Skip to main content
NCBI home page
Cerebral Cortex (New York, NY) logoLink to Cerebral Cortex (New York, NY)
. 2023 Feb 22;33(12):7428–7435. doi: 10.1093/cercor/bhad049

Genetic overlap between cortical brain morphometry and frontotemporal dementia risk

Santiago Diaz-Torres 1,2,#, Natalia Ogonowski 3,4,#, Luis M García-Marín 5,6, Luke W Bonham 7,8, Claudia Duran-Aniotz 9,10, Jennifer S Yokoyama 11,12,13, Miguel E Rentería 14,15,✉
PMCID: PMC10267623  PMID: 36813468

Abstract

Frontotemporal dementia (FTD) has a complex genetic etiology, where the precise mechanisms underlying the selective vulnerability of brain regions remain unknown. We leveraged summary-based data from genome-wide association studies (GWAS) and performed LD score regression to estimate pairwise genetic correlations between FTD risk and cortical brain imaging. Then, we isolated specific genomic loci with a shared etiology between FTD and brain structure. We also performed functional annotation, summary-data-based Mendelian randomization for eQTL using human peripheral blood and brain tissue data, and evaluated the gene expression in mice targeted brain regions to better understand the dynamics of the FTD candidate genes.

Pairwise genetic correlation estimates between FTD and brain morphology measures were high but not statistically significant. We identified 5 brain regions with a strong genetic correlation (rg > 0.45) with FTD risk. Functional annotation identified 8 protein-coding genes. Building upon these findings, we show in a mouse model of FTD that cortical N-ethylmaleimide sensitive factor (NSF) expression decreases with age. Our results highlight the molecular and genetic overlap between brain morphology and higher risk for FTD, specifically for the right inferior parietal surface area and right medial orbitofrontal cortical thickness. In addition, our findings implicate NSF gene expression in the etiology of FTD.

Keywords: frontotemporal dementia, gene expression, genetic overlap, MRI, neuroimaging

Introduction

Frontotemporal dementia (FTD) is a heterogeneous spectrum comprising a cluster of neurodegenerative clinical syndromes (Goedert et al. 2012). It is classified based on its 3 clinical subconditions, including the behavioral, language, and motor variants (Leroy et al. 2021). FTD is commonly underdiagnosed, and its point prevalence has been estimated between 15 and 22 per 100,000 people, representing between 10% and 20% of all dementia cases (Onyike and Diehl-Schmid 2013). Although its clinical manifestations, such as psychiatric prodrome, neuropsychiatric symptoms, and language difficulties, are well known, these come in various presentations and severity, ultimately representing a challenge for diagnosis (Bott et al. 2014).

MRI is a noninvasive imaging technology that delineates structural and functional alterations of the brain, which has been proposed as a tool for diagnosing FTD (Cajanus et al. 2018; Bruun et al. 2019; Yu et al. 2021). Several observational and clinical studies using brain imaging techniques have reported associations between brain morphology and FTD (Bruun et al. 2019; Yu et al. 2021). For instance, lesions on the parietal and frontal cortex, which are commonly affected by FTD, have been linked with cognitive decline (Bisbing et al. 2015). Furthermore, it has been demonstrated that atrophy in the orbitofrontal cortex, the medial and lateral prefrontal cortex, the anterior cingulate, and the insula cortex is associated with behavioral symptoms observed in FTD, such as apathy, disinhibition, loss of empathy, and aggression (Whitwell 2019).

Neuroimaging genetics studies, which integrate brain imaging and individual-level genetic data to elucidate the genetic factors influencing brain structure (Mufford et al. 2017), have shed light on the pathophysiological characteristics of these FTD subtypes (Convery et al. 2019; Häkkinen et al. 2020). For instance, findings from neuroimaging genetics studies suggest that abnormalities in functional connectivity networks, gray matter volume, and white matter integrity are detectable before FTD onset and influenced by genetic variation (Häkkinen et al. 2020). Nonetheless, the genetic and neurobiological mechanisms underlying FTD and its subtypes remain largely unknown.

In the present study, we sought to identify brain regions with shared genetic and molecular basis. To that end, we leveraged summary data of genome-wide association studies (GWAS) of individuals of European ancestry to investigate the genetic overlap between 216 brain imaging phenotypes and FTD. We conducted pairwise GWAS (GWAS-PW) analyses to identify specific shared genomic regions for FTD and brain regions of interest and performed functional annotation analysis to provide meaningful insights into the underlying biological processes involved in the disease etiology.

Materials and methods

FTD GWAS data

We leveraged publicly available GWAS summary statistics for FTD, including 2,154 cases and 4,308 controls of European ancestry. Briefly, patients were genotyped using Illumina Human 370 K, 550 K, and 660 K Quad Beadchips and Omni Express chips (Illumina Inc, CA, United States). The GWAS was performed under strict quality control methods (Ferrari et al. 2014). A detailed description of these GWAS summary statistics is available in their corresponding publication (Ferrari et al. 2014).

Neuroimaging GWAS data

We leveraged publicly available GWAS summary statistics from the Oxford Brain Imaging Genetic Server—BIG40 (Smith et al. 2021). This database contains hemispheric independent MRI brain measurements of 33,224 individuals of European ancestry based on UK Biobank data. Notably, the BIG40 dataset presents genetic association analyses performed independently for brain regions of interest located in each hemisphere of the brain, which is relevant for FTD-related analyses, given that damage to the right or left temporal lobes results in different clinical manifestations (Irwin et al. 2018). Furthermore, the UK Biobank is a prospective epidemiological study of approximately half a million individuals between 40 and 69 years old (Bycroft et al. 2018). In the present study, we included MRI brain measurements, as processed by WIN FMRIB using FMRIB’s Automated Segmentation Tool (Zhang et al. 2001; Alfaro-Almagro et al. 2018; Smith et al. 2021), corresponding to the FreeSurfer segmentation pipeline that includes volume, thickness, and surface area parcellation data based on the Desikan-Killiany aparc parcellation (n = 216).

Linkage Disequilibrium Score Regression genetic correlation

We used Linkage Disequilibrium Score Regression (LDSC) (Bulik-Sullivan et al. 2015) to measure the genetic overlap between FTD and FreeSurfer DK morphological measurements. The LDSC method accounts for possible confounding due to sample overlap. In addition, we used a “Bonferroni” correction to account for multiple testing with a corrected P-value for a statistically significant significance of 0.0002 (0.05/216 FreeSurfer DK segmentation measurements). We also used a K-means algorithm integrated into the R package “pheatmap,” rstudio v.4.0.2, to cluster the regions according to their correlation with FTD.

GWAS-PW analyses

We used the GWAS-PW method on highly correlated regions with FTD (rg > 0.45) to identify specific genomic regions shared between FTD and brain morphology measurements. Briefly, GWAS-PW evaluates the genetic overlap over specific genomic regions by splitting the genome into 1,703 segments and estimating the posterior probability of association (PPA). To account for potential confounding due to sample overlap between GWASs of brain MRIs and FTD, we used the correlation between effect sizes of single-nucleotide polymorphisms (SNPs) with a PPA < 0.2, which was calculated in fGWAS (Huang et al. 2017) as a proxy for sample overlap.

For each brain imaging phenotype and genomic region, GWAS-PW estimates the posterior probability of 4 parallel models: (i) the region is unique to FTD, (ii) the region is unique to the imaging phenotype, (iii) both phenotypes share the region through the same genetic variants, and (iv) the region is associated with both phenotypes but through different genetic variants.

Functional annotation and eQTL

SNPs identified in shared regions in GWAS-PW (model 3 with PPA > 0.05) were subjected to functional annotation as implemented in the FUMA platform (Watanabe et al. 2017) v1.3.6 and were mapped to protein-coding genes using MAGMA v1.08. We applied a Bonferroni multiple testing correction that was defined as 0.05/(total number of genes in shared regions).

Causal genes present on shared loci across traits were further tested against eQTL data from the peripheral blood of 2,765 individuals from the Consortium for the Architecture of Gene Expression (Lloyd-Jones et al. 2017) and brain tissue of 449 individuals from the GTEx consortium (GTEx Consortium et al. 2017) using summary-data-based Mendelian Randomization (SMR), a method commonly used to identify the association between gene expression and complex traits using GWAS summary statistics (Zhu et al. 2016). We used a Bonferroni correction for multiple testing (0.05/total number of genes).

Human gene expression

Genes identified through functional annotation and a causal association between FTD and brain morphology in SMR were tested using the Allen Brain Atlas, which comprises an extensive dataset that links gene expression with neuroanatomical information (Sunkin et al. 2013). We used the brain model of a 57-year-old male, given that it is the model with the most similar age to the mean age in the UK Biobank (56.5 years) and the onset age for FTD (between 40 and 65 years). To accurately estimate differences in gene expression between males and females and between different ethnic groups, we conducted a chi-squared homogeneity test that included the gene expression for N-ethylmaleimide sensitive factor (NSF) in 39 regions of the brain from 5 male individuals between the ages of 24 and 57, 2 white Europeans, 2 African Americans, and 1 female of 49 years with a Hispanic ethnicity (see Supplementary Table 2). Nine brain regions with missing expression data were excluded from the analysis. We ran 2 separate models to test the homogeneity of NSF expression among different ethnic groups and between males and females. The first model included 3 categorical variables for ethnicity (i.e. Hispanic, White European, and African American), and the second model included 2 categories for sex (male or female). In instances where >1 measurement was present for a category, the mean estimate was included in the analysis. The Allen Brain Atlas project quantifies gene expression by fragment counts of RNA-Seq in a quantitative PCR; values are based on fluorescence or intensity measurements obtained from RNA microarrays. Results were based on probe number 1053553.

Mouse gene expression

To better understand the dynamics of the candidate FTD genes in both healthy aging and FTD, we compared expression from selected in a transgenic mouse model of tauopathy against wild-type (WT) mice. We utilized data from the Mouseac project, which includes brain tissue samples from the P301L-tau transgenic mouse model of FTLD (Hutton et al. 1998; Terwel et al. 2005) and WT mice of the same background strain at varying ages. The Mouseac project has been described in detail elsewhere (Matarin et al. 2015). Briefly, samples were collected from 3 brain regions (cortex, hippocampus, and cerebellum) from WT and P301L-tau transgenic mice at ages 8, 16, 32, and 72 weeks. Gene expression was measured using the Illumina Ref8 v2 microarray and was processed by the Mouseac project staff according to previously described and validated techniques (Matarin et al. 2015). Briefly, raw expression levels were normalized using a log2 transformation, and quantile normalization was performed for all samples. An individual probe was excluded if the P-value for detection was >0.05 in >50% of the WT and transgenic group's samples at any age. Additionally, samples were excluded if <95% of the probes for a given gene were detected. Tau pathology was quantified histologically (Matarin et al. 2015). Expression data were inspected for outlier data points and normality. ANOVA was used to test for gene expression differences between the P301L-tau transgenic mouse model and WT mice brain tissue with respect to both ages across all brain regions.

Results

LDSC genetic correlation

To elucidate the genetic factors influencing FTD risk and brain structure (Mufford et al. 2017), first we performed LD score regression and estimated pairwise genetic correlations (Supplementary Table 1). We did not observe any statistically significant genetic correlations between FTD and brain morphology measurements; this is most likely due to the limited sample size of the FTD GWAS. However, 4 brain MRI measurements were highly correlated with FTD (rg > 0.45). For instance, the left frontal pole area (rg = 0.62, s.e. = 0.41), right inferior parietal area (rg = 0.52, s.e. = 0.27), right caudal middle frontal thickness (rg = 0.48, s.e. = 0.27), and right medial orbitofrontal thickness (rg = 0.45, s.e. = 0.32), as shown in Fig. 1.

Fig. 1.

Fig. 1

Open in a new tab

Brain regions used in the GWAS-pairwise analyses. A) Lateral image of the right hemisphere highlighting the caudal middle frontal region in red, B) internal image of the right hemisphere highlighting the medial orbitofrontal in yellow, C) fronto-lateral image of the brain highlighting the frontal pole region in yellow, and D) frontal image of the brain with the frontal pole and medial orbitofrontal in yellow and the caudal middle frontal region in red.

GWAS-Pairwise

We further investigated the overlap between highly correlated brain regions with FTD (rg > 0.45). We identified the region chr17:43057141–45874715 (GRCh37-hg19) as shared between FTD and 2 brain phenotypes, the right inferior parietal surface area, and the right medial orbitofrontal cortical thickness.

Functional annotation and eQTL

We mapped the SNPs in the region 43057141 to 45874715 on chromosome 17 for the right inferior parietal area, right medial orbitofrontal thickness, and FTD. Eight protein-coding genes (ARMGAP27, CRHR1, KANSL1, NSF, PLEKHM1, SPPL2C, STH, and WNT3) were shared between the 2 brain regions and FTD; however, these genes were only nominally significant after multiple testing correction. Results from SMR using peripheral blood eQTL data confirmed a common causal association between NSF expression, the right inferior parietal area (b = 0.11, s.e. = 0.03), right medial orbitofrontal thickness (b = −0.19, s.e. =0.04), and FTD (b = −0.61, s.e. = 0.21) (Fig. 2). Results from brain tissue eQTL data were inconclusive, given the small sample size (n = 449). NSF is a protein-coding gene in chromosome 17, which is highly expressed in brain tissue as compared to other tissues (Fan et al. 2020) and has been previously associated with red cell distribution, hemoglobin levels, balding, brain morphology, Parkinson's disease, blood pressure, and developmental and epileptic encephalopathy, among other traits (Belluzzi et al. 2016).

Fig. 2.

Fig. 2

Open in a new tab

Causal association between NSF and FTD based on eQTL data of peripheral blood. P-value corrected threshold (pSMR) = 0.0062.

Human gene expression

Results regarding gene expression in the Allen Brain Atlas confirm the expression of NSF in the inferior parietal area and a medial orbitofrontal with a mean expression value of 12.8 and 12.7, respectively. The results of chi-squared tests of homogeneity showed no statistically significant differences in gene expression between ethnic groups or between males and females (P > 0.05). This suggests that there are no major variations in gene expression between the different groups. However, the data available are not a representative sample of the overall population, and further research is necessary to determine if these results are consistent under a larger sample size.

Mouse gene expression

To further characterize and explore the above-described findings in humans, we next used gene expression data from P301L transgenic and WT mice to explore the relationships between NSF and MAPT expression at varying ages across the brain. We started by comparing the main effects of a brain region, age, and transgene status. ANOVA analyses revealed significant differences in NSF (P = 1.96E-04) and MAPT (P = 0.01) expression in P301L transgenic mice relative to WT mice when compared by brain region across age (Table 1). Given this, we next explored the expression of NSF and MAPT within the cortex, hippocampus, and cerebellum to characterize the expression changes across time within each brain region that were driving our findings. Interestingly, NSF (P = 4.68E-04), but not MAPT (P > 0.05), demonstrated significantly different expression over time in P301L mice relative to WT mice despite increasing tau pathology in the cerebral cortex (Fig. 3A–C). In the hippocampus, both NSF (P = 4.72E-04) and MAPT (P = 1.56E-05) expression decreased over time in P301L mice relative to WT mice as tau pathology increased (Fig. 3D–F). Finally, in the cerebellum, there was no temporal relationship between NSF (P > 0.05) and MAPT (P > 0.05) expression in P301L mice relative to WT mice(Fig. 3G and H), and there was no significant burden of tau pathology in this region (Fig. 3I).

Table 1.

ANOVA results from the tau P301L mouse model of frontotemporal lobar degeneration are shown when compared by transgene status, age, and brain region (cortex, hippocampus, or cerebellum). Significant findings (P-value < 0.05) are shown in bold. Data used in these analyses were obtained from the Mouseac Project (http://www.mouseac.org/) (Matarin et al. 2015).

Gene P301L Age Region P301L:age P301L:region Age:region P301L: Age:region
Across brain regions NSF 0.83 0.67 <2E-16 0.04 0.11 0.15 1.96E-04
MAPT 0.55 0.11 5.18E-04 2.83E-03 0.97 0.28 0.01
Within cortex NSF 0.93 0.77 - 4.68E-04 - - -
MAPT 0.78 0.76 - 0.48 - - -
Within hippocampus NSF 0.07 0.07 - 4.72E-04 - - -
MAPT 0.82 0.04 - 1.56E-05 - - -
Within cerebellum NSF 0.36 0.42 - 0.10 - - -
MAPT 0.62 0.26 - 0.80 - - -
Open in a new tab

Fig. 3.

Fig. 3

Open in a new tab

NSF and MAPT expression in both pathologic and normal aging. Line plots with mean expression at each time point (+/− the standard error) for NSF and MAPT expression in the P301L mouse model of FTD relative to WT mice across the lifespan (data courtesy of the Mouseac Project, http://www.mouseac.org/) (Matarin et al. 2015). Expression profiles are shown for NSF and MAPT by brain region with across multiple ages stratified by transgene status. A and B) In the cortex, there was a significant association between NSF expression (P = 4.68E-04) and P301L transgene status over time which was not observed in MAPT (P > 0.05). C) Accumulation of tau pathology at varying ages in the cortex is provided as a reference. D and E) By contrast, in the hippocampus, both NSF (P = 4.72E-04) and MAPT (P = 1.56E-05) expression was significantly different in P301L transgenic mice relative to controls at varying ages when compared to WT. F) Accumulation of tau pathology at different ages in the hippocampus is provided as a reference. Finally, no association was observed for NSF (P > 0.05) or MAPT (P > 0.05) expression across time when comparing P301L transgenic mice relative to controls. I) Neither P301L nor WT mice demonstrate significant tau pathology at any age.

Discussion

In this work, we sought to advance our understanding of the genetic and molecular underpinnings of the relationship between FTD and brain morphology. We conducted LD score regression and GWAS-PW analyses. We identified 1 shared genomic region in chromosome 17 between FTD risk and 2 brain regions (right inferior parietal surface area and right medial orbitofrontal cortical thickness). In addition, we provide evidence for a putative causal association between brain morphology, FTD, and the NSF gene.

We note that although the genetic correlations between brain morphology and FTD did not reach statistical significance, they do not necessarily indicate a lack of an association between these phenotypes, especially when considering the high correlation values between FTD, the small sample size of FTD, and brain structural measurements in the frontal and parietal regions. This claim is supported by the significant correlation of the parietal and orbitofrontal brain regions with FTD through GWAS-PW.

Studies seeking to investigate behavioral changes in FTD have demonstrated the crucial role of the orbitofrontal cortex in the determination of social and emotional behavior (Viskontas et al. 2007). Individuals with lesions in the orbitofrontal cortex have shown impairment, to different extents, of stimulus–reward reversal learning, response inhibition, and the ability to judge the appropriateness of their social behavior (Viskontas et al. 2007). In addition, individuals with progressive atrophy in the orbitofrontal cortex experience detriment in emotion recognition, empathy, and their capacity to process complex stimulus–reward contingencies (Viskontas et al. 2007). It has been suggested that the substantial (50%–75%) and selective loss of Von Economo neurons, which is not as prominent among other types of dementia, could be an early major event shaping the manifestation of FTD symptoms (Viskontas et al. 2007; Santillo et al. 2013). Therefore, measurements of the orbitofrontal cortex have been proposed to discriminate between subtypes of FTD (Hornberger et al. 2010).

Similarly, neuroimaging studies have suggested that atrophy in the inferior parietal lobe, which is associated with anosognosia and mild cognitive impairment (Sedaghat et al. 2008), could be a potential marker for the diagnosis of Alzheimer’s disease (AD) (Jacobs et al. 2011). In addition, previous studies have linked metabolic dysfunctions in the inferior parietal area with reduced autonoetic consciousness in FTD (Muñoz et al. 2019).

We identified a common potential causal association between NSF and FTD and between NSF and 2 brain morphological measurements (the right inferior parietal area and right medial orbitofrontal thickness). NSF is a protein-coding gene that has been associated with protein kinase binding (Belluzzi et al. 2016) which in turn contributes to adaptive homeostasis mechanisms working as signaling pathways of the cellular response to stress (Pomatto and Davies 2017). Notably, a decrease in the signaling through the kinase sensor mechanisms has been associated with age-related neurodegeneration (Martinez et al. 2021). NSF has also been associated with AD (Fan et al. 2020) and Parkinson’s disease (Cheng et al. 2020; Fan et al. 2020) and FTD, as per the results of this study. However, NSF expression has been associated with MAPT haplotype (Soto-Beasley et al. 2020). We further assessed the expression of MAPT through eQTL of brain tissue with inconclusive results due to a low sample size (n = 449). Therefore, it is unclear if there is a causal association between MAPT and FTD and the degree in which NSF expression in brain tissue could be related to MAPT.

Gene expression data in the brain atlas confirmed the expression of NSF in the inferior parietal and medial orbitofrontal regions. That suggests an underlying biological mechanism between neurodegenerative conditions which may affect specific brain regions. Studies investigating the role of NSF in the etiology of tauopathies report that a reduction in NSF expression levels reflects extensive neurodegeneration in the brain (Mackenzie et al. 2006).

Furthermore, genetic studies have demonstrated that the NSF/rs199533 minor allele is protective against AD risk, and its homozygosity has been shown to delay the age at onset of AD (Fan et al. 2020), which comes in contrast with its role in the etiology of PD where it increases risk for the disease (Fan et al. 2020). Building upon these findings, we show in a mouse model of frontotemporal lobar degeneration that cortical NSF, but not MAPT, expression decreases with increasing age—corresponding closely with the steadily increasing burden of tau pathology in the cortex. These findings may help to elucidate how NSF expression changes uniquely contribute to risk for and the pathophysiology of frontotemporal lobar degeneration independent of its association with the MAPT locus.

We must acknowledge several limitations of this study. For instance, the small sample size for FTD used in the present study resulted in low statistical power, which was further limited by strict multiple testing corrections. Therefore, we note that a nonsignificant association in this study does not necessarily reflect a lack of association.

Similarly, eQTL data also had a constrained sample size. Future studies should leverage larger sample sizes to unveil other potential causal associations between specific genes, brain morphology, and FTD. In addition, GWAS data used here represent individuals of European ancestry. Results should not be generalized to other populations until findings are confirmed using data for different ancestry populations.

Our results highlight a genetic correlation between FTD risk and brain structure (right inferior parietal surface area and right medial orbitofrontal cortical thickness) at the genome-wide level and implicate the NSF gene in the etiology of FTD. This study advances our understanding of the relationship between FTD, brain morphology, and genetic factors despite the relatively modest sample size of the existing FTD GWAS. Further research is required to elucidate further the role of the NSF gene in the etiology of FTD and whether it could be targeted in neuroprotective treatments.

Supplementary Material

Suppl_FTD_bhad049
Click here for additional data file. (16.5KB, xlsx)

Acknowledgments

We thank the participants and investigators involved in the original genetic studies that made this study possible.

Contributor Information

Santiago Diaz-Torres, Mental Health & Neuroscience Program, QIMR Berghofer Medical Research Institute, Brisbane, QLD, Australia; School of Biomedical Sciences, Faculty of Medicine, The University of Queensland, Brisbane, QLD, Australia.

Natalia Ogonowski, Latin American Brain Health Institute (BrainLat), Universidad Adolfo Ibáñez, Santiago, Chile; Centro de Neurociencias Cognitivas (CNC), Universidad de San Andrés, Buenos Aires, Argentina.

Luis M García-Marín, Mental Health & Neuroscience Program, QIMR Berghofer Medical Research Institute, Brisbane, QLD, Australia; School of Biomedical Sciences, Faculty of Medicine, The University of Queensland, Brisbane, QLD, Australia.

Luke W Bonham, Memory and Aging Center, University of California, San Francisco, CA, United States; Department of Radiology and Biomedical Imaging, University of California, San Francisco, CA, United States.

Claudia Duran-Aniotz, Latin American Brain Health Institute (BrainLat), Universidad Adolfo Ibáñez, Santiago, Chile; School of Psychology, Center for Social and Cognitive Neuroscience (CSCN), Universidad Adolfo Ibanez, Santiago, Chile.

Jennifer S Yokoyama, Memory and Aging Center, University of California, San Francisco, CA, United States; Department of Radiology and Biomedical Imaging, University of California, San Francisco, CA, United States; Department of Neurology, Weill Institute of Neurosciences, University of California, San Francisco, CA, United States.

Miguel E Rentería, Mental Health & Neuroscience Program, QIMR Berghofer Medical Research Institute, Brisbane, QLD, Australia; School of Biomedical Sciences, Faculty of Medicine, The University of Queensland, Brisbane, QLD, Australia.

Authors’ contributions

Santiago Diaz-Torres (Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Visualization, Writing—original draft), Natalia Ogonowski (Data curation, Investigation, Methodology, Writing—original draft, Writing—review & editing), Luis M. García-Marín (Formal analysis, Software, Writing—review & editing), Luke W. Bonham (Formal analysis, Methodology, Writing—review & editing), Claudia Duran-Aniotz (Conceptualization, Writing—review & editing), Jennifer S. Yokoyama (Conceptualization, Methodology, Writing—review & editing), and Miguel E. Rentería (Conceptualization, Investigation, Methodology, Project administration, Supervision, Writing—review & editing)

Funding

This work was supported by the Australian National Health and Medical Research Council and Australian Research Council through an NHMRC-ARC Dementia Research Development Fellowship (GNT1102821 to MER); ANID/FONDEF IDEA (ID20I10152 and ID22I10029 to CDA); ANID/FONDECYT Regular (1210622 to CDA); ANID/PIA/ANILLOS (ACT210096 to CDA); the MULTI-PARTNER CONSORTIUM TO EXPAND DEMENTIA RESEARCH IN LATIN AMERICA (ReDLat supported by National Institutes of Health and National Institutes of Aging, R01 AG057234 to CDA; Alzheimer’s Association, SG-20-725707 to CDA; Rainwater Charitable foundation-Tau Consortium; and Global Brain Health Institute); National Institutes of Health and National Institute of Aging (R01 AG062588, R01 AG057234, and P30 AG062422; NIH-NINDS U54 NS123985 to JSY); Rainwater Charitable Foundation; the Alzheimer’s Association; the Global Brain Health Institute; and the Mary Oakley Foundation.

Conflict of interest statement

None declared.

Data availability

Data and code associated with this study may be made available upon request to the authors.

Ethics declaration

This study used summary-based statistics from GWAS. Ethics approval for individual studies is described in the original publications where the GWAS are described.

Consent to participate

No individual-based data were used in this study. Participants in all cohorts from these studies gave written informed consent, and the sites involved obtained approval from local research ethics committees or Institutional Review Boards.

Consent for publication

Not applicable.

Competing interests

JSY serves in the scientific advisory board for the Epstein Family Alzheimer’s Research Collaboration. All remaining authors declare no competing interests.

References

  1. Alfaro-Almagro  F, Jenkinson  M, Bangerter  NK, Andersson  JLR, Griffanti  L, Douaud  G, Sotiropoulos  SN, Jbabdi  S, Hernandez-Fernandez  M, Vallee  E, et al.  Image processing and quality control for the first 10,000 brain imaging datasets from UK Biobank. NeuroImage. 2018:166:400–424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Belluzzi  E, Gonnelli  A, Cirnaru  M-D, Marte  A, Plotegher  N, Russo  I, Civiero  L, Cogo  S, Carrion  MP, Franchin  C, et al.  LRRK2 phosphorylates pre-synaptic N-ethylmaleimide sensitive fusion (NSF) protein enhancing its ATPase activity and SNARE complex disassembling rate. Mol Neurodegener. 2016:11:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bisbing  TA, Olm  CA, McMillan  CT, Rascovsky  K, Baehr  L, Ternes  K, Irwin  DJ, Clark  R, Grossman  M. Estimating frontal and parietal involvement in cognitive estimation: a study of focal neurodegenerative diseases. Front Hum Neurosci. 2015:9:317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bott  NT, Radke  A, Stephens  ML, Kramer  JH. Frontotemporal dementia: diagnosis, deficits and management. Neurodegener Dis Manag. 2014:4:439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bruun  M, Koikkalainen  J, Rhodius-Meester  HFM, Baroni  M, Gjerum  L, van  Gils  M, Soininen  H, Remes  AM, Hartikainen  P, Waldemar  G, et al.  Detecting frontotemporal dementia syndromes using MRI biomarkers. Neuroimage Clin. 2019:22:101711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bulik-Sullivan  BK, Loh  P-R, Finucane  HK, Ripke  S, Yang  J, Schizophrenia Working Group of the Psychiatric Genomics Consortium, Patterson  N, Daly  MJ, Price  AL, Neale  BM. LD score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015:47:291–295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bycroft  C, Freeman  C, Petkova  D, Band  G, Elliott  LT, Sharp  K, Motyer  A, Vukcevic  D, Delaneau  O, O’Connell  J, et al.  The UK Biobank resource with deep phenotyping and genomic data. Nature. 2018:562:203–209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cajanus  A, Hall  A, Koikkalainen  J, Solje  E, Tolonen  A, Urhemaa  T, Liu  Y, Haanpää  RM, Hartikainen  P, Helisalmi  S, et al.  Automatic MRI quantifying methods in behavioral-variant frontotemporal dementia diagnosis. Dement Geriatr Cogn Dis Extra. 2018:8:51–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cheng  W-W, Zhu  Q, Zhang  H-Y. Identifying risk genes and interpreting pathogenesis for Parkinson’s disease by a multiomics analysis. Genes. 2020:11(9). [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Convery  R, Mead  S, Rohrer  JD. Review: clinical, genetic and neuroimaging features of frontotemporal dementia. Neuropathol Appl Neurobiol. 2019:45(1):6–18. [DOI] [PubMed] [Google Scholar]
  11. Fan  K-H, Feingold  E, Rosenthal  SL, Demirci  FY, Ganguli  M, Lopez  OL, Kamboh  MI. Whole-exome sequencing analysis of Alzheimer’s disease in non-APOE*4 carriers. J Alzheimers Dis. 2020:76:1553–1565. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Ferrari  R, Hernandez  DG, Nalls  MA, Rohrer  JD, Ramasamy  A, Kwok  JBJ, Dobson-Stone  C, Brooks  WS, Schofield  PR, Halliday  GM, et al.  Frontotemporal dementia and its subtypes: a genome-wide association study. Lancet Neurol. 2014:13:686–699. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Goedert  M, Ghetti  B, Spillantini  MG. Frontotemporal dementia: implications for understanding Alzheimer disease. Cold Spring Harb Perspect Med. 2012:2:a006254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. GTEx Consortium, Laboratory, Data Analysis &Coordinating Center (LDACC)—Analysis Working Group, Statistical Methods Groups—Analysis Working Group, Enhancing GTEx (eGTEx) Groups, NIH Common Fund, NIH/NCI, NIH/NHGRI, NIH/NIMH, NIH/NIDA, Biospecimen Collection Source Site—NDRI  et al.  Genetic effects on gene expression across human tissues. Nature. 2017:550:204–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Häkkinen  S, Chu  SA, Lee  SE. Neuroimaging in genetic frontotemporal dementia and amyotrophic lateral sclerosis. Neurobiol Dis. 2020:145:105063, 105063. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hornberger  M, Savage  S, Hsieh  S, Mioshi  E, Piguet  O, Hodges  JR. Orbitofrontal dysfunction discriminates behavioral variant frontotemporal dementia from Alzheimer’s disease. Dement Geriatr Cogn Disord. 2011:30(6):547–552. [DOI] [PubMed] [Google Scholar]
  17. Huang  C, Thompson  P, Wang  Y, Yu  Y, Zhang  J, Kong  D, Colen  RR, Knickmeyer  RC, Zhu  H, Alzheimer’s Disease Neuroimaging Initiative . FGWAS: functional genome wide association analysis. NeuroImage. 2017:159:107–121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hutton  M, Lendon  CL, Rizzu  P, Baker  M, Froelich  S, Houlden  H, Pickering-Brown  S, Chakraverty  S, Isaacs  A, Grover  A, et al.  Association of missense and 5′-splice-site mutations in tau with the inherited dementia FTDP-17. Nature. 1998:393:702–705. [DOI] [PubMed] [Google Scholar]
  19. Irwin  DJ, McMillan  CT, Xie  SX, Rascovsky  K, Van Deerlin  VM, Coslett  HB, Hamilton  R, Aguirre  GK, Lee  EB, Lee  VMY, et al.  Asymmetry of post-mortem neuropathology in behavioural-variant frontotemporal dementia. Brain. 2018:141:288–301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jacobs  HI, Van Boxtel  MP, Uylings  HB, Gronenschild  EH, Verhey  FR, Jolles  J. Atrophy of the parietal lobe in preclinical dementia. Brain Cogn. 2011:75(2):154–163. [DOI] [PubMed] [Google Scholar]
  21. Leroy  M, Bertoux  M, Skrobala  E, Mode  E, Adnet-Bonte  C, Le Ber  I, Bombois  S, Cassagnaud  P, Chen  Y, Deramecourt  V, et al.  Characteristics and progression of patients with frontotemporal dementia in a regional memory clinic network. Alzheimers Res Ther. 2021:13:19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lloyd-Jones  LR, Holloway  A, McRae  A, Yang  J, Small  K, Zhao  J, Zeng  B, Bakshi  A, Metspalu  A, Dermitzakis  M, et al.  The genetic architecture of gene expression in peripheral blood. Am J Hum Genet. 2017:100:371. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mackenzie  IR, Baker  M, West  G, Woulfe  J, Qadi  N, Gass  J, Cannon  A, Adamson  J, Feldman  H, Lindholm  C, et al.  A family with tau-negative frontotemporal dementia and neuronal intranuclear inclusions linked to chromosome 17. Brain. 2006:129:853–867. [DOI] [PubMed] [Google Scholar]
  24. Martinez  NW, Gómez  FE, Matus  S. The potential role of protein kinase R as a regulator of age-related neurodegeneration. Front Aging Neurosci. 2021:13:638208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Matarin  M, Salih  DA, Yasvoina  M, Cummings  DM, Guelfi  S, Liu  W, Nahaboo Solim  MA, Moens  TG, Paublete  RM, Ali  SS, et al.  A genome-wide gene-expression analysis and database in transgenic mice during development of amyloid or tau pathology. Cell Rep. 2015:10:633–644. [DOI] [PubMed] [Google Scholar]
  26. Mufford  MS, Stein  DJ, Dalvie  S, Groenewold  NA, Thompson  PM, Jahanshad  N. Neuroimaging genomics in psychiatry—a translational approach. Genome Med. 2017:9:1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Muñoz-Neira  C, Tedde  A, Coulthard  E, Thai  NJ, Pennington  C. Neural correlates of altered insight in frontotemporal dementia: a systematic review. Neuroimage Clin. 2019:24:102066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Onyike  CU, Diehl-Schmid  J. The epidemiology of frontotemporal dementia. Int Rev Psychiatry. 2013:25:130. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Pomatto  LCD, Davies  KJA. The role of declining adaptive homeostasis in ageing. J Physiol. 2017:595:7275–7309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Santillo  AF, Nilsson  C, Englund  E. von Economo neurones are selectively targeted in frontotemporal dementia. Neuropathol Appl Neurobiol. 2013:39:572. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Sedaghat  F, Gotzamani-Psarrakou  A, Dedousi  E, Costa  V, Dimitriadis  AS, Baloyannis  SJ. Right inferior parietal region hypoperfusion as a reflection of anosognosia in dementia. Ann General Psychiatry. 2008:7:1–1. [Google Scholar]
  32. Smith  SM, Douaud  G, Chen  W, Hanayik  T, Alfaro-Almagro  F, Sharp  K, Elliott  LT. An expanded set of genome-wide association studies of brain imaging phenotypes in UK Biobank. Nat Neurosci. 2021:24:737–745. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Soto-Beasley  AI, Walton  RL, Valentino  RR, Hook  PW, Labbé  C, Heckman  MG, Johnson  PW, Goff  LA, Uitti  RJ, McLean  PJ, et al.  Screening non-MAPT genes of the Chr17q21 H1 haplotype in Parkinson’s disease. Parkinsonism Relat Disord. 2020:78:138–144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Sunkin  SM, Ng  L, Lau  C, Dolbeare  T, Gilbert  TL, Thompson  CL, Hawrylycz  M, Dang  C. Allen Brain Atlas: an integrated spatio-temporal portal for exploring the central nervous system. Nucleic Acids Res. 2013:41:D996–D1008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Terwel  D, Lasrado  R, Snauwaert  J, Vandeweert  E, Van Haesendonck  C, Borghgraef  P, Van Leuven  F. Changed conformation of mutant Tau-P301L underlies the moribund tauopathy, absent in progressive, nonlethal axonopathy of Tau-4R/2N transgenic mice. J Biol Chem. 2005:280:3963–3973. [DOI] [PubMed] [Google Scholar]
  36. Viskontas  IV, Possin  KL, Miller  BL. Symptoms of frontotemporal dementia provide insights into orbitofrontal cortex function and social behavior. Ann N Y Acad Sci. 2007:1121(1):528–545. [DOI] [PubMed] [Google Scholar]
  37. Watanabe  K, Taskesen  E, van  Bochoven  A, Posthuma  D. Functional mapping and annotation of genetic associations with FUMA. Nat Commun. 2017:8(1):1826. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Whitwell  JL. Neuroimaging across the FTD spectrum. Prog Mol Biol Transl Sci. 2019:165:187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Yu  Q, The National Alzheimer’s Coordinating Center, The Alzheimer’s Disease Neuroimaging Initiative, Mai  Y, Ruan  Y, Luo  Y, Zhao  L, Fang  W, Cao  Z, Li  Y, et al.  An MRI-based strategy for differentiation of frontotemporal dementia and Alzheimer’s disease. Alzheimer’s Res Ther. 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Zhang  Y, Brady  M, Smith  S. Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Trans Med Imaging. 2001:20(1):45–57. [DOI] [PubMed] [Google Scholar]
  41. Zhu  Z, Zhang  F, Hu  H, Bakshi  A, Robinson  MR, Powell  JE, Montgomery  GW, Goddard  ME, Wray  NR, Visscher  PM, et al.  Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet. 2016:48:481–487. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Suppl_FTD_bhad049
Click here for additional data file. (16.5KB, xlsx)

Data Availability Statement

Data and code associated with this study may be made available upon request to the authors.

Articles from Cerebral Cortex (New York, NY) are provided here courtesy of Oxford University Press

RESOURCES