A bibliometric review of peripartum cardiomyopathy compared to other cardiomyopathies using artificial intelligence and machine learning.

As developments in artificial intelligence and machine learning become more widespread in healthcare, their potential to transform clinical outcomes also increases. Peripartum cardiomyopathy is a rare and poorly-characterised condition that presents as heart failure in the last trimester prior to delivery or within 5-6 months postpartum. The lack of a definitive understanding of the molecular causes and clinical progress of this condition suggests that bibliometrics will be well-suited to creating new insights into this serious clinical problem. We examine similarities and differences between peripartum and its closely related familial dilated cardiomyopathy and idiopathic dilated cardiomyopathy. Using PubMed as the source of bibliometric data, we apply artificial intelligence-supported natural language processing to compare extracted data and genes association with these cardiomyopathies. Gene data were enhanced with additional metadata from third-party datasets and then analysed for their impact and specificity for peripartum cardiomyopathy. Artificial intelligence identified 14 genes that distinguished peripartum from both dilated and familial dilated cardiomyopathy. They are as follows: CTSD, RLN2, MMP23B*, SLC17A5, ST2*, PTHLH, CFH*, CFI, GPT, MR1, Rln1, SRI, STAT5A* and THBD. We then used the Human Protein Atlas website that uses affinity-purified rabbit polyclonal antibodies to identify genes that are expressed at the protein level (bold), or as RNA transcripts (*) in healthy human left ventricles. Additional analysis focussed on the full set of peripartum genes on linkage and specificity to cardiomyopathy yielded a different set of thirteen genes (bold font indicates those expressed in cardiomyocytes: PRL, RLN2, PLN, ST2, CTSD, F2, ACE, STAT3, TTN, SPP1, LGALS3, miR-146a, GNB3, SRI). This type of analysis can highlight new avenues for research, aimed at improving genomics-driven peripartum cardiomyopathy diagnosis as well as potential pathological and clinical sub-classification. We expect that this will allow for future improvements in identification, treatment and management of this condition. The first step in the application of these bibliometric-based artificial intelligence methods is to understand the current knowledge, and it is the aim of this paper to show how this might be achieved.

Correction to: The Sydney Heart Bank: improving translational research while eliminating or reducing the use of animal models of human heart disease.

In the original version of this article, the name of one of the authors is not correct. The correct name should be W. A. Linke, which is shown correctly in the authorgroup section above.

The Sydney Heart Bank: improving translational research while eliminating or reducing the use of animal models of human heart disease.

The Sydney Heart Bank (SHB) is one of the largest human heart tissue banks in existence. Its mission is to provide high-quality human heart tissue for research into the molecular basis of human heart failure by working collaboratively with experts in this field. We argue that, by comparing tissues from failing human hearts with age-matched non-failing healthy donor hearts, the results will be more relevant than research using animal models, particularly if their physiology is very different from humans. Tissue from heart surgery must generally be used soon after collection or it significantly deteriorates. Freezing is an option but it raises concerns that freezing causes substantial damage at the cellular and molecular level. The SHB contains failing samples from heart transplant patients and others who provided informed consent for the use of their tissue for research. All samples are cryopreserved in liquid nitrogen within 40 min of their removal from the patient, and in less than 5-10 min in the case of coronary arteries and left ventricle samples. To date, the SHB has collected tissue from about 450 failing hearts (>15,000 samples) from patients with a wide range of etiologies as well as increasing numbers of cardiomyectomy samples from patients with hypertrophic cardiomyopathy. The Bank also has hearts from over 120 healthy organ donors whose hearts, for a variety of reasons (mainly tissue-type incompatibility with waiting heart transplant recipients), could not be used for transplantation. Donor hearts were collected by the St Vincent's Hospital Heart and Lung transplantation team from local hospitals or within a 4-h jet flight from Sydney. They were flushed with chilled cardioplegic solution and transported to Sydney where they were quickly cryopreserved in small samples. Failing and/or donor samples have been used by more than 60 research teams around the world, and have resulted in more than 100 research papers. The tissues most commonly requested are from donor left ventricles, but right ventricles, atria, interventricular system, and coronary arteries vessels have also been reported. All tissues are stored for long-term use in liquid N or vapor (170-180 °C), and are shipped under nitrogen vapor to avoid degradation of sensitive molecules such as RNAs and giant proteins. We present evidence that the availability of these human heart samples has contributed to a reduction in the use of animal models of human heart failure.

Quantification of thyroxine and 3,5,3'-triiodo-thyronine in human and animal hearts by a novel liquid chromatography-tandem mass spectrometry method.

Assaying tissue T3 and T4 would provide important information in experimental and clinical investigations. A novel method to determine tissue T3 and T4 by HPLC coupled to mass spectrometry is described. The major difference vs. previously described methods lies in the addition of a derivatization step, that is, to convert T3 and T4 into the corresponding butyl esters. The yield of esterification was Ì´ 100% for T3 and 80% for T4. The assay was linear (r>0.99) in the range of 0.2-50 ng/ml, accuracy was in the order of 70-75%, and the minimum tissue amount needed was in the order of 50 mg, that is, about one order of magnitude lower than observed with the same equipment (AB Sciex API 4000 triple quadrupole mass spectrometer) if derivatization was omitted. The method allowed detection of T3 and T4 in human left ventricle biopsies yielding concentrations of 1.51±0.16 and 5.94±0.63 pmol/g, respectively. In rats treated with different dosages of exogenous T3 or T4, good correlations (r>0.90) between plasma and myocardial T3 and T4 concentrations were observed, although in specific subsets different plasma T4 concentrations were not associated with different tissue content in T4. We conclude that this method could provide a novel insight into the relationship between plasma and tissue thyroid hormone levels.

Cluster of differentiation antibody microarrays on plasma immersion ion implanted polycarbonate.

Plasma immersion ion implantation (PIII) modifies the surface properties of polymers, enabling them to covalently immobilize proteins without using linker chemistry. We describe the use of PIII treated polycarbonate (PC) slides as a novel platform for producing microarrays of cluster of differentiation (CD) antibodies. We compare their performance to identical antibody microarrays printed on nitrocellulose-coated glass slides that are currently the industry standard. Populations of leukocytes are applied to the CD microarrays and unbound cells are removed revealing patterns of differentially immobilized cells that are detected in a simple label-free approach by scanning the slides with visible light. Intra-slide and inter-slide reproducibility, densities of bound cells, and limits of detection were determined. Compared to the nitrocellulose-coated glass slides, PIII treated PC slides have a lower background noise, better sensitivity, and comparable or better reproducibility. They require three-fold lower antibody concentrations to yield equivalent signal strength, resulting in significant reductions in production cost. The improved transparency of PIII treated PC in the near-UV and visible wavelengths combined with superior immobilization of biomolecules makes them an attractive platform for a wide range of microarray applications.

A comparison of covalent immobilization and physical adsorption of a cellulase enzyme mixture.

This paper reports the first use of a linker-free covalent approach for immobilizing an enzyme mixture. Adsorption from a mixture is difficult to control due to varying kinetics of adsorption, variations in the degree of unfolding and competitive binding effects. We show that surface activation by plasma immersion ion implantation (PIII) produces a mildly hydrophilic surface that covalently couples to protein molecules and avoids these issues, allowing the attachment of a uniform monolayer from a cellulase enzyme mixture. Atomic force microscopy (AFM) showed that the surface layer of the physically adsorbed cellulase layer on the mildly hydrophobic surface (without PIII) consisted of aggregated enzymes that changed conformation with incubation time. The evolution observed is consistent with the existence of transient complexes previously postulated to explain the long time constants for competitive displacement effects in adsorption from enzyme mixtures. AFM indicated that the covalently coupled bound layer to the PIII-treated surface consisted of a stable monolayer without enzyme aggregates, and became a double layer at longer incubation times. Light scattering analysis showed no indication of aggregates in the solution at room temperature, which indicates that the surface without PIII-treatment induced enzyme aggregation. A model for the attachment process of a protein mixture that includes the adsorption kinetics for both surfaces is presented.

An actin-myosin functional assay for analysis of smooth muscle (anti-microfilament) autoantibodies in human plasma.

The detection of serum autoantibodies to smooth muscle (SMA) on rodent gastric mucosa by indirect immunofluorescence (IIF) has long been an immunodiagnostic marker for autoimmune hepatitis type 1 (AIH-1). The reactive antigenic moieties are cytoskeletal proteins which include polymeric F-actin as judged by the staining of microfilaments of tissue by IIF. However, their specificity for actin in AIH-1 can be and usually is uncertain. Using an in vitro functional assay, we compared the effects of Fab fragments of immunoglobulin (IgG) prepared from SMA-positive plasma from two patients with the effects of Fabs from 10 healthy subjects. Fabs are incorporated into an assay where actin (the putative antigen) activates skeletal muscle heavy meromyosin (HMM) ATPase activity. The data from these functional assays provide new insights into the significance of anti-microfilament assays in the diagnosis, and perhaps also pathogenesis, of AIH-1.

Plasma immersion ion implantation treatment of polyethylene for enhanced binding of active horseradish peroxidase.

Robust attachment of active proteins to synthetic surfaces underpins the development of biosensors and protein arrays. This paper presents the results of experiments in which energetic ions, extracted from an inductively coupled nitrogen plasma, are used to modify the surface of ultrahigh molecular weight polyethylene (UHMWPE). The ability of the surface to bind active horseradish peroxidase (HRP) is significantly enhanced by the plasma treatment. The amide signal in infrared spectroscopy indicates an increased quantity of surface-attached protein on the modified surface. The activity of the bound HRP remains high compared with that of protein attached to the untreated surface, after repeated washing in buffer solution. Although Tween 20 was an effective blocking agent for the unmodified polyethylene surface, binding of HRP to the modified surface is not inhibited by its presence. We propose that the treatment produces new binding sites on the surface and that the function of the HRP is retained because the treated surface is substantially more hydrophilic.

Cofilin, actin and their complex observed in vivo using fluorescence resonance energy transfer.

Actin is the principal component of microfilaments. Its assembly/disassembly is essential for cell motility, cytokinesis, and a range of other functions. Recent evidence suggests that actin is present in the nucleus where it may be involved in the regulation of gene expression and that cofilin binds actin and can translocate into the nucleus during times of stress. In this report, we combine fluorescence resonance energy transfer and confocal microscopy to analyze the interactions of cofilin and G-actin within the nucleus and cytoplasm. By measuring the rate of photobleaching of fluorescein-labeled actin in the presence and absence of Cy5-labeled cofilin, we determined that almost all G-actin in the nucleus is bound to cofilin, whereas approximately (1/2) is bound in the cytoplasm. Using fluorescence resonance energy transfer imaging techniques we observed that a significant proportion of fluorescein-labeled cofilin in both the nucleus and cytoplasm binds added tetramethylrhodamine-labeled G-actin. Our data suggest there is significantly more cofilin-G-actin complex and less free cofilin in the nucleus than in the cytoplasm.

A novel biosensor for mercuric ions based on motor proteins.

We explored the potential of contractile proteins, actin and myosin, as biosensors of solutions containing mercuric ions. We demonstrate that the reaction of HgCl2 with myosin rapidly inhibits actin-activated myosin ATPase activity. Mercuric ions inhibit the in vitro analog of contraction, namely the ATP-initiated superprecipitation of the reconstituted actomyosin complex. Hg reduces both the rate and extent of this reaction. Direct observation of the propulsive movement of actin filaments (10 nm in diameter and 1 microm long) in a motility assay driven by a proteolytic fragment of myosin (heavy meromyosin or HMM) is also inhibited by mercuric ions. Thus, we have demonstrated the biochemical, biophysical and nanotechnological basis of what may prove to be a useful nano-device.

[Conformational changes of actin induced by strong or weak myosin subfragment-1 binding]. / Konformatsionnye izmeneniia aktina, vyzvannye sil'nym i slabym sviazyvaniem subfragmenta-1 miozina.

Movements of different areas of polypeptide chains within F-actin monomers induced by S1 or pPDM-S1 binding were studied by polarized fluorimetry. Thin filaments of ghost muscle were reconstructed by adding G-actin labeled with fluorescent probes attached alternatively to different sites of actin molecule. These sites were: Cys-374 labeled with 1,5-IAEDANS, TMRIA or 5-IAF; Lys-373 labeled with NBD-Cl; Lys-113 labeled with Alexa-488; Lys-61 labeled with FITC; Gln-41 labeled with DED and Cys-10 labeled with 1,5-IAEDANS, 5-IAF or fluorescein-maleimid. In addition, we used TRITC-, FITC-falloidin and e-ADP that were located, respectively, in filament groove and interdomain cleft. The data were analysed by model-dependent and model-independent methods (see appendixes). The orientation and mobility of fluorescent probes were significantly changed when actin and myosin interacted, depending on fluorophore location and binding site of actomyosin. Strong binding of S with actin leads to 1) a decrease in the orientation of oscillators of derivatives of falloidin (TRITC-falloidin, FITC-falloidin) and actin-bound nucleotide (e-ADP); 2) an increase in the orientation of dye oscillators located in the "front' surface of the small domain (where actin is viewed in the standard orientation with subdomains 1/2 and 3/4 oriented to the right and to the left, respectively); 3) a decrease in the angles of dye oscillators located on the "back" surface of subdomain-1. In contrast, a weak binding of S1 to actin induces the opposite effects in orientation of these probes. These data suggest that during the ATP hydrolysis cycle myosin heads induce a change in actin monomer (a tilt and twisting of its small domain). Presumably, these alterations in F-actin conformation play an important role in muscle contraction.

Fluorescence depolarization of actin filaments in reconstructed myofibers: the effect of S1 or pPDM-S1 on movements of distinct areas of actin.

Fluorescence polarization measurements were used to study changes in the orientation and order of different sites on actin monomers within muscle thin filaments during weak or strong binding states with myosin subfragment-1. Ghost muscle fibers were supplemented with actin monomers specifically labeled with different fluorescent probes at Cys-10, Gln-41, Lys-61, Lys-373, Cys-374, and the nucleotide binding site. We also used fluorescent phalloidin as a probe near the filament axis. Changes in the orientation of the fluorophores depend not only on the state of acto-myosin binding but also on the location of the fluorescent probes. We observed changes in polarization (i.e., orientation) for those fluorophores attached at the sites directly involved in myosin binding (and located at high radii from the filament axis) that were contrary to the fluorophores located at the sites close to the axis of thin filament. These altered probe orientations suggest that myosin binding alters the conformation of F-actin. Strong binding by myosin heads produces changes in probe orientation that are opposite to those observed during weak binding.

Apparatus for exposing cell membranes to rapid temperature transients.

We seek to determine whether cell membranes contain sensors that trigger a downstream response to temperature excursions. To do this, we have developed a novel apparatus for exposing a cell membrane to an extremely rapid temperature excursion in the nanosecond range. Cells are plated on a gold surface that is back-heated by a pulsed laser and cooled by conduction of heat into the glass substrate and the liquid medium. Analysis using the heat diffusion equation shows that the greatest temperature rise is localized within a region tens of nanometres thick, suitable for specifically heating a cell membrane without heating the remainder of a cell. We refer to this device as a nanosecond hotplate.

Genomics, proteomics and bioinformatics of human heart failure.

Unraveling the molecular complexities of human heart failure, particularly end-stage failure, can be achieved by combining multiple investigative approaches. There are several parts to the problem. Each patient is the product of a complex set of genetic variations, different degrees of influence of diets and lifestyles, and usually heart transplantation patients are treated with multiple drugs. The genomic status of the myocardium of any one transplant patient can be analysed using gene arrays (cDNA- or oligonucleotide-based) each with its own strengths and weaknesses. The proteins expressed by these failing hearts (myocardial proteomics) were first investigated over a decade ago using two-dimensional polyacrylamide gel electrophoresis (2DGE) which promised to resolve several thousand proteins in a single sample of failing heart. However, while 2DGE is very successful for the abundant and moderately expressed proteins, it struggles to identify proteins expressed at low levels. Highly focused first dimension separations combined with recent advances in mass spectrometry now provide new hope for solving this difficulty. Protein arrays are a more recent form of proteomics that hold great promise but, like the above methods, they have their own drawbacks. Our approach to solving the problems inherent in the genomics and proteomics of heart failure is to provide experts in each analytical method with a sample from the same human failing heart. This requires a sufficiently large number of samples from a sufficiently large pool of heart transplant patients as well as a large pool of non-diseased, non-failing human hearts. We have collected more than 200 hearts from patients undergoing heart transplantations and a further 50 non-failing hearts. By combining our expertise we expect to reduce and possibly eliminate the inherent difficulties of each analytical approach. Finally, we recognise the need for bioinformatics to make sense of the large quantities of data that will flow from our laboratories. Thus, we plan to provide meaningful molecular descriptions of a number of different conditions that result in terminal heart failure.

Cardiomyocyte apoptosis is associated with increased wall stress in chronic failing left ventricle.

AIMS: We examined cardiomyocyte apoptosis in chronic heart failure (HF) and its possible link to elevated wall stress. METHODS AND RESULTS: Moderate HF was produced in sheep by sequential coronary microembolization. Six months later, the animals remained in a stable compensated haemodynamic state of HF. Apoptosis of cardiomyocytes in left ventricles was verified using Western blotting based on increased expression of: the apoptosis-associated death receptor Fas (1.5-fold); its ligand (FasL, 2.0-fold); and an upstream protease caspase-8 (2.7-fold) as well as its active cleavage peptide, p20 (5.6-fold). Previously we have reported the elevated expression of caspase-3 in the same animal model. The occurrence of apoptotic cardiomyocytes (0.3%) was quantified by TUNEL assays. Haemodynamic analysis indicated that ventricular dilatation, without wall thickening, caused a 2-fold increase in LV wall stress which, together with LV end-diastolic pressure, was linearly correlated with expression of Fas/FasL. Immunohistochemical studies localized FasL and caspase-8 to intercalated discs, suggesting that wall stress may play a role in initiating cardiomyocyte apoptosis. CONCLUSION: Apoptosis of cardiomyocytes in chronic HF is associated with increased wall stress, which may be responsible for the activation of a Fas/FasL and caspase-8 interaction in the region of intercalated discs.

Actin binding proteins: regulation of cytoskeletal microfilaments.

The actin cytoskeleton is a complex structure that performs a wide range of cellular functions. In 2001, significant advances were made to our understanding of the structure and function of actin monomers. Many of these are likely to help us understand and distinguish between the structural models of actin microfilaments. In particular, 1) the structure of actin was resolved from crystals in the absence of cocrystallized actin binding proteins (ABPs), 2) the prokaryotic ancestral gene of actin was crystallized and its function as a bacterial cytoskeleton was revealed, and 3) the structure of the Arp2/3 complex was described for the first time. In this review we selected several ABPs (ADF/cofilin, profilin, gelsolin, thymosin beta4, DNase I, CapZ, tropomodulin, and Arp2/3) that regulate actin-driven assembly, i.e., movement that is independent of motor proteins. They were chosen because 1) they represent a family of related proteins, 2) they are widely distributed in nature, 3) an atomic structure (or at least a plausible model) is available for each of them, and 4) each is expressed in significant quantities in cells. These ABPs perform the following cellular functions: 1) they maintain the population of unassembled but assembly-ready actin monomers (profilin), 2) they regulate the state of polymerization of filaments (ADF/cofilin, profilin), 3) they bind to and block the growing ends of actin filaments (gelsolin), 4) they nucleate actin assembly (gelsolin, Arp2/3, cofilin), 5) they sever actin filaments (gelsolin, ADF/cofilin), 6) they bind to the sides of actin filaments (gelsolin, Arp2/3), and 7) they cross-link actin filaments (Arp2/3). Some of these ABPs are essential, whereas others may form regulatory ternary complexes. Some play crucial roles in human disorders, and for all of them, there are good reasons why investigations into their structures and functions should continue.

Heart failure and apoptosis: electrophoretic methods support data from micro- and macro-arrays. A critical review of genomics and proteomics.

The multiple causes and multiple consequences of mammalian heart failure make it an attractive proposition for analysis using gene array technology, especially where the failure is idiopathic in nature. However, gene arrays also hold potential artefacts, particularly when gene expression levels are low, and where changes in expression levels are modest. Also, at present, the number of genes available on arrays is not large enough to prevent potential sampling deficiencies. Thus, it may not be wise to place too much reliance on quantitative interpretations of gene array data. Also, recently doubts were raised about the qualitative reliability of array genes. Electrophoretic methods are slow, cumbersome and complex but they can provide confirmation that the trends and numbers arising from the new gene arrays are reliable. In this overview, we compare gene array data with data from protein activity assays such as zymograms, Western blots, two-dimensional electrophoresis, and immunohistochemistry. Similar or complementary data from the same heart tissues analyzed by either microarrays or macroarrays can be reassuring to those interested in reliable molecular analyses of normal and failing hearts. Similar principles will apply to other tissues and cells.