Ribosomal RNA sequence phylogeny is not congruent with ascospore morphology among species in Ceratocystis sensu stricto

The genus Ceratocystis sensu stricto includes important fungal pathogens of woody and herbaceous plants. This genus is distinguished from species in Ceratocystis sensu lato by the presence of Chalara anamorphs. Ascospore shape has been used extensively in delineating Ceratocystis taxa, which show a large variety of ascospore shapes. Sequence analysis of one region of he 18S ribosomal RNA subunit and two regions of the 28S ribosomal RNA subunit showed that there was a majority of multiple substitutions at nucleotide sites and that there was a low transition/transversion ratio, T = 0.72. Both of these results suggest that these are well established, old species. Ascospore morphology, for the most part, was not congruent with the molecular phylogeny, and the use of morphological characters may be misleading in the taxonomy of these species.      

Giemsa-based cytological assessment of area,shape and nucleus:cytoplasm ratio of goblet cells of rabbit bulbar conjunctiva

Goblet cells were visualized in impression cytology specimens from bulbar conjunctiva of the rabbit eye using Giemsa staining. Highly magnified images were used to generate outlines of the goblet cells and their characteristic eccentric nuclei. Using sets of 10 cells from 15 cytology specimens, I found that the longest dimension of the goblet cells averaged 16.7 ± 2.3 μm, the shortest dimension averaged 14.4 ± 1.8 μm and the nucleus averaged 6.3 ± 0.8 μm. The goblet cells were ellipsoid in shape and the longest:shortest cell dimension ratio averaged 1.169 ± 0.091. The goblet cell areas ranged from 108 to 338 μm² (average 193 ± 50 μm²). The area could be predicted reliably from the longest and shortest dimensions (r² = 0.903). The areas of goblet cell nuclei were 15–58 μm² (average 33 ± μm²) and the nucleus:cytoplasm area fraction was predictably greater in smaller goblet cells and less in the larger goblet cells (Spearman correlation = 0.817). The nuclei were estimated to occupy an average of 9.5% of the cell volume. The differences in size, shape and nucleus:cytoplasm ratio may reflect differences in goblet cell maturation.     

Cell stiffness and receptors: evidence for cytoskeletal subnetworks

Viscoelastic models of cells often treat cells as homogeneous objects. However, studies have demonstrated that cellular properties are local and can change dramatically on the basis of the location probed. Because membrane receptors are linked in various ways to the intracellular space, with some receptors linking to the cytoskeleton and others diffusing freely without apparent linkages, the cellular physical response to mechanical stresses is expected to depend on the receptor engaged. In this study, we tested the hypothesis that cellular mechanical stiffness as measured via cytoskeletally linked receptors is greater than stiffness measured via receptors that are not cytoskeletally linked. We used a magnetic micromanipulator to apply linear stresses to magnetic beads attached to living cells via selected receptors. One of the receptor classes probed, the dystroglycan receptors, is linked to the cytoskeleton, while the other, the transferrin receptors, is not. Fibronectin-coated beads were used to test cellular mechanical properties of the cytoskeleton without membrane dependence by allowing the beads to endocytose. For epithelial cells, transferrin-dependent stiffness and endocytosed bead-dependent stiffness were similar, while dystroglycan-dependent stiffness was significantly lower. For smooth muscle cells, dystroglycan-dependent stiffness was similar to the endocytosed bead-dependent stiffness, while the transferrin-dependent stiffness was lower. The conclusion of this study is that the measured cellular stiffness is critically influenced by specific receptor linkage and by cell type and raises the intriguing possibility of the existence of separate cytoskeletal networks with distinct mechanical properties that link different classes of receptors.     

Caspase-14 but not caspase-3 is processed during the development of fetal mouse epidermis

The activation of caspases is a central step in apoptosis and may also be critical for terminal differentiation of epidermal keratinocytes (KC). In particular, caspase-3 has been implicated in the differentiation of embryonic KC as well as in programmed cell death of KC, and caspase-14 has been suggested to function in the formation or homeostasis of the stratum corneum (SC). To test the putative roles of these proteases, we determined their expression level and activation status during development of fetal mouse epidermis. The level of procaspase-3 did not change significantly during epidermal development, and enzyme activation was undetectable at any timepoint investigated. Despite the lack of active caspase-3, the newly formed stratum granulosum and the regressing periderm contained cells positive in the terminal deoxynucleotidyl transferase-mediated fluorescein-dUTP nick end labeling assay, indicating that nuclear DNA was degraded without activation of caspase-3, thereby arguing against a proteolytic function of caspase-3 in embryonic KC differentiation. By contrast, caspase-14 increased in abundance from embryonic day 14.5 (E14.5) onwards and consistently localized to the suprabasal layers of fetal epidermis. The caspase-14 pro-enzyme was processed into its catalytic subunits, a step required for enzyme activity, on day E17.5, coinciding with SC formation. Thus, processing of procaspase-14 is not confined to air-exposed mature skin but also occurs during epidermal development in utero. In summary, this study demonstrates that caspase-14, but not caspase-3 activation coincides temporally and spatially with embryonic KC differentiation, suggesting a role for caspase-14 in terminally differentiated KC.     

Hedgehog regulated Slit expression determines commissure and glial cell position in the zebrafish forebrain

Three major axon pathways cross the midline of the vertebrate forebrain early in embryonic development: the postoptic commissure (POC), the anterior commissure (AC) and the optic nerve. We show that a small population of Gfap+ astroglia spans the midline of the zebrafish forebrain in the position of, and prior to, commissural and retinal axon crossing. These glial ;bridges' form in regions devoid of the guidance molecules slit2 and slit3, although a subset of these glial cells express slit1a. We show that Hh signaling is required for commissure formation, glial bridge formation, and the restricted expression of the guidance molecules slit1a, slit2, slit3 and sema3d, but that Hh does not appear to play a direct role in commissural and retinal axon guidance. Reducing Slit2 and/or Slit3 function expanded the glial bridges and caused defasciculation of the POC, consistent with a ;channeling' role for these repellent molecules. By contrast, reducing Slit1a function led to reduced midline axon crossing, suggesting a distinct role for Slit1a in midline axon guidance. Blocking Slit2 and Slit3, but not Slit1a, function in the Hh pathway mutant yot (gli2DR) dramatically rescued POC axon crossing and glial bridge formation at the midline, indicating that expanded Slit2 and Slit3 repellent function is largely responsible for the lack of midline crossing in these mutants. This analysis shows that Hh signaling helps to pattern the expression of Slit guidance molecules that then help to regulate glial cell position and axon guidance across the midline of the forebrain.     

LARGE can functionally bypass alpha-dystroglycan glycosylation defects in distinct congenital muscular dystrophies

Several congenital muscular dystrophies caused by defects in known or putative glycosyltransferases are commonly associated with hypoglycosylation of alpha-dystroglycan (alpha-DG) and a marked reduction of its receptor function. We have investigated changes in the processing and function of alpha-DG resulting from genetic manipulation of LARGE, the putative glycosyltransferase mutated both in Large(myd) mice and in humans with congenital muscular dystrophy 1D (MDC1D). Here we show that overexpression of LARGE ameliorates the dystrophic phenotype of Large(myd) mice and induces the synthesis of glycan-enriched alpha-DG with high affinity for extracellular ligands. Notably, LARGE circumvents the alpha-DG glycosylation defect in cells from individuals with genetically distinct types of congenital muscular dystrophy. Gene transfer of LARGE into the cells of individuals with congenital muscular dystrophies restores alpha-DG receptor function, whereby glycan-enriched alpha-DG coordinates the organization of laminin on the cell surface. Our findings indicate that modulation of LARGE expression or activity is a viable therapeutic strategy for glycosyltransferase-deficient congenital muscular dystrophies.     

Immunohistochemical expression of HGF, c-MET and transcription factor STAT3 in colorectal tumors

By immunohistochemistry, we have investigated the expression of hepatocyte growth factor (HGF), HGF-R or c-met and the transcritor factor STAT3 in a series of 80 colorectal tumours (40 adenomas and 40 adenocarcinomas). The expression of HGF, c-met and STAT3 was revealed in 40/40 (100%) of adenomas and in 26/40 (65%) of adenocarcinomas; the remaining 14/40 (35%) carcinomas expressed c-met but failed to express HGF and STAT3. Positive immunoreaction score was defined through the number of stained cells: low (1-10%), moderate (11-50%) and high (>51%). In adenomas, the HGF immunoreaction was high in 33 (82.5%) and moderate in 7 (17.5%); the c-met staining was high in 3 (7.5%) and moderate in 37 (92.5%); and the STAT3 reactivity was high in 25 (62.5%) and moderate in 15(37.5%). In carcinomas, the HGF immunoreaction was moderate in 21 (80.7%) and low in 5 (19.2%); the c-met staining was high in 14 (35%), moderate in 25 (62.5) and low in 1 (2.5%); and the STAT3 reactivity was moderate in 17 (65.3%) and low in 9 (34.6%). In both type of lesions, HGF and c-met showed a membranous and cytoplasmic location. In adenomas, STAT3 was detected in cytoplasm and nucleus and in carcinomas it was limited to cytoplasm. While the HGF/c-met/STAT3 expression in adenomas was significantly different from carcinomas (c2 = 17, p < 0.0001), no correlation was found among HGF, c-met, or STAT3 immunostaining with histotype or degree of dysplasia in adenomas and the same for histotype, grading or staging in carcinomas. These features, suggesting a role of the HGF/c-met/STAT3 signal in colon tumorigenesis, indicate that a reduced expression of HGF and c-met is associated to progression of adenoma into carcinoma.