Altered neural activity of magnitude estimation processing in adults with the fragile X premutation

Abstract

Mutations of the fragile X mental retardation 1 (FMR1) gene are the genetic cause of fragile X syndrome (FXS). Expanded CGG trinucleotide repeat (>200 repeats) result in transcriptional silencing of the FMR1 gene and deficiency/absence of the FMR1 protein (FMRP). Carriers with a premutation allele (55-200 CGG repeats) are often associated with mildly reduced levels of FMRP and/or elevated levels of FMR1 mRNA, and are associated with the risk of developing a neurodegenerative disorder known as fragile X-associated tremor/ataxia syndrome (FXTAS). While impairments in numerical processing have been well documented in FXS, recent behavioral research suggests that premutation carriers also present with subtle but significant impairments in numerical processing. Using fMRI, the current study examined whether asymptomatic adults with the premutation would show aberrant neural correlates of magnitude estimation processing in the fronto-parietal area. Using a magnitude estimation task, we demonstrated that activity in the intraparietal sulcus and inferior frontal gyrus, associated with magnitude estimation processing, was significantly attenuated in premutation carriers compared to their neurotypical counterparts despite their comparable behavioral performance. Further, multiple regression analysis using CGG repeat size and FMR1 mRNA indicated that increased CGG repeat size is a primary factor for the decreased fronto-parietal activity, suggesting that reduced FMRP, rather than a toxic gain-of-function effect from elevated mRNA, contributes to altered neural activity of magnitude estimation processing in premutation carriers. In conclusion, we provide the first evidence on the aberrant neural correlates of magnitude estimation processing in premutation carriers accounted for by their FMR1 gene expression.

Introduction

Fragile X syndrome (FXS) is the most common inherited form of intellectual disability, resulting from a trinucleotide repeat expansion in the 5′ untranslated region of the fragile X mental retardation 1 (FMR1) gene located at Xq27.3, affecting approximately 1 in 2500 individuals (Fernandez-Carvajal et al., 2009). When the CGG expansion exceeds 200 repeats (i.e., full mutation), it results in hypermethylation of the FMR1 gene and subsequent loss of the gene protein product, fragile X mental retardation protein (FMRP). Individuals with expanded, but unmethylated, repeats (55–200 repeat) are categorized as fragile X premutation carriers, and are associated with elevated levels of FMR1 mRNA and mildly reduced levels of FMRP (Garcia-Arocena and Hagerman, 2010, Hagerman and Hagerman, 2004, Kenneson et al., 2001, Tassone et al., 2000b). The prevalence of the premutation is relatively common (1 of 130–250 females and 1 of 250–800 males (Hagerman et al., 2009)), and about 40% of male and 8–16% of female adults with premutation allele (>50 years old) may develop a late-onset neurodegenerative disorder known as fragile X-associated tremor/ataxia syndrome (FXTAS) (Jacquemont et al., 2004), which is associated with tremors, gait ataxia, and executive function impairments (Bourgeois et al., 2009).

Until recently, it has been widely regarded that asymptomatic (i.e., non-FXTAS), young adults with the premutation are unaffected in their cognitive processing (Snow et al., 1993). However, recent studies have reported evidence indicating that, although the effect may be very subtle, presence of the premutation allele might modify neural development and/or cognitive function similar to that which is seen in individuals with FXS or FXTAS. For example, Keri and Benedek, 2009, Keri and Benedek, 2012 have found that, like in the case of FXS, premutation carriers present with mild dysfunction in motion perception, detection of spatial location, and visuomotor coordination. Further, these researchers recently demonstrated that subtle visual dysfunction in individuals with the premutation was significantly explained by FMRP level, suggesting possible neurodevelopmental changes in the low-level visual processing in the premutation with genetic contributions—i.e., expanded CGG length (Keri and Benedek, 2012). Furthermore, consistent with evidence on deficits in dorsal-stream visual processing in people with FXS (e.g., Kogan et al., 2004a, Kogan et al., 2004b), Hocking et al. (2012) demonstrated that asymptomatic male premutation carriers with high CGG repeat size (100 < CGG < 200) performed significantly worse than normal controls on a dot test of visuospatial working memory (WM) after accounting for effects of age and IQ on the performance. Such findings indicate that individuals with premutation alleles may have subtle cognitive dysfunctions which are a much milder form of those seen in FXS, and which have little, if any, effect on their everyday cognitive functioning.

Like deficits in dorsal-stream visual processing, quantitative and magnitude estimation processing have also been identified as one of the core cognitive functions affected in people with the full mutation (FXS). Specifically, females with FXS often reveal arithmetic difficulties in IQ tests (Bennetto et al., 2001; Grigsby et al., 1990, Kemper et al., 1986, Miezejeski et al., 1986), and a previous study showed impairments on complex arithmetic tests (e.g., 3-operamd arithmetic equations) in females with FXS both at behavioral and neural levels (Rivera et al., 2002). Interestingly, recent studies have also reported subtle impairments in arithmetic performance in females with the premutation. For example, using the Wide Range Achievement test, Lachiewicz et al. (2006) found that females with the premutation showed significant difficulty in arithmetic tests compared to tests for reading and spelling. Although the results from Lachiewicz et al. (2006) must be considered with a caution because the study was a retrospecitve review of the clinical experience rather than a prospective controlled study, it is still worth noting that the reported deficit in mathematics were positively correlated with CGG repeat length in premutation carriers, indicating a significant dose–response of FMR1 gene on arithmetic processing in the premutation. Likewise, recent studies from our group also found a dosage effect of the FMR1 gene (indexed by CGG repeat size) on basic magnitude processing in females with the premutation, although no significant performance differences were found in either enumeration or magnitude estimation task between premutation and control groups (Goodrich-Hunsaker et al., 2011a, Goodrich-Hunsaker et al., 2011b).

Despite such recent behavioral evidence on effect of FMR1 gene expression on behavioral performance of quantitative processing in asymptomatic premutation carriers, to date there has been no published study on the neural substrates of numerical/magnitude processing in premutation carriers. In fact, aberrant neural correlates responding to other cognitive functions have been found in asymptomatic adults with the premutation. Structurally, it has been found that asymptomatic adult carriers reveal significantly reduced gray matter density in the amygdala–hippocampal complex (Moore et al., 2004), reduced hippocampal volumes (Jakala et al., 1997), and a negative correlation between CGG repeat length and brain volume (Cohen et al., 2006). In functional brain imaging studies, researchers have demonstrated reduced activity in brain regions, such as hippocampus, in premutation carriers relative to age and IQ matched controls while performing a recall task, despite comparable behavioral performance on the task between the two groups (Koldewyn et al., 2008). Motivated by aforementioned studies reporting atypical brain activity for some of cognitive functions and also inspired by recent behavioral findings on the effect of FMR1 gene expression on magnitude processing in premutation carriers, the present study investigated whether asymptomatic adult carriers show atypical neural correlates of magnitude estimation processing modulated by FMR1 gene expressions.

To examine neural correlates of basic magnitude estimation processing, we used a well-established magnitude comparison task developed by Ansari and colleagues (Ansari and Dhital, 2006, Holloway and Ansari, 2010, Holloway et al., 2010, Price et al., 2007). The task requires magnitude comparison processing for non-symbolic numerical quantities. The “numerical distance effect” in this task is measured by comparing performance in a smaller numerical difference condition to that of a larger difference condition. Previous neuroimaging studies using this task have found strong fronto-parietal activation in response to the magnitude comparison processing, with great emphasis on activation in intraparietal sulcus (IPS) and inferior frontal gyrus (IFG) (Ansari and Dhital, 2006, Holloway and Ansari, 2009, Holloway et al., 2010). Based on these previous findings, we expected to find a distance effect in the fronto-parietal areas including bilateral IPS and IFG, and hypothesized that asymptomatic adults with the premutation would show attenuated brain activity in such areas associated with the distance effect compared to neurotypical counterparts. Further, we expected to find a dosage effect of FMR1 gene expression on reduced fronto-parietal activation associated with distance effect in premutation carriers.