**6. Human endogenous retroviruses in neurodevelopmental disorders**

The first article pioneering the hypothesis of the possible involvement of HERVs in the aetiopathogenesis of neurodevelopmental disorders was published by us 10 years ago. Specifically, we hypothesized a link between HERV activity and ASD highlighting a distinct expression profile in Italian autistic children in which HERV-H was highly expressed in peripheral lymphocytes, when compared with controls, with higher levels in younger children, supporting the hypothesis that HERV-H overexpression might be regarded as a potential early biomarker of ASD. This view was even more supported by the fact that ASD children with more severe impairments in Communication and Motor Psychoeducational Profile-3 showed the highest expression levels of HERV-H [124]. The analysis of HERV expression profile was then replicated in a cohort of Albanian ASD children who showed HERV-H high expression level in peripheral lymphocytes as already found in Italian ASD children. This allowed us to conclude that HERV-H could be considered as a molecular signature of the disease unrelated to ethnicity [125]. These findings opened a challenging scenario to extend the study to a cohort of attention deficit hyperactivity disorder (ADHD) children in order to verify whether the peculiar HERV expression profile could also be identified in a 'twin disorder' of ASD. Indeed, the two disorders are highly correlated [126], and ADHD often occurs in conjunction with ASD [127]. ADHD children showed the highest expression levels of HERV-H in peripheral lymphocytes correlated with inattention and hyperactivity symptoms, suggesting HERV-H a molecular biomarker also for ADHD [128]. All the evidence emerged from these initial papers was always obtained by studying drug-naϊve children considering that, as indicated at the time by other research groups, HERV activity could be strongly influenced by drugs exposure [90]. And it is from all of this that the subsequent work arose, in which the intention was to evaluate HERV-H as a potential biomarker of response to treatment in ADHD patients undergoing methylphenidate (MPH) therapy. As such, the reduction of HERV-H expression levels H in peripheral lymphocytes from ADHD children was found after only 1 week of MPH therapy with a further decrease at 24 weeks of treatments in parallel with improvement in symptoms. These findings suggested HERV-H as a predictive marker of the response to MPH therapy despite the awareness that the absence of a non-responsive patient group is a major limitation of this research preventing definitive conclusions [129]. The validation of early results about HERV expression profile in ASD children and new evidence came from a paper we published in 2019 in which we also included the parents of children in order to investigate the parent-of-origin effects in ASD in terms of HERVs and immune deregulation. ASD children and their mothers shared common expression levels of HERV-H and HEMO and of the cytokines such as TNF-α, IFN-γ and IL-10 in peripheral lymphocytes. Therefore, the abnormal expression of HERVs and cytokines was not an exclusive trait of autistic patients but also of their mothers, suggesting a close mother-child association within the ASD families. Taken together, these findings support the potential use of selected HERVs and cytokines in a set of biomarkers that accounts for the multifaceted nature of the disorder and can complement existing clinical methods [130]. A subgroup of this cohort was also used to conduct a proof of concept study in which lymphocytes from ASD children and their parents were exposed to stimulating factors (Interleukin-2/ Phytohaemagglutinin) or drugs, such as the antiepileptic drug valproic acid (VPA) and the antiretroviral drug efavirenz (EFV) with the intent to investigate whether the expression level of HERVs and cytokines could be modulated. Lymphocytes from ASD children and their mothers share intrinsic responsiveness to stimulating factors and VPA in expressing HERVs and cytokines. EFV specifically restored the HERV activity with a concomitant modulation of cytokines, in particular lowering the pro-inflammatory while maintaining high regulatory ones. This evidence provided new insights into the potential role of HERVs as biomarkers of ASD and raising the possibility of using HERV expression as a therapeutic target for a tailored approach to patient care [86].

With the intent of deciphering other factors linked to HERV activity that could contribute to ASD pathogenesis, the expression of epigenetic effectors known to regulate HERV expression and brain functions has been evaluated in ASD children. The authors found a correlation among the overexpression of these

### *Human Endogenous Retroviruses in Autism Spectrum Disorders: Recent Advances… DOI: http://dx.doi.org/10.5772/intechopen.108671*

elements and several HERVs suggesting their involvement in pathogenetic mechanisms leading to ASD [131]. All the studies above described were conducted using peripheral lymphocytes from ASD cohorts in agreement with the findings of various research studies that there is a shared gene expression profile between whole blood and brain tissues, suggesting that the cautious and thoughtful use of peripheral gene expression may be a useful surrogate for analysis in the brain. Of course, the brain remains the district of choice for studying neurodevelopmental alterations and to circumvent the issue, different preclinical models of ASD have been developed enabling studies on the aetiology, pathogenesis and possible prevention and treatment modalities of ASD. The main categories of models comprise genetic animal models, idiopathic strain, models of infection/ inflammation and chemically induced animal models displaying robust and well-replicated social deficits and repetitive behaviours [132]. The idiopathic strain BTBR T+tf/J (BTBR) and the prenatally CD-1 VPA-induced models of ASD have been extensively studied to characterize the expression profile of ERV and immune mediators. As such, whole embryos at about half of gestation, brain and blood tissues at different postnatal ages have been analysed. Both ASD models showed higher expression levels of ERVs beginning from intrauterine life and up to adulthood (**Figure 2**). Moreover, the aberrant expression of some ERV families correlated with expression levels of pro-inflammatory cytokines and TLR-3 and TLR-4 in embryos and brain tissues, supporting the interplay between ERVs and neuroinflammation as contributing factors in the appearance of ASD-like phenotype [133]. Subsequently, we conceived a study to evaluate the multigenerational impact of prenatal VPA exposure, demonstrating transgenerational changes in both behaviour and ERV expression that last, with fading of epigenetic memories across generations, till the third one that lacks a direct exposure

#### **Figure 2.**

*The abnormal ERV, cytokines and TLRs expression in mice with autistic-like phenotype prenatally exposed to valproic acid (created by BioRender).*

to the drug (**Figure 2**). Of note, offspring from maternal lineages showed more marked transcriptional effects compared with paternal lineages both in the second and in third generations suggesting the hypothesis of maternal imprinting as a contributing factor in increasing susceptibility to ASD [134]. Moreover, these findings also suggest ERV blood transcriptional levels as a stable peripheral biomarker, even at early life stages, of derailed brain development [135]. In our last paper, we demonstrated that MIA induced abnormal expression of ERVs and immune mediators in mouse off-spring in a sex-dependent fashion. Specifically, we demonstrated that the prenatal exposure to Poly I:C in C57BL6/J mice induced a tissue-specific expression of several ERVs, ERV-related genes and inflammatory mediators with ERVs as the main carriers of the peculiar profile found in brain areas from Poly I:C mice. In addition, Poly I:C induced larger effects on the expression of some retroviral elements only in pre-frontal cortex from female offspring reinforcing the view on sex bias as a possible risk factor for ASD [136].

Taken together, these findings candidate ERV activation as common feature shared by several risk factors of appearance of ASD and suggest ERVs as main carriers of changes occurring in brain from autistic mice, primarily in female offspring. These papers could represent the starting point to set up a new preclinical experimental design to verify the chance to inhibit ERV activity to rescue ERV activation, immune dysregulation and ASD-like phenotype observed in offspring trying to figure out cause and effect in this complex interplay. Moreover, a deep characterization of the molecular mechanisms by which gender differences could affect the neurodevelopment will help in identifying gender-specific diagnosis and personalized treatment strategies.
