A Physician-in-the-Loop Approach by Means of Machine Learning for the Diagnosis of Lymphocytosis in the Clinical Laboratory.

CONTEXT.­: The goal of the lymphocytosis diagnosis approach is its classification into benign or neoplastic categories. Nevertheless, a nonnegligible percentage of laboratories fail in that classification. OBJECTIVE.­: To design and develop a machine learning model by using objective data from the DxH 800 analyzer, including cell population data, leukocyte and absolute lymphoid counts, hemoglobin concentration, and platelet counts, besides age and sex, with classification purposes for lymphocytosis diagnosis. DESIGN.­: A total of 1565 samples were included from 10 different lymphoid categories grouped into 4 diagnostic categories: normal controls (458), benign causes of lymphocytosis (567), neoplastic lymphocytosis (399), and spurious causes of lymphocytosis (141). The data set was distributed in a 60-20-20 scheme for training, testing, and validation stages. Six machine learning models were built and compared, and the selection of the final model was based on the minimum generalization error and 10-fold cross validation accuracy. RESULTS.­: The selected neural network classifier rendered a global 10-class classification validation accuracy corresponding to 89.9%, which, considering the aforementioned 4 diagnostic categories, presented a diagnostic impact accuracy corresponding to 95.8%. Finally, a prospective proof of concept was performed with 100 new cases with a global diagnostic accuracy corresponding to 91%. CONCLUSIONS.­: The proposed machine learning model was feasible, with a high benefit-cost ratio, as the results were obtained within the complete blood count with differential. Finally, the diagnostic impact with high accuracies in both model validation and proof of concept encourages exploration of the model for real-world application on a daily basis.

Genomic characterization of the nucleus for conservation of the Chino Santandereano breed using SNP markers.

The Chino Santandereano (CHS) creole cattle breed has undergone an adaptation process of more than 500 years to the Colombian mountainous tropics. Despite its advantages, the breed has been replaced by specialized cattle that put it at risk of extinction. The aim of this work was to estimate the genomic diversity and population structure of the nucleus of conservation of the CHS breed. Thirty-seven CHS and 20 Brahman animals were genotyped due to the possible introgression of genes with the GGP Bovine LD v3 chip. Quality control was performed, and linkage disequilibrium (LD), effective population size (Ne), ROH segments, homozygosity, and genomic inbreeding in the breed were estimated. Subsequently, 50 K genomic information of the Holstein (n = 30) and Romosinuano (n = 8) breeds were included to estimate the minor allele frequency (MAF) with common markers and constructing the graphs of the principal component analysis (PCA). Pairwise FSTs were estimated and a neighbor-joining tree was constructed using the IBS matrix. Admixture was used with k = 2 to 10 for the racial composition. LD (r2) was found up to a distance of 0.13 Mb, r2 > 0.3 at a distance of 340.3 kb, and Ne of 32 ± 1. ROH inbreeding was 5.36 ± 0.86%, with a higher contribution from recent inbreeding (4.55%). The PCA showed that the creole breeds were closer together, and the Brahman was more distant. The admixture analysis suggested k = 5 possible ancestral groups and shows that within the CHS breed, there seem to be two different groups with little Holstein and Brahman introgression. The genetic diversity parameters obtained in this work show minimal diversity in this breed and reinforce the need to protect this resource and the conservation banks.

Abnormal characteristic "round bottom flask" shape volume-based scattergram as a trigger to suspect persistent polyclonal B-cell lymphocytosis.

BACKGROUND AND AIMS: The diagnosis of persistent polyclonal B-cell lymphocytosis (PPBL) is often challenging because of the lack of features and the overlap with the peripheral expression of splenic marginal zone lymphomas (SMZL). To obtain new clues for PPBL detection and diagnosis, all data provided by the DxH 800 analyzer (including scatter and cell population data (CPD)) was exploited and combined using a machine learning (ML) approach. MATERIALS AND METHODS: A total 211 samples from 101 normal controls and 110 patients (PPBL and SMZL) were assessed. Age, gender, full blood count, CPD, scatter, flags and CellaVision differentials were also considered. A ML model was built for true classification purposes. RESULTS: PPBL and SMZL shared increased absolute lymphoid counts, atypical lymphoid flag presence and CPD values (8 out of 14). A typical "round-bottom-flask" shape scattergram was described for the first time for PPBL which was also present in 51.4% of SMZL cases. The developed ML model render a global classification accuracy of 93.4%, allowing the detection of all pathological cases, with mean misclassification rates of 12% among PPBL and SMZL. CONCLUSION: Our ML model represents a new unbiased tool than can be widely applied in the laboratory as an aid for detection of PPBL.

Machine learning algorithms for the detection of spurious white blood cell differentials due to erythrocyte lysis resistance.

AIMS: Red blood cell (RBC) lysis resistance interferes with white blood cell (WBC) count and differential; still, its detection relies on the identification of an abnormal scattergram, and this is not clearly adverted by specific flags in the Beckman-Coulter DXH-800. The aims were to analyse precisely the effect of RBC lysis resistance interference in WBC counts, differentials and cell population data (CPD) and then to design, develop and implement a novel diagnostic machine learning (ML) model to optimise the detection of samples presenting this phenomenon. METHODS: WBC counts, differentials and CPD from 232 patients (anaemia or liver disease) were compared with 100 healthy controls (HC) using analysis of variance. The data were analysed after a corrective action, and the analyser differentials were also compared with the digital leucocyte differentials. The ML support vector machine (SVM) algorithm was trained with 70% of the samples (n=233) and the 30% remaining (n=99) were employed exclusively during the validation phase. RESULTS: We identified that impedance WBC was not affected by the RBC lysis resistance interference while the DXH-800 differentials overestimated lymphoid subpopulations (17.6%), sometimes even yielding spurious lymphocytosis, and the latter were corrected when sample dilution was performed. The ML-SVM algorithm allowed the classification of the pathological groups when compared with HC with validation accuracies corresponding to 97.98%, 100% and 88.78% for the global, anaemia and liver disease groups, respectively. CONCLUSIONS: The proposed algorithm has an impressive discriminatory potential and its application would be a valuable support system to detect spurious results due to RBC lysis resistance.

Lectin-Binding Specificity of the Fertilization-Relevant Protein PDC-109 by Means of Surface Plasmon Resonance and Carbohydrate REcognition Domain EXcision-Mass Spectrometry.

Seminal plasma proteins are relevant for sperm functionality and some appear responsible for establishing sperm interactions with the various environments along the female genital tract towards the oocyte. In recent years, research has focused on characterizing the role of these proteins in the context of reproductive biology, fertility diagnostics and treatment of related problems. Herein, we focus on the main protein of bovine seminal plasma, PDC-109 (BSP-A1/-A2), which by virtue of its lectin properties is involved in fertilization. By means of surface plasmon resonance, the interaction of PDC-109 with a panel of the most relevant glycosidic epitopes of mammals has been qualitatively and quantitatively characterized, and a higher affinity for carbohydrates containing fucose has been observed, in line with previous studies. Additionally, using the orthogonal technique of Carbohydrate REcognition Domain EXcision-Mass Spectrometry (CREDEX-MS), the recognition domain of the interaction complexes between PDC-109 and all fucosylated disaccharides [(Fuc-α1,(3,4,6)-GlcNAc)] has been defined, revealing the specific glycotope and the peptide domain likely to act as the PDC-109 carbohydrate binding site.

Lysyl oxidase-like 2 (LOXL2) oxidizes trimethylated lysine 4 in histone H3.

Methylation of histone H3 lysine 4 is linked to active transcription and can be removed by LSD1 or the JmjC domain-containing proteins by amino-oxidation or hydroxylation, respectively. Here we describe that its deamination can be catalyzed by lysyl oxidase-like 2 protein (LOXL2), presenting an unconventional chemical mechanism for H3K4 modification. Infrared spectroscopy and mass spectrometry analyses demonstrated that recombinant LOXL2 specifically deaminates trimethylated H3K4. Moreover, by regulating H3K4me3 deamination, LOXL2 activity is linked with the transcriptional control of the CDH1 gene. These results reveal the existence of further H3 modification as well as a novel mechanism for H3K4me3 demethylation. DATABASE: The GEO accession number for the data referred to this paper is GSE35600.

Identification of Bovine Sperm Surface Proteins Involved in Carbohydrate-mediated Fertilization Interactions.

Glycan-protein interactions play a key role in mammalian fertilization, but data on the composition and identities of protein complexes involved in fertilization events are scarce, with the added complication that the glycans in such interactions tend to differ among species. In this study we have used a bovine model to detect, characterize and identify sperm lectins relevant in fertilization. Given the complexity of the sperm-toward-egg journey, two important aspects of the process, both primarily mediated by protein-sugar interactions, have been addressed: (1) formation of the sperm reservoir in the oviductal epithelium, and (2) gamete recognition (oocyte-sperm interaction). Using whole sperm cells and a novel affinity capture method, several groups of proteins with different glycan specificities, including 58 hitherto unreported as lectins, have been identified in sperm surface, underscoring both the efficacy of our selective approach and the complex composition and function of sperm. Based on these results and previous data, we suggest that sperm surface proteins play significant roles in fertilization events such as membrane remodeling, transport, protection and function, thus supporting the hypothesis that rather than a simple lock-and-key model, mammalian fertilization relies on a complex interactome involving multiple ligands/receptors and recognition/binding events.

SUMOylation of the brain-predominant Ataxin-3 isoform modulates its interaction with p97.

BACKGROUND: Machado-Joseph Disease (MJD), a form of dominantly inherited ataxia belonging to the group of polyQ expansion neurodegenerative disorders, occurs when a threshold value for the number of glutamines in Ataxin-3 (Atx3) polyglutamine region is exceeded. As a result of its modular multidomain architecture, Atx3 is known to engage in multiple macromolecular interactions, which might be unbalanced when the polyQ tract is expanded, culminating in the aggregation and formation of intracellular inclusions, a unifying fingerprint of this group of neurodegenerative disorders. Since aggregation is specific to certain brain regions, localization-dependent posttranslational modifications that differentially affect Atx3 might also contribute for MJD. METHODS: We combined in vitro and cellular approaches to address SUMOylation in the brain-predominant Atx3 isoform and assessed the impact of this posttranslational modification on Atx3 self-assembly and interaction with its native partner, p97. RESULTS: We demonstrate that Atx3 is SUMOylated at K356 both in vitro and in cells, which contributes for decreased formation of amyloid fibrils and for increased affinity towards p97. CONCLUSIONS AND GENERAL SIGNIFICANCE: These findings highlight the role of SUMOylation as a regulator of Atx3 function, with implications on Atx3 protein interaction network and self-assembly, with potential impact for further understanding the molecular mechanisms underlying MJD pathogenesis.

Detection and differentiation of 22 kDa and 20 kDa Growth Hormone proteoforms in human plasma by LC-MS/MS.

Human growth hormone (GH) is suspected to be widely and illegally used in sport to improve athletes' performance. For the detection of GH abuse, blood samples are screened for abnormal ratios between the 22 and 20 kDa GH proteoforms that demonstrate the administration of the synthetic hormone. Current detection methods are based on classical immunoassays as they provide sufficient sensitivity for the detection of GH proteoforms. These antibody based methods, however, suffer from unclear selectivity and potential cross-reactivity towards similar proteins. For unambiguous GH detection, we report a Mass Spectrometry ImmunoAssay (MSIA) that first enriches GH from plasma with an antibody of relatively low specificity, and subsequently quantifies the 22 and 20 kDa proteoforms by Selected Reaction Monitoring (SRM) LC-MS/MS analysis. This method proved superior to an antibody-free strategy based on GH purification by protein precipitation. Using GH-MSIA we successfully quantified the 22/20 kDa GH ratio in post-exercise capillary plasma extracted from two individuals. This GH-MSIA is applicable to anti-doping and GH-related disease analysis.

Increased N-glycosylation efficiency by generation of an aromatic sequon on N135 of antithrombin.

The inefficient glycosylation of consensus sequence on N135 in antithrombin explains the two glycoforms of this key anticoagulant serpin found in plasma: α and ß, with four and three N-glycans, respectively. The lack of this N-glycan increases the heparin affinity of the ß-glycoform. Recent studies have demonstrated that an aromatic sequon (Phe-Y-Asn-X-Thr) in reverse ß-turns enhances N-glycosylation efficiency and stability of different proteins. We evaluated the effect of the aromatic sequon in this defective glycosylation site of antithrombin, despite of being located in a loop between the helix D and the strand 2A. We analyzed the biochemical and functional features of variants generated in a recombinant cell system (HEK-EBNA). Cells transfected with wild-type plasmid (K133-Y-N135-X-S137) generated 50% of α and ß-antithrombin. The S137T, as previously reported, K133F, and the double mutant (K133F/S137T) had improved glycosylation efficiency, leading to the secretion of α-antithrombin, as shown by electrophoretic and mass analysis. The presence of the aromatic sequon did not significantly affect the stability of this conformationally sensitive serpin, as revealed by thermal denaturation assay. Moreover, the aromatic sequon hindered the activation induced by heparin, in which is involved the helix D. Accordingly, K133F and particularly K133F/S137T mutants had a reduced anticoagulant activity. Our data support that aromatic sequons in a different structural context from reverse turns might also improve the efficiency of N-glycosylation.

The leukocyte activation receptor CD69 controls T cell differentiation through its interaction with galectin-1.

CD69 is involved in immune cell homeostasis, regulating the T cell-mediated immune response through the control of Th17 cell differentiation. However, natural ligands for CD69 have not yet been described. Using recombinant fusion proteins containing the extracellular domain of CD69, we have detected the presence of a ligand(s) for CD69 on human dendritic cells (DCs). Pulldown followed by mass spectrometry analyses of CD69-binding moieties on DCs identified galectin-1 as a CD69 counterreceptor. Surface plasmon resonance and anti-CD69 blocking analyses demonstrated a direct and specific interaction between CD69 and galectin-1 that was carbohydrate dependent. Functional assays with both human and mouse T cells demonstrated the role of CD69 in the negative effect of galectin-1 on Th17 differentiation. Our findings identify CD69 and galectin-1 to be a novel regulatory receptor-ligand pair that modulates Th17 effector cell differentiation and function.

Structural requirements of glycosaminoglycans for their interaction with HIV-1 envelope glycoprotein gp120.

Heparan sulfate proteoglycans are known to assist HIV-1 entry into host cells, mediated by the viral envelope glycoprotein gp120. We aimed to determine the general structural features of glycosaminoglycans that enable their binding to gp120, by surface plasmon resonance. Binding was found to be dependent on sequence type, size and sulfation patterns. HIV-1 gp120 prefers heparin and heparan sulfate (with at least 16 monomers in length) over chondroitin and dermatan. Sulfate groups were essential to promote this interaction. These results advance the understanding of the molecular-level requirements for virus attachment and cell entry.

Analysis of urinary human growth hormone (hGH) using hydrogel nanoparticles and isoform differential immunoassays after short recombinant hGH treatment: preliminary results.

Successful application clinical-grade human growth hormone (hGH) immunoassays to the discovery of illegal doping cases has been rare. Indeed, the preferred biological matrix in doping control is urine, where the estimated baseline concentration of hGH falls well below the linear range and sensitivity threshold of all commercially available immunoassays, including hGH isoform differential immunoassays which can discriminate pituitary endogenous hGH from recombinant hGH. We employed hydrogel nanoparticles as a pre-processing step that concentrate urinary hGH into the linear range of isoform differential immunoassays. We explored the characteristics of immunoassays in urine spiked with both phGH or rhGH, after pre-treatment with the nanoparticles. Subsequently, pre-treatment was applied to urine obtained from 3 healthy volunteers administered during three days with daily subcutaneous injections of 0.026 mg/kg/day rhGH, Genotonorm(®). Linearity between both rhGH and phGH concentrations in urine measured by a chemoluminescent assay (Immulite) and in the particle eluate was evident for differential immunoassays (R square higher than 0.999). In case of treated individuals the recombinant/pituitary concentration ratios remained above the established World Anti-Doping Agency (WADA) criterion for hGH misuse up to 24h after the last administration dose, using both assays for volunteer 1 and 2 while in case of volunteer 3 results were inconclusive. The use of nanoparticles appears to open the possibility of assessing rhGH misuse in urine.

Identification of antithrombin-modulating genes. Role of LARGE, a gene encoding a bifunctional glycosyltransferase, in the secretion of proteins?

The haemostatic relevance of antithrombin together with the low genetic variability of SERPINC1, and the high heritability of plasma levels encourage the search for modulating genes. We used a hypothesis-free approach to identify these genes, evaluating associations between plasma antithrombin and 307,984 polymorphisms in the GAIT study (352 individuals from 21 Spanish families). Despite no SNP reaching the genome wide significance threshold, we verified milder positive associations in 307 blood donors from a different cohort. This validation study suggested LARGE, a gene encoding a protein with xylosyltransferase and glucuronyltransferase activities that forms heparin-like linear polysaccharides, as a potential modulator of antithrombin based on the significant association of one SNPs, rs762057, with anti-FXa activity, particularly after adjustment for age, sex and SERPINC1 rs2227589 genotype, all factors influencing antithrombin levels (p = 0.02). Additional results sustained this association. LARGE silencing inHepG2 and HEK-EBNA cells did not affect SERPINC1 mRNA levels but significantly reduced the secretion of antithrombin with moderate intracellular retention. Milder effects were observed on α1-antitrypsin, prothrombin and transferrin. Our study suggests LARGE as the first known modifier of plasma antithrombin, and proposes a new role for LARGE in modulating extracellular secretion of certain glycoproteins.

Large-scale metabolome analysis and quantitative integration with genomics and proteomics data in Mycoplasma pneumoniae.

Systems metabolomics, the identification and quantification of cellular metabolites and their integration with genomics and proteomics data, promises valuable functional insights into cellular biology. However, technical constraints, sample complexity issues and the lack of suitable complementary quantitative data sets prevented accomplishing such studies in the past. Here, we present an integrative metabolomics study of the genome-reduced bacterium Mycoplasma pneumoniae. We experimentally analysed its metabolome using a cross-platform approach. We explain intracellular metabolite homeostasis by quantitatively integrating our results with the cellular inventory of proteins, DNA and other macromolecules, as well as with available building blocks from the growth medium. We calculated in vivo catalytic parameters of glycolytic enzymes, making use of measured reaction velocities, as well as enzyme and metabolite pool sizes. A quantitative, inter-species comparison of absolute and relative metabolite abundances indicated that metabolic pathways are regulated as functional units, thereby simplifying adaptive responses. Our analysis demonstrates the potential for new scientific insight by integrating different types of large-scale experimental data from a single biological source.

Dissecting the energy metabolism in Mycoplasma pneumoniae through genome-scale metabolic modeling.

Mycoplasma pneumoniae, a threatening pathogen with a minimal genome, is a model organism for bacterial systems biology for which substantial experimental information is available. With the goal of understanding the complex interactions underlying its metabolism, we analyzed and characterized the metabolic network of M. pneumoniae in great detail, integrating data from different omics analyses under a range of conditions into a constraint-based model backbone. Iterating model predictions, hypothesis generation, experimental testing, and model refinement, we accurately curated the network and quantitatively explored the energy metabolism. In contrast to other bacteria, M. pneumoniae uses most of its energy for maintenance tasks instead of growth. We show that in highly linear networks the prediction of flux distributions for different growth times allows analysis of time-dependent changes, albeit using a static model. By performing an in silico knock-out study as well as analyzing flux distributions in single and double mutant phenotypes, we demonstrated that the model accurately represents the metabolism of M. pneumoniae. The experimentally validated model provides a solid basis for understanding its metabolic regulatory mechanisms.

Unique thrombin inhibition mechanism by anophelin, an anticoagulant from the malaria vector.

Anopheles mosquitoes are vectors of malaria, a potentially fatal blood disease affecting half a billion humans worldwide. These blood-feeding insects include in their antihemostatic arsenal a potent thrombin inhibitor, the flexible and cysteine-less anophelin. Here, we present a thorough structure-and-function analysis of thrombin inhibition by anophelin, including the 2.3-Å crystal structure of the human thrombin·anophelin complex. Anophelin residues 32-61 are well-defined by electron density, completely occupying the long cleft between the active site and exosite I. However, in striking contrast to substrates, the D50-R53 anophelin tetrapeptide occupies the active site cleft of the enzyme, whereas the upstream residues A35-P45 shield the regulatory exosite I, defining a unique reverse-binding mode of an inhibitor to the target proteinase. The extensive interactions established, the disruption of thrombin's active site charge-relay system, and the insertion of residue R53 into the proteinase S(1) pocket in an orientation opposed to productive substrates explain anophelin's remarkable specificity and resistance to proteolysis by thrombin. Complementary biophysical and functional characterization of point mutants and truncated versions of anophelin unambiguously establish the molecular mechanism of action of this family of serine proteinase inhibitors (I77). These findings have implications for the design of novel antithrombotics.