Hybridization cross-reactivity within homologous gene families on glass cDNA microarrays.

Glass cDNA microarrays can be used to profile the expression of thousands of gene targets in a single experiment. However, the potential for hybridization cross-reactivity needs to be considered when interpreting the results. Here, we describe hybridization experiments with a model array representing four distinct functional classes (families): chemokines, cytochrome P-450 isozymes, G proteins, and proteases. The cDNA clones selected for this array exhibited pairwise sequence identities ranging from 55% to 100%, as determined by a homology scoring algorithm (LALIGN). Targets for microarraying were amplified by PCR and spotted in 4-fold replication for signal averaging. One designated target from each family was further amplified by PCR to incorporate a T7 promoter sequence for the production of synthetic RNA transcripts. These transcripts were used to generate fluorescent hybridization probes by reverse transcription at varying input concentrations. As expected, hybridization signals were highest at the matching target elements. Targets containing less than 80% sequence identity relative to the hybridization probe sequences showed cross-reactivities ranging from 0.6% to 12%. Targets containing greater than 80% identity showed higher cross-reactivities (26%-57%). These cross-reactive signals were analyzed for statistical correlation with the length of sequence overlap, percent sequence identity, and homology score determined by LALIGN. Overall, percent sequence identity was the best predictor of hybridization cross-reactivity. These results provide useful guidelines for interpreting glass cDNA microarray data.

Genomic analysis of differentially expressed genes in liver and biliary epithelial cells of patients with primary biliary cirrhosis.

The characterization of differentially expressed genes provides a powerful tool for identifying molecules that may be involved in the pathogenesis of disease. We have used two independent techniques to identify overexpressed transcripts in bile duct cells and in liver from patients with primary biliary cirrhosis (PBC). In the first method, we used suppressive subtractive hybridization to compare mRNA from isolated PBC bile duct epithelial cells (BECs) to normal BECs and identified 71 clones as transcribed at higher levels in PBC-BECs. Amongst these clones, 62/71 had matches in a non-redundant nucleotide database and 9/71 had matches in an EST database. Of the 62 clones, 51/62 include a complexity of genes involved in cell proliferation, signal transduction, transcription regulation, RNA processing, carbohydrate metabolism and hypothetical/unknown proteins; 4/62 were identified as interstitial collagenase and collagenase precursors, 4/62 as ribosomal proteins, 3/62 as mitochondrial DNA. The mitochondrial cDNA sequences included cytochrome c oxidase, Wnt-13, and the pHL gene, a c-myc oncogene containing coxIII sequence. In the second method, we constructed cDNA libraries from three different PBC livers and sequenced a total of 12,324 independent clones. These 12,324 clones underwent virtual subtraction with 2,814,148 independent clones from Incyte LifeSeq libraries. Twenty one sequences were identified as unique to PBC liver. Collectively, these approaches identified a number of genes involved in signalling, RNA processing, mitochondrial function, inflammation, and fibrosis. Interestingly, both Wnt-13 and Notch transcripts are overexpressed in PBC liver. Further studies are needed to focus on the significance of these genes during the natural history of disease.

Similar organization of the lipopolysaccharide-binding protein (LBP) and phospholipid transfer protein (PLTP) genes suggests a common gene family of lipid-binding proteins.

The transfer of lipids in aqueous environments such as serum has been attributed to a recently characterized class of proteins. Abnormal regulation of serum lipids by these proteins is thought to be a key event in the pathophysiology of cardiovascular diseases. Lipopolysaccharide (endotoxin) binding protein (LBP) was identified by virtue of its ability to bind bacterial lipid A. We have analyzed the exon-intron organization of the LBP gene and the nucleotide sequence of its approximately 20 kb spanning 5'- and 3'-untranslated regions. When comparing the genomic organization of LBP with that of two other genes coding for lipid transfer proteins, significant homologies were found. The LBP gene includes 15 exons, and the 2-kb promoter contains recognition elements of acute phase-typical reactants and a repetitive 12-mer motif with an as yet unknown protein-binding property. Detailed sequence comparison revealed a closer relatedness of LBP with PLTP than with CETP as demonstrated by an almost identical intron positioning. This high degree of similarity supports functional studies by others suggesting that like LBP, PLTP may also be able to bind and transport bacterial lipopolysaccharide.

Synthesis and secretion of wild-type and mutant human plasma cholesteryl ester transfer protein in baculovirus-transfected insect cells: the carboxyl-terminal region is required for both lipoprotein binding and catalysis of transfer.

Functional plasma cholesteryl ester transfer protein (CETP; 476 amino acids) has been expressed in baculovirus-transfected Sf9 insect cells by using a full-length cDNA derived from a human placental library. The product bound to each major plasma lipoprotein class, and it catalyzed the transfer of both cholesteryl esters and triglyceride. CETP species with overlapping deletions were generated in the carboxyl-terminal region. These mutants were defective in cholesteryl ester and triglyceride transfer. Structural and functional analysis suggests that normal lipoprotein binding and effective catalysis may require the carboxyl-terminal sequence -Phe-Leu-Leu-Leu- (residues 454-457), possibly with the involvement of other sequences in the carboxyl-terminal region. A similar sequence is contained in several other proteins whose functions involve binding nonpolar lipids, including lecithin: cholesterol acyltransferase, lipopolysaccharide-binding protein, bactericidal permeability-increasing protein, cholesterol 7 alpha-hydroxylase, cholesterol esterase, and hormone-sensitive lipase. These data suggest that a conserved neutral lipid-binding sequence may be one important factor in the activity of CETP and possibly in several other proteins of plasma and cellular lipid metabolism.

Isolation of a chitin synthase gene (CHS1) from Candida albicans by expression in Saccharomyces cerevisiae.

Chitin synthase activity was studied in yeast and hyphal forms of Candida albicans. pH-activity profiles showed that yeast and hyphae contain a protease-dependent activity that has an optimum at pH 6.8. In addition, there is an activity that is not activated by proteolysis in vitro and which shows a peak at pH 8.0. This suggests there are two distinct chitin synthases in C. albicans. A gene for chitin synthase from C. albicans (CHS1) was cloned by heterologous expression in a Saccharomyces cerevisiae chs1 mutant. Proof that the cloned chitin synthase is a C. albicans membrane-bound zymogen capable of chitin biosynthesis in vitro was based on several criteria. (i) the CHS1 gene complemented the S. cerevisiae chs1 mutation and encoded enzymatic activity which was stimulated by partial proteolysis; (ii) the enzyme catalyses incorporation of [14C]-GlcNAc from the substrate, UDP[U-14C]-GlcNAc, into alkali-insoluble chitin; (iii) Southern analysis showed hybridization of a C. albicans CHS1 probe only with C. albicans DNA and not with S. cerevisiae DNA; (iv) pH profiles of the cloned enzyme showed an optimum at pH 6.8. This overlaps with the pH-activity profiles for chitin synthase measured in yeast and hyphal forms of C. albicans. Thus, CHS1 encodes only part of the chitin synthase activity in C. albicans. A gene for a second chitin synthase in C. albicans with a pH optimum at 8.0 is proposed. DNA sequencing revealed an open reading frame of 2328 nucleotides which predicts a polypeptide of Mr 88,281 with 776 amino acids. The alignment of derived amino acid sequences revealed that the CHS1 gene from C. albicans (canCHS1) is homologous (37% amino acid identity) to the CHS1 gene from S. cerevisiae (sacCHS1).

The S. cerevisiae structural gene for chitin synthase is not required for chitin synthesis in vivo.

The chitin synthase of Saccharomyces is a plasma membrane-bound zymogen. Following proteolytic activation, the enzyme synthesizes insoluble chitin that has chain length and other physical properties similar to chitin found in bud scars. We isolated mutants lacking chitin synthase activity (chs1) and used these to clone CHS1. The gene has an open reading frame of 3400 bases and encodes a protein of 130 kd. The fission yeast S. pombe lacks chitin synthase and chitin. When a plasmid encoding a CHS1-lacZ fusion protein is introduced into S. pombe, both enzymatic activities are expressed in the same ratio as in S. cerevisiae, demonstrating that CHS1 encodes the structural gene of chitin synthase. Three CHS1 gene disruption experiments were performed. In all cases, strains with the disrupted gene have a recognizable phenotype, lack measurable chitin synthase activity in vitro but are viable, contain normal levels of chitin in vivo, and mate and sporulate efficiently.

Modification of yeast plasma membrane density by concanavalin A attachment. Application to study of chitin synthetase distribution.

Yeast protoplasts were coated with different amounts of concanavalin A. Upon subsequent lysis and centrifugation in isopycnic density gradients, it was found that the buoyant density of plasma membranes was progressively increased from 1.125 to 1.21, according to the amount of bound concanavalin A. Enzymes that are attached to the plasma membrane showed the same density modifications and could thus be distinguished from constituents of intracellular membranes and organelles. With this methodology, it was confirmed that about two-thirds of yeast chitin synthetase is associated with the plasma membrane. The remainder of the enzyme was found in a peak at a lower density. Vanadate-sensitive ATPase showed a similar pattern, whereas GDP-mannose dolichyl-phosphate mannosyltransferase, an enzyme attached to the endoplasmic reticulum, remained in the same position in the gradients, irrespective of the amount of concanavalin A associated with the plasma membrane. Potential applications of this technique to the determination of plasma membrane markers and to the separation of subcellular organelles are discussed.

Serologic analysis of antigen-specific reactivity in patients with systemic candidiasis.

Antibody responses to candidal polypeptides and mannans were studied in patients with systemic candidiasis, candiduria, and other fungal and bacterial infections, and in healthy laboratory personnel to determine the diagnostic value of these immunologic responses. When tested by immunoblot analysis, sera from 15 patients with systemic candidiasis frequently contained antibodies to three antigens: 15 of 15 sera from patients with invasive disease reacted to a molecular species having a molecular weight (Mr of 90-200 kd, 13 of 15 reacted with a 45-kd polypeptide, and 12 of 15 reacted with a 17-kd polypeptide. Lesser reactivity was observed in 11 of 15 sera with a 28-kd candidal antigen and in 9 of 15 to a 57-kd candidal antigen. Quantitation of antibody titers against the 45-kd candidal polypeptide demonstrated much higher immunoreactivity in patients with systemic candidiasis than in patients with superficial candidal infections, bacterial infections, other systemic mycoses, and healthy individuals. Antimannan antibody titers were measured by an enzyme-linked immunosorbent assay (ELISA) and these titers were also higher in patients with systemic candidiasis than in patients in the other categories. These differences, however, were less than those observed with the anti-45-kd polypeptide antibody. Therefore, the ability to detect systemic candidiasis is improved by testing sera for immunoreactivity to polypeptide and to mannan antigens from Candida albicans. Detection of polypeptide antibodies improves the serodiagnosis of systemic candidiasis.

Isolation of chitin synthetase from Saccharomyces cerevisiae. Purification of an enzyme by entrapment in the reaction product.

Chitin synthetase, in the zymogen form, was extracted with digitonin from a particulate fraction from Saccharomyces cerevisiae and converted into active form by treatment with immobilized trypsin. When the activated enzyme was incubated with UDP-GlcNAc and other components of an assay mixture, a chitin precipitate formed, trapping a large portion of the synthetase. The enzyme was easily extracted frm the chitin gel with a recovery of approximately 50% and an enrichment of approximately 100-fold. Further purification was obtained by repeating the chitin step. After polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate, the purified synthetase showed a major band corresponding to Mr 63,000, a weaker band at Mr 74,000, and some other minor bands. Under nondenaturing conditions, an Mr of 570,000 was calculated for the enzyme from Stokes radius and sedimentation coefficient determinations. After electrophoresis in a nondenaturing gel and incubation with the components of the standard assay, chitin was formed and precipitated in the gel, yielding an opaque band. Soluble oligosaccharides were not precursors for insoluble chitin, suggesting that synthesis of chitin chains takes place by a processive mechanism. N-Acetylglucosamine stimulated the purified synthetase only slightly and did not participate as a primer in the reaction. The same chain length, somewhat more than 100 units of GlcNAc, was determined in samples of chitin that had been synthesized either in vivo, or with a membrane preparation or with purified synthetase. These results suggest that chitin synthetase itself is capable both of initiating chitin chains without a primer and of determining their length.