Rare missense variants in Tropomyosin‐4 (TPM4) are associated with platelet dysfunction, cytoskeletal defects, and excessive bleeding

A significant challenge is faced for the genetic diagnosis of inherited platelet disorders in which candidate genetic variants can be found in more than 100 bleeding, thrombotic, and platelet disorder genes, especially within families in which there are both normal and low platelet counts. Genetic variants of unknown clinical significance (VUS) are found in a significant proportion of such patients in which functional studies are required to prove pathogenicity.


| INTRODUC TI ON
Inherited platelet disorders comprise an extremely heterogeneous group making genetic diagnosis challenging, especially in those with normal platelet counts. 1,2 Some patients experience frequent and debilitating bleeding episodes throughout their lifetimes; therefore, a precise genetic diagnosis provides significant clinical benefit for the patients and offers the possibility of accurate counseling and clinical management. Current next generation sequencing (NGS) technologies allow the detection of both point mutations and copy number variants (CNVs) with a single platform and workflow, thus allowing a comprehensive molecular diagnosis. [3][4][5] Currently, 93 bleeding, thrombotic, and platelet disorder (BTPD) Tier 1 genes exist (https://www.isth.org/page/GinTh_GeneL ists) for which a well-curated and evidence-based catalogue of gene-disease associations for use in diagnostic genetic screening of patients has been developed by the International Society on Thrombosis and Haemostasis (ISTH) Scientific and Standardization Committee (SSC) on Genomics in Thrombosis and Haemostasis (GinTH). 6 A further 8 Tier 2 genes exist (PRKACG, TRPM7, TPM4, EPHB2, PTPRJ, NFE2, BLOC1S5, PTGS1) that require further pedigrees in order to reach the tier 1 threshold of three unrelated families with causal variants in the same gene. 7 Furthermore, a recent initiative to investigate pathogenic variants, variants of unknown clinical significance (VUS), and benign variants was set out by the SSC-GinTH and a tool developed (Gold variants) to capture the genomic variants detected in patients with BTPD. 8 In this study we describe and elucidate the genetic and functional basis of two unrelated families with three affected individuals with missense variants in TPM4 who suffer from a significant bleeding diathesis with previously unknown genetic diagnosis and therefore make TPM4 a Tier 1 diagnostic gene.

| Patient recruitment and testing
Patients were consented and recruited to the Genotyping and Phenotyping of Platelets (GAPP) study from multiple collaborating hemophilia centers across the UK and Ireland as previously

| Sequencing
Whole exome sequencing (WES) and targeted analysis was performed in patient genomic DNA as previously reported. 10 WES filtering of candidate genetic variants was performed to identify rare variants classified as below a frequency of 0.0001 and as performed previously. 10 P1 and P2 were sequenced using the R90 bleeding and platelet disorders targeted gene panel performed at Oxford Regional Genetics Laboratories. The final sequence variants were confirmed in patients using Sanger sequencing.

| Platelet phenotyping
Peripheral blood was collected from patients and platelet phenotyping using lumiaggregometry or flow cytometry was performed on platelet-rich plasma (PRP) as previously described. 10,12,13 Platelet spreading, transmission electron microscopy (TEM), and western blotting were performed as previously described. 7,14 genetic screening for patients with significant bleeding and undiagnosed platelet disorders, particularly for those with a normal platelet count.

K E Y W O R D S
bleeding, cytoskeleton, next generation sequencing, platelet disorder, platelet dysfunction, TPM4

Essentials
• Identifying genetic variants in platelet disorders is challenging due to their heterogenous nature.
• We performed detailed genetic and functional analysis in patients with missense variants in TPM4.
• TPM4 missense variants are associated with reduced platelets secretion, spreading, and cytoskeletal defects.

| RE SULTS AND D ISCUSS I ON
Both families (F1 and F2) presented here have a strong family history of clinical bleeding including easy bruising, prolonged bleeding after cuts, and poor healing. The patients (P1-P3) were recruited to the UK GAPP study for platelet function testing and genetic studies ( Figure 1A, Table 1). 15 The index case (F1: P1) was a 14-year-old female recruited to the GAPP study with a strong history of a "probable" hereditary platelet disorder with epistaxis, cutaneous bleeding, bleeding from minor wounds, oral cavity bleeding, and menorrhagia. She had an initial platelet count of 153 × 10 9 /L and MPV of 12.1 fL. Her ISTH BAT score was 14. 11 Her mother (F1: P2) was 49 years old when recruited to the study with a clinical history of epistaxis, cutaneous bleeding, bleeding from minor wounds, oral cavity bleeding and bleeding following tooth extraction, menorrhagia, and postpartum hemorrhage (BAT score 22). She had a platelet count of 244 × 10 9 /L and MPV of 11.1 fL (Table 1). F2: P3 suffers from easy bruising and prolonged bleeding after minor cuts and dental extractions. She had a mild thrombocytopenia (last recorded platelet count 119 × 10 9 /L) and initial local investigations showed impaired aggregation to low concentrations of ADP, a prolonged lag phase with collagen, and reduced release of ADP and ATP using platelet nucleotide studies. All other hematological parameters for all three patients were within normal ranges.  Figure 2C). F I G U R E 1 Platelet functional analysis of TPM4 families with an inherited platelet disorder. A, Pedigree of family (F1) showing affected individuals (shaded) and platelet counts. B, Platelet spreading on fibrinogen, labeled with phalloidin-Alexa488 and imaged on a Zeiss Axio Observer7 microscope with a 63x 1.4NA oil objective, Colibri 7 LED light source, Zeiss filter set 38 for GFP/FITC and a Hamamatsu ORCA Flash 4 LT sCMOS camera. i, Representative images of spread platelets from P1, P2, and control; scale bar: 10 µm. ii, Platelet surface area and platelet type calculated from all platelets in 10 fields of view for each condition (total >2500 platelets measured) using the Knime workflow. 14 Data on platelet surface area analyzed using the non-parametric Kruskal-Wallis test with multiple comparisons. iii, Platelet type analysis. Type 1: adhered, type 2: filopodia, type 3: filopodia and lamellipodia, type 4: spread with stress fibres. Scale bar: 5 µm. Significance is shown against control. ****P < .0001. C, Platelet transmission electron microscopy (TEM) showed reduced granule content in TPM4 patients and enlarged vacuole spaces indicating increased platelet fragility. i, Representative images of platelet TEM sections in control and TPM4 patient platelets. Alpha (α) granules (yellow arrow), dense (δ) granules (yellow arrowhead). Other platelet ultrastructure features include: mitochondria (M), lysosomes (L), open canalicular system (OCS) glycogen (green arrow), and large vacuoles only present in TPM4patient platelets (V). Scale bar 500 nm. ii, Quantification of platelet α and δ granules from TEM images. Data presented is mean ± standard error of the mean (SEM) from 30-60 platelets per sample. Significance assessed by one-way analysis of variance with correction for multiple comparisons. **P = .009, ****P < .0001. D, In vitro assessment of platelet glycoprotein receptors and activation in P1, P2 and control. i, Resting platelet surface glycoprotein expression levels. Median fluorescence intensity (MFI). Mean ±SEM of two technical replicates per sample; significance assessed with t-test comparing each patient to experimental control. P-value adjusted to correct for multiple comparisons (Holm-Sidak). *P < .05, **P < .01, ***P < .001, ****P < .0001. ii, Activated P-selectin and (iii) activated αIIbβ3 (PAC-1) expression on control/patient platelets in response to indicated agonist stimulation. The percentage of platelets expressing activation markers was assessed. Data presented is mean ±SEM of two technical replicates per sample. Significance assessed with t-test comparing each patient to experimental control for each agonist. P-value adjusted to correct for multiple comparisons (Holm-Sidak) *p < .05, **P < .01, ***P < .001, ****P < .0001. E, Patients show abnormal aggregation responses to major platelet agonists. Platelet responses to (i) collagen 3 µg/mL (ii) collagen 10 µg/mL (iii) PAR-1 peptide 30 µM and (iv) PAR-1 peptide 100 µM Tropomyosins are dimers of coiled-coil proteins that polymerize end-to-end along the major groove in most actin filaments and provide stability to the filaments, regulating access of other actin-binding proteins. 17 In mammals, more than 40 tropomyosin isoforms can be generated through alternative splicing from four tropomyosin genes. Different isoforms display non-redundant functions and partially non-overlapping localization patterns, for example within the stress fiber network. One such stress fiberassociated tropomyosin Tpm4.2 has previously been associated with a platelet disorder and a macrothrombocytopenia with a nonsense mutation in TPM4. 18 To further investigate whether the p.R182C variant in this study disrupted tropomyosin-4 structures and the associated actin filaments, immunofluorescence was performed in platelets from P1 and P2 to assess phalloidin staining together with a TPM4-specific antibody. Washed platelets from P1 and P2 spread onto fibrinogen displayed disordered tropomyosin-4 staining compared to a healthy control TA B L E 1 Patient demographics and platelet function analysis in patients P1-P3 at time of recruitment  In a separate family 2 (P3) an amino acid substitution was detected (p.A183V). 12 The previously reported nonsense variant (p.R105*) is shown in red font. 18 The tropomyosin domain spans from amino acids 48 to 283. Genetic alterations are numbered according to positions in the NM_001145160.1 transcript for TPM4 (largest isoform). D, TPM4 immunofluorescence showing disordered localization in patient platelets spread on fibrinogen. 11 Representative cropped images from platelets derived from P1, P2, and control samples, seeded onto fibrinogen, fixed and co-stained for TPM4 (magenta), phalloidin (white), and imaged in 3D using AiryScan confocal microscope. Scale bar 10 µm, zoomed panel 5 µm. E, TPM4 protein expression was measured from platelet lysates of TPM4 patients compared to normal controls. No difference was observed in protein expression of TPM4. Expression was normalized to the α-tubulin loading control. Images were analyzed in ImageJ software and significance assessed in Prism. N=3 experiments with 3 control samples used in each experiment. Data presented is mean expression relative to α-tubulin ± standard error of the mean ( Figure 2D). In addition, platelet TPM4 protein levels did not appear to be affected by the missense mutations present in P1 and P2 and were thus comparable to controls assessed by western blot ( Figure 2E).
This report describes three patients from two unrelated families with an inherited platelet disorder and history of excessive bleeding. Both families harbor rare conserved missense variants