An Automated Pipeline for Differential Cell Counts on Whole-Slide Bone Marrow Aspirate Smears.

The pathologic diagnosis of bone marrow disorders relies in part on the microscopic analysis of bone marrow aspirate (BMA) smears and the manual counting of marrow nucleated cells to obtain a differential cell count (DCC). This manual process has significant limitations, including the analysis of only a small subset of optimal slide areas and nucleated cells, as well as interobserver variability due to differences in cell selection and classification. To address these shortcomings, we developed an automated machine learning-based pipeline for obtaining 11-component DCCs on whole-slide BMAs. This pipeline uses a sequential process of identifying optimal BMA regions with high proportions of marrow nucleated cells, detecting individual cells within these optimal areas, and classifying these cells into 1 of 11 DCC components. Convolutional neural network models were trained on 396,048 BMA region, 28,914 cell boundary, and 1,510,976 cell class images from manual annotations. The resulting automated pipeline produced 11-component DCCs that demonstrated a high statistical and diagnostic concordance with manual DCCs among a heterogeneous group of testing BMA slides with varying pathologies and cellularities. Additionally, we demonstrated that an automated analysis can reduce the intraslide variance in DCCs by analyzing the whole slide and marrow nucleated cells within all optimal regions. Finally, the pipeline outputs of region classification, cell detection, and cell classification can be visualized using whole-slide image analysis software. This study demonstrates the feasibility of a fully automated pipeline for generating DCCs on scanned whole-slide BMA images, with the potential for improving the current standard of practice for utilizing BMA smears in the laboratory analysis of hematologic disorders.

Flagellar Structures from the Bacterium Caulobacter crescentus and Implications for Phage ϕ CbK Predation of Multiflagellin Bacteria.

Caulobacter crescentus is a Gram-negative alphaproteobacterium that commonly lives in oligotrophic fresh- and saltwater environments. C. crescentus is a host to many bacteriophages, including ϕCbK and ϕCbK-like bacteriophages, which require interaction with the bacterial flagellum and pilus complexes during adsorption. It is commonly thought that the six paralogs of the flagellin gene present in C. crescentus are important for bacteriophage evasion. Here, we show that deletion of specific flagellins in C. crescentus can indeed attenuate ϕCbK adsorption efficiency, although no single deletion completely ablates ϕCbK adsorption. Thus, the bacteriophage ϕCbK likely recognizes a common motif among the six known flagellins in C. crescentus with various degrees of efficiency. Interestingly, we observe that most deletion strains still generate flagellar filaments, with the exception of a strain that contains only the most divergent flagellin, FljJ, or a strain that contains only FljN and FljO. To visualize the surface residues that are likely recognized by ϕCbK, we determined two high-resolution structures of the FljK filament, with and without an amino acid substitution that induces straightening of the filament. We observe posttranslational modifications on conserved surface threonine residues of FljK that are likely O-linked glycans. The possibility of interplay between these modifications and ϕCbK adsorption is discussed. We also determined the structure of a filament composed of a heterogeneous mixture of FljK and FljL, the final resolution of which was limited to approximately 4.6 Å. Altogether, this work builds a platform for future investigations of how phage ϕCbK infects C. crescentus at the molecular level.IMPORTANCE Bacterial flagellar filaments serve as an initial attachment point for many bacteriophages to bacteria. Some bacteria harbor numerous flagellin genes and are therefore able to generate flagellar filaments with complex compositions, which is thought to be important for evasion from bacteriophages. This study characterizes the importance of the six flagellin genes in C. crescentus for infection by bacteriophage ϕCbK. We find that filaments containing the FljK flagellin are the preferred substrate for bacteriophage ϕCbK. We also present a high-resolution structure of a flagellar filament containing only the FljK flagellin, which provides a platform for future studies on determining how bacteriophage ϕCbK attaches to flagellar filaments at the molecular level.

Caulobacter chromosome segregation is an ordered multistep process.

Despite its fundamental nature, bacterial chromosome segregation remains poorly understood. Viewing segregation as a single process caused multiple proposed mechanisms to appear in conflict and failed to explain how asymmetrically dividing bacteria break symmetry to move only one of their chromosomes. Here, we demonstrate that the ParA ATPase extends from one cell pole and pulls the chromosome by retracting upon association with the ParB DNA-binding protein. Surprisingly, ParA disruption has a specific effect on chromosome segregation that only perturbs the latter stages of this process. Using quantitative high-resolution imaging, we demonstrate that this specificity results from the multistep nature of chromosome translocation. We propose that Caulobacter chromosome segregation follows an ordered pathway of events with distinct functions and mechanisms. Initiation releases polar tethering of the origin of replication, distinction spatially differentiates the two chromosomes, and commitment irreversibly translocates the distal centromeric locus. Thus, much as eukaryotic mitosis involves a sequence of distinct subprocesses, Caulobacter cells also segregate their chromosomes through an orchestrated series of steps. We discuss how the multistep view of bacterial chromosome segregation can help to explain and reconcile outstanding puzzles and frame future investigation.

Growth conditions regulate the requirements for Caulobacter chromosome segregation.

Growth environments are important metabolic and developmental regulators. Here we demonstrate a growth environment-dependent effect on Caulobacter chromosome segregation of a small-molecule inhibitor of the MreB bacterial actin cytoskeleton. Our results also implicate ParAB as important segregation determinants, suggesting that multiple distinct mechanisms can mediate Caulobacter chromosome segregation and that their relative contributions can be environmentally regulated.

Germ Cell-less Promotes Centrosome Segregation to Induce Germ Cell Formation.

The primordial germ cells (PGCs) specified during embryogenesis serve as progenitors to the adult germline stem cells. In Drosophila, the proper specification and formation of PGCs require both centrosomes and germ plasm, which contains the germline determinants. Centrosomes are microtubule (MT)-organizing centers that ensure the faithful segregation of germ plasm into PGCs. To date, mechanisms that modulate centrosome behavior to engineer PGC development have remained elusive. Only one germ plasm component, Germ cell-less (Gcl), is known to play a role in PGC formation. Here, we show that Gcl engineers PGC formation by regulating centrosome dynamics. Loss of gcl leads to aberrant centrosome separation and elaboration of the astral MT network, resulting in inefficient germ plasm segregation and aborted PGC cellularization. Importantly, compromising centrosome separation alone is sufficient to mimic the gcl loss-of-function phenotypes. We conclude Gcl functions as a key regulator of centrosome separation required for proper PGC development.

The response regulator PhoP4 is required for late developmental events in Myxococcus xanthus.

Phosphate regulation is complex in the developmental prokaryote Myxococcus xanthus, and requires at least four two-component systems (TCSs). Here, the identification and characterization of a member of one TCS, designated PhoP4, is reported. phoP4 insertion and in-frame deletion strains caused spore viability to be decreased by nearly two orders of magnitude, and reduced all three development-specific phosphatase activities by 80-90 % under phosphate-limiting conditions. Microarray and quantitative PCR analyses demonstrated that PhoP4 is also required for appropriate expression of the predicted pstSCAB-phoU operon of inorganic phosphate assimilation genes. Unlike the case for the other three M. xanthus Pho TCSs, the chromosomal region around phoP4 does not contain a partner histidine kinase gene. Yeast two-hybrid analyses reveal that PhoP4 interacts reciprocally with PhoR2, the histidine kinase of the Pho2 TCS; however, the existence of certain phenotypic differences between phoP4 and phoR2 mutants suggests that PhoP4 interacts with another, as-yet unidentified, histidine kinase.

Mutations affecting predation ability of the soil bacterium Myxococcus xanthus.

Myxococcus xanthus genetic mutants with characterized phenotypes were analysed for the ability to prey on susceptible bacteria. Quantification of predatory ability was scored by a newly developed method under conditions in which prey bacteria provided the only source of nutrients. These results were corroborated by data derived using a previously published protocol that measures predation in the presence of limited external nutrients. First, early developmental regulatory mutants were examined, because their likely functions in assessing the local nutrient status were predicted to be also important for predation. The results showed that predation efficiency is reduced by 64-80 % for mutants of three A-signalling components, AsgA, AsgC and AsgE, but not for AsgB. This suggests that an Asg regulon function that is separate from A-signal production is needed for predation. Besides the Asg components, mutations in the early developmental genes sdeK and csgA were also consistently observed to reduce predatory efficacy by 36 and 33 %, respectively. In contrast, later developmental components, such as DevRS, 4406 and PhoP4, did not appear to play significant roles in predation. The predatory abilities of mutants defective for motility were also tested. The data showed that adventurous, but not social, motility is required for predation in the assay. Also, mutants for components in the chemotaxis-like Frz system were found to be reduced in predation efficiency by between 62 and 85 %. In sum, it was demonstrated here that defects in development and development-related processes affect the ability of M. xanthus to prey on other bacteria.

BrgE is a regulator of Myxococcus xanthus development.

We report here the identification and characterization of a member of the Myxococcus xanthus SdeK signal transduction pathway, BrgE. This protein was identified as an SdeK-interacting component using a yeast two-hybrid screen, and we further confirmed this interaction by the glutathione S-transferase (GST) pulldown assay. Additional yeast two-hybrid analyses revealed that BrgE preferentially interacts with the putative amino-terminal sensor domain of SdeK, but not with the carboxy-terminal kinase domain. A brgE insertion strain was shown to be blocked in development between aggregation and mound formation, and decreased by 50-fold in viable spore production compared with the parental wild type. These phenotypes are similar to those of sdeK mutants. The brgE mutation also altered expression of a sample of Tn5 lac developmental markers that are also SdeK regulated. Finally, we demonstrated that a brgE sdeK double mutant has a more severe sporulation defect than either of the two single mutants, suggesting that BrgE and SdeK act synergistically to regulate wild-type levels of sporulation. In sum, these data suggest that BrgE operates as an auxiliary factor to stimulate the SdeK signal transduction pathway by directly binding to the amino-terminal sensor domain of SdeK.

RasA is required for Myxococcus xanthus development and social motility.

An insertion in the rasA gene entirely blocked developmental aggregation and sporulation in Myxococcus xanthus while also reducing swarm expansion on a 0.3% agar surface. Data presented here demonstrate that rasA is required for extracellular fibril formation and social gliding motility.