Evidence for a partial redundancy of the fibronectin-binding proteins for the transfer of mycoloyl residues onto the cell wall arabinogalactan termini of Mycobacterium tuberculosis

Mycobacterium tuberculosis produces a series of major secreted proteins, the fibronectin-binding proteins (Fbps), also known as the antigen 85 complex, that are believed to play an essential role in the pathogenesis of tuberculosis through their mycoloyltransferase activity required for maintaining the integrity of the bacterial cell envelope. Four different fbp genes are found in the genome of M. tuberculosis, but the reason for the existence of these Fbps sharing the same substrate specificity in vitro in mycobacteria is unknown. We have shown previously that, in the heterologous host, Corynebacterium glutamicum, FbpA, FbpB and FbpC can all add mycoloyl residues to the cell wall arabinogalactan and that, in M. tuberculosis, the cell wall mycoloylation decreases by 40% when fbpC is knocked out. To investigate whether the remaining 60% mycoloylation came from the activity of FbpA and/or FbpB, fbpA- and fbpB-inactivated mutant strains were biochemically characterized and compared with the previously studied fbpC-disrupted mutant. Unexpectedly, both mutants produced normally mycoloylated cell walls. Overproduction of FbpA, FbpB or FbpC, but not FbpD, in the fbpC-inactivated mutant strain of M. tuberculosis restored both the cell wall-linked mycolate defect and the outer cell envelope permeability barrier property. These results are consistent with all three enzymes being involved in cell wall mycoloylation and FbpC playing a more critical role than the others or, alternatively, FbpC is able to compensate for FbpA and FbpB in ways that these enzymes cannot compensate for FbpC, pointing to a partial redundancy of Fbps. In sharp contrast, FbpD does not appear to be an active mycoloyltransferase enzyme, as it cannot complement the fbpC-inactivated mutant. Most importantly, application of Smith degradation to the cell walls of transformants demonstrated that the multiple Fbp enzymes are redundant rather than specific for the various arabinogalactan mycoloylation regions. Neither FbpA nor FbpB attaches mycoloyl residues to specific sites but, like FbpC, each enzyme transfers mycoloyl residues onto the four sites present in the arabinogalactan non-reducing end hexaarabinosides.     

Inactivation of the antigen 85C gene profoundly affects the mycolate content and alters the permeability of the Mycobacterium tuberculosis cell envelope

The antigen 85 complex of Mycobacterium tuberculosis consists of three abundantly secreted proteins. The recent characterization of a mycoloyltransferase activity associated in vitro with each of these antigens suggested that they are potentially important for the building of the unusual cell envelope of mycobacteria. To define the physiological role of these proteins, the gene coding for antigen 85C was inactivated by transposon mutagenesis. The resulting mutant was shown to transfer 40% fewer mycolates to the cell wall with no change in the types of mycolates esterifying arabinogalactan or in the composition of non-covalently linked mycolates. As a consequence, the diffusion of the hydrophobic chenodeoxycholate and the hydrophilic glycerol, but not that of isoniazid, was found to be much faster through the cell envelope of the mutant than that of the parent strain. Taken together, these data demonstrate that: (i) antigen 85C is involved directly or indirectly in the transfer of mycolates onto the cell wall of the whole bacterium; (ii) the enzyme is not specific for a given type of mycolate; and (iii) the cell wall-linked mycolate layer may represent a barrier for the diffusion of small hydrophobic and hydrophilic molecules.     

Expression of biologically active mouse ciliary neutrophic factor (CNTF) and soluble CNTFRalpha in Escherichia coli and characterization of their functional specificities

Ciliary neurotrophic factor (CNTF) is a neuroprotective cytokine initially identified in chick embryo. It has been evaluated for the treatment of neurodegenerative diseases. CNTF also acts on non-neuronal cells such as oligodendrocytes, astrocytes, adipocytes and skeletal muscles cells. CNTF has regulatory effects on body weight and is currently in clinical trial for the treatment of diabetes and obesity. CNTF mediates its function by activating a tripartite receptor comprising the CNTF receptor alpha chain (CNTFRalpha), the leukemia inhibitory factor receptor beta chain (LIFRbeta) and gp130. Human, rat and chicken CNTF have been expressed as recombinant proteins, and most preclinical studies in murine models have been performed using rat recombinant protein. Rat and human CNTF differ in their fine specificities: in addition to CNTFR, rat CNTF has been shown to activate the LIFR (a heterodimer of LIFRbeta and gp130), whereas human CNTF can bind and activate a tripartite receptor comprising the IL-6 receptor alpha chain (IL-6Ralpha) and LIFR. To generate tools designed for mouse models of human diseases; we cloned and expressed in E. coli both mouse CNTF and the CNTFRalpha chain. Recombinant mouse CNTF was active and showed a high level of specificity for mouse CNTFR. It shares the arginine residue with rat CNTF which prevents binding to IL-6Ralpha. It did not activate the LIFR at all concentrations tested. Recombinant mouse CNTF is therefore specific for CNTFR and as such represents a useful tool with which to study CNTF in mouse models. It appears well suited for the comparative evaluation of CNTF and the two additional recently discovered CNTFR ligands, cardiotrophin-like cytokine\cytokine-like factor-1 and neuropoietin.     

Expression of the hepatitis delta virus large and small antigens in transgenic mice.

Simultaneous infection with hepatitis delta virus (HDV) and hepatitis B virus (HBV) in humans is often associated with severe viral liver disease including fulminant hepatitis. Since HBV is thought to be noncytopathic to the hepatocyte, the enhanced disease severity observed during dual infection has been attributed to either simultaneous immune responses against the two viruses or direct cytotoxic effects of HDV products on the hepatocyte or both. To examine these alternate possibilities, we produced transgenic mice that express the small and large delta antigens (HDAg) in hepatocyte nuclei at levels equal to those observed during natural HDV infection. No biological or histopathological evidence of liver disease was detectable during 18 months of observation, suggesting that neither the large nor small form of HDAg is directly cytopathic to the hepatocyte in vivo.     

Identification of the missing trans-acting enoyl reductase required for phthiocerol dimycocerosate and phenolglycolipid biosynthesis in Mycobacterium tuberculosis

Phthiocerol dimycocerosates (DIM) and phenolglycolipids (PGL) are functionally important surface-exposed lipids of Mycobacterium tuberculosis. Their biosynthesis involves the products of several genes clustered in a 70-kb region of the M. tuberculosis chromosome. Among these products is PpsD, one of the modular type I polyketide synthases responsible for the synthesis of the lipid core common to DIM and PGL. Bioinformatic analyses have suggested that this protein lacks a functional enoyl reductase activity domain required for the synthesis of these lipids. We have identified a gene, Rv2953, that putatively encodes an enoyl reductase. Mutation in Rv2953 prevents conventional DIM formation and leads to the accumulation of a novel DIM-like product. This product is unsaturated between C-4 and C-5 of phthiocerol. Consistently, complementation of the mutant with a functional pks15/1 gene from Mycobacterium bovis BCG resulted in the accumulation of an unsaturated PGL-like substance. When an intact Rv2953 gene was reintroduced into the mutant strain, the phenotype reverted to the wild type. These findings indicate that Rv2953 encodes a trans-acting enoyl reductase that acts with PpsD in phthiocerol and phenolphthiocerol biosynthesis.     

Osteopontin is upregulated by BCR-ABL

Chronic myelogenous leukemia (CML) is characterized by its hallmark oncogene BCR-ABL and the progression from a chronic phase toward an acute leukemia, with a differentiation arrest of the leukemic clone. In the present study, we conducted a microarray analysis using an inducible model of BCR-ABL expression based on the TET-OFF system, and we found that osteopontin (OPN), a component of stem cell niche, is overexpressed in BCR-ABL-expressing cells. Studies using mutant forms of BCR-ABL demonstrated that the BCR-ABL-induced OPN overexpression was a tyrosine kinase-dependent event. Furthermore, OPN concentration was significantly increased in the serum of leukemic mice generated by transplantation of BCR-ABL-expressing bone marrow cells. Most importantly, a significant increase of OPN concentration was observed in the serum of CML patients as compared to controls. Overall these results show that OPN is deregulated by BCR-ABL oncogene and suggest that OPN could be involved in CML stem cell biology.     

Karyotypes of 33 patients with clonal aberrations in chronic lymphocytic leukaemia. Review of 216 abnormal karyotypes in chronic lymphocytic leukaemia

We report on 33 unpublished patients with clonal anomalies in chronic lymphocytic leukaemia. The literature was thoroughly reviewed in order 1) to quantify the frequency of anomalies found in chronic lymphocytic leukaemia and to give new status to the rarest, 2) to determine whether a given anomaly was an additional anomaly and/or a primary anomaly, and 3) to find out whether strong associations between different anomalies exist in this disease.     

Genomics in cereals: from genome-wide conserved orthologous set (COS) sequences to candidate genes for trait dissection

Recent updates in comparative genomics among cereals have provided the opportunity to identify conserved orthologous set (COS) DNA sequences for cross-genome map-based cloning of candidate genes underpinning quantitative traits. New tools are described that are applicable to any cereal genome of interest, namely, alignment criterion for orthologous couples identification, as well as the Intron Spanning Marker software to automatically select intron-spanning primer pairs. In order to test the software, it was applied to the bread wheat genome, and 695 COS markers were assigned to 1,535 wheat loci (on average one marker/2.6 cM) based on 827 robust rice–wheat orthologs. Furthermore, 31 of the 695 COS markers were selected to fine map a pentosan viscosity quantitative trait loci (QTL) on wheat chromosome 7A. Among the 31 COS markers, 14 (45%) were polymorphic between the parental lines and 12 were mapped within the QTL confidence interval with one marker every 0.6 cM defining candidate genes among the rice orthologous region.