Detection of dysmorphic cells in peripheral blood (PB) smears is essential in diagnostic screening of hematological diseases. Myelodysplastic syndromes (MDS) are hematopoietic neoplasms characterized by dysplastic and ineffective hematopoiesis, which diagnosis is mainly based on morphological findings of PB and bone marrow. We developed an automated diagnostic support system of MDS by combining an automated blood cell image-recognition system using a deep learning system (DLS) powered by convolutional neural networks (CNNs) with a decision-making system using extreme gradient boosting (XGBoost). The DLS of blood cell image-recognition has been trained using datasets consisting of 695,030 blood cell images taken from 3,261 PB smears including hematopoietic malignancies. The DLS simultaneously classified 17 blood cell types and 97 morphological features of such cells with >93.5% sensitivity and >96.0% specificity. The automated MDS diagnostic system successfully differentiated MDS from aplastic anemia (AA) with high accuracy; 96.2% of sensitivity and 100% of specificity (AUC 0.990). This is the first CNN-based automated initial diagnostic system for MDS using PB smears, which is applicable to develop new automated diagnostic systems for various hematological disorders.

Early and accurate diagnosis is critical in reducing the morbidity and mortality associated with malaria. Microscopy (MI) is the current diagnostic gold standard in the field; however, it requires expert personnel, is time-consuming, and has limited sensitivity. Although rapid diagnostic tests for antigen detection (RDTs) are an alternative to diagnosis, they also have limited sensitivity and produce false positive results in detecting recent past infection. The automated hematology analyzer XN-31 prototype (XN-31p) (Sysmex Corporation, Kobe, Japan) is able to identify plasmodium-infected erythrocytes, count parasitemia and perform complete blood-cell counts within one minute. The performance of the XN-31p in diagnosing malaria was evaluated and compared with real-time polymerase chain reaction (qPCR), MI and RDT in an endemic area of Colombia where Plasmodium falciparum and Plasmodium vivax are present. Acute febrile patients were enrolled from July 2018 to April 2019 in Quibdó, Colombia. Malaria diagnoses were obtained from MI and RDT in the field and later confirmed by qPCR. Venous blood samples in EDTA were processed with an XN-31p in the field. Sensitivity, specificity, positive/negative predictive values, and the likelihood ratios of positive and negative tests were calculated with respect to the results from qPCR, MI and RDT. The intraclass correlation coefficient (ICC) and Bland–Altman plot were used to evaluate the concordance in the parasitemia with respect to MI. A total of 1,754 subjects were enrolled. The mean age was 27.0 years (IQR 14–44); 89.6% were Afro-Colombians, 94.3% lived in urban areas and 0.91% were pregnant. With respect to qPCR, the XN-31p showed a sensitivity of 90% (95% CI 87.24–92.34) and a specificity of 99.83% (95% CI 99.38–99.98) in detecting Plasmodium spp.; both parameters were equivalent to those for MI and RDT. Using MI as the reference, the XN-31p showed a sensitivity of 98.09% (95% CI 96.51–99.08), a specificity of 99.83% (95% CI 99.4–99.98), an ICC of 0.85 (95% CI 0.83–0.87) and an average difference of − 3096 parasites/µL when compared with thick-smear MI and an ICC of 0.98 (95% CI 0.97–0.98) and an average difference of − 0.0013% when compared with thin-smear MI. The XN-31p offers a rapid and accurate alternative method for diagnosing malaria in clinical laboratories in areas where P. falciparum and P. vivax cocirculate.

Nylon hydrolase (NylC) is initially expressed as an inactive precursor (36 kDa). The precursor is cleaved autocatalytically at Asn266/Thr267 to generate an active enzyme composed of an α subunit (27 kDa) and a β subunit (9 kDa). Four αβ heterodimers (molecules A-D) form a doughnut-shaped quaternary structure. In this study, the thermostability of the parental NylC was altered by amino acid substitutions located at the A/D interface (D122G/H130Y/D36A/L137A) or the A/B interface (E263Q) and spanned a range of 47 °C. Considering structural, biophysical, and biochemical analyses, we discuss the structural basis of the stability of nylon hydrolase. From the analytical centrifugation data obtained regarding the various mutant enzymes, we conclude that the assembly of the monomeric units is dynamically altered by the mutations. Finally, we propose a model that can predict whether the fate of the nascent polypeptide will be correct subunit assembly, inappropriate protein-protein interactions causing aggregation, or intracellular degradation of the polypeptide.

Arthrobacter sp. strain KI72 grows on a 6-aminohexanoate oligomer, which is a by-product of nylon-6 manufacturing, as a sole source of carbon and nitrogen. We cloned the two genes, nylD and nylE , responsible for 6-aminohexanoate metabolism on the basis of the draft genomic DNA sequence of strain KI72. We amplified the DNA fragments that encode these genes by polymerase chain reaction using a synthetic primer DNA homologous to the 4-aminobutyrate metabolic enzymes. We inserted the amplified DNA fragments into the expression vector pColdI in Escherichia coli, purified the His-tagged enzymes to homogeneity, and performed biochemical studies. We confirmed that 6-aminohexanoate aminotransferase (NylD) catalyzes the reaction of 6-aminohexanoate to adipate semialdehyde using α-ketoglutarate, pyruvate, and glyoxylate as amino acceptors, generating glutamate, alanine, and glycine, respectively. The reaction requires pyridoxal phosphate (PLP) as a cofactor. For further metabolism, adipate semialdehyde dehydrogenase (NylE) catalyzes the oxidative reaction of adipate semialdehyde to adipate using NADP as a cofactor. Phylogenic analysis revealed that NylD should be placed in a branch of the PLP-dependent aminotransferase sub III, while NylE should be in a branch of the aldehyde dehydrogenase superfamily. In addition, we established a NylD/NylE coupled system to quantify the aminotransferase activity and to enable the conversion of 6-aminohexaoate to adipate via adipate semialdehyde with a yield of > 90%. In the present study, we demonstrate that 6-aminohexanoate produced from polymeric nylon-6 and nylon oligomers (i.e., a mixture of 6-aminohexaoate oligomers) by nylon hydrolase (NylC) and 6-aminohexanoate dimer hydrolase (NylB) reactions are sequentially converted to adipate by metabolic engineering technology.

We report here the 4.6-Mb genome sequence of a nylon oligomer-degrading bacterium, Arthrobacter sp. strain KI72. The draft genome sequence of strain KI72 consists of 4,568,574 bp, with a G+C content of 63.47%, 4,372 coding sequences (CDSs), 54 tRNAs, and six rRNAs.