Detection of circulating transthyretin amyloid aggregates in plasma: a novel biomarker for transthyretin amyloidosis

Transthyretin amyloidosis (ATTR) is a highly underdiagnosed disease caused by the systemic deposition of amyloid fibrils composed of misfolded transthyretin (TTR)¹. Difficulties in diagnosing ATTR arise from phenotypic variability and lack of sensitive and specific blood biomarkers for ATTR^1,2. We hypothesized that the conformational change of TTR to amyloid fibrils during pathogenesis can be exploited to develop a novel disease biomarker³. Targeting segments of TTR that are only exposed in pathogenic conformations, we developed a first-generation transthyretin aggregate detector (TAD1) probe for the detection of ATTR fibrils and aggregates in patient tissues and blood.

Raw data, patient data and analytical methods can be made available for reproduction of results and procedures upon request. Human tissue and proprietary probes cannot be made available because of legal constraints. The Office of the Human Research Protection Program at UTSW granted exemption from Internal Review Board review because specimens were anonymized. Experiments involving blood samples were blinded during analysis.

TAD1 development was performed by rational design as previously published peptide inhibitors targeting two amyloid-driving segments of TTR³. To assess TAD1 binding activity, samples of interest were added to nitrocellulose membrane and incubated with TAD1 after blocking with bovine serum albumin in tris-buffered saline. TAD1 binding to each sample was measured through fluorescence of the N-terminal fluorescein isothiocyanate tag.

We first validated the ability of TAD1 to bind TTR fibrils in heart lysates from ATTR patients or purified as previously described⁴. We found that 5 μM TAD1 detects 5 μg purified fibrils regardless of patient TTR genotype whereas there is no recognition of negative control samples, which include other conformations of TTR and other fibrillar proteins (Figure A). We were also able to detect fibrils present in 5 μg crude tissue lysates obtained from these patients (Figure B). TAD1 detects patient-derived ATTR fibrils and ATTR fibrils in lysates with high specificity, thereby validating our design strategy.

Figure.

(A) Dot blot and quantification of TAD1 fluorescence intensity of binding to extracted wild type ATTR (ATTRwt) fibrils, mutant ATTR fibrils (ATTR-P24S, ATTR-V30M, ATTR-T60A, ATTR-I84S, ATTR-V122I), and controls. Controls include recombinant wild-type transthyretin (WT TTR), recombinant TTR with a T119M mutation that stabilizes the tetramer (T119M TTR), recombinant monomeric transthyretin (MTTR), recombinant serum amyloid a protein (SAP), recombinant collagenase IV, which is used in the fibril extraction process, recombinant tau protein implicated in Alzheimer’s Disease (AD), brain lysate from an AD patient, and a synthetic peptide derived from TAD1 but lacking fluorescent tags (Tab3-12). 5 μg of ex-vivo ATTR cardiac fibrils or controls were dotted onto a nitrocellulose membrane (Bio-Rad 1620112). The membrane was blocked for 30 minutes in 1X bovine serum albumin (BSA, Thermo Fisher Scientific 37525), 1X tris-buffered saline (Boston Bioproducts BM-300), 0.1% Tween-20 (Thermo Fisher Scientific J20605.AP) (TBST). After washing in TBST three times for five minutes each, samples were probed with 5 μM TAD1 in 1x BSA/TBST for 1 hour. After washing in TBST three times for ten minutes each, the fluorescence intensity of TAD1 binding was measured in an Azure Biosystems C600 imaging system through excitation of the membrane at 472 nm and reading emission at 513 nm, with autoexposure. TAD1 fluorescence intensity was quantified using the ImageJ software. The signal was normalized to the highest fluorescence intensity on the membrane. (B) Dot blot and quantification of TAD1 fluorescence intensity of binding to ATTR in cardiac lysates and controls. Methods and samples are described as in A. For preparation of crude tissue lysates, 5 mg of patient heart tissue was suspended in 500 μL tris-calcium buffer containing a protease inhibitor cocktail (Sigma Aldrich 11836170001). The sample was homogenized using a biomasher (Polysciences 25533-100) for 5 minutes. The sample was then placed into a bath sonicator where it was pulsed for 5 seconds on, 5 seconds off for ten minutes at amplitude 80. The protein concentration within the lysate was determined using the Pierce Micro BCA protein assay kit (Thermo Scientific 23235). (C) Dot blot and quantification of TAD1 fluorescence intensity of binding to titrated ATTRwt fibrils. Methods are described as in A except ATTRwt fibrils were titrated onto the membrane. A four-parameter logistic sigmoidal curve was fitted to calculate the binding constant K_D. (D) Precision assay to determine whether 10 μM TAD1 can detect small variations in ATTR fibrils randomly loaded onto membrane. Right, quantification of TAD1 fluorescence intensity of binding to each dot in three independent repeats. Significant differences between all groups were established using a non-parametric Kruskal-Wallis test. A non-parametric Mann-Whitney t-test was used to establish significant differences between each pair of groups. Data is plotted as violin plots showing the frequency distribution of the data. Methods are described as in A with the following modifications: the membrane was probed with 10 μM TAD1 overnight instead of one hour. The concentration of fibrils loaded onto the membrane was titrated (1.5 ng, 3 ng, or 6 ng) and randomly placed. (E) TAD1 fluorescence intensity of 30 μL EDTA plasma from negative controls, including one light chain amyloidosis (AL) patient, carriers of pathogenic TTR alleles with no disease phenotype, ATTR patients before starting treatment, and ATTR patients on treatment. Methods are described as in A with the following modifications: 30 μL of each sample was loaded onto the membrane before being probed with 5 μM TAD1 overnight. Sample numbers are included in the labels. All statistical analysis was performed as in D. (F) TAD1 fluorescence intensity comparing non-amyloid controls, ATTR patients, and AL patients in an independent assay. Sample numbers are included in the labels. Statistical analysis was performed as in D.

We evaluated the sensitivity and precision of our TAD1 dot blotting system by titrating ATTR fibrils on a membrane and incubating fibrils with TAD1. TAD1 displays high sensitivity, recognizing approximately 20 ng of purified ATTR fibrils with a calculated K_D of 26.1 ng at a concentration of 5 μM (Figure C). To assess the precision of the fluorescence readouts, we loaded different amounts of ATTR fibrils (1.5 ng, 3 ng, and 6 ng) onto a membrane and probed it with 10 μM TAD1. Quantification of the relative fluorescence signal shows that TAD1 can distinguish between small variations in the amount of fibrils with statistical significance (Figure D). Our probe and experimental setup display high sensitivity and precision.

Previous studies indicate that TTR can adopt non-native conformations in plasma of neuropathic (non-cardiomyopathic) ATTR-V30M patients⁵ and we hypothesized that TAD1 could detect these and/or similar species. Using the previous workflow, we tested the binding of 5 μM TAD1 to 30 μL of ethylenediaminetetraacetic acid (EDTA) plasma from ATTR patients pre- and post-treatment (including wild-type and hereditary cases), carriers of pathogenic TTR alleles without symptoms, immunoglobulin light chain amyloidosis (AL) patients, and healthy non-amyloid controls. TAD1 signal is significantly higher in ATTR patients (pre-treatment) when compared with controls (Figure E). TAD1 did not discriminate ATTR genotypes, evidencing TAD1-positive species as a novel biomarker of ATTR amyloidosis. We also found statistically significant differences in the TAD1 signal between ATTR patients pre- and post-treatment (Figure E), supporting the hypothesis that this biomarker is influenced by ATTR-specific therapies. Additionally, we detected TAD1 signal in some carriers of pathogenic TTR alleles who do not yet manifest disease, indicating that TAD1 may be used as a tool for early detection of these unique ATTR species in plasma (Figure E). TAD1 signal for AL patients is significantly lower compared with ATTR patients, further demonstrating the specificity of TAD1 for ATTR (Figure F).

Possible limitations include a limited sample size, as the current sample size does not have the power to adjust for confounding variables. There is a selection bias related to location, disease prevalence, treatment, and sample source. Our study also needs independent validation. Sensitivity and precision assays were completed in different experimental conditions than those done to assess TAD1 activity with patient samples.

In the present study, we use known aggregation-driving segments of transthyretin to design a peptide for the detection of ATTR aggregates in cardiac tissues and plasma with high specificity and sensitivity. Our results suggest that ATTR plasma contains a unique biomarker that decreases in response to treatment and appears in the blood prior to symptoms. This biomarker may represent a useful detection tool with possible applications for early ATTR detection and monitoring treatment response. Our findings can open new avenues of study of the biology and pathogenesis of ATTR, which may result in the identification of new targets for therapeutic development.

Abbreviations:

ATTR

Transthyretin amyloidosis

TTR

transthyretin

TAD1

transthyretin aggregate detector 1

EDTA

ethylenediaminetetraacetic acid

AL

immunoglobulin light chain amyloidosis