Androgen Aggravates Chorioamnionitis-Induced White Matter Brain Injury and Neurobehavioral Impairments in Males

Autism spectrum disorder (ASD) is a neurodevelopmental condition marked by repetitive behaviors, and difficulties in social interaction and communication across multiple contexts [1, 2]. ASD affects 1 in 100 children, and is typically identified between ages three and six, with males being four times more likely than females to be diagnosed [3]. Many genes related to neurodevelopmental disorders have been identified; however, no single gene has been found to be the cause of the most common forms of these disorders. In total, these genes explain only about 40% of ASD cases [4]. This highlights the role of environmental factors and their contribution to the expression of ASD-related traits. Epidemiological studies have uncovered a link between infection-induced maternal immune activation (MIA) and subsequent neurodevelopmental disorders in the progeny [5‒7]. This correlation is not surprising given that the brain is particularly vulnerable to changes in the prenatal environment during this highly sensitive period of development. In addition to epidemiological studies supporting this notion, animal models have also provided robust evidence linking gestational immune dysregulation to neurodevelopmental impairments [8‒12].

Group B Streptococcus (GBS) is a leading infectious pathogen inducing MIA. This gram-positive bacterium colonizes the gastrointestinal and/or lower genital tract in approximately 20% of pregnant women [13, 14]. While GBS colonization is mostly asymptomatic, it can lead to placental inflammation, known as chorioamnionitis (CA), urinary tract infections and life-threatening neonatal infections including pneumonia, sepsis, and meningitis [15]. Our laboratory developed a preclinical model of GBS-induced CA, where we observed that male rats exhibited higher placental inflammation compared to their female littermates, resulting in ASD-like traits in males [9, 10, 16]. This finding is significant considering the skewed sex ratio toward males observed in ASD. Given the well-established link between GBS-induced CA and brain injuries in our model, this study takes a novel approach by targeting androgen signaling as a potential mediator of the innate immune response. We hypothesized that androgens drive sex-dependent differences in the innate immune response. As such, we found that administering an androgen receptor antagonist, flutamide, to dams affected by GBS-induced CA reduced pro-inflammatory activity in the male placenta [17]. More specifically, androgens produced in male fetuses, via the hypothalamic-pituitary-gonadal axis, modulated immune activity by exacerbating key pro-inflammatory cytokines such as interleukin (IL)-1β, IL-6 and TNF-α, while also promoting polymorphonuclear cell infiltration in both maternal and fetal compartments of the placenta [17]. This highlights the significant role of androgens in increasing placental inflammation during GBS-induced CA. Steroid hormones, derived from cholesterol, play a crucial role in many physiological processes including sex organ development, secondary sexual characteristics and importantly in the brain [18]. Many regions in the central nervous system express receptors for multiple steroid hormones including estrogen, progestogens, and androgen suggesting a role for sex hormones in brain modulation [18]. Thus, expanding on this understanding beyond the placenta, we investigated whether innate modulation by androgens during GBS infection affects neurobehavioral development in males.

GBS serotype Ia (strain #16955) was stored at −80°C in Brain Heart Infusion (BHI; Sigma-Aldrich, ON, Canada) broth supplemented with 15% glycerol and was used for experiments, as per our validated laboratory protocol [9]. In summary, GBS was inoculated and incubated in BHI broth at 37°C with shaking for 18 h, then reinoculated and incubated with shaking until the optical density at 600 nm reached 0.5–0.7. The culture was centrifuged at 4,000 rotations per minute, inactivated with 10% formaldehyde, washed, and resuspended in 0.9% sterile saline to 10⁹ colony forming units/100 μL. Inactivation was confirmed by plating on CHROMID Strepto B plates (BioMérieux, Saint-Laurent, QC, Canada). The control group was injected with 100 μL sterile saline.

Flutamide, a competitive androgen receptor antagonist, was prepared as previously described [17]. Flutamide (Sigma-Aldrich, ON, Canada) was dissolved in 5% dimethyl sulfoxide and corn oil. The control group was injected with 150 μL 5% dimethyl sulfoxide/corn oil.

Primiparous Lewis dams were sourced from Charles River Laboratories (Raleigh, NC, USA) and transported to our animal facility (RI-MUHC Glen Site, Montreal, QC, Canada) on gestation day (G)13. They were housed separately under controlled conditions and given unrestricted access to food and water. Flutamide and inactivated GBS injections were administered based on established experimental frameworks developed in our laboratory (Fig. 1) [9, 17]. Only results from male offspring were included in this study and the four experimental groups (n litters per group) were: (1) vehicle/saline (V/S, n = 5), (2) vehicle/GBS (V/GBS, n = 6), (3) flutamide/saline (F/S, n = 6), and (4) flutamide/GBS (F/GBS, n = 6).

Two male rats per dam from 4 to 6 litters were used for each behavioral test and experiments were conducted with a minimum of 5 days apart (Fig. 1). Tests were done as previously described [9, 16]. A blinded experimenter analyzed all parameters. Apparatus was cleaned and dried in between each animal.

Pups are dependent on their olfactory systems for maternal communication during the first 2 weeks after birth, as their auditory and visual systems are underdeveloped. The nest-seeking test on postnatal day (P) 9 assesses this developmental process, making it a relevant tool for assessing ASD-like traits [9, 19, 20]. A standard cage was divided into three equal sections, with home cage and clean bedding on either side. Each pup underwent two 3-min trials, scored on home bedding, clean bedding or no movement. Success rate to reach familiar odor (%) was measured, with home bedding used for at least 5 days before the test.

The open field test is used to assess spontaneous locomotion, exploratory, and anxiety-like behavior. On P20, rats were acclimated to the testing room for 30 min and then placed in the center of the arena for 5 min. The arena was divided into equal virtual squares and movement was automatically tracked and recorded using the Any-Maze Video Tracking System (Stoelting Co., IL, USA). Outcomes included total distance travelled, duration of mobility, and number of virtual lines crossed.

To investigate deficits in social behavior, a hallmark of ASD, we used the social interactions test on P40. Two days before the test, rats were acclimated in the open testing arena for 10 min each day. On the test day, 2 rats of the same sex and experimental condition but from different litters were placed in the arena for a total of 5 min. We recorded the length of active social interactions, namely behaviors like grooming, sniffing, playing, and chasing. Passive physical contact was excluded. If, during the adaptation phase, an animal displayed aggressive or excessively passive behavior, it was excluded from the test.

To assess anxiety-like behavior and locomotion, we used the elevated plus maze test on P50. Rats were acclimated to the test room for 10 min on the day of the test. Rats were placed on the center of the maze, consisting of two enclosed arms and two open arms elevated above the ground. Outcomes were automatically quantified by Any-Maze Video Tracking System (Stoelting Co.) and included total distance travelled and time spent in open arms.

Rats used in behavioral experiments were sacrificed on P50, a developmental stage when most brain maturation is complete. The brains were collected as previously described [9], bisected sagittally and fixed in 4% buffered formaldehyde (with 0.1% glutaraldehyde, pH 7.4), then processed and embedded in paraffin.

Paraffin-embedded forebrains were sectioned at 5 μm thickness and two consecutive slides were used for histological analysis. Coronal sections of the brain were cut at three zones from bregma: −0.3, −1.0, and 2.8 mm for immunohistochemical experiments performed as previously described [9]. MBP staining allowed for the measurement of the lateral ventricle area, thickness of corpus callosum, and external capsule (coronal plane, at the level of the interhemispheric sulcus). The average of two thickness measurements from two distinct bregma zones (−0.3 mm and −1.0 mm) was measured and calculated by an experimenter blinded to conditions. Antibodies against CC-1, Iba-1, and GFAP were used to identify mature oligodendrocytes, microglia, and astrocytes, respectively. Cell counts were averaged from three fields for CC-1 (∼0.04 mm²) and Iba-1 (∼0.1 mm²), and stained area (%) was quantified for GFAP (∼0.1 mm²) within the external capsule at −2.9 mm from the bregma.

Slides were scanned using the Axio Scan. Z1 equipment (Zeiss, Oberkochen, Germany) with a ×20 objective, and slides were analyzed using the ZEN Lite software (Zeiss) and ImageJ. The same forebrains were used for all histological analyses, and all images were examined and analyzed by an experimenter blinded to conditions.

Littermates are regarded as pseudo-replicates as all pups come from a single dam treated prenatally, thus they should not be considered as completely independent subjects in statistical evaluations. The average of two males per dam was used as n = 1 per litter to account for litter effects and prevent artificial sample size inflation. This approach ensured that variability between litters, rather than within litters, was the primary source of comparison. Based on the parameters studied in our preliminary results, a sample size of 5–10 animals per group depending on the experimental end point will allow us to detect a 30% reduction (in means) between the groups with 80% power, a two-sided alpha of 5%, and a coefficient of variations of 15% in each group. Statistical analyses and representation were done using Graph Pad Prism software version 10.2.3 (San Diego, CA, USA). All data were normally distributed thus two-way ANOVA was applied with Bonferroni’s post hoc test. Data are displayed as the mean ± SEM; p ≤ 0.05 was considered statistically significant.

No sickness related behavior or mortality occurred in dams. Mean dam weight did not differ between experimental groups prior to injections (Fig. 2a). Both flutamide- and GBS-treated dams displayed reduced weight gain compared to their respective controls (Fig. 2a). There was no difference in total litter size between experimental groups indicating that neither of the treatments affected mortality (Fig. 2b). All male pups gained a similar amount of weight from P1 to 9 (Fig. 2c). Altogether, our results showed that there was no significant effect of flutamide and GBS on perinatal general morphometric outcomes.

To investigate the effect of treatments on ASD traits, male rats were assessed in the nest-seeking, open-field, social interactions, and elevated plus maze tests at four time points: P9, 20, 40, and 50, respectively. The success rate to reach familiar odor in the nest-seeking test was reduced in GBS-exposed versus non-exposed male pups (p < 0.01, Fig. 3a). A trend toward improved success rate was observed in the flutamide/GBS versus GBS-exposed male offspring (Fig. 3a). In GBS condition, flutamide maintained at the control level the success rate to reach familiar odor (Fig. 3a).

Flutamide significantly improved social interaction length in GBS infected males (p = 0.03, Fig. 4a) done during late puberty. There was a trend toward reduced length of interaction in the GBS-exposed versus nonexposed males (Fig. 4a).

No significant interaction was observed for any outcomes measured in the open field nor elevated plus maze tests (Fig. 3b, 4b; online suppl. Fig. S1, S2; for all online suppl. material, see https://doi.org/10.1159/000545074). Overall, these results showed an amelioration of GBS-induced neurobehavioral impairments by flutamide and importantly, that the differences in social interactions were not due to decreased spontaneous motor activity or anxiety.

Neuroanatomical assessments were done on forebrains by MBP staining at P50 (Fig. 5a). There was no difference in brain weight between experimental groups (online suppl. Fig. S3). GBS CA reduced thickness of the external capsule (15.4% reduction) compared to control (p = 0.03, Fig. 5b). In GBS CA, flutamide preserved the normal thickness of the external capsule, i.e., at the same level as control males (Fig. 5b). Similarly, area of the lateral ventricle – reflecting forebrain atrophy – was significantly enlarged (85% increase) in GBS-exposed versus nonexposed males (p = 0.04, Fig. 5c). Flutamide/GBS males showed no difference in lateral ventricle area compared to flutamide controls (Fig. 5c). Thickness of the corpus callosum was significantly reduced in GBS-exposed versus non-exposed males, but there was no androgen effect on this outcome (Fig. 5d).

The damaged external capsule was characterized by a reduced number of CC-1+ oligodendrocytes in GBS-exposed versus unexposed males (p = 0.001, Fig. 6a). A significant interaction was observed (p = 0.01) between GBS and flutamide treatment. Specifically, flutamide significantly increased mature CC-1+ oligodendrocytes in GBS-exposed males (p = 0.05, Fig. 6a). While flutamide increased microglial density in the external capsule (Fig. 6b), there was no effect of flutamide in GBS condition on microglial and astrocyte densities within the forebrain white matter (Fig. 6b; online suppl. Fig. S4). In summary, neuroanatomical assessments revealed that GBS exposure caused structural changes in the brain, which flutamide partially prevented in GBS-exposed males. GBS exposure also significantly reduced the number of mature CC-1+ oligodendrocytes, a deficit that was also reversed by flutamide.

It is well established that ASD shows a male sex bias, yet the factors behind this phenomenon remain unclear. The predominance of males in this condition cannot be attributed solely to genetics, including X-linked defects [4]. Instead, it can be influenced by hormones, which are genetically regulated but also interact with environmental factors, highlighting a complex gene-environment interplay. The perinatal period is a critical time when testosterone, partially through its conversion to estradiol, initiates masculinization, and defeminization of the male brain [21]. These processes have an organizational impact on brain development and function, influencing the activation and effects of sex hormones throughout life [22]. In addition, glial cells, such as oligodendrocytes and astrocytes, are targeted by neuroactive steroids due, in part, to their expression of androgen and estrogen receptors [23].

Our results demonstrate that while GBS-induced CA does not significantly alter general morphometric outcomes, it induced neurobehavioral deficits and white matter injuries in male offspring, both of which were partially improved by flutamide treatment. Flutamide can cross both the placental and blood-brain barriers; this suggests that it can influence fetal brain development and function [24, 25]. Interestingly, these observations are supported by studies showing the neuroprotective effects of finasteride, a drug that inhibits the formation of dihydrotestosterone, the active metabolite of testosterone, in a propionic acid-induced autism rat model [26]. Although this model is not a MIA model, Sever et al. [26] highlight the role of androgen pathways in directly influencing neurodevelopmental deficits. The neuroprotective effects of finasteride were demonstrated by biochemical, histopathological, and behavioral analyses. These benefits were attributed to its antiandrogenic properties, consistent with the improvements observed with flutamide in social interactions test and the trends seen in the nest-seeking test in our study.

Oligodendrocytes are glial cells that support and maintain integrity of axons and neuronal connectivity within the central nervous system [27]. We observed that androgen blockade protected against CC-1+ oligodendrocyte loss in the external capsule of CA-exposed males. Caruso et al. [28] investigated testosterone’s effect on the excitotoxic death of cultured oligodendrocytes, explaining some more severe forms of multiple sclerosis observed in males. The study showed that flutamide prevented both the direct toxicity of testosterone and its ability to exacerbate excitotoxic damage induced by glutamate receptors in cultured oligodendrocytes. However, other studies also indicate that testosterone is generally associated with increased myelination and neuroprotection [29, 30], suggesting that its effects in the context of MIA and other pathological conditions are likely more complex. For instance, Fanaei et al. [31] demonstrated that testosterone can have both neuroprotective and harmful effects in a rat model of cerebral ischemia – with low levels being protective and high levels detrimental. Flutamide reduced infarct volumes, mitigated brain edema, and caused improved neurological outcomes [31]. Interestingly, blocking testosterone’s conversion to estradiol with letrozole increased damage, consistent with previous studies highlighting estradiol as a neuroprotective factor against cerebral ischemia [32, 33].

Plasma levels of luteinizing hormone and testosterone significantly increase after acute flutamide treatment [34, 35]. This is because flutamide, as an androgen receptor antagonist, reduces the negative feedback that testosterone and other androgens have on the hypothalamic-pituitary-gonadal axis [36]. A potential interpretation of this in the context of our study is that flutamide’s effect on hormonal balance may have led to higher conversions of androgens into estrogens in brain tissues by aromatase [37]. This could be due to the accumulation of upstream androgen precursors, such as testosterone. In the absence of normal androgen receptor activation, excess androgens may have been more readily converted into estrogens. Estradiol plays a key role in regulating oligodendrocyte progenitor cells (OPCs) and their development into mature oligodendrocytes [38]. It has been reported that estradiol delays the exit of OPCs from the cell cycle [39]. Thus, the observed increase in CC-1+ oligodendrocytes may result from estradiol’s role in prolonging the cell cycle of OPCs, thereby increasing the pool of mature oligodendrocytes in flutamide/GBS groups. A limitation of our study is that we did not distinguish the specific contributions of androgens and estrogens. Future studies could address this by using letrozole, and aromatase inhibitor, to better delineate the respective role of each hormone family. For astrocytes and microglia, neither fetal exposure to GBS-induced CA nor flutamide treatment significantly altered the density of these cell populations. Previous research in our lab noted that flutamide caused a 2.3-fold decrease in polymorphonuclear cell infiltration into the decidua of GBS CA, highlighting that androgens exacerbate the placental innate immune response [17]. It is possible that the observed decrease in cellular infiltration and cytokine levels was not substantial enough to impact microglia or astrocyte populations in the brain.

The observed decrease in CC-1+ oligodendrocytes in the forebrain white matter likely resulted in the observed enlargement of lateral ventricles in GBS-infected groups. Oligodendrocytes are essential for maintaining axonal integrity and supporting white matter structure; their loss can lead to dysmyelination and reduced axonal density, ultimately resulting in brain tissue atrophy. This shrinkage of surrounding white matter may have led to the expansion of the lateral ventricles [40]. While glucocorticoids are often used to treat cerebral edema by reducing inflammation, they can also lead to adverse effects such as brain atrophy [41, 42]. This raises the question of whether a decrease in androgen activity or an increase in estradiol might have contributed to these changes. In our study, the enlargement of the lateral ventricles may likely be a result of the underdevelopment of neuronal connections, resulting in smaller bundles of axons within the forebrain white matter [43].

Paternal genetic and epigenetic influences may also shape placental function and offspring development [44]. Interestingly, paternal epigenetic modifications have been shown to influence both placental development and function, as well as fetal brain abnormalities associated with ASD [45]. Our results add androgen signaling, innate immune signaling, or their combined effects to the list of potential mechanisms involved in these processes.

While we hypothesized that changes in androgen and estrogen activity could influence brain structure and function in the context of GBS-induced CA, our study did not directly measure hormone levels or the activity of relevant signaling pathways associated with the androgen blockade. Additionally, while we noted a potential relationship between decreased oligodendrocytes and enlarged lateral ventricles, we did not directly address neuronal loss or cortical and basal gray nuclei thickness. Our findings underscore the significant role of hormones in driving sex-dependent differences in the immune response and its direct or indirect impact on brain development. Specifically, these results provide new insights into how androgen-mediated immune modulation during GBS-induced CA affects neurobehavioral development in male offspring, thereby advancing our understanding of sex differences in ASD.

We would like to thank Taghreed Ayash (RI-MUHC, Montreal) for her assistance with animal work.

This study protocol was reviewed and approved by the Research Institute of McGill University Health Centre (RI-MUHC) Animal Care Committee (RI-MUHC-7675).

The authors have no competing interests to declare.

This work was supported by a grant from Canadian Institutes of Health Research (CIHR 423146) and a grant from Pierre Lachapelle (RI-MUHC, Montreal). S.V. received a doctoral training scholarship from Fonds de recherche du Québec (FRQS). B.R. is a James McGill Professor.

S.V.: investigation, writing – original draft, visualization, and formal analysis. M.C.: assisting investigation, project administration, and writing – review and editing. M.-J.A.: methodology and writing – review and editing. N.B.: assisting investigation. B.R.: conceptualization, supervision, and funding acquisition. G.S.: conceptualization, methodology, supervision, and funding acquisition. All authors have read and agreed to the published version of the manuscript.

Bernard Robaire and Guillaume Sébire should be considered as the co-senior authors.

All data are available upon request. The data can be found in figshare at 10.6084/m9.figshare.27969564.