Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jun 1.
Published in final edited form as: Amyloid. 2021 Dec 22;29(2):110–119. doi: 10.1080/13506129.2021.2018678

The integration of genetically-regulated transcriptomics and electronic health records highlights a pattern of medical outcomes related to increased hepatic Transthyretin expression

Gita A Pathak 1,2,*, Antonella De Lillo 1,3,*, Frank R Wendt 1,2, Flavio De Angelis 1,2,3, Dora Koller 1,2, Brenda Cabrera Mendoza 1,2, Daniel Jacoby 4, Edward J Miller 4, Joel N Buxbaum 5, Renato Polimanti 1,2,#
PMCID: PMC9213571  NIHMSID: NIHMS1814492  PMID: 34935565

Abstract

Transthyretin (TTR) is the precursor of the fibrils that compromise organ function in familial and sporadic systemic amyloidoses (ATTR). RNA-interference and anti-sense therapeutics targeting TTR hepatic transcription have been shown to reduce TTR amyloid formation. In the present study, we leveraged genetic and phenotypic information from the UK Biobank and transcriptomic profiles from the Genotype-Tissue Expression project to test the association of genetically regulated TTR gene expression with 7,149 traits assessed in 420,531 individuals. We conducted a multi-tissue analysis of TTR transcription and identified an association with a operational procedure related to bone fracture (p=5.46×10−6). Using tissue-specific TTR expression information, we demonstrated that the association is driven by the genetic regulation of TTR hepatic expression (odds ratio [OR] = 3.46, p = 9.51×10−5). Using the UK Biobank electronic health records (EHRs), we investigated the comorbidities affecting individuals undergoing this surgical procedure. Excluding bone fracture EHRs, we identified a pattern of health outcomes previously associated with ATTR manifestations. These included osteoarthritis (OR=3.18, p=9.18×10−8), carpal tunnel syndrome (OR=2.15, p=0.002), and a history of gastrointestinal diseases (OR=2.01, p=8.07×10−4). In conclusion, our study supports that TTR hepatic expression can affect health outcomes linked to physiological and pathological processes presumably related to the encoded protein.

Keywords: amyloidosis, gene expression, phenome-wide association study, retinol, thyroxine, TTR, UK Biobank

Introduction

Transthyretin (TTR) is a multi-function protein involved in a range of physiological processes [1, 2]. The bulk of circulating TTR is produced in the liver, but its synthesis has been reported in other tissues (e.g., choroid plexus epithelial cells, retinal epithelial cells, Schwann cells, pancreatic alpha cells, and skin) [1, 2]. The TTR tetramer binds thyroxine (T4) and retinol (bound to the retinol-binding protein 4, holo-RBP4) and plays a role in their transportation through the circulatory system and into the central nervous system [1, 2]. TTR knockout mouse models showed a consistent reduction of T4 (50%) and holo-RBP4 (95%) in blood [3, 4]. This suggests that TTR regulation may impact health outcomes linked to retinol and thyroid homeostasis. In addition to its transport function, TTR is a metallopeptidase with three recognized substrates: apolipoprotein A-I (apoA-I), neuropeptide Y (NPY), and amyloid-beta peptide (Aβ) [1]. TTR can cleave the C-terminus of apoA-I, decreasing cholesterol efflux and potentially affecting lipid metabolism and the development of atherosclerosis [5]. In vitro TTR has been shown to cleave different forms of Aβ, but many studies have shown that it can prevent the formation of new aggregates, oligomers and fibrils with little or no evidence that proteolysis is required for its anti-Aβ role in vitro or in vivo [6].

TTR is best known for its role as a fibril precursor involved in the pathogenesis of human systemic amyloidoses (e.g., familial amyloidotic polyneuropathy, familial amyloid cardiomyopathy) [7]. TTR amyloidogenesis is initiated by the dissociation of the tetramer into monomers, which is followed by misfolding and aggregation into non-native oligomers, and finally the highly structured amyloid fibrils [8]. The accumulation of the TTR amyloid fibrils in multiple organs causes different symptoms depending on the site of the fibrils deposition. TTR-associated amyloidosis (ATTR) can be familial or sporadic. The familial forms of ATTR (ATTRv) are caused by coding mutations in the TTR gene with an autosomal dominant inheritance [9]. ATTRv patients have multiple organ involvement with peripheral and autonomic neuropathy, cardiomyopathy, gastrointestinal impairment, nephropathy, and ocular deposition with vitreous opacities [10]. Wild-type TTR (ATTRwt; i.e. TTR protein without amyloidogenic mutations) can also form amyloid fibrils depositing primarily in the heart [11]. Although ATTRwt is generally associated with late-onset cardiomyopathy, there are other sites of deposition which may be associated with clinical symptoms, most prominently carpal tunnel syndrome and lumbar spinal stenosis [12]. Gastrointestinal deposits of TTRwt are also common in the elderly, being found in 8–20% of gastrointestinal biopsies in such individuals [13]. In addition, many patients undergoing hip and knee replacements have ATTRwt deposition in the excised abnormal joints [14].

Due to their clinical heterogeneity, ATTRv and ATTRwt can be often misdiagnosed as other, more common disorders which present similar symptomatology. The diagnosis of ATTRv is frequently only made many years after the onset of symptoms [15]. With respect to ATTRwt, its prevalence has increased in recent years, largely because of heightened awareness in the medical community and improved diagnostic tools, but the disorder is still underdiagnosed [16]. Additionally, it is now known that ATTRwt can be identified in individuals less than 50 years of age, challenging the concept that this is a disease exclusively of the elderly [12]. Although TTR coding mutations are the cause of ATTRv, other factors contribute to the complex genotype-phenotype correlation (the same point mutation may be associated with different phenotypic combinations even within the same family) [17]. Non-coding variants regulating TTR transcription have been suggested to modify the phenotypic presentation in carriers of amyloidogenic mutations [18–20]. Similarly, transcriptomic regulation appears to have a role in the onset of ATTRwt [21, 22]. Other non-coding regulatory mechanisms such as epigenetic changes seem to affect ATTR genotype-phenotype correlation [23, 24].

RNA-interference (RNAi) and anti-sense therapeutics capable of silencing TTR gene expression in the liver have been shown to decrease total TTR production by as much as 80%, with subsequent reduction of symptoms associated with the deposits [25–27]. Recently, CRISPR-Cas9 in vivo gene editing targeting TTR gene expression in liver showed a reduction of TTR protein hepatic production in ATTRv patients [28]. To explore the relationship between physiological and pathological processes linked to TTR transcriptomic regulation, we conducted a phenome-wide investigation, applying transcriptome prediction models to 7,149 traits assessed in up to 420,531 individuals and subsequently investigated TTR tissue-specific contributions using a Mendelian randomization (MR) framework.

Methods

Study Design

Leveraging genome-wide and transcriptomic data, we tested the effect of TTR transcriptomic variation across the human phenotypic spectrum. Specifically, we estimated the joint effect sizes of genetic variants for TTR gene expression. This permitted us to evaluate the genetically-regulated transcriptomic association of TTR with human traits and diseases, independent of the transcriptomic changes induced by the environment and the phenotypes investigated [29]. Subsequently, we conducted a two-sample cis-MR analysis to assess the contributions of the tissue-specific gene expression and investigated electronic health records (EHR) of UK Biobank (UKB) participants to identify the spectrum of medical outcomes related to TTR gene.

Datasets

The UKB cohort is an open-access resource available to investigate a wide range of severe and life-threatening illnesses as well as normal-range traits [30]. This initiative has recruited more than 500,000 people (54% females; age at recruitment: 37–73 years) collecting information on their diet, cognitive function, work history, health status, and other relevant phenotypes. More than 7,000 phenotypic traits are available to be investigated in this cohort so far (full list available at https://docs.google.com/spreadsheets/d/1AeeADtT0U1AukliiNyiVzVRdLYPkTbruQSk38DeutU8/edit#gid=511623409). Leveraging UKB genome-wide association statistics, we derived information regarding genetic variants regulating TTR gene expression and data related to the health and disease status of its participants. These data are available at https://pan.ukbb.broadinstitute.org/downloads. A detailed description of the methods used to generate these data is available at https://pan.ukbb.broadinstitute.org/. Briefly, the genome-wide association analysis was conducted using the SAIGE (Scalable and Accurate Implementation of GEneralized) mixed model [31] and including a kinship matrix as a random effect and covariates as fixed effects. The covariates included age, sex, age×sex, age2, age2×sex, and the top-10 within-ancestry principal components. We investigated association statics derived from UKB participants of European descent (N=420,531), because no large-scale multi-tissue transcriptomic datasets are available to investigate other ancestries.

Information regarding the genetic regulation of TTR gene expression across multiple tissues was derived from the Genotype-Tissue Expression project (GTEx) [32]. This is a comprehensive public resource to study tissue-specific gene expression and its regulation. Samples were collected from 54 non-diseased tissue sites across nearly 1000 individuals, primarily for molecular assays including whole-genome sequencing, whole-exome sequencing, and RNA sequencing (RNA-seq). We used the GTEx V8 [33], which includes 17,382 RNA-Seq samples from 948 donors (312 women and 636 men). Supplementary Figure 1 shows TTR transcriptomic profiles in the 54 human tissues available in GTEx V8. The cis-regulatory effect of transcriptomic variation was assessed investigating the association of genetic variants located within 1 Mb up- and downstream of the transcription start site (for TTR gene, hg19 chr18:28,171,730–30,178,986). A detailed description of the GTEx methods (i.e., preprocessing, expression quantification, and association analysis) used is available at https://gtexportal.org/. Briefly, the cis expression quantitative locus (eQTL) mapping was conducted using FastQTL and considering as covariate the top-five genotyping principal components, PEER (Probabilistic Estimation of Expression Residuals) factors, sequencing platform, sequencing protocol, and sex. The eQTL mapping was conducted in 49 GTEx tissues that were assessed in at least 70 samples.

Owing to the use of previously collected, deidentified data, this study did not require institutional review board approval.

Gene Expression Prediction from TTR Association Statistics

To evaluate the genetically-regulated TTR transcriptomic variation, we used pre-trained prediction models available in S-PrediXcan method [34]. These gene expression weights were derived from GTEx v8 transcriptomic data using the fine-mapping software DAP-G [35] with a biologically informed prior, Multivariate Adaptive Shrinkage in R (MASHR) [36]. With respect to TTR transcriptomic variation, MASHR-based models were available for eight GTEx tissues: adrenal gland, aorta artery, nucleus accumbens (basal ganglia), putamen (basal ganglia), esophagus (gastroesophageal junction), minor salivary gland, skin (lower leg), and whole blood. These models were tested with respect to TTR association statistics available from the UKB analysis using S-PrediXcan [34]. Since the MASHR-based models were derived from samples collected from individuals of European descent, we used UKB association statistics calculated from UKB participants of European descent (N=420,531) to avoid population stratification biases. To boost the statistical power, we combined tissue-specific information using S-MultiXcan, which permitted us to perform a joint multi-tissue analysis accounting for transcriptomic correlation across the tissues tested [37]. S-MultiXcan analysis of TTR gene was conducted across 7,149 traits. A false discovery rate correction (FDR q<0.05) was applied to account for the number of traits tested.

Two-sample Mendelian Randomization

The results obtained from the S-MultiXcan analysis were investigated further using a two-sample MR analysis. This approach permitted us to estimate the putative causal effect of the tissue-specific genetically-regulated TTR transcriptomic variation on the traits of interest. To maximize the statistical power, we included all linkage disequilibrium (LD) independent TTR cis eQTL in genetic instruments derived from the 49 GTEx tissues available. For each tissue, we conducted a clumping considering an LD cutoff of R2=0.001 within a 10,000-kilobase window and using the 1,000 Genomes Project Phase 3 reference panel for European populations. Since we considered LD-independent TTR cis eQTL independently from their individual statistical significance, we conducted an MR analysis using the robust adjusted profile score (MR-RAPS) approach [38]. This is a method specifically designed to conduct causal inference analysis based on weak genetic instruments, accounting for the widespread pleiotropy of complex traits.

Co-morbidity Analysis using UKB Electronic Health Records

To investigate the comorbidities of the traits identified as associated with genetically determined TTR transcriptomic variation through S-MultiXcan and MR-RAPS analyses, we used EHRs available from UKB. Specifically, we evaluated the occurrence of ICD-10 (International Classification of Disease, 10th revision) and OPCS-4 (Office of Population Censuses and Surveys Classification of Interventions and Procedures, 4th revision) codes between cases presenting S-MultiXcan/MR-RAPS-identified traits and controls matched by ancestry, age, sex, Townsend deprivation index, and UKB recruiting center. Based on the matching criteria, we maximized the number of controls for each case to boost the statistical power of our analysis. Townsend deprivation index is a measure of material deprivation incorporating unemployment, non-car ownership, and non-home ownership [39]. At recruitment, each UKB participant is assigned a deprivation score corresponding to their address location Comparing the occurrence of ICD-10 and OPCS-4 codes between cases vs. matched controls, we calculated odds ratios (OR) and the corresponding 95% confidence intervals (95%CI) and applied an FDR correction (q<0.05) accounting for the number of medical outcomes tested.

Results

Leveraging MASHR-based models within the S-MultiXcan approach, we conducted a phenome-wide analysis across 7,149 traits and identified one association surviving FDR correction (q < 0.05; Figure 1): a specific operational procedure recorded in hospital inpatient records (Phenotype ID 41200-W231; OPCS-4: W23.1 – Secondary open reduction of fracture of bone and intramedullary fixation; p=5.46×10−6, FDR q=0.039). To confirm this finding, we applied the MR-RAPS approach to estimate the effect of the tissue-specific transcriptomic regulation of TTR gene on the odds of presenting the phenotype 41200-W231 among UKB participants. In liver, one standard deviation increase in the genetically-regulated transcriptomic profile of TTR gene was associated with a 3.45-fold increase in the odds of presenting the phenotype 41200-W231 (95%CI=1.85–6.43, p=9.51×10−5; Figure 2). Among the other tissues, we observed the second strongest effect in the pancreas (OR=2.83, 95%CI=1.2–6.69, p=0.017). This is in line with the TTR expression where liver and pancreas present the highest levels among the tissues available from GTEx (median transcripts per million: 2,734 and 202.8, respectively; see GTEx data in Supplementary Figure 1). Because TTR gene is mainly expressed in the liver, we conducted a phenome-wide analysis to identify health outcomes not detected by the S-MultiXcan analysis. We identified a small negative effect of hepatic TTR expression on phenotype 1697 (“comparative height size at age 10”; beta=−0.024, p=5.15×10−5). In the S-MultiXcan analysis, this phenotype was only nominally associated with the genetically-regulated transcriptomic profile of TTR gene (S-MultiXcan p=0.008). Beyond these two traits, we found limited concordance between these approaches (Supplementary Figure 2). However, in addition to phenotypes 41200-W231 and 1697, we identified fifteen phenotypes that were nominally significant in both S-MultiXcan and MR-RAPS analyses (Table 1). Among the strongest effects, decreased genetically-regulated TTR transcription was associated with increased odds of emergency excision of abnormal appendix and drainage (OPCS4: H01.1; S-MultiXcan p=0.030; MR-RAPS 95%CI=0.24–0.74), unspecified endoscopic ultrasound examination of pancreas (OPCS4: J74.9; S-MultiXcan p=0.49; MR-RAPS 95%CI=0.44–0.94), and chronic thyroiditis (S-MultiXcan p=0.037; MR-RAPS 95%CI=0.35–0.96). Conversely, increased genetically-regulated TTR transcription was associated with increased odds of cardiac computed tomography angiography (OPCS4: U10.2; S-MultiXcan p=0.034; MR-RAPS 95%CI=1.24–4.08) and malignant neoplasm of other and unspecified parts of biliary tract (ICD-10: C24; S-MultiXcan p=0.008; MR-RAPS 95%CI=1.03–3.11).

Figure 1:

Figure 1:

Open in a new tab

Phenome-wide association statistics of the joint multi-tissue analysis of the genetically regulated TTR transcriptomic variation across the 7,149 traits tested. The association statistics (i.e., −log10(P value)) are reported for each trait investigated. Traits are color-coded based on phenotypic domains. Red color indicates association statistics surviving FDR multiple testing correction.

Figure 2:

Figure 2:

Open in a new tab

Tissue-specific effect of genetically regulated TTR transcriptomic variation on the 41200-W231 phenotype calculated via an MR analysis using the robust adjusted profile score. Odds ratios and their corresponding 95% confidence intervals (95%CI) are reported for tissue-specific genetically-regulated TTR expression in 49 tissues available from GTEx.

Table 1:

Concordant results between S-MultiXcan and MR-RAPS analyses, testing multi-tissue and hepatic genetic regulation of TTR transcription, respectively. We report phenotypes showing at least nominally associations with both approaches.

ID Description S-MultiXcan MR-RAPS
P value OR 95% CI
41200-W231 OPCS4: W23.1 Secondary open reduction of fracture of bone and intramedullary fixation HFQ 5.46E-06 3.45 1.85–6.43
1697 Comparative height size at age 10 0.008 0.98 0.97–0.99
20116–0 Smoking status: Never 0.005 1.03 1.00–1.06
20084–484 Vitamin and/or mineral supplement use: Selenium 0.005 0.75 0.57–0.97
Smoking-Ever_Never Smoking status: ever vs never 0.012 0.99 0.98–1.00
41200-W362 OPCS4: W36.2 Needle biopsy of lesion of bone NEC 0.014 0.58 0.36–0.92
icd10-C24 ICD-10: C24 Malignant neoplasm of other and unspecified parts of biliary tract 0.008 1.79 1.03–3.11
41200-U051 OPCS4: U05.1 Computed tomography of head 0.013 1.09 1.01–1.17
41200-H011 OPCS4: H01.1 Emergency excision of abnormal appendix and drainage HFQ 0.030 0.42 0.24–0.74
104460 Banana intake 0.009 0.97 0.93–1.00
100160 Low calorie drink intake 0.023 1.08 1.01–1.15
41200-U102 OPCS4: U10.2 Cardiac computed tomography angiography 0.034 2.25 1.24–4.08
966 Poisoning by anticonvulsants and anti-Parkinsonism drugs 0.040 0.70 0.52–0.93
41250–1003 Discharged on clinical advice/consent: Transfer to another provider 0.043 0.75 0.59–0.95
245.2 Chronic thyroiditis 0.037 0.58 0.35–0.96
6179–6 Mineral and other dietary supplements: Selenium 0.039 0.91 0.83–0.99
41200-J749 OPCS4: J74.9 Unspecified endoscopic ultrasound examination of pancreas 0.049 0.64 0.44–0.94
Open in a new tab

Abbreviations: International Classification of Disease, 10th revision (ICD-10); Office of Population Censuses and Surveys Classification of Interventions and Procedures, 4th revision (OPCS-4); Odds ratio (OR); Confidence interval (CI); Not Elsewhere Classified (NEC); HFQ, however further qualified.

We observed a large effect of the genetically-regulated TTR transcriptomic variation in the liver with the phenotype 41200-W231 (OR=3.46; 95%CI=1.85–6.44) that is reported only for 290 UKB participants of European descent (0.07%). None of them is a carrier of one of the TTR coding mutations defined as amyloidogenic in the curated list available at http://www.amyloidosismutations.com/. We verified whether the strong case-control imbalance of the phenotype 41200-W231 affects the association statistics. Conducting a S-MultiXcan analysis across 22,196 genes, we confirmed that TTR is the gene with the strongest association and that there is no inflation in the transcriptome-wide association statistics (λgc=0.996; Supplemental Figure 3). To investigate the comorbidities of 41200-W231 cases, we verified whether this group of individuals present phenotypic presentations linked to physiological and pathological processes that may be affected by a genetic predisposition to a high hepatic expression of TTR gene. To improve power of detection for 41200-W231 comorbidities, we tested ICD-10 and OPCS-4 codes that were present in at least 5% of the 41200-W231 cases. Comparing 41200-W231 cases with 25,229 controls matched by ancestry, age, sex, Townsend deprivation index, and UKB recruiting center (Supplementary Table 1; controls did not include carriers of TTR amyloidogenic mutations), we identified several associated medical outcomes after applying an FDR multiple testing correction (q<0.05; Supplementary Tables 2 and 3). Excluding outcomes that can be directly reconducted to 41200-W231 phenotype (e.g., ICD-10 codes related to injuries and factors influencing contact with health services), we observed that 41200-W231 cases present a pattern of medical outcomes that may be due to hepatic TTR transcriptomic regulation (Table 2). The strongest 41200-W231 association was with osteoarthritis (ICD-10: M19.9, OR=3.18, 95%CI=1.93–4.25, p=9.18×10−8). Considering ICD-10 and OPCS-4 codes, we observed a convergence related to medical outcomes involving gastrointestinal system (ICD-10: Z87.1 “history of gastrointestinal diseases”, OR=2.01, 95%CI=1.33–3.01, p=8.07×10−4; OPCS-4: G45.9 “Unspecified diagnostic fiberoptic endoscopic examination of upper gastrointestinal tract”, OR=1.89, 95%CI=1.37–2.61 p=9.75×10−5) and carpal tunnel (ICD-10: G56.0 “Carpal tunnel syndrome”, OR=2.15, 95%CI=1.33–3.48, p=0.002; OPCS-4: A65.1 “Carpal tunnel release”, OR=2.2, 95%CI=1.36–3.56, p=0.001). In addition to outcomes related to physical health, we also observed associations with traits related to mental health: depressive episode (ICD-10: F32.9; OR=2.86, 95%CI=1.93–4.25, p=1.9×10−7), personal history of psychoactive substance abuse (ICD-10: Z86.4; OR=2, 95%CI=1.46–2.73, p=3.5×10−5), tobacco use (ICD-10: Z72.0, OR=2.46, 95%CI=1.45–4.15, p=5.51×10−4), and harmful nicotine use (ICD-10: F17.1, OR=2.13, 95%CI=1.32–3.44, p=0.002).

Table 2:

Significant association of 41200.W231 status with diagnostic and operational codes (FDR q<0.05). Outcomes directly related to 41200.W231 were not reported. Full results are listed in Supplementary Table 2 and 3.

Code 41200.W231 cases (%) matched controls (%) OR 95% CI
ICD-10 M19.9 Arthrosis, unspecified 0.083 0.028 3.18 2.08–4.86
F32.9 Depressive episode, unspecified 0.097 0.036 2.86 1.93–4.25
E11.9 Without complications 0.117 0.049 2.57 1.79–3.69
M81.99 Osteoporosis, unspecified (Site unspecified) 0.052 0.014 3.93 2.31–6.68
N17.9 Acute renal failure, unspecified 0.059 0.017 3.63 2.20–5.98
J45.9 Asthma, unspecified 0.159 0.08 2.18 1.58–3.00
Z86.4 Personal history of psychoactive substance abuse 0.166 0.09 2 1.46–2.73
D64.9 Anemia, unspecified 0.076 0.035 2.28 1.47–3.54
R11 Nausea and vomiting 0.079 0.037 2.22 1.44–3.42
I10 Essential (primary) hypertension 0.321 0.233 1.56 1.22–2.00
Z72.0 Tobacco use 0.055 0.023 2.46 1.48–4.10
Z87.1 Personal history of diseases of the digestive system 0.09 0.047 2.01 1.34–3.02
J44.9 Chronic obstructive pulmonary disease, unspecified 0.052 0.022 2.45 1.45–4.15
G56.0 Carpal tunnel syndrome 0.062 0.03 2.15 1.32–3.49
E78.0 Pure hypercholesterolemia 0.152 0.097 1.66 1.20–2.29
F17.1 Harmful use 0.062 0.03 2.13 1.32–3.44
Z88.0 Personal history of allergy to penicillin 0.093 0.052 1.87 1.26–2.78
K52.9 Non-infective gastro-enteritis and colitis, unspecified 0.076 0.041 1.91 1.23–2.97
N39.0 Urinary tract infection, site not specified 0.072 0.04 1.9 1.21–2.97
E66.9 Obesity, unspecified 0.069 0.038 1.9 1.20–3.00
I20.9 Angina pectoris, unspecified 0.076 0.043 1.84 1.18–2.87
R69 Unknown and unspecified causes of morbidity 0.083 0.048 1.79 1.17–2.73
K62.5 Hemorrhage of anus and rectum 0.059 0.033 1.82 1.11–2.99
K59.0 Constipation 0.059 0.034 1.75 1.07–2.87
I25.9 Chronic ischemic heart disease, unspecified 0.055 0.032 1.76 1.06–2.92
K21.0 Gastro-esophageal reflux disease with esophagitis 0.052 0.03 1.77 1.04–3.00
R19.4 Change in bowel habit 0.062 0.038 1.66 1.02–2.69
OPCS-4 G45.9 Unspecified diagnostic fiberoptic endoscopic examination of upper gastrointestinal tract 0.155 0.088 1.89 1.37–2.60
A65.1 Carpal tunnel release 0.062 0.029 2.2 1.38–3.52
Z92.6 Abdomen NEC 0.062 0.035 1.82 1.13–2.93
J18.3 Total cholecystectomy NEC 0.066 0.039 1.72 1.08–2.75
Z28.2 Caecum 0.114 0.078 1.52 1.05–2.19
Z27.4 Duodenum 0.141 0.102 1.45 1.04–2.02
Open in a new tab

Abbreviations: International Classification of Disease, 10th revision (ICD-10); Office of Population Censuses and Surveys Classification of Interventions and Procedures, 4th revision (OPCS-4); Odds ratio (OR); Confidence interval (CI); Not Elsewhere Classified (NEC).

Discussion

Leveraging genomic and transcriptomic datasets, we have demonstrated that increased hepatic TTR transcription due to genetic regulation is linked to a set of clinical phenotypes. We hypothesize that the effect of genetically-regulated TTR transcriptomic variation on TTR protein production. RNAi and anti-sense therapeutics targeting TTR transcriptomic regulation in liver showed a reduction in the hepatic production of TTR protein [25–27]. Recently, an in vivo gene-editing therapeutic agent has been shown to induce an 87% reduction of serum TTR protein concentration by altering the TTR gene sequence in liver (i.e., targeted knockout) to reduce its hepatic transcription [28]. Accordingly, the association between genetically-regulated TTR transcriptomic variation and certain health outcomes can be due to two mechanisms: i) direct effect – increased TTR production with the protein being toxic (as in amyloid); ii) altered homeostasis of molecules carried by TTR (e.g., T4).

In our phenome-wide analysis, genetically-regulated TTR transcriptomic variation was positively associated with increased odds of having a secondary open reduction of fracture of bone and intramedullary fixation (phenotype: 41200-W231; OPCS-4: W23.1). As mentioned, this effect is due to the transcriptomic variation regulated by genetic variation and not transcriptomic changes induced by environmental factors or physiological and pathological conditions of the individuals investigated. Additionally, the association was consistent across two analytic approaches based on different genetic estimators of TTR gene expression (i.e., joint-tissue gene expression modeling and tissue-specific cis-regulatory mediation). In line with TTR tissue distribution, hepatic gene regulation was the primary driver of this effect. As expected, individuals having W23.1 code in their OPCS-4 records also report ICD-10 codes related to bone fractures and fallings (from slipping, tripping, and stumbling) on same level and on/from stairs and steps (Supplementary Table 2). The susceptibility to bone fractures requiring open surgeries can be linked to the role of TTR protein in retinol and T4 transport. Altered retinol homeostasis has been associated with an increased risk of hip fractures [40–42]. In mice, excess vitamin A intake (high dosages for a short period of time and clinically relevant dosages over prolonged times) has been associated with decreased cortical bone thickness and increased fracture risk [43]. There is also a well-established relationship between T4 levels and the risk of bone fracture [44]. The mechanism is related to the effect of T4 on adult bone turnover and maintenance [45]. This effect is particularly evident in both hyperthyroidism and hypothyroidism. In hyperthyroidism, individuals have an excessive amount of the hormone T4, causing bone loss among several other symptoms [45]. In hypothyroidism, individuals are usually treated with levothyroxine (a synthetic form of T4) that can lead to a bone density reduction as a side effect [46]. As demonstrated by knockout models, TTR plays an important role in maintaining retinol and T4 homeostasis [3, 4]. Accordingly, individuals predisposed to having high hepatic TTR expression could have an increased risk of bone fracture and subsequent corrective surgeries due to the altered blood levels of retinol and/or T4. The same mechanism could also explain the association of the genetically-regulated TTR expression in the liver with the likelihood of reporting to be shorter than average at 10 years of age (phenotype 1697 “comparative height size at age 10”). In particular, TTR could affect the regulatory effect of T4 on endochondral ossification, which is essential for skeletal development and linear growth [45].

Since the W23.1 code is present in the OPCS-4 records of 290 out of the 426,884 UKB participants investigated in the present study, we decided to verify whether these individuals present co-morbidities that cannot be attributed directly to a bone fracture and its subsequent surgeries. Comparing 41200-W231 cases with 25,229 matched controls, we identified a pattern of medical outcomes. The strongest evidence was observed for osteoarthritis (ICD-10: M19.9). TTR amyloid deposits have been reported in the articular cartilage of patients affected by osteoarthritis, suggesting a role in the pathogenesis of the disease [47]. 41200-W231 cases have also increased odds of having codes for carpal tunnel syndrome (ICD-10: G56.0) and carpal tunnel release (OPCS-4: A65.1) in their health records. Carpal tunnel syndrome is a recognized early sign of ATTR [48]. In the general population, carpal tunnel release has a lifetime prevalence of 3.1% [49]. This is in line with its frequency in our matched controls (2.9%) while it was 6.2% in 41200-W231 cases. TTR amyloid deposition has been documented in the transverse carpal ligament in between 10 and 35% of patients undergoing carpal tunnel release and in as many as 60% of cases with TTR amyloidogenic mutations [50]. Gastrointestinal symptoms were also supported by concordant evidence from ICD-10 and OPCS-4 codes. Compared to matched controls, 41200-W231 cases showed higher odds of presenting codes related to diseases of the digestive system (ICD-10: Z87.1), nausea (ICD-10: R11), non-infective gastro-enteritis (ICD-10: K52.9), Constipation (ICD-10: K59.0), gastro-esophageal reflux disease (ICD-10: K21.0), change in bowel habit (ICD-10: R19.4), endoscopic examination of the upper gastrointestinal tract (OPCS-4: G45.9), and different types of gastrointestinal tract surgery (OPCS-4: Z92.6, Z28.2, and Z27.4). In the Transthyretin Amyloidosis Outcomes Survey (THAOS), gastrointestinal symptoms were reported in 59% of the patients enrolled, indicating that this is a common manifestation for ATTR [51]. Other 41200-W231-associated health outcomes also overlap with known ATTR symptoms including renal failure (ICD-10: N17.9) [52], urinary tract infection (ICD-10: N39.0) [53], and angina pectoris (ICD-10: I20.9) [54]. Beyond the medical codes that can be linked to ATTR symptoms, we observed certain outcomes that may be linked to other functions of TTR protein. In particular, codes related to mental health (ICD-10: F32.9 depressive episode, Z86.4 psychoactive substance abuse, and F17.1 harmful nicotine use) may be linked to the role of TTR in the physiology of the nervous system. In knockout mice, TTR showed multiple effects on the central nervous system [2]. In particular, the absence of TTR was associated with a reduction in depressive-like behavior and an increase in exploratory activity [55]. In human subjects, low cerebrospinal fluid TTR levels were associated with major depression, suicidal ideation, and low serotonin function [56]. In our study, we focused on TTR hepatic expression, observing a possible link with depressive symptoms and substance use. Further studies focused on choroid plexus TTR synthesis will be needed to investigate the potential role of TTR in mental health.

Although we demonstrate that the combination of genomic information and electronic health records may expose previously unsuspected TTR-associated diseases manifestations, there is still much to learn about the hepatic regulation of TTR transcription and its relationship to the disorders associated with TTR deposition as well as those in which at present the relationship is unexpected. We also acknowledge several limitations. While our main finding (i.e., the association of genetically-regulated TTR expression with 41200-W231 phenotype) was confirmed by two independent methods (i.e., S-MultiXcan and MR-RAPS), there was a limited convergence between these two approaches (Supplementary Figure 2). This is most likely because the current transcriptomic datasets are underpowered to investigate TTR gene regulation. Larger datasets will be needed to conduct more powerful studies of genetic regulation of TTR transcriptomic variation. Although multiple studies have shown that suppression of TTR hepatic transcription leads to a reduction in the concentration of TTR protein in the circulation [25–28], our study lacks information related to the amount of TTR protein synthesized and secreted in response to the increase in hepatic transcription. Hence, it is unclear whether the increased transcription results in an increase in amyloid precursor available for aggregation with subsequent tissue compromising fibrillogenesis. Similarly, we do not have data from the subjects regarding the circulating levels of the TTR ligands (i.e., T4 and holo-RBP4), abnormalities of which have been associated with susceptibility to fracture. Hence, we cannot ascertain the molecular mechanism relating increased hepatic transcription to identified clinical outcomes. Finally, our comorbidity analysis of 41200-W231 cases showed evidence consistent with symptoms that can be related to the pathologic effects of the TTR protein. However, we do not have information whether the symptoms observed in 41200-W231 cases are due to TTR amyloid deposits. Although previous ATTR studies showed the importance to investigate diverse populations [20, 22, 57], we investigated only UKB participants of European descent, because no large-scale genomic and transcriptomic datasets are currently available to investigate other ancestries. This represents a major limitation for our study and for the entire field. Indeed, as other biomedical areas, certain minority groups are under-investigated in ATTR studies [58]. Similarly, we were not able to investigate sex-specific effects, because of the lack of sex-stratified eQTL statistics from GTEx. Since TTR non-coding variation can be associated with different health outcomes between sexes [22], future studies will need to explore sex-specific transcriptomic changes in the context of ATTR pathogenesis. Acknowledging these limitations, our findings should be considered a proof of concept regarding the integration of omics information and EHRs to investigate the complexity of multi-function proteins such as TTR suggesting further studies leading to greater knowledge of their involvement in human diseases.

Conclusions

Our results suggest that the hepatic expression of TTR gene potentially affect several health outcomes due to its impact on physiological and pathological processes. With respect to the TTR-related physiological processes, we provide evidence supporting that variation in TTR gene regulation could lead to medical outcomes due to the altered homeostasis of retinol and T4. Beyond the role of TTR in ligand transport, genetic regulation of TTR transcription may be also related to certain psychiatric traits in line with previous evidence about TTR function in brain physiology.

Supplementary Material

Supplemental Material

Acknowledgements

This research has been conducted using the UK Biobank Resource (application reference no. 58146). We thank the research groups contributing to the GTEx Project and the Pan-UKB analysis for making their data publicly available.

Funding

This study was funded by a research grant from the Amyloidosis Foundation. The authors also acknowledge support from the National Institutes of Health (R21 DC018098, R33 DA047527, and F32 MH122058), the European Commission (H2020 Marie Sklodowska-Curie Individual Fellowship 101028810), and Pfizer (ATTR-PN grant 61031847). The funders had no role in the study design, data analysis, and data interpretation of the present study.

Disclosure statement

ADL is supported by a grant from Pfizer. DJ has served as consultant and steering committee member for MyoKardia, Inc. EJM reports grants from Bracco and Eidos, and consulting for General Electric, Alnylam, and Pfizer. RP received a research grant from Pfizer. The other authors declare no conflict of interest.

Abbreviations:

95%CI

95% confidence intervals

Aβ

Amyloid-Beta Peptide

apoA-I

Apolipoprotein A-I

EHRs

Electronic Health Records

eQTL

Expression Quantitative Locus

FDR

False Discovery Rate

ATTR

Transthyretin-related Amyloidosis

ATTRv

familial ATTR

ATTRwt

wild-type TTR

GTEx

Genotype-Tissue Expression project

ICD-10

International Classification of Disease, 10th revision

LD

Linkage Disequilibrium

MR

Mendelian randomization

MR-RAPS

MR analysis using the robust adjusted profile score

MASHR

Multivariate Adaptive Shrinkage in R

NPY

Neuropeptide Y

OR

Odds Ratio

OPCS-4

Office of Population Censuses and Surveys Classification of Interventions and Procedures, 4th revision

PEER

Probabilistic Estimation of Expression Residuals

RBP4

Retinol-Binding Protein 4

RNA-seq

RNA Sequencing

RNAi

RNA-interference

SAIGE

Scalable and Accurate Implementation of GEneralized

T4

Thyroxine

TTR

Transthyretin

UKB

UK Biobank

Data Availability

All data discussed in this study are provided in the article and in the Supplementary Material.

References

  • 1.Vieira M, Saraiva MJ. Transthyretin: a multifaceted protein. Biomol Concepts. 2014;5(1):45–54. [DOI] [PubMed] [Google Scholar]
  • 2.Liz MA, Coelho T, Bellotti V, et al. A Narrative Review of the Role of Transthyretin in Health and Disease. Neurol Ther. 2020;9(2):395–402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.van Bennekum AM, Wei S, Gamble MV, et al. Biochemical basis for depressed serum retinol levels in transthyretin-deficient mice. J Biol Chem. 2001;276(2):1107–1113. [DOI] [PubMed] [Google Scholar]
  • 4.Episkopou V, Maeda S, Nishiguchi S, et al. Disruption of the transthyretin gene results in mice with depressed levels of plasma retinol and thyroid hormone. Proc Natl Acad Sci U S A. 1993;90(6):2375–2379. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Liz MA, Gomes CM, Saraiva MJ, et al. ApoA-I cleaved by transthyretin has reduced ability to promote cholesterol efflux and increased amyloidogenicity. J Lipid Res. 2007;48(11):2385–2395. [DOI] [PubMed] [Google Scholar]
  • 6.Ghadami SA, Chia S, Ruggeri FS, et al. Transthyretin Inhibits Primary and Secondary Nucleations of Amyloid-beta Peptide Aggregation and Reduces the Toxicity of Its Oligomers. Biomacromolecules. 2020;21(3):1112–1125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Adams D, Koike H, Slama M, et al. Hereditary transthyretin amyloidosis: a model of medical progress for a fatal disease. Nat Rev Neurol. 2019;15(7):387–404. [DOI] [PubMed] [Google Scholar]
  • 8.Si JB, Kim B, Kim JH. Transthyretin Misfolding, A Fatal Structural Pathogenesis Mechanism. Int J Mol Sci. 2021;22(9). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.D’Aguanno V, Ralli M, Artico M, et al. Systemic Amyloidosis: a Contemporary Overview. Clin Rev Allergy Immunol. 2020;59(3):304–322. [DOI] [PubMed] [Google Scholar]
  • 10.Koike H, Katsuno M. Transthyretin Amyloidosis: Update on the Clinical Spectrum, Pathogenesis, and Disease-Modifying Therapies. Neurol Ther. 2020;9(2):317–333. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Sekijima Y, Yazaki M, Ueda M, et al. First nationwide survey on systemic wild-type ATTR amyloidosis in Japan. Amyloid. 2018;25(1):8–10. [DOI] [PubMed] [Google Scholar]
  • 12.Grogan M, Scott CG, Kyle RA, et al. Natural History of Wild-Type Transthyretin Cardiac Amyloidosis and Risk Stratification Using a Novel Staging System. J Am Coll Cardiol. 2016;68(10):1014–1020. [DOI] [PubMed] [Google Scholar]
  • 13.Freudenthaler S, Hegenbart U, Schonland S, et al. Amyloid in biopsies of the gastrointestinal tract-a retrospective observational study on 542 patients. Virchows Arch. 2016;468(5):569–577. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Rubin J, Alvarez J, Teruya S, et al. Hip and knee arthroplasty are common among patients with transthyretin cardiac amyloidosis, occurring years before cardiac amyloid diagnosis: can we identify affected patients earlier? Amyloid. 2017;24(4):226–230. [DOI] [PubMed] [Google Scholar]
  • 15.Sekijima Y, Ueda M, Koike H, et al. Diagnosis and management of transthyretin familial amyloid polyneuropathy in Japan: red-flag symptom clusters and treatment algorithm. Orphanet J Rare Dis. 2018;13(1):6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Pinney JH, Whelan CJ, Petrie A, et al. Senile systemic amyloidosis: clinical features at presentation and outcome. J Am Heart Assoc. 2013;2(2):e000098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Ando Y, Coelho T, Berk JL, et al. Guideline of transthyretin-related hereditary amyloidosis for clinicians. Orphanet J Rare Dis. 2013;8:31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Iorio A, De Lillo A, De Angelis F, et al. Non-coding variants contribute to the clinical heterogeneity of TTR amyloidosis. Eur J Hum Genet. 2017;25(9):1055–1060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Norgren N, Olsson M, Nystrom H, et al. Gene expression profile in hereditary transthyretin amyloidosis: differences in targeted and source organs. Amyloid. 2014;21(2):113–119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Polimanti R, Nunez YZ, Gelernter J. Increased Risk of Multiple Outpatient Surgeries in African-American Carriers of Transthyretin Val122Ile Mutation Is Modulated by Non-Coding Variants. J Clin Med. 2019;8(2). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Sikora JL, Logue MW, Chan GG, et al. Genetic variation of the transthyretin gene in wild-type transthyretin amyloidosis (ATTRwt). Hum Genet. 2015;134(1):111–121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.De Lillo A, De Angelis F, Di Girolamo M, et al. Phenome-wide association study of TTR and RBP4 genes in 361,194 individuals reveals novel insights in the genetics of hereditary and wildtype transthyretin amyloidoses. Hum Genet. 2019;138(11–12):1331–1340. [DOI] [PubMed] [Google Scholar]
  • 23.Pathak GA, Wendt FR, De Lillo A, et al. Epigenomic Profiles of African-American Transthyretin Val122Ile Carriers Reveals Putatively Dysregulated Amyloid Mechanisms. Circ Genom Precis Med. 2021;14(1):e003011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.De Lillo A, Pathak GA, De Angelis F, et al. Epigenetic profiling of Italian patients identified methylation sites associated with hereditary transthyretin amyloidosis. Clin Epigenetics. 2020;12(1):176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Coelho T, Adams D, Silva A, et al. Safety and efficacy of RNAi therapy for transthyretin amyloidosis. N Engl J Med. 2013;369(9):819–829. [DOI] [PubMed] [Google Scholar]
  • 26.Butler JS, Chan A, Costelha S, et al. Preclinical evaluation of RNAi as a treatment for transthyretin-mediated amyloidosis. Amyloid. 2016;23(2):109–118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Adams D, Slama M. Hereditary transthyretin amyloidosis: current treatment. Curr Opin Neurol. 2020;33(5):553–561. [DOI] [PubMed] [Google Scholar]
  • 28.Gillmore JD, Gane E, Taubel J, et al. CRISPR-Cas9 In Vivo Gene Editing for Transthyretin Amyloidosis. N Engl J Med. 2021. [DOI] [PubMed] [Google Scholar]
  • 29.Gamazon ER, Wheeler HE, Shah KP, et al. A gene-based association method for mapping traits using reference transcriptome data. Nat Genet. 2015;47(9):1091–1098. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Bycroft C, Freeman C, Petkova D, et al. The UK Biobank resource with deep phenotyping and genomic data. Nature. 2018;562(7726):203–209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Zhou W, Nielsen JB, Fritsche LG, et al. Efficiently controlling for case-control imbalance and sample relatedness in large-scale genetic association studies. Nat Genet. 2018;50(9):1335–1341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Consortium GT. The Genotype-Tissue Expression (GTEx) project. Nat Genet. 2013;45(6):580–585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Consortium GT. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. 2020;369(6509):1318–1330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Barbeira AN, Dickinson SP, Bonazzola R, et al. Exploring the phenotypic consequences of tissue specific gene expression variation inferred from GWAS summary statistics. Nat Commun. 2018;9(1):1825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Wen X, Lee Y, Luca F, et al. Efficient Integrative Multi-SNP Association Analysis via Deterministic Approximation of Posteriors. Am J Hum Genet. 2016;98(6):1114–1129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Urbut SM, Wang G, Carbonetto P, et al. Flexible statistical methods for estimating and testing effects in genomic studies with multiple conditions. Nat Genet. 2019;51(1):187–195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Barbeira AN, Pividori M, Zheng J, et al. Integrating predicted transcriptome from multiple tissues improves association detection. PLoS Genet. 2019;15(1):e1007889. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Qingyuan Z, Jingshu W, Gibran H, et al. Statistical inference in two-sample summary-data Mendelian randomization using robust adjusted profile score. The Annals of Statistics. 2020;48(3):1742–1769. [Google Scholar]
  • 39.Townsend P, Phillimore P, Beattie A. Health and Deprivation: Inequality and the North: Croom Helm; 1988.
  • 40.Zhang X, Zhang R, Moore JB, et al. The Effect of Vitamin A on Fracture Risk: A Meta-Analysis of Cohort Studies. Int J Environ Res Public Health. 2017;14(9). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Wu AM, Huang CQ, Lin ZK, et al. The relationship between vitamin A and risk of fracture: meta-analysis of prospective studies. J Bone Miner Res. 2014;29(9):2032–2039. [DOI] [PubMed] [Google Scholar]
  • 42.Promislow JH, Goodman-Gruen D, Slymen DJ, et al. Retinol intake and bone mineral density in the elderly: the Rancho Bernardo Study. J Bone Miner Res. 2002;17(8):1349–1358. [DOI] [PubMed] [Google Scholar]
  • 43.Lionikaite V, Gustafsson KL, Westerlund A, et al. Clinically relevant doses of vitamin A decrease cortical bone mass in mice. J Endocrinol. 2018;239(3):389–402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Delitala AP, Scuteri A, Doria C. Thyroid Hormone Diseases and Osteoporosis. J Clin Med. 2020;9(4). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Bassett JH, Williams GR. Role of Thyroid Hormones in Skeletal Development and Bone Maintenance. Endocr Rev. 2016;37(2):135–187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Karimifar M, Esmaili F, Salari A, et al. Effects of Levothyroxine and thyroid stimulating hormone on bone loss in patients with primary hypothyroidism. J Res Pharm Pract. 2014;3(3):83–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Akasaki Y, Reixach N, Matsuzaki T, et al. Transthyretin deposition in articular cartilage: a novel mechanism in the pathogenesis of osteoarthritis. Arthritis Rheumatol. 2015;67(8):2097–2107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Karam C, Dimitrova D, Christ M, et al. Carpal tunnel syndrome and associated symptoms as first manifestation of hATTR amyloidosis. Neurol Clin Pract. 2019;9(4):309–313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Pourmemari MH, Heliovaara M, Viikari-Juntura E, et al. Carpal tunnel release: Lifetime prevalence, annual incidence, and risk factors. Muscle Nerve. 2018;58(4):497–502. [DOI] [PubMed] [Google Scholar]
  • 50.Milandri A, Farioli A, Gagliardi C, et al. Carpal tunnel syndrome in cardiac amyloidosis: implications for early diagnosis and prognostic role across the spectrum of aetiologies. Eur J Heart Fail. 2020;22(3):507–515. [DOI] [PubMed] [Google Scholar]
  • 51.Wixner J, Mundayat R, Karayal ON, et al. THAOS: gastrointestinal manifestations of transthyretin amyloidosis - common complications of a rare disease. Orphanet J Rare Dis. 2014;9:61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Lobato L, Rocha A. Transthyretin amyloidosis and the kidney. Clin J Am Soc Nephrol. 2012;7(8):1337–1346. [DOI] [PubMed] [Google Scholar]
  • 53.Bentellis I, Amarenco G, Game X, et al. Diagnosis and treatment of urinary and sexual dysfunction in hereditary TTR amyloidosis. Clin Auton Res. 2019;29(Suppl 1):65–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Yamamoto H, Yokochi T. Transthyretin cardiac amyloidosis: an update on diagnosis and treatment. ESC Heart Fail. 2019;6(6):1128–1139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Sousa JC, Grandela C, Fernandez-Ruiz J, et al. Transthyretin is involved in depression-like behaviour and exploratory activity. J Neurochem. 2004;88(5):1052–1058. [DOI] [PubMed] [Google Scholar]
  • 56.Sullivan GM, Mann JJ, Oquendo MA, et al. Low cerebrospinal fluid transthyretin levels in depression: correlations with suicidal ideation and low serotonin function. Biol Psychiatry. 2006;60(5):500–506. [DOI] [PubMed] [Google Scholar]
  • 57.Parker MM, Damrauer SM, Tcheandjieu C, et al. Association of the transthyretin variant V122I with polyneuropathy among individuals of African ancestry. Sci Rep. 2021;11(1):11645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Buxbaum JN, Ruberg FL. Transthyretin V122I (pV142I)* cardiac amyloidosis: an age-dependent autosomal dominant cardiomyopathy too common to be overlooked as a cause of significant heart disease in elderly African Americans. Genet Med. 2017;19(7):733–742. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplemental Material
NIHMS1814492-supplement-Supplemental_Material.docx (1.5MB, docx)

Data Availability Statement

All data discussed in this study are provided in the article and in the Supplementary Material.

RESOURCES