Finding and following: a deep learning-based pipeline for tracking platelets during thrombus formation in vivo and ex vivo.

The last decade has seen increasing use of advanced imaging techniques in platelet research. However, there has been a lag in the development of image analysis methods, leaving much of the information trapped in images. Herein, we present a robust analytical pipeline for finding and following individual platelets over time in growing thrombi. Our pipeline covers four steps: detection, tracking, estimation of tracking accuracy, and quantification of platelet metrics. We detect platelets using a deep learning network for image segmentation, which we validated with proofreading by multiple experts. We then track platelets using a standard particle tracking algorithm and validate the tracks with custom image sampling - essential when following platelets within a dense thrombus. We show that our pipeline is more accurate than previously described methods. To demonstrate the utility of our analytical platform, we use it to show that in vivo thrombus formation is much faster than that ex vivo. Furthermore, platelets in vivo exhibit less passive movement in the direction of blood flow. Our tools are free and open source and written in the popular and user-friendly Python programming language. They empower researchers to accurately find and follow platelets in fluorescence microscopy experiments.

In this paper we describe computational tools to find and follow individual platelets in blood clots recorded with fluorescence microscopy. Our tools work in a diverse range of conditions, both in living animals and in artificial flow chamber models of thrombosis. Our work uses deep learning methods to achieve excellent accuracy. We also provide tools for visualizing data and estimating error rates, so you don't have to just trust the output. Our workflow measures platelet density, shape, and speed, which we use to demonstrate differences in the kinetics of clotting in living vessels versus a synthetic environment. The tools we wrote are open source, written in the popular Python programming language, and freely available to all. We hope they will be of use to other platelet researchers.

PI3KC2α inhibition is antithrombotic in blood from hypercholesterolemic mice.

BACKGROUND: Current antiplatelet agents exhibit reduced antithrombotic efficacy in high-risk populations such as populations with hypercholesterolemia. The class II PI3-kinase, PI3KC2α, is a recently discovered target for novel antiplatelet therapy. PI3KC2α inhibition is antithrombotic in healthy mouse models, but whether this is preserved in hypercholesterolemia remains unknown. OBJECTIVES: This study aimed to examine whether genetic deficiency or pharmacologic inhibition of PI3KC2α provides antithrombotic effects in blood from hypercholesterolemic mice. METHODS: Hypercholesterolemic PI3KC2α-deficient mice were generated by breeding into an ApoE-/- background. Thrombosis was examined using an ex vivo whole blood thrombosis assay. The effect of pharmacologic inhibition of PI3KC2α was examined in whole blood from ApoE-/- mice treated with the PI3KC2α inhibitor MIPS-21335. RESULTS: ApoE-/- mice exhibited the anticipated prothrombotic effect of hypercholesterolemia, with a 1.5-fold increase in thrombus volume in blood from ApoE-/- vs wild-type mice. This prothrombotic phenotype in blood from hypercholesterolemic mice was significantly reduced with PI3KC2α deficiency. Acute pharmacologic inhibition of PI3KC2α with MIPS-21335 similarly reduced thrombosis in blood from ApoE-/- mice. CONCLUSION: These findings demonstrate that targeting PI3KC2α results in a potent antithrombotic effect in hypercholesterolemic mice and suggest that PI3KC2α is a promising target for antithrombotic therapy in patients with hypercholesterolemia at a high risk of thrombotic events.

Determination of PAR4 numbers on the surface of human platelets: no effect of the single nucleotide polymorphism rs773902.

The thrombin receptor, protease-activated receptor 4 (PAR4), is important for platelet activation and is the target of emerging anti-thrombotic drugs. A frequently occurring single nucleotide polymorphism (SNP; rs773902) causes a function-altering PAR4 sequence variant (NC_000019.10:p.Ala120Thr), whereby platelets from Thr120-expressing individuals are hyper-responsive to PAR4 agonists and hypo-responsive to some PAR4 antagonists than platelets from Ala120-expressing individuals. This altered pharmacology may impact PAR4 inhibitor development, yet the underlying mechanism(s) remain unknown. We tested whether PAR4 surface expression contributes to the altered receptor function. Quantitative flow cytometry was used to determine the absolute number of PAR4 on platelets from individuals subsequently genotyped at rs773902. We detected 539 ± 311 PAR4 per platelet (mean ± SD, n = 84). This number was not different across rs773902 genotypes. This first determination of cellular PAR4 numbers indicates variations in platelet surface expression do not explain the altered pharmacology of the rs773902 PAR4 sequence variant.

Analysis of the F2LR3 (PAR4) Single Nucleotide Polymorphism (rs773902) in an Indigenous Australian Population.

The F2RL3 gene encoding protease activated receptor 4 (PAR4) contains a single nucleotide variant, rs773902, that is functional. The resulting PAR4 variants, Thr120, and Ala120, are known to differently affect platelet reactivity to thrombin. Significant population differences in the frequency of the allele indicate it may be an important determinant in the ethnic differences that exist in thrombosis and hemostasis, and for patient outcomes to PAR antagonist anti-platelet therapies. Here we determined the frequency of rs773902 in an Indigenous Australian group comprising 467 individuals from the Tiwi Islands. These people experience high rates of renal disease that may be related to platelet and PAR4 function and are potential recipients of PAR-antagonist treatments. The rs773902 minor allele frequency (Thr120) in the Tiwi Islanders was 0.32, which is similar to European and Asian groups and substantially lower than Melanesians and some African groups. Logistic regression and allele distortion testing revealed no significant associations between the variant and several markers of renal function, as well as blood glucose and blood pressure. These findings suggest that rs773902 is not an important determinant for renal disease in this Indigenous Australian group. However, the relationships between rs773902 genotype and platelet and drug responsiveness in the Tiwi, and the allele frequency in other Indigenous Australian groups should be evaluated.

Using PAR4 Inhibition as an Anti-Thrombotic Approach: Why, How, and When?

Protease-activated receptors (PARs) are a family of four GPCRs with a variety of cellular functions, yet the only advanced clinical endeavours to target these receptors for therapeutic gain to date relates to the impairment of platelet function for anti-thrombotic therapy. The only approved PAR antagonist is the PAR1 inhibitor, vorapaxar-the sole anti-platelet drug against a new target approved in the past 20 years. However, there are two PARs on human platelets, PAR1 and PAR4, and more recent efforts have focused on the development of the first PAR4 antagonists, with first-in-class agents recently beginning clinical trial. Here, we review the rationale for this approach, outline the various modes of PAR4 inhibition, and speculate on the specific therapeutic potential of targeting PAR4 for the prevention of thrombotic conditions.

The Influence of Differentially Expressed Tissue-Type Plasminogen Activator in Experimental Autoimmune Encephalomyelitis: Implications for Multiple Sclerosis.

Tissue type plasminogen activator (t-PA) has been implicated in the development of multiple sclerosis (MS) and in rodent models of experimental autoimmune encephalomyelitis (EAE). We show that levels of t-PA mRNA and activity are increased ~4 fold in the spinal cords of wild-type mice that are mice subjected to EAE. This was also accompanied with a significant increase in the levels of pro-matrix metalloproteinase 9 (pro-MMP-9) and an influx of fibrinogen. We next compared EAE severity in wild-type mice, t-PA-/- mice and T4+ transgenic mice that selectively over-express (~14-fold) mouse t-PA in neurons of the central nervous system. Our results confirm that t-PA deficient mice have an earlier onset and more severe form of EAE. T4+ mice, despite expressing higher levels of endogenous t-PA, manifested a similar rate of onset and neurological severity of EAE. Levels of proMMP-9, and extravasated fibrinogen in spinal cord extracts were increased in mice following EAE onset regardless of the absence or over-expression of t-PA wild-type. Interestingly, MMP-2 levels also increased in spinal cord extracts of T4+ mice following EAE, but not in the other genotypes. Hence, while the absence of t-PA confers a more deleterious form of EAE, neuronal over-expression of t-PA does not overtly protect against this condition with regards to symptom onset or severity of EAE.

Coeliac disease patients do not produce antibodies to a common cerebellar epitope.

BACKGROUND: Most adult-onset sporadic ataxias are unexplained, and the claim that many of these may be a result of gluten sensitivity has led to uncertainty as to whether to test for anti-gliadin antibodies (αGAb) and, if present, whether to recommend a gluten-free diet or continue searching for other causes of ataxia. This uncertainty arises in part from the frequency of αGAb in the population (about 1 in 10), but recent work delineating transglutaminase 6 as the target antigen in gluten ataxia has clarified the situation somewhat. Our aim was to determine whether there is molecular mimicry between cerebellar Purkinje cell antigens and gluten in subjects selected for recent diagnosis of CD rather than for ataxia. RESULTS: High titre αGAb sera from 11 newly-diagnosed CD patients and normal sera from 10 healthy controls were used to detect cross-reacting antibodies to cerebellar and cerebral cortex antigens in mouse, monkey and human tissue. None of the CD patients displayed ataxia. Mouse and human cerebellar and cerebral cortex extracts were analysed by Western blot probed with CD and control sera. Immunofluorescence microscopy was used on mouse and monkey cerebellar sections immunostained with CD and control sera to detect cross-reacting IgG antibodies. Western blot analysis of cerebellar and cerebral cortex extracts probed with CD sera did not demonstrate any specific immunoreactivity unique to the cerebellum. An identical twin pair with CD produced different patterns of reactivity. Immunofluorescence staining of mouse and monkey cerebellar sections showed most control and CD sera reacted non-specifically, with the exception of two CD and one control sera, each having a unique staining pattern. CONCLUSIONS: CD patient sera with high titre αGAb do not detect a common Purkinje cell or cerebellar-specific epitope. The pattern of reactivity is not solely dependent on genetic background.

Assessing the efficacy of specific cerebellomodulatory drugs for use as therapy for spinocerebellar ataxia type 1.

Spinocerebellar ataxias are autosomal dominant diseases, associated in some types with a CAG repeat expansion, and characterised by a progressive loss of motor function. Currently, as there is no cure for most ataxias, treatment predominantly involves physical therapy. Various symptomatic drug treatments have been tried; however, published clinical studies have provided inconsistent results, likely due to small sample sizes, mixed patient populations and insensitive or subjective assessment scales. SCA1(154Q) transgenic mice display motor function impairments and ultimately a reduced number of cerebellar Purkinje neurons-characteristics comparable to most forms of sporadic and hereditary ataxias. We monitored motor function in SCA1(154Q) mice from 5 to 20 weeks of age and assessed the efficacy of four potential cerebellar modulatory drugs in attenuating deficits in rotor-rod performance. The drugs riluzole, amantadine, zolpidem and buspirone were selected based on their different mechanisms of action and their Food and Drug Administration (FDA)/Australian Therapeutic Goods Administration approval for other indications. SCA1(154Q) and C57/Bl6 wild-type mice were administered with four ascending acute doses of each drug, over 2 days. Following each dose, mice were assesed for motor function on the accelerating rotor-rod. None of the four drugs attenuated motor deficts in SCA1(154Q) mice at any dose; at FDA equivalent and higher dose administration of zolpidem and buspirone led to sedation in both strains. Our results suggest that the aforementioned drugs are likely to be ineffective for symptomatic treatment of SCA1 and most other ataxic patients and emphasise the need for comphrehensive drug studies prior to clinical use.

HLA-DR3-DQ2 mice do not develop ataxia in the presence of high titre anti-gliadin antibodies.

Recently, it has been suggested that anti-gliadin antibodies (αGAb) may produce "gluten ataxia", even in the absence of celiac disease enteropathy. αGAb are reportedly present in 12-50 % of patients with sporadic ataxia, but also in 12 % of the general population, such that the importance of αGAb as a cause of sporadic ataxia is not conclusively settled. We aimed to determine whether mice transgenic for HLA-DR3-DQ2 and immunised with gliadin to achieve high titres of αGAb would develop ataxia and/or cerebellar damage. From 6 weeks of age, HLA-DR3-DQ2 transgenic mice were immunised fortnightly with gliadin (n = 10) or a saline control (n = 6) in adjuvant. Serum titres were measured by αGAb enzyme-linked immunosorbent assay. At 24 weeks of age, mice were tested for locomotor function using the accelerating rotarod, ledged beam, ink-paw gait, and several neurological severity score subtests. Brains were then collected and processed for immunohistochemistry. Sections were analysed for lymphocytic infiltration, changes in morphology and Purkinje cell (PC) dendritic volume and the number of PCs counted via unbiased stereology. Gliadin-immunised mice developed high αGAb titres while controls did not. There was no statistically significant difference between the gliadin and sham-immunised HLA-DR3-DQ2 mice on any of the tests of motor coordination, in lymphocytic infiltration, PC number or in dendritic volume. High levels of αGAb are not sufficient to produce ataxia or cerebellar damage in HLA-DR3-DQ2 transgenic mice.

Reprint of: Expression and purification of ataxin-1 protein.

Ataxin-1 is part of a larger family of polyglutamine-containing proteins that is linked to nine distinct neurodegenerative disorders. There are no known effective therapies for any of these expanded polyglutamine tract disorders. One possible reason for this is the lack of sufficient amounts of pure polyglutamine-containing proteins suitable for biochemical and conformational studies. Here, we show that we were able to successfully purify a non-pathological, wild-type human ataxin-1 protein containing a 30-glutamine repeat sequence. This ataxin-1 protein was expressed in Escherichia coli as a fusion protein with a GST tag at the N-terminus and a double (His)(6) tag at the C-terminus. The devised dual affinity tag strategy allowed successful purification of the full-length ataxin-1 fusion protein to 90% homogeneity as confirmed by Western blot analysis using the two monoclonal ataxin-1 antibodies developed in our laboratory. In addition, the GST tag was successfully removed from the purified ataxin-1 fusion protein by treatment with Tobacco etch virus (TEV) protease. Since polyglutamine-containing proteins tend to aggregate, solvents/buffers that minimize aggregation have been used in the purification process. This dual affinity purification protocol could serve as a useful basis for purifying aggregation-prone proteins that are involved in other neurodegenerative diseases.

Expression and purification of ataxin-1 protein.

Ataxin-1 is part of a larger family of polyglutamine-containing proteins that is linked to nine distinct neurodegenerative disorders. There are no known effective therapies for any of these expanded polyglutamine tract disorders. One possible reason for this is the lack of sufficient amounts of pure polyglutamine-containing proteins suitable for biochemical and conformational studies. Here, we show that we were able to successfully purify a non-pathological, wild-type human ataxin-1 protein containing a 30-glutamine repeat sequence. This ataxin-1 protein was expressed in Escherichia coli as a fusion protein with a GST tag at the N-terminus and a double (His)(6) tag at the C-terminus. The devised dual affinity tag strategy allowed successful purification of the full-length ataxin-1 fusion protein to 90% homogeneity as confirmed by Western blot analysis using the two monoclonal ataxin-1 antibodies developed in our laboratory. In addition, the GST tag was successfully removed from the purified ataxin-1 fusion protein by treatment with Tobacco etch virus (TEV) protease. Since polyglutamine-containing proteins tend to aggregate, solvents/buffers that minimize aggregation have been used in the purification process. This dual affinity purification protocol could serve as a useful basis for purifying aggregation-prone proteins that are involved in other neurodegenerative diseases.

Inclusion formation by ataxins -1, -2, -3, and -7.

The authors studied inclusion formation in vitro using transiently transfected PC12 cells, with epitope-tagged and untagged full-length and truncated wild-type and expanded ataxins -1, -2, -3, and -7. At 72 hours, no inclusions were seen with wild-type full-length or truncated ataxins -2, -3, or -7, and only one with ataxin-1. Truncation abolished nuclear localization of ataxins -1 and -7, and allowed nuclear entry of ataxin-2. Of the expanded ataxins, only -1 and -2 formed inclusions, and those of ataxin-2 were rare and exclusively cytoplasmic. Truncation resulted in inclusion formation by ataxins -3 and -7, increased ataxin-1 inclusions, and enabled formation of nuclear ataxin-2 inclusions. There was no recruitment of wild-type ataxin-1 to expanded ataxin-1 inclusions.

Different ataxin-2 antibodies display different immunoreactive profiles.

We have developed a monoclonal antibody (4A7) directed against the C-terminus of the ataxin-2 protein that is involved in the polyglutamine neurodegenerative disorder spinocerebellar ataxia type 2. Comparison with other ataxin-2 antibodies showed that 4A7 specifically recognized ataxin-2. In contrast, a previously reported ataxin-2 antibody (15F6) did not appear to recognize full-length ataxin-2 in our systems. Immunocytochemical and subcellular fractionation studies using 4A7 confirmed previous reports that ataxin-2 is localized to both the cytoplasm and the trans-Golgi network in rat PC12 cells and rat brain tissue. In contrast, 4A7 failed to label the trans-Golgi network in the three primate cell lines examined. Cytoplasmic ataxin-2 was not associated with mitochondria, lysosomes, endoplasmic reticulum, peroxisomes, proteasomes, clathrin-coated pits or vesicles, or F-actin. Ataxin-2 was found to be phosphorylated but not glycosylated, and exhibited an estimated half-life of not less than 21 h. Interestingly, another commercially available ataxin-2 antibody did not detect ataxin-2 localized to the trans-Golgi network. This antibody was also found to immunoprecipitate fewer proteins/protein partners than 4A7. Although cross-reactivity of the 4A7 antibody with other protein(s) cannot be ruled out, it appears likely that the interaction of ataxin-2 with other cell components is dependent on both the host cell type and its subsequent subcellular localization.