Communication deficits in a case of a deletion in 7q31.1-q31.33 encompassing FOXP2

ABSTRACT Copy number variants (CNVs) found in individuals with communication deficits provide a valuable window to the genetic causes of problems with language and, more generally, to the genetic foundation of the human-specific ability to learn and use languages. This paper reports on the language and communication problems of a patient with a microduplication in 22q11.23 and a microdeletion in 7q31.1-q1.33 encompassing FOXP2. The proband exhibits severe speech problems and moderate comprehension deficits, whereas her pragmatic abilities are a relative strength, as she uses gestures quite competently to compensate for her expressive issues. This profile is compatible with the deficiencies found in patients with similar CNVs, particularly with people bearing microdeletions in 7q31.1-q31.33.


Introduction
Copy number variants (CNVs) are chromosomal deletions or duplications that alter the number of genes, and ultimately gene dosage within the affected genomic fragments.When they are found in individuals with communication deficits, they provide a valuable window to the genetic causes of language problems and more generally, to the genetic foundations of the human-specific ability to learn and use languages.To date, dozens of CNVs with an impact on language have been described (see, Gialluisi et al., 2016 for review).Delving into the deficits and the strengths of patients in the language domain enables to achieve more robust links between specific genes and aspects of language processing, and ultimately, to optimise strategies aimed to ameliorate symptoms and improve the communication skills of the affected people.In this paper, we report in detail on the language and communication problems of a girl with a microduplication in 22q11.23 and a microdeletion in 7q31.1-q31.33encompassing the well-known FOXP2 gene.
The 7q31.1-q31.33region is part of a hotspot for language problems (Schneider et al., 2012).However, a precise characterisation of the language deficits exhibited by individuals bearing deletions in this region is still pending.Some studies suggest that deletions of the 7q31.1-q31.33region result in broad neurodevelopmental disorders impacting language (Gimelli et al., 2014), whereas others point to the occurrence of verbal production difficulties only (Lennon et al., 2007;Palka et al., 2011;Zeesman et al., 2006) and/or delayed speech and language (Rieger et al., 2020).Regarding the genetic causes of the observed deficits, most studies have linked the speech and language problems to the haploinsufficiency (i.e. the inactivation or the absence of one copy) of the FOXP2 gene, located at 7q31.1.FOXP2 is a robust candidate for speech and language impairment (Graham et al., 2015;Kurt et al., 2012;Vargha-Khadem et al., 2005).This gene plays a role in neuron proliferation, differentiation, and migration in selected brain areas important for language processing (Chiu et al., 2014;García-Calero et al., 2016;Tsui et al., 2013).More specifically, the gene has been suggested to contribute to auditory-motor association learning (Kurt et al., 2012), as well as to the transition from declarative (i.e.explicit) to procedural (i.e.automatised) performance (Chandrasekaran et al., 2015;Schreiweis et al., 2014).Some research has proposed the involvement of other genes besides FOXP2 in the language deficits exhibited by subjects with deletions of the 7q31.1-q31.33region (Zhao et al., 2016).Nonetheless, no studies have investigated the impact of microdeletions in 7q31.1-q31.33 in the broader area of communication, here understood as the fundamental sphere for social interaction, entailing the holistic interaction between cognitive, linguistic, and social skills.
The main contribution of our work is thus the detailed characterisation of the communicative profile of our patient (beyond her language deficits and strengths), which we find important, as noted, for a better understanding of the interactional problems exhibited by people bearing deletions of this chromosomal hotspot for language problems.

Language and cognitive assessment
The global developmental profile of the proband was evaluated with the Spanish version of the Portage Developmental Guide (PDG; Bluma et al., 2004) and the BOEHM test of basic concepts (BTBC-3;Boehm, 2003).The proband's global language skills were evaluated with the Peabody Picture Vocabulary Test (PPVT-3;Dunn et al., 2006).Her comprehension abilities were assessed with the Token Test Short-Form (TT;De Renzi & Faglioni, 1978), and with narrative stories designed ad hoc.The Registro Fonológico Inducido test [Elicited Phonological Register] (RFI; Monfort & Juárez, 1988) was used to gain a better understanding of her expressive abilities, and specifically, to evaluate the proband's phonological awareness.Her voice profile was examined with the Estill Voiceprint Plus Program.Finally, to evaluate her communication abilities and how language knowledge is put into use, we conducted a pragmatic analysis of three conversational exchanges involving our participant.For this, we relied on the Protocolo Rápido de Evaluación Pragmática (PREP-R; Martínez-Urquiza et al., 2015).Because of the severe speech deficits exhibited by our participant, we also employed the motor expression subtest of The Illinois Test of Psycholinguistic Abilities (ITPA; Kirk et al., 2009) to evaluate her ability to use gestures for communicating.We now provide a brief characterisation of the tests and tools mentioned above.

The Portage Developmental Guide (PGD)
This is an assessment guide aimed to determine the general abilities of children from 0 to 6 years of age.It is organised around five developmental areas: socialisation, language, self-development, cognition, and motor skills.These domains are evaluated by a mixture of direct observation and direct instructions to the child aimed to elicit specific behaviours.This is not a standardised test, but it provides some basic information about the developmental stage achieved by the child.

The BOEHM test of basic concepts (BTBC-3)
This test evaluates some of the basic abilities that are necessary for attending school, including language comprehension skills and the overall cognitive development.It is designed for children between 3 and 6 years of age.It examines 50 basic items via coloured pictures and drawings.The evaluator first provides a verbal description to the child, and she must then point to the correct image among three of them.This test is mostly a screening tool, not a diagnostic test, so that the resulting scales just provide a rough indication of the knowledge by the child of basic concepts, such as left-right or up-down.

The Peabody Picture Vocabulary Test (PPVT-3)
This standardised test is aimed to assess the acquisition of receptive vocabulary by individuals from 2.5 to 90 years of age.It consists of 192 items.After hearing a word, the participant must point to one of four different coloured drawings.This test also enables to screen problems with language comprehension.

Token Test Short-Form (TT)
This standardised test aims to identify subtle auditory comprehension deficits in non-verbal subjects.This is a gestural test.In our analysis, we employed the short version of the test, which comprises just 36 items.During the evaluation, the participant must attend and follow different verbal instructions, from simple to complex commands, that involve tokens that differ in colour, form (e.g.squares vs circles) and size (e.g.large vs small).

Illinois test of Psycholinguistic Abilities (ITPA)
As noted, we only used the motor expression subtest of the ITPA, because our proband has notable problems with expressive language.This is a standardised test aimed to evaluate the proband's ability to convey meanings related to actions through gestures and body changes.Therefore, this subtest demands a good command of gross psychomotor skills.The subtest can be administered to children from 3 to 10 years of age.Overall, 14 concepts related to specific gestures and motor actions are evaluated.During the test, the experimenter asks the child to reproduce the gesture or movement that she usually associates with a particular concept (for example, the gesture of combing the hair after hearing "hair comb").

Narrative stories
Three stories were created ad hoc to evaluate the proband's ability to understand an oral narration.A reason was that the available standarized tests might contain linguistic structures that were too complex for the child; in other cases, the structures we were interested in lacked appropiate companion pictures.Each narrative consisted of around 45 words.The story was read aloud by the experimenter using a drawing as ancillary material.Texts contained simple sentences only, with just one single verb per sentence.The first story was descriptive by nature, reporting on characters and their interests.The second story was a genuine narration, reporting on two children that are playing together.The third story was more complex, describing how a child gets lost in a forest.For each story, two questions were asked to the proband.Overall, six different types of questions were tested: textual (i.e.questions about the characters or the plot, like 'Who went to the forest?' or 'Where were the children playing?'), inferential (e.g.'Why did Eric start to cry?'), yes/no questions (e.g.'Were there some plates on the table?'),polar questions (e.g.'Did Tomás like the cakes?'), identification questions (i.e.questions aimed to identify particular characters, like 'Who is Ana?'), questions with two possible answers (e.g.'What did the children drink: juice or water?'), and open questions (e.g.'How did Ana feel when she found her house?').

Elicited Phonological Register (RFI)
This test assesses phonological and articulatory abilities of Spanish-speaking children between 3 and 7 years of age.This is an elicited (word) production and repetition test.Word production is triggered by showing the child several pictures that she must name.If the child cannot generate the correct name by herself, the experimenter provides the appropriate name and then asks the child to repeat it.The test comprises every possible sound occurring in Spanish, thus enabling to detect dyslalia (i.e.phonetic difficulties) and phonological impairment (i.e.inappropriate understanding of sound use for conveying meaning).

Estill Voiceprint Plus Program
Estill Voiceprint Plus is a real-time spectral analysis program that allows to display, record, and playback the human voice.It provides meaningful information about the participant's pitch and voice quality.Voice samples from the proband were collected from three different conversational settings: during a free-play activity, while speaking with her parents or the speech therapist, and during an elicited task in which she was asked to repeat several utterances.

Protocolo Rápido de Evaluación Pragmática (PREP-R)
This protocol is used to assess the participant's broad communication skills.The test evaluates whether the proband can produce utterances with the appropriate illocutive force (i.e. that can fulfil the intended objective), and if she has acquired the needed skills to be a competent interlocutor (e.g. if she knows how turn-taking works during a conversation).Three different interactions were examined, each lasting 10 minutes and each representing a different communication setting: playing at home, attending a session with the speech therapist, and playing with a close conversational partner.Interactions were videorecorded to also analyse paralinguistic (i.e.voice effects) and non-linguistic elements (i.e.gestures and body language).PREP-R is not a standardised test, but it provides a good characterisation and understanding of the proband's communicative behaviour.

Microarrays for whole-genome CNVs search and chromosome aberrations analysis
The DNA from the patient was hybridised on a CGH platform (Agilent Technologies).The derivative log ratio spread (DLRS) value was 0.25.The platform included 750.000 probes.Data were analysed with the Chromosome Analysis Suite (ChAS) v. 4.0.0.385 from Applied Biosystems© using CytoScan750K_Array Simple Analysis, NA33.2.The SNP QC was > 15, and the Waviness SD < 0.12.Only CNVs > 100 kb and ROHs (Regions of Homozygosity) > 5 Mb were considered.Breakpoints were not resequenced, nor an exome analysis was conducted to validate the CNVs.The purported pathogenic nature of the detected CNVs was determined according to the pathogenicity criteria by the Database for Genomic Variants (DGVs) (http://dgv.tcag.ca/dgv/app/home)and The International Standards for Cytogenetic Array Consortium (ISCA) (http://dbsearch.clinicalgenome.org/).Both the microarray and the preliminary analysis and clinical interpretation of the findings were conducted by the geneticists from Igenomix (https://www.igenomix.es/),a CLIA-certified laboratory in Paterna, Spain.One of the authors (ABB), who has expertise in genetics and molecular biology, conducted the subsequent in-depth analysis of the data, as shown and discussed in the paper.

Clinical history
The proband is a girl born from healthy parents that are monolingual in Spanish.The baby was delivered by caesarean section after 37 weeks of pregnancy.At birth, her weight was 3450 g (65 th percentile), her height was 50 cm (54 th percentile), and her occipitofrontal circumference was 37 cm (98 th percentile); the Apgar scores were 9 (at 1') and 10 (at 5').The pregnancy was uneventful.At 12 months of age, the child was diagnosed with facial dysmorphism (i.e.abnormal ear size), as well as congenital clinodactyly (i.e.abnormal curvature of the index finger of her right hand).At the age of 18 months, she was diagnosed with developmental delay, mostly impacting on her speech, motor (fine and gross) control, and intellectual development.The proband started to walk at the age of 22 months.Due to her swallowing difficulties, she suffered from malnutrition and growing problems.At the age of 24 months, her weight was 8350 g (<50 th percentile) and her height was 85 cm (<50 th percentile).At the age of 3 years, the girl started to attend school, and 2 months later, she also began attending a speech therapy clinic.At that time, she was reported to suffer from a severe developmental delay.Expressive language was almost absent: she could barely utter sounds [a] and [e] only, and she mostly interacted with others using gestures.She also exhibited notable behavioural disturbances and problems for maintaining sustained attention during prolongued periods, even when playing.Furthermore, she had reduced tolerance to frustration, as well as tantrums and disruptive behaviours.
At the time of our assessment, the proband was 5 years and 9 months old.She was attending a Primary School and mainstreamed with her age peers in a non-adapted classroom, although she was receiving external aid from a speech therapist 3 days a week.

General cognitive development
The proband's scores in the tests aimed to evaluate her general developmental profile suggested that she was delayed in all the assessed domains.According to PGD test (Figure 1), the girl exhibited a 1-2-year delay in the self-development domain: e.g.she was able to undress by herself, perform self-cleaning, or feed herself, but only with adult guidance.
In the cognitive sphere, she presented with a 2-3-year delay.For instance, she succeeded in categorising elements that are similar, but failed in tasks demanding a link between conceptual items and motor movements (e.g.reproduce the movements needed for handwashing).Likewise, she was delayed in the motor domain, with gross motor skills being less impaired than fine motor skills.However, the most affected area was the communication domain.The girl had not gone beyond the first stage of socialisation, which is normally achieved at the age of 2-3 years.She did not interact with their peers while playing.She was unable to maintain the joint attention needed for successfully performing everyday tasks.In the domain of language, she exhibited a delay of 3-4-years, and she only uttered a limited set of holophrases (usually disyllabic, poorly articulated sequences, such as [pa.pa]), as well as rough attempts at mimicking the intonation profile of some common words (like in [aaa-AAAA], for papa [daddy]).In the BOEHM test, the proband obtained a direct score of 15 (1 st percentile) (Supplemental file 1).Although, she scored typically in tasks related to spatial and numerical cognition, she failed in tasks involving temporal concepts.Overall, the proband's ability to understand abstract concepts and to make inferences seemed seriously impaired.

Speech and language development
Expressive language.We conducted a detailed acoustic analysis of the proband's speech.The mean fundamental frequency of the proband's voice was 450 Hz.No interharmonic noise was detected, with noise/harmonics ratio (NHR) values being above normal values.By contrast, we observed signs of glottal onset and increased closed phase during the phonatory cycle, with a maximum phonatory time of 1.5 seconds (see, Figure 2 for details).
An endoscopy examination confirmed that the proband's difficulties for initiating phonation were due to an abnormal prephonatory activity, resulting in an excessive contraction and closure of the glottis and the supraglottic spaces.Overall, these anomalies Besides these phonatory problems, the proband exhibited notable articulatory problems.In the RFI test, she was only able to utter some isolated sounds, but not whole words as found in the test (see Supplemental file 2).Specifically, she only succeeded in articulating the voiceless velar stop [k], the two labial stops [p] and [b], and the alveolar and labial nasals [n] and [m].Likewise, she barely managed to combine any of these consonants with the vowels [a] and [o].Overall, the proband's profile was suggestive of dyslalia, resulting from severe form of apraxia caused by a poor control of orofacial movements.Finally, our analysis of the proband's spontaneous oral productions (including song reproduction) revealed that the girl also had severe problems with prosody.Accordingly, she was unable to confer proper intonational patterns to her vocalisations, either spoken or sung.For instance, when asking, she did not confer the typical rising pitch to the end of the utterance, as commonly found in questions; instead, pitch fell abruptly.The same difficulties with melodic patterns were observed when she tried to reproduce the music of popular songs, such as "Let it go", from Frozen, the Disney film.Overall, these abnormal features match the American Speech-Language-Hearing Association's characterisation of Childhood Apraxia of Speech (CAS), an "impairment in planning and/or programming spatiotemporal parameters of movement sequences that results in errors in speech sound production and prosody" (American Speech-Language-Hearing Association, 2007; the emphasis is ours).
Receptive language.Because of the severe expressive problems exhibited by the proband, we also evaluated in detail her verbal comprehension abilities, to know whether they were similarly impaired.According to the scores obtained in the PPVT-3 test, she had an associated intellectual quotient (IQ) of 55, which is suggestive of a 2-year delay.In the Token Test, she performed correctly in tasks evaluating shape-colour associations (e.g.'touch a blue square'), as well as shape-colour-size associations (e.g.'touch the small yellow circle').However, she failed in tasks demanding more complex sensory-motor integration, particularly in tasks involving some sort of spatial displacement or rotation (e.g.'place the big white triangle on a square').On average, she obtained a raw score of 23 in the test (Supplemental file 3), which is suggestive of moderate comprehension problems.Finally, in the story comprehension task, the proband correctly answered 83% of the questions about characters, but only 33% of the questions about the plot.She was unable to answer any of the inferential questions.Likewise, she provided correct answers to closed questions, polar questions, and questions aimed to identify characters.By contrast, she failed in more complex questions, like questions with two possible answers, or open questions.

Language in use.
As noted, we were interested not only in determining the proband's deficits (and strengths) with regards to the structural components of language, but particularly, her broader communicative (dis)abilities.A reason is that, as mentioned in the introduction, less is known about how language knowledge is put into use by people bearing similar CNVs to the ones found in our proband.Another reason is that despite the proband's marked deficits with oral language, she was able to convey some basic information via gestures.
Overall, our proband turned out to be quite a competent interlocutor, able to inform others about her basic needs.She had mastered some rudimentary rules for social interaction and compensated her problems with oral language through non-verbal strategies.For this, she did not rely on any learned augmentative or alternative communication system, but on her own gestures, mostly deictic cues pointing to surrounding objects (Table 1).In our in-depth analysis of the proband's conversational interactions, we found that she got the expected feedback from her conversational partners in nearly 90% of cases, although this is certainly explained by the circumstance that she normally interacts with familiar people who  knows her very well.At the same time, failed communication attempts normally resulted in disruptive behaviours by the girl, such as crying or been aggressive towards the receiver.As noted above, the proband's expressive language was reduced to combinations of sounds mimicking some common words and functioning as holophrases.Hence, the most frequent utterances were 'papa [dad]", 'mama [mum]' and 'no [no, not]'.Nonetheless, the first two vocatives were also used for requesting, usually for demanding the completion of specific tasks (e.g. the proband could be playing with her mother and say 'papa' to mean that she wants to bring the activity to its end).By contrast, she had not mastered yet the conversational uses of pauses or silences, so that her metapragmatic awareness was minimal.As also noted, these shortcomings were partially alleviated by the pragmatic uses of gestures.Gestures were mostly used for qualifying or reinforcing the illocutive force of utterances.This is a notable achievement in view of her noteworthy motor problems, as suggested by the low scores obtained in the ITPA test.Some of the intended meanings and uses of the proband's gestures, often accompanied by the prolonged vocalisation of [a], are summarised in Table 1.Additionally, gestures were also used to facilitate turn-taking during conversational exchanges.We also found that the girl was able to maintain some eye contact with her interlocutor during conversation, mostly to confirm that her partner was actually listening to her.

Discussion
The detailed evaluation of the language and communication (dis)abilities exhibited by our proband suggests that she suffers from severe language problems, mostly resulting from phonatory and articulatory deficits, but also from her cognitive impairment.At the same time, she can communicate quite efficiently via gestures, particularly when settled in a familiar environment.
In the speech domain, the proband presents a severe apraxia of speech that hinders the intelligibility of her discourse.The lack of voice effects with linguistic and paralinguistic functions is also a notable feature of her oral production.Verbal expressions are reduced to strings of a few sounds functioning as holophrases.Nonetheless, the proband's problems with language are not restricted to the expressive domain.On the contrary, she also exhibits notable problems with receptive language.Hence, although she exhibits a good command on basic aspects of lexical semantics, like understanding single words referring to specific objects, or some ability for categorising, she has notable problems with abstract words, and particularly, with sensory-motor association learning.
Overall, these problems can be hypothesised to result from her cognitive and behavioural problems, particularly from her mild-to-severe cognitive delay (she is 2 years below their peers matched on chronological age), with this delay severely impacting on language acquisition and use.Accordingly, although she seems to be able to gain knowledge from previous experience via inference and generalisation, she is mostly insensitive to the others' feelings, this inability impacting negatively on her pragmatic and conversational abilities, and ultimately, on language acquisition through interaction.Still, the way in which she uses gestures to enrich the intended meaning of her oral productions suggests that her pragmatic abilities are relatively spared compared to other domains.
Table 2 shows a comparison between the abnormal features found in our proband (column A) and the features commonly associated to mutations in the FOXP2 gene (column B) and to deletions of the 7q31.1-q33.3region (column C).Table 3 shows a comparison of the features found in our proband with the features observed in DECIPHER patients with deletions in 7q of a similar size to the one found in our proband.Overall, the deficits exhibited by our proband resemble the problems found in patients with (similar) deletions in the 7q7q31.1q31.3region, and specifically in patients with mutations in FOXP2, particularly, in the domain of speech.For example, similarly to our proband, the patient analysed by García-Bellido et al. ( 2009) also exhibited a 2-year delay in the verbal domain.Interestingly too, Kurt et al. (2012) linked FOXP2 mutations to problems with sensory-motor association learning, which is a hallmark of our proband's cognitive profile too.Finally, Shriberg et al. (2006) also found speech anomalies in two cases of mutations in FOXP2, although some differences exist with our proband (she does not exhibit spastic features, or incur in assimilations, as observed in these patients).In summary, we regard the articulatory and expressive problems exhibited by our proband as compatible with the central FOXP2 phenotype (Laffin et al., 2012;Turner et al., 2016, which is characterised by oral apraxia and phonological deficits, which are also core symptoms in cases of deletions of the 7q31 region (Morgan et al., 2016).
That said, our proband exhibits more severe cognitive and behavioural problems than patients bearing mutations in FOXP2 only.This is seemingly due to the large size of her deletion in 7q31 (around 12 Mb), which encompasses many other genes besides FOXP2.As noted by Girirajan et al. (2011), the larger size of the CNVs, the greater risk of intellectual disability.Some authors (e.g.Lennon et al., 2007;Palka et al., 2011;Zeesman et al., 2006) have argued in favour of a 7q31 Deletion Syndrome (resulting from large deletions encompassing FOXP2), featuring impaired speech and language abilities, but also spared pragmatic abilities in socially comfortable situations.The phenotypic profile of our proband matches this profile.That said, the proband's microduplication in the 22q11.23 region could contribute, even if slightly, to some of her clinical features, but mostly to her dysmorphic features.Although, as noted, the pathogenicity of this duplication is uncertain, it overlaps, even if minimally, with the region deleted in the 22q11.2Distal Deletion Syndrome, featuring facial dysmorphism, developmental delay, and mild skeletal abnormalities (Ben-Shachar et al., 2008).Beyond the overlapping region with the 22q11.2Distal Deletion Syndrome, the region duplicated in our proband contains 3 genes that are particularly sensitive to haploinsufficiency: BCR, SMARCB1, and SPECC1L, resulting in increased susceptibility to conditions like rhabdoid tumour or Coffin-Siris Syndrome.However, there is no evidence of triplosensitivity for any of them.
Our proband illustrates a common outcome of clinical practice, that is, the finding of patients with CNVs that cannot be ascribed to any known genomic syndrome.By contrast, what one usually finds is patients with partially overlapping CNS and partially overlapping clinical features.In such cases, a detailed characterisation of each subject is particularly desirable in order to achieve stronger bridging hypotheses between the genotype and the phenotype.That said, we acknoledge that an important limitation of our study is the need of a more detailed genetic analysis of our patient aimed to know if FOXP2 and/or other genes with a known role in brain development and function level aredysregulated, thus accounting for her language and cognitive deficits, as reported in the paper.

Figure 1 .
Figure 1.Proband's scores in the PGD test.Note.The Y-axes represents PGD's scores.The X-axes represents the areas assessed

Figure 2 .
Figure 2. Proband's sample voice analysis by Estill Voiceprint Plus Program.Note.The Y-axes represents the frequency (in Hertz).The X-axes shows a sample of proband's oral production the hand with open palm + closing of fingers with fist upwards.More Repeated movement of the hand with open palm extending the arm and bringing it closer to the body.Give me Extend index finger and bring it close to an object.Point to and then point to an object This I want Repeated movement of bringing and bumping fists together To work

Figure 3 .
Figure 3. Chromosomal alterations found in our proband.Note. A. Left.Screen capture of the array-CGH of the proband's chromosome 7 showing the microdeletion 7q31.Right.Screen capture of the Decipher Genome Browser (https://decipher.sanger.ac.uk/browser/) showing the genes deleted in the proband.B. Left.Screen capture of the array-CGH of the proband's chromosome 22 showing the microduplication at 22q11.2.Right.Screen capture of the Decipher Genome Browser (https://decipher.sanger.ac.uk/ browser/) showing the genes duplicated in the proband.

Table 2 .
Clinical features in cases of genetic alterations similar to the ones found in our proband.

Table 3 .
A comparison between the phenotypic profile of our proband and the profiles of DECIPHER patients with similar deletions in the 7q31.1q31.3region.Note.For each subject, the DECIPHER ID, the GRCh38 coordinates, and the size of the deletion are provided.