Social behavior in 16p11.2 and 22q11.2 copy number variations: Insights from mice and humans

1 INTRODUCTION

Copy number variations (CNVs) are submicroscopic deletions or duplications of DNA segments. CNVs have consistently been implicated as vulnerability factors for neurodevelopmental disorders.1 CNVs genetic alterations can range from 102 to 106 base pairs and consist of complex alterations of homologous sequences at multiple sites in the genome or of simple gains or losses of genetic information.2 Consequently, they can impact multiple protein-coding genes and regulatory regions.3

The total prevalence of recurrent neurodevelopmental disorders associated with CNVs (duplications and deletions) has been estimated in live-born children to 0.48% that is, ~1 in 200 newborns has either a deletion or duplication.4, 5 Noteworthy, there are hotspots for CNVs in the human genome.6 In this review, we will focus on 16p11.2 and 22q11.2 genetic loci that are among the most common and most studied CNVs with high penetrance to neuropsychiatric disorders.5, 7 CNVs on 16p11.2 and 22q11.2 genetic loci are associated with increased risk of developing schizophrenia, attention-deficit/hyperactivity disorder (ADHD) and autism spectrum disorder (ASD), compared with the general population. More specifically, CNVs at the proximal region of the 16p11.2 locus refer to the deletion or duplication of a segment of 600 kb with breakpoints (BP) at 29.5 and 30.1 Mb (BP4-BP5).8 22q11.2 microdeletions (also named as Di George syndrome and velocardiofacial syndrome), refer to the deletions of a segment of 1.5–3 Mb on the long arm of chromosome 22.9-11 Duplication cases are less frequently reported and still less described, with reported phenotypes that can range from mild cognitive impairment to apparent normal cognition.12, 13

Here, we will concentrate on social behavioral alterations associated with the 16p11.2 and 22q11.2 hemi-deletion syndromes, and we will contrast the inherent available data collected in humans and in mouse models. Our goals are to (i) better integrate human observational studies with the biological findings derived from animal research; (ii) highlight similarities and distinctions between different CNVs conditions. See Figure 1 for a summary. We believe this might ultimately serve as a powerful platform to advance our knowledge of social processes and suggest more efficient strategies to ameliorate social dysfunctions.

image Summary of the main findings in humans and in mice models of 16p11.2 and 22q11.2 deletions. For humans, arrows indicate presence (↓) or absence (=) of problems, identified through the mentioned instruments. The arrows are therefore not directly referring to higher or lower raw scoring of the instrument. For mice, symbols represent the trend of mice performances during the tasks used for assessing the social behavior, in comparison with their controls. Specifically, we used (↓) or (=) to indicate, respectively, the decrease or the comparability of tests results deriving from 16p11.2 and 22q11.2 deletion mice with those shown by control mice. Notably, the variability of results within each social item could be a consequence of genetic and/or environmental factors. Refer to the text for more details. (Pictures created with BioRender.com)

A recent review summarized the genotype–phenotype correlations in proximal 16p11.2 deletions and duplications in humans and mice.14 A comprehensive description of the behavioral features of mouse models and of the human phenotype from a diagnostic perspective were reported, but the social component was only briefly described. Similarly, other reviews have compared ASD- and schizophrenia-relevant phenotypes in humans and mouse models of 22q11.2 CNVs,15-17 or the impact of specific single genes contributing to these developmental disorders.18 Moreover, Jonas et al. have provided an overview of neuropsychiatric implications of 22q11.2 deletion describing human and mouse studies.19 However, in all cases, the social phenotypes have only been marginally characterized and addressed. We aim to provide a critical review of the literature on 16p11.2 and 22q11.2 CNVs and their influence on distinct aspects of social processes that might inform on future studies beyond more strict psychiatric classification.

2 METHODS

We collected the most relevant literature papers from PubMed by using the following search string (“16p11.2” OR “22q11.2”) AND “deletion” AND (“social” OR “sociability” OR “emotion” OR “emotional” OR “face” OR “recognition” OR “cognition” OR “cognitive” OR “behavior” OR “behavior”). For description of the selection process, please see Figure 2. We included all search results covering human and mouse studies. Case-reports, commentaries and reviews, were excluded. Only studies including information about 16p11.2 and 22q11.2 deletions and social or behavioral measures were included.

image

Flow chart summarizing the selection process of papers included in the review. The full process of identification, screening, eligibility and inclusion of the literature papers relevant for the purposes of this review is summarized in the flow chart. For each step, the number of papers that were included or excluded is reported, for both humans and mouse studies. The main reasons for exclusion are mentioned for each step

2.1 Preliminary considerations for studies in humans

Even though 16p11.2 and 22q11.2 CNVs are associated with a high risk of developing specific neurodevelopmental disorders and other psychiatric conditions, not all carriers with these CNVs have impairments that reach the threshold for a psychiatric diagnosis. In fact, carriers of CNVs can present sub-threshold abnormalities in specific social behavioral sub-domains, which do not require clinical attention. In depth characterization of abnormalities in these sub-constructs could improve our understanding of their psychopathological trajectories, which may lead to a better understanding of the underlying biological mechanisms, and consequently, to an improved design of animal studies.

Attempting to describe how each social sub-construct could be impaired in CNVs carriers comes with methodological issues. The biggest problem, which we cannot overcome in the context of this review, is that studies on CNVs have a high risk of an ascertainment bias. The genetic testing is usually performed only on patients with a clinical indication, usually developmental delay. This means that CNVs carriers without strong clinical impairments could be underrepresented, or represented in a different manner in each cohort, depending on the criteria for the access to genetic testing. The only way to solve this issue would be screening for CNVs all newborn children prior to the identification of impairments.

Secondly, CNVs carriers have rarely been studied with the specific aim of characterizing their social profile: in the majority of cases, the assessment of social and behavioral impairments in CNVs is made with the aim of identifying disorders with a categorical diagnostic approach. Thus, most of the available information on social impairments is derived from studies using assessment tools designed for screening and diagnostic purposes. Some of the exceptions are the cognitive batteries often used in the context of magnetic resonance imaging studies, or scales such as The Awareness of Social Inference Test (TASIT), designed to study theory of mind.20 Consequently, the most commonly used assessments tools usually individuate thresholds in macro-areas of impairments in order to support or disprove the diagnosis, rather than describe a social profile of the tested subjects.

The most frequently used tools for characterizing CNVs that carry information on social impairments are the Autism Diagnostic Observation Schedule (ADOS),21 the Autism Diagnostic Interview–Revised (ADI-R),22 the Social Responsiveness Scale (SRS),23 and the Social Communication Questionnaire (SCQ).24 It is important to note that the majority of these are employed in the context of screening and assessment of ASD: this is because diagnosing ASD (but not schizophrenia, or ADHD) requires assessment of social impairments. Typical social impairments observed in ASD involve three social constructs: (1) Socio-emotional reciprocity (sharing of emotions or interests, initiating or responding to social interactions, such as conversations and social approach); (2) Nonverbal communication used for social interaction (e.g., expressing or understanding social cues such as facial expressivity, eye contact, body language and integrating verbal and nonverbal communication); (3) Ability to develop, maintain and understand relationships (adjusting behavior in a context-appropriate manner, showing interest in peers, sharing imaginative play with friends).25 Consequently, sub-scales of the diagnostic tools often contain evaluations of the abovementioned social constructs. Another source of indirect information are the scales used for the dimensional assessment of emotional and behavioral problems or for the evaluation of the patient's overall ability to function in neuropsychiatric disorders. Some examples are the Child Behavior Checklist (CBCL)26 and Vineland Adaptive Behavior Scale (VABS).27 See Table 1 for a summary of the properties of these tools. Overall, even if the purpose of the abovementioned scales is not the assessment of social domain impairments of CNVs carriers, their sub-scales can provide useful information on specific social constructs which may be impaired. It is important to keep in mind that diagnostic criteria of neuropsychiatric disorders have evolved over time, and that instruments for the assessment have changed as well. For example, while in DSM-IV and ICD-10 the “Social Interaction” and the “Communication” constructs were considered separate contributors to the diagnosis of ASD, these two dimensions were collapsed in the DSM-528 and in the ADOS as well.21 This means that the reporting of some sub-constructs might be obscured in the most recent studies. To overcome these limitations, we grouped results from studies using similar diagnostic assessment tools, to improve homogeneity when describing domains of social impairments.

TABLE 1. Summary of scales, and their sub-scales, with information on social impairments in ADS Instrument Sub-scales Examples of social features evaluated Autism Diagnostic Observation Schedule (ADOS)

Social-affective

Repetitive behaviors

Social overtures

Social responses

Initiating and maintaining behavior

Eye-contact

Facial expression

Gestures

Quality of social response

Speech and conversation abnormalities

Amount of reciprocal communication

Imagination

Autism Diagnostic Interview-Revised (ADI-R)

Social interaction

Communication

Repetitive behaviors

Social interaction

Peer relationships

Shared enjoyment

Socio-emotional reciprocity

Gesture

Imitation and imagination

Social Responsiveness Scale (SRS)

Social awareness

Social cognition

Social communication

Social motivation

Mannerisms

Awareness: perceiving social cues

Cognition: interpreting social cues

Communication: expression of social communication

Motivation: motivation to initiate and maintain a social contact

Social Communication Questionnaire (SCQ)

Reciprocal interaction

Communication

Restricted interests

Speech and conversation abnormalities

Body language expression

Facial expression of emotion

Interest in other children

Social play

Child Behavior Checklist (CBCL)

Total problems

Externalizing

Internalizing

Aggressive behavior

Anxiety/depression

Attention problems

Rule breaking problems

Social problems

Somatic complaints

Thought problems

Withdrawal

Withdrawal: Shyness, Preference for loneliness, refusal to talk, secretiveness

Social Problems: sociability with peers, quality of social interaction (being not liked by peers or teased, preferring company of younger peers), being dependent, being lonely, speech problems, clumsiness

Vineland Adaptive Behavior Scale (VABS)

Communication

Socialization

Daily living skills

Motor skills

Communication: Receptive, written, expressive communication

Socialization: interpersonal relationships, play and leisure, ability to cope with social rules

2.2 Social impairments in human 16p11.2 proximal deletion carriers

Carriers of the proximal 16p11.2 deletion (16p-DEL) often have significant impairments in the social domain.29 In children with this deletion, social impairments can include deficits in imitation, showing interest, playing, and expressing social cues when interacting with peers.30 This CNV has been strongly linked with ASD and around 0.50% of the probands with ASD receiving a genetic test carry the 16p11.2 proximal microdeletion.31 On the other hand, among 16p-DEL the prevalence of ASD has been reported to vary from 11%8 to 27%29, 30 when assessed with both the ADOS and the ADI-R, and to vary from 22%32 to 43%33 when assessed with the ADI-R alone.

Within the 16p-DEL population, specific construct alterations have been described. Excluding the 16p-DELs diagnosed with ASD, 60% of the remaining 16p-DEL presented impairments in the ADI-R Social Interaction sub-domain, and 73% in the Communication sub-domain.33 Similar data have been reported in independent cohorts29, 30 where 16p-DEL carriers showed higher Social Interaction and Communication impairments compared with siblings.29 Socialization was significantly impaired when measured with the VABS as well but not consistently.30, 34, 35 Interestingly, Moreno-De-Luca et al. reported that the SRS score and the VABS Socialization score of 16p-DEL were significantly correlated with those of their parents and siblings respectively, suggesting that parental ability to perceive, understand, communicate and their motivation to interact socially could be potential contributors to social impairments in their children.34 Similarly, Hudac et al.36 described a significant correlation between the VABS total score and perinatal events, with impaired adaptive abilities linked to perinatal complications (e.g., preterm labor, abnormal presentation, C-section, respiratory distress, low APGAR score and/or low birthweight). A similar trend was observed for the ADOS social-affective score (ADOS-SA), measuring both Social Interaction and Communication.36 Overall, these findings show the presence of social impairments in a subset of 16p-DEL carriers, involving both communication and social interaction, albeit with milder presentation in 16p-DEL than in idiopathic autism (iASD) without CNVs.33, 37 Notably, the presence of social deficits in 16p-DEL could be linked with perinatal complications and with the level of social responsiveness in their parents.

2.3 Social impairments in human 22q11.2 deletion carriers

Children with 22q11.2 deletion have been described as having poor social skills, including difficulties in initiation and attaining social relationships, shyness and poor relationships with peers38, 39 and these impairments exacerbate in late childhood.40 Rates of ASD in carriers of the 22q11.2 deletion (22q-DEL) vary from 23%33 to 42%9 across studies, however higher percentages were reported in studies using the SCQ, varying from 26%41 to 91%.42 More reliable assessments for ASD,43 using both the ADI-R and the ADOS, reported a lower 28% of ASD in 22q-DEL.44 When the clinician evaluation was added to the interviews, as recommended by guidelines, a further reduction in ASD prevalence (16%–18%) was reported.45, 46

Taken together, these data indicate the presence of social and behavioral impairments in 22q-DEL that may overlap with symptoms of ASD. Within the 22q-DEL group, subjects with ASD (22q-DEL + ASD) had more severe Socialization problems (measured with the VABS) than 22q-DEL without ASD, and more social problems assessed with the CBCL, indicating poor quality of the interaction with peers.9 The ADOS-SA differed between children with 22q-DEL + ASD and 22-DEL without ASD as well, indicating more severe impairments in 22q-DEL + ASD than in 22q-DEL without ASD in the sub-domains of social interaction and communication.45 To better characterize these impairments and their link with 22q-DEL (with or without ASD), carriers of this CNV were compared with patients with iASD. A less impaired social profile was observed in 22q-DEL than in iASD, confirmed by both the ADI-R and ADOS, and by the CBCL Withdrawal.45 Similarly, the ADOS-SA was higher in the iASD group compared with the 22q-DEL + ASD, who displayed relative strengths especially in nonverbal communication, including facial expression and gestures.45 However, the ADI-R did not confirm these differences, with similar ADI-R social interaction and communication scores in the iASD group and in the 22q-DEL + ASD.45 Thus, such deficits might be classified as ASD symptoms, but they are milder and without the typical presentation. Taken together, these findings suggest that the 22q-DEL + ASD could represent an intermediate social phenotype between iASD and 22q-DEL (iASD >22q-DEL + ASD > 22q-DEL > general population).

To understand if some social sub-domains are more impaired than others in 22q-DEL, even without a clinical diagnosis of ASD, we looked for studies reporting the percentages of carriers scoring above the cut-offs in the most relevant sub-scales of the diagnostic tools (communication and social interaction). The sub-domain of Communication measured with the ADI-R was impaired in 32% of subjects with 22q-DEL45, 46 and in 47% of the 22q-DEL without ASD.33 Among 22q-DEL with high total ADOS scores, impairments were mostly related to the Communication sub-domain rather than social interaction.47

Impairments in the sub-domain of Social Interaction were reported in 22q-DEL children in percentages varying from 36%46 to 40%,45 as measured with the ADI-R. Even after excluding subjects with a diagnosis of ASD from the sample, 46% of the remaining 22q-DEL without ASD still presented impairments at the ADI-R Social Interaction.33 Presence of Social Problems, assessed with the CBCL, were often reported in children with 22q-DEL38, 48-50 ranging from 27% to 37%.51-53 Clinical elevations in the CBCL Withdrawal subscale was reported as a feature in 12%–14% of cases of 22q-DEL children as well.49, 53 However, despite the deficits in this sub-domain, it was suggested that 22q-DEL children may still understand emotions and thoughts from others and may also be interested in social interactions.44 Notably, 22q-DEL social impairments seem to intensify when growing older50, 54-56 and to correlate with deterioration of intellectual functioning, with greater deficits observed for lower IQ57 and for executive functions impairments.58 Thus, while the ability to adapt behavior to the context becomes increasingly important with age, deficits in the capability to correctly perceive,40 contextualize, process social information,59 and give adequate social responses, could determine the social problems observed in 22q-DEL. These abilities are connected to the social-cognitive domain of theory of mind, meaning the ability of perspective-taking and abstracting feelings, thoughts, intentions of others from the context, during a social interaction.60 Deficits in theory of mind, assessed through the TASIT, have been consistently described in 22q-DEL61-63 and have been linked to their risk of developing schizophrenia.64, 65 Considering that about 25% of 22q-DEL carriers develop schizophrenia later in life,66 the link between social impairments and cognitive dysfunctions, and the progressive decline of social functioning after childhood, the presence of social deficits in 22-DEL that are less severe than in iASD could represent a manifestation of vulnerability for psychotic disorders. Overall, the findings reported so far suggest more socio-cognitive deficits in 22q-DEL which tend to become exacerbated throughout development.

Although not limited to ASD, face identification and emotion recognition are other social-cognition sub-constructs extensively studied in 22q-DEL, with contrasting results.40, 67, 68 Some studies have described impairments in recognition of faces in 22q-DEL.40, 69 Moreover, 22q-DEL were reported to have more problems properly identifying emotions compared with controls. Deficits in the recognition of angry, disgusted, fearful and neutral faces but not of happy, surprised and sad faces were described,70, 71 even though the results are not consistently replicated.72-74 Accuracy and reaction times in emotion recognition were measured in some studies, with opposing results.54, 75-77 Such discrepancies could be because of differences in the intensity of emotions presented during the tests, as 22q-DEL carriers require more intense emotional cues to properly perform in these tasks.78 Moreover, they display less fixation events over face emotional stimuli79 and they seem to focus their gaze more on face regions less relevant for emotion recognition (such as mouth) rather than exploring more relevant regions such as eyes.80 Interestingly, 22q-DEL show less impairment in face emotion recognition than iASD, but have poorer performance than people affected by schizophrenia, subjects with high psychotic risk and healthy family members.69, 81 Together, these results indicate that emotion recognition might add important information about social impairments in 22q-DEL and on the longitudinal evolution of their psychopathology. We hypothesize that this could be true for 16p-DEL as well, given that—to our knowledge—there are no studies on this. Indeed, a full social evaluation of CNVs carriers at risk of developing for neuropsychiatric disorders, which includes an emotion recognition assessment, could add relevant information to the specific profiles of social impairment for each CNV.

2.4 Comparisons of 16p11.2 and 22q11.2 social profiles

Comparing social profiles of different CNVs could aid our understanding of how different genetic variants contribute to specific deficits. Two studies have done this so far.33, 82 Cunningham et al.82 investigated the presence of emotional and behavioral abnormalities in 13 CNVs, including 16p-DEL and 22q-DEL, through the Developmental Behavior Checklist, an instrument specifically created to evaluate problems in intellectual disabilities. Although assessing social abnormalities in those two CNVs was not a specific purpose of the study (and the results were not discussed), the social phenotype of these two CNVs, compared with the others, seemed to present the mildest impairments, among all, in Communication, Social Interaction and self-adsorption. Chawner et al.33 compared deletions and duplications at the 16p and 22q loci, with the purpose of characterizing the ASD profile of these CNVs and comparing ASD-relevant impairments in the CNVs with iASD. 22q-DEL seemed to be the least impaired in the Social Interaction and Communication sub-domains of the ADI-R. Generally, their profile was comparable with the one of 16p-DEL, with the exception of socio-emotional reciprocity and socially sharing of enjoyment, in which they displayed relative healthy functioning compared with 16p-DEL. In any case, both the 16p-DEL and 22q-DEL were significantly less impaired in all the social constructs relevant for ASD, compared with the carriers of the duplications in the same loci, as well as compared with iASD. Interestingly, the Social Interaction impairments were less severe in 22q-DEL when compared with other diagnostic groups displaying social deficits, such as patients with schizotypal personality disorder83 and comparable with those of people with idiopathic intellectual disabilities.44 The same comparisons were not available for 16p-DEL.

The promising recent studies comparing CNVs reviewed above will hopefully be followed by more research using this approach. Indeed, as mentioned above, describing social profiles in CNVs carriers can lead to important insights on the interaction between genetics and behavior, and could address the determinants bridging the gap from genotypes to phenotypes. In fact, all the studies reported so far confirm that not all carriers of a CNV display social impairments. The cause of this variability of phenotypical effects within the same CNV is still unknown. It has been suggested that neurodevelopmental disorders could arise when presence of CNVs is conjoined with a second perturbing factor that adds its effect to the CNV-related vulnerability.84, 85 Perturbing factors can be genetic “second-hits,” such as simultaneous presence of another CNV or single nucleotide variations, or environmental factors such as maternal immune activation, perinatal events, and parental functioning. We previously mentioned that the presence of perinatal events and parental deficits in social functioning was correlated with higher social problems in 16p-DEL.34, 36 Moreover, in 22q-DEL, the presence of pro-inflammatory cytokines has been linked with increased ASD-related impairments, indicating a potential role of inflammation.86 In our opinion, mouse studies on CNV represent a promising model to investigate the genotype–phenotype gap, because they would enable the examination of in-depth genetics and environmental factors as well as the biological mechanisms implicated in distinct social processes.

2.5 Preliminary considerations for studies in mice

Mice are the most commonly used animal models to study genetic alterations87-89 and several mouse lines with deletions of the syntenic 16p11.2 and 22q11.2 human regions have been generated (Table 2). The main goal of these mouse lines is to correlate altered behavioral phenotypes with brain circuits functioning and specific molecular/cellular pathways. However, studies on social profiling are still limited, especially in relationship to the assessments reported above in humans. In this section, we review the existing literature investigating social behavior in 16p11.2 and 22q11.2 deleted mouse lines, compare their profile, and highlight similarities and discrepancies with the literature describing human studies.

TABLE 2. Mouse models of 16p11.2 and 22q11.2 deletions reported in studies 16p11.2 deletion Model Deleted region Features Genetic Background of the original mouse line Genetic Background Tested in the experiment Mills90 Slx1b-Sept1 interval on mouse chromosome 7F4

- Locomotor activity alterations (hyperactivity)

- Affected brain volume size and synaptic defects

- Sleep disorder

- Wider deletion than in humans (includes four genes outside the human BP4-BP5 interval)

C57BL/6N:129Sv

- B6129S-Del(7Slx1b-Sept1)4Aam/J are generated from frozen sperm of df/+ mice mated to C57BL/6J mice for a minimum of two generations and used to fertilize B6129SF1/J oocytes.91

- C57BL/6N129Sv92

- C57BL/6 N93

Dolmetsch94 Coro1a-Spn interval on mouse chromosome 7F3 - Hearing deficit (no particular hearing defect has been reported in humans with 16p11.2 CNVs)95 C57BL/6N mixed background: 67% of C57BL/6N, 30% of 129P2/Ola, and 3% CD-195-97 Herault98 Sult1a1-Spn interval on mouse chromosome 7F3

- Homologous to human 16p11.2 BP4-BP5

- Created to overcome the limits of the two previous models

- Inbred C57BL/6N genetic background

- C57BL/6N

- F1 C57BL/6NxC3B

- C57BL/6N

- F1 C57BL/6NxC3B98

22q11.2 deletion Model Deleted region Features Genetic Background Df(16)A+/−99 Dgcr2-Hira interval (1.3-Mb) on mouse chromosome 16

- Syntenic to the 1.5-Mb human 22q11.2 microdeletion

- Smaller dendritic spines

- Impaired sensorimotor gating

- Acquisition of a spatial working memory–dependent

C57BL/6J C57BL/6J100 Df1/+101

Es2-Ufd1l interval (∼1.2 Mb) on mouse chromosome 16

- Syntenic to 1.5 Mb region in human chromosome 22

- Cardiovascular defects (are similar to those seen in patients)

- Hearing difficulty

- Impaired sensorimotor gating

- Deficits learning and remembering of complex cues102

- 129SvEvBrd (129S5) × C57BL/6

留言 (0)

沒有登入
gif