Professor Juan Lerma’s group, from the UMH-CSIC Institute of Neurosciences, in Alicante (Spain), has identified the gene called GRIK1, fundamental in the balance between excitation and inhibition in the brain, as one of the causes for people with Down syndrome having spatial orientation problems.
The GRIK1 gene is located on chromosome 21, which is triplicated in people with Down syndrome, who therefore have a higher of dose of many genes.
This particular gene plays a very important role in communication between neurons, regulating the release of the main inhibitory neurotransmitter in the brain, called GABA.
To maintain proper brain function, a good regulation of communication between neurons is necessary.
This communication is done through neurotransmitters and can be excitatory or inhibitory, which would be the equivalent to the accelerator and the brake, respectively, of the nervous system.
The release of excitatory or inhibitory neurotransmitters takes place at the points of contact between neurons, called synapses.
The correct balance of the excitatory and inhibitory neurotransmission makes it possible for the neural circuits to function properly.
Pathologies apparently as different as anxiety, depression, schizophrenia, bipolar disorder or autism, all have in common the breakdown of the excitation/inhibition balance in the brain.
“We have seen in mouse models of Down syndrome that there is a problem of imbalance between excitation and inhibition of certain neuronal circuits of the hippocampus, a brain structure related to memory and orientation in space.
This imbalance depends on the dose of GRIK1. Through genetic manipulation techniques, we have normalized the dose of GRIK1 in our Down syndrome model, reverting this imbalance between excitation and inhibition,” explains Professor Lerma.
When the dose of this gene is normalized in this transgenic mouse, spatial memory problems disappear,” explains Sergio Valbuena, study’s first author.
Professor Lerma points out that the changes caused by this imbalance in inhibition are subtle and had gone unnoticed throughout the years of study on Down Syndrome.
Last year, Professor Lerma’s group already demonstrated how slight changes in the intensity of synaptic transmission cause significant modification of behavior, depending on the brain structure affected.
When changes in the intensity of synaptic transmission take place in the brain amygdala, fear or anxiety responses can be affected.
If alterations in the intensity of transmission occur in the prefrontal cortex, it can lead to problems in personal relationships or an increase in aggressiveness, says Professor Lerma.
And now this new study, published in Nature Communications, shows that a further inhibitory imbalance in the hippocampus results in alterations related to spatial memory, which likely accounts for orientation deficits in people with Down syndrome.
Down syndrome is the most frequent genetic cause of cognitive disability, so unraveling the physiological mechanisms responsible for these deficits is an important challenge. Described by John Langdon Down for the first time in 1866, Down syndrome is characterized by the existance of an extra copy of chromosome 21, which leads to several medical problems.
Down syndrome (DS) is the most common genetic cause of intellectual disability. A diagnosis of DS is given when an error in cell development results in an extra copy of chromosome 21, so there are 47 chromosomes rather than the usual 46. DS can also be the result of mosaicism, when only some cells include an extra copy of this chromosome, or translocation, when part of chromosome 21 attaches to another chromosome.
The majority of individuals with DS have a moderate intellectual disability ( Chapman & Hesketh, 2001); however, IQ scores can span from the severe to the average range ( Roizen, 2007). Language difficulties in children with DS are well documented, particularly those affecting vocabulary, phonology, morphology, and simple sentence structures ( Dodd & Thomspson, 2001; Eadie et al., 2002; Laws & Bishop, 2003; Price et al., 2007).
However, information regarding these children’s understanding of complex syntax is very limited. In addition, although children with DS have increased risk of a number of difficulties likely to influence their language development (involving limited cognitive ability, hearing level and memory skills) the relationship between these factors and language competence is not straightforward and has never been investigated in relation to the complex syntactic abilities of this population. We aim to address this gap in the literature.
Our previous research has shown that children’s performance on language comprehension tests can be heavily influenced by task demands ( Frizelle et al., 2017). Previous studies have tended to use multiple-choice tasks that have a heavy cognitive load and make demands beyond the linguistic. Given that people with DS have a cognitive impairment, we anticipated that these tasks may underestimate their comprehension abilities. For the current study, we used a novel method of assessment, designed to minimize non-linguistic demands. We hypothesised that this may reveal a greater level of syntactic understanding than when using a traditional multiple-choice format.
The term ‘complex sentence’ is used to refer to constructions that have more than one clause, linked in specific ways. This can be done through co-ordination (using connectors such as and or but) or subordination, where there is a main clause in which an element is embedded or expanded into a subordinate clause. Subordination is of particular interest as it allows for the expression of thoughts that involve hierarchical relationships between ideas, rather than just chaining them together. There are three distinct types of subordinate clause; complement clauses, adverbial clauses and relative clauses, and all three are the focus of the current paper.
Examples of each clause type are given in Table 3. Complement clauses are the earliest developing form of complex sentence ( Diessel, 2004) and are often used with mental state verbs such as know and think. In a complement clause, the embedded sentence serves as one of the arguments of the verb in the matrix clause ( Quirk et al., 1985). The complement clause can therefore be the subject, object or indirect object of the main verb. In this paper we are concerned with complements that serve as the object of the main clause.
In adverbial constructions the two clauses are linked semantically, most commonly using temporal (e.g. when) or causal (e.g. because) connectives. Finally, a relative clause serves to post-modify the noun in the main clause. They are usually defined according to (a) the sentential position of the modified noun phrase and (b) the role of the relativized noun phrase in the embedded clause. In this study, in keeping with children’s early production of relative clauses (see Diessel, 2004), we focused on relatives that modify the main clause object. In addition, we included relative clauses, where the relativized noun phrase realizes a range of syntactic roles, such as subject, object, oblique and indirect object.
Language characteristics of children with DS
The characteristic profile of language abilities in those with DS suggests that receptive language is typically better than expressive language ( Chapman et al., 2002; Laws & Bishop, 2003) and that vocabulary is stronger than syntax. The latter is evident in both receptive and expressive modalities ( Abbeduto et al., 2003; Berglund et al., 2001; Chapman et al., 1991).
Much of the work in relation to syntax has focussed on expressive language and primarily on spontaneous language production. Given the marked differences between children with DS and their age-matched peer group, it is customary to compare their language profiles with that of younger, typically developing (TD) children.
This makes it possible to see whether language development is merely following a typical, but markedly delayed, course, or whether there is a distinctive profile with strengths and weaknesses in specific aspects of language. Individuals with DS have been reported to produce fewer complex noun phrases, verb phrases, sentence structures, questions and negations than TD individuals of a similar non-verbal age ( Price et al., 2008).
A limited production of passives has also been reported ( Bridges & Smith, 1984; Fowler, 1990; Ring & Clahsen, 2005). In relation to complex syntax specifically, Thordardottir et al. (2002) analysed 12-minute narrative samples from 24 adolescents with DS (mean age 16.5 years) and a control group of younger TD children matched on mean length of utterance (MLU).
Co-ordinated sentences, clausal complements and relative clauses were all noted in the narrative samples, with no significant differences between the groups in either the proportion or the diversity of complex sentences used. However, Thordartottir and colleagues did highlight the degree of variability in the group with DS.
More recently, Christodoulou & Grohmann (2018) reported on the comprehension of syntactically complex subjunctives (e.g. The cat wants to dance) in 30 Greek Cypriot bilectal adolescents with DS. Using an act out priming task followed by a picture selection task they found a high rate of comprehension accuracy in their DS participants and posit that their results contradict previous suggestions of an overall syntactic impairment in people with DS. However, a close look at the distractor items suggests that the participants could respond correctly by understanding key words in the sentence and did not need to understand the complex syntax i.e. for the sentence above, from an array of 4 pictures there was only one image depicting a cat dancing, therefore if the participant understands the words ‘cat’ and ‘dance’ they are likely to choose the correct picture.
Other studies have reported deficits in syntactic comprehension in individuals with DS, (see Fortunato-Tavares et al., 2015; Michael et al., 2012; Perovic, 2006; Ring & Clahsen, 2005) however, the range of structures investigated is narrow and complex syntax has been given little or no attention. When complex syntax is involved, it has been in the context of standardized measures, in which different syntactic structures (both simple and complex) are grouped together and a composite score is reported (for example using The Test for Auditory Comprehension of Language-Revised (TACL-R) ( Carrow-Woolfolk, 1985) or the Test for the Reception of Grammar (TROG-2) ( Bishop, 2003). It is, accordingly, not possible to tease out the potential contribution of the complex constructions included in these tests to the scores achieved.
In addition, we have found that children’s performance on language comprehension tests can be heavily influenced by the specific demands of the assessment method employed ( Frizelle et al., 2017).
The format used in the TACL-R and TROG-2 is the traditional multiple-choice sentence picture-matching presentation, where the goal is to select from an array the picture that matches a spoken sentence. The other images are distractors that represent alternative interpretations of the sentence and the child is required to rule them out in order to respond correctly.
These competing interpretations are presented so that only children with a deep understanding of the construction will chose the correct item. However, this format is likely to lead to children failing for reasons other than a lack of linguistic knowledge.
In particular, it can disadvantage children (such as those with DS) who are inattentive and impulsive, and those who do not appreciate the need to scan the array carefully to choose between similar-looking items. We developed a new test (TECS-E: Test of Complex Syntax-Electronic) that was designed to minimise such demands by using a format where the child sees a specific animation and has to judge whether it matches a spoken sentence.
Because this is in effect a two-choice test, it is necessary to give at least eight items per structure to distinguish chance performance from understanding. Using this approach we found that TD children as young as 3;06 years showed understanding of some complex constructions that they had found difficult when tested using the more traditional multiple choice picture-pointing approach (see Frizelle et al., 2018a). Of course, no method is completely free of task demands or item-specific influences on performance, but our experience of the TECS-E with young children raised the possibility that traditional approaches to assessing comprehension may underestimate understanding in children with DS.
Receptive language and cognition in DS
Although it is tempting to consider cognitive ability as a core factor in explaining receptive language differences between those with DS and other groups, the literature is not consistent in this regard, particularly in relation to vocabulary. Several studies suggest that the receptive vocabulary of those with DS is in keeping with that of cognitively matched children with typical development, ( Chapman et al., 1991; Laws & Bishop, 2003; Miller, 1995) while other studies suggest a lower performance from those with DS ( Caselli et al., 2008; Hick et al., 2005; Price et al., 2007).
The literature regarding syntactic comprehension appears to be more homogenous with the majority of studies showing that those with DS have a lower than expected understanding of syntax relative to their non-verbal cognitive skills ( Abbeduto et al., 2003; Chapman et al., 1991; Joffe & Varlokosta, 2009; Laws & Bishop, 2003; Price et al., 2007; Rosin et al., 1988). However, it is important to note that most of these studies have used the same assessment measures, with a significant focus on morphology and simple syntax and few embedded sentences.
Some studies have compared those with DS to mental-age-matched TD controls, while others have matched cognitive ability with other cognitively impaired groups such as those with Williams syndrome, Fragile X syndrome and those with specific language impairment.
The aim of these comparisons is to see whether there is a distinctive profile specific to those with DS relative to other groups who have a language and or cognitive impairment. While previous reports appear mixed and are somewhat dependant on the comparison group under scrutiny, existing literature suggests that children with DS perform at a similar level to those with Williams syndrome and specific language impairment. A summary of the findings comparing those with DS with other groups, on their understanding of syntax is shown in Table 1.
Memory characteristics of children with DS
Individuals with DS show particular difficulties with verbal short-term or working memory tasks ( Jarrold & Baddeley, 2010; Jarrold et al., 2002; Laws, 2002) even when compared to other groups with cognitive delay, who do not have DS ( Bower & Hayes, 1994; Chapman, 2006; Laws, 2004). In contrast, their visual memory skills are often superior to, or at least in keeping with these groups, ( Bower & Hayes, 1994; Chapman, 2006; Rowe et al., 2006) suggesting that their memory deficits are specific to language.
Laws (1998); Laws (2004) reported a strong correlation between verbal short-term memory and a reduced mean length of utterance, as well as language comprehension difficulties. A strong relationship between memory and syntax has also emerged. Chapman et al. (2002) took a number of measures, at four time intervals over a 6-year period, from 31 individuals with DS between the ages of 5 and 20 years. They reported that along with chronological age, both verbal and visual working memory, were significant predictors of syntactic comprehension ability.
In addition, Chapman & Hesketh (2001) reported these factors to be key predictors of expressive syntax at the onset of their study. The connection between memory ability and syntactic difficulties in those with DS is also evident in work by Michael et al. (2012).
Michael and colleagues took a number of memory measures from individuals with DS, and a TD group matched on vocabulary, including digit span, word span, a spatial memory task and a sentence repetition task. Both groups performed similarly on all measures, with the exception of the sentence repetition task. They suggested that when compared to digit and word span, the syntactic processing load of a sentence was particularly difficult for the individuals with DS to parse and recall.
Because the current study focuses particularly on complex sentences, and these constructions require the parsing of different clauses over a more lengthy time span ( Martin & McElree, 2009; Marton et al., 2006), we might predict that short term and working memory would be particularly relevant to the ability of someone with DS to understand them. However, to our knowledge, this has never been investigated in relation to this population.
In sum, given the limited scope of previous research on comprehension, in terms of both methods and linguistic structures, we are uninformed about how individuals with DS process and understand specific complex structures as well as how cognitive ability, memory and hearing level impact the ability of those with DS to deal with these sentence types.
Our first aim was to investigate how well individuals with DS can understand complex structures such as relative clauses, complement clauses and adverbial clauses. Based on findings that individuals with DS produced relative and complement clauses in their narrative samples ( Thordardottir et al., 2002), we hypothesised that many of those with DS would be able to understand these constructions, although, on the basis of prior literature, we anticipated considerable performance variation. We compared strengths or weaknesses seen in those with DS to two other groups: (a) those with intellectual disability but of unknown aetiology (matched to those with DS on non-verbal mental age), and (b) a group of TD younger children at the same non-verbal mental age.
This allowed us to identify whether those with DS have a characteristic syntactic profile relative to the other two groups. Based on previous findings (such as those reported by Abbeduto et al., 2003; Chapman, 2006; Laws & Bishop, 2004) we hypothesised that those with DS would perform similarly to those with an intellectual disability of unknown origin but more poorly than the TD group matched on non-verbal ability.
Based on data from TECS-E with TD children aged from 3;06 to 5 years (see Frizelle et al., 2018a) we anticipated an order of difficulty within each family of constructions (relative, complement and adverbial). Within the five types of relative clause we expected children to have the least difficulty with intransitive subject relatives, with other relative clause types being of a similar level of difficulty.
Within adverbial clauses we expected causal adverbials to be the least demanding, followed by those that are temporal, with conditional adverbials causing the greatest difficulty. Finally, we anticipated that sentences using the verb pretend would be the least difficult complement clause items and that those using the cognitive state verb think would be the most difficult to understand. We based this expectation on previous (unpublished) data collected from young typically children between 3;06 and 4;11 years.
We also examined how children performed on TECS-E relative to a standardized test of grammar using the multiple-choice format. In the standardized measure (TROG-2; Bishop, 2003) children must show an understanding of the syntactically simple constructions before they progress on to those that are more complex.
Therefore, by applying the discontinue rule, if a given number of items are failed, children will not be tested on complex sentences. Here, in order to compare test administrations between TECS-E and TROG-2, we always administered block S (relative clauses) from TROG-2 at the end of the test, even if the stopping criterion was reached.
This block of four items uses relative clauses attached to a main clause object, two of which are similar in construction to those used in the TECS-E (albeit with some lexical differences—a noun rather that a pronoun in the head noun position: The girl chases the dog that is jumping) and two of which incorporate prepositional phrases ( The cup that is on the box is red).
Finally we considered how far comprehension difficulties in those with DS were associated with cognitive ability, verbal short-term or working memory abilities and hearing thresholds, and whether these associations differed according to the assessment format used. We predicted correlations of comprehension scores with all three variables, though performance on particular clause types would differ, as discussed above.
Given the additional cognitive load involved in a multiple-choice format, we hypothesised that children’s performance on this task would correlate more highly than TECS-E with overall cognitive and memory abilities.
Our pre-registered hypotheses https://osf.io/5ntvc/ were as follows:
- 1)Individuals with DS will be able to understand a range of the complex sentences tested, although we expect considerable individual variation.
- 2)Those with DS will perform more poorly overall than TD controls but at a similar level to those with cognitive impairment of unknown origin.
- 3)Children will have greater difficulty understanding comparable constructions on the multiple-choice test than on the animation task.
- 4)Cognitive ability, verbal memory, working memory and hearing level will predict performance in the DS group.
- 5)Cognitive, verbal and working memory abilities will account for more variance on the multiple-choice than on the animation task.
More information: Sergio Valbuena et al, Unbalanced dendritic inhibition of CA1 neurons drives spatial-memory deficits in the Ts2Cje Down syndrome model, Nature Communications (2019). DOI: 10.1038/s41467-019-13004-9
Journal information: Nature Communications
Provided by Spanish National Research Council