further investigation.

First, the sample size was small for VPA females. It is unclear if the lower accuracy in discriminating between VPA, and control females based on the developmental composite scores was due to lower power or due to the higher individual variability in female VPA rats as compared to male VPA rats. Second, our focus was on sensory and motor development and at the time of the study, we did not include tests for early social development in our test battery. Future studies should consider including social development assessment, such as ultrasonic vocalizations test and social-odor discrimination task (Potasiewicz et al., [89]; Schneider & Przewłocki, [101]), in their early developmental test battery for additional relevance for ASD models.

Third, these rat pups were not followed up and tested for core features of autism (social interaction deficits and repetitive behavior) later in life. These rat pups had gone through additional handling during their early developmental period as part of the assessment of developmental milestones and studies indicate that additional handling of neonatal rodents impacts their performance in social interaction test and other behavioral measures in adulthood (Heyser, [51]; Schneider et al., [102]; Todeschin et al., [119]). To avoid this confound, we refrained from testing these pups later in life. We have previously provided robust evidence of the presence of ASD like phenotype, such as reduced social interaction, altered attentional processing, and impaired sensorimotor gating in a separate cohort of naïve VPA adult rats that had not undergone detailed developmental assessment (Anshu et al, [5]) and we replicated these results across multiple cohorts. Nonetheless, this is an important point to consider and it would be valuable to carry out a future study in the VPA model that does follow up testing of core features after detailed evaluation of developmental trajectory, and compares these findings with a separate cohort of rats that have not undergone this additional handling.

The present study aimed to characterize the acquisition of different milestones during early postnatal development in the VPA rat model of autism. Our findings demonstrate significantly deviated developmental trajectories in both male and female VPA rats in comparison to typically developing control rats and are in line with clinical findings showing delayed development in children with ASD. The evidence of sex-specific differences in developmental milestones in VPA rats underscore the need for carrying out autism related studies in both sexes. Our findings suggest an altered developmental programming of prenatal brain development process after prenatal VPA exposure in these rats. Future studies are needed to pinpoint the underlying mechanisms leading to early developmental delay followed by autism like phenotype later in life. Overall, the present study provides a detailed characterization of early development in the prenatal VPA model of autism in both sexes and presents a developmental composite score (combining a range of tests for different physical, sensory, and motor developmental landmarks) as a robust measure for evaluating early development in different animal models of autism and as a potential outcome measure for early intervention studies.

This work was supported by University Grants Commission (UGC), Government of India for the fellowship to K.A. and National Institute of Mental Health and Neurosciences (NIMHANS), Bengaluru for the infrastructure facilities and support to carry out the work. We thank Dr. U.D. Kumaresan for helpful discussions.

KA, AKN, SS and TRL contributed to the study conception and design. Data collection was performed by KA. Data analysis was performed by KA and AKN. The first draft of the manuscript was written by KA and all authors provided feedback. All authors read and approved the final manuscript.

The authors have no conflicts of interests to declare.

Below is the link to the electronic supplementary material.

Graph: Supplementary file1 (DOCX 1326 kb)

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.