Late rather than early responses of human dendritic cells highlight selective induction of cytokines, chemokines and growth factors by probiotic bacteria.

The probiotic properties of commensal bacteria including lactobacilli and bifidobacteria are likely to be determined at least in part by their effects on dendritic cells. Like traditional immune stimulants such as lipopolysaccharides (LPS), probiotic bacteria promote maturation of cultured human dendritic cells (DC) by inducing elevated expression of MHC-II and co-stimulatory molecules. Different effects have been reported on cytokine induction, especially of major regulatory cytokines such as TNF-α, IL-12 and IL-10. Yet, these previous analyses have failed to reveal consistent differences between such effects of probiotics on the one hand, and of LPS on the other. Selective response markers for probiotics, however, would be important for our understanding of their biological properties and for a rational selection of strains for in vivo studies. In this study, we compared in detail both early and late effects on cultured human DC of 4 different probiotics with those of LPS. At the early stages of stimulation, all stimuli induced qualitatively very similar responses in DC at the level of surface markers and secretion of cytokines and chemokines. A lower immune stimulatory effect was observed by Bifidobacterium animalis BB-12 as compared to lactobacilli. Late responses, on the other hand, tended to diverge. Microarray transcript profiling for 268 cytokines, chemokines, growth factors and their receptors after 2 days of culture revealed various transcripts to be selectively induced by certain probiotics but not LPS. Our data indicate that late rather than early DC responses may be helpful to clarify the divergent biological effects of probiotics on human innate immune responses.

Identification of secreted proteins of Aspergillus oryzae associated with growth on solid cereal substrates.

Filamentous growth of Aspergillus oryzae on solid cereal substrates involves secretion of substrate converting enzymes and a solid substrate specific polarised hyphal growth phenotype. To identify proteins produced under these specific conditions, the extracts of A. oryzae grown on wheat-based media were analysed using N-terminal sequence analysis. In a submerged wheat-based growth medium of A. oryzae, besides alpha-amylase, also an arabinosidase and xylanase were abundantly produced. In the extracts of A. oryzae grown on wheat-based solid substrate besides alpha-amylase and chitinase, two new proteins of 16 and 27 kDa were identified. These hypothetical proteins showed only close homologies to filamentous fungal proteins.

Branching mutants of Aspergillus oryzae with improved amylase and protease production on solid substrates.

To study the relation between the number of hyphal tips and protein secretion during growth on a solid substrate, we have constructed two mutant strains of Aspergillus oryzae with increased hyphal branching. We have analysed hydrolytic enzyme activities during growth on wheat kernels (WK) of A. oryzae strains carrying the disrupted allele of the pclA gene encoding a secretion pathway specific (KEX2-like) endo-protease and the disrupted allele of the pg/pi-tp gene encoding a phosphatidylglycerol/phosphatidylinositol transfer protein. The biomass levels produced by the pclA and pg/pi-tp disrupted strains on wheat-based solid media were similar as found for the wild-type strain. However, the pclA disrupted strain showed much more compact colony morphology than the other two strains. Sporulation of the pclA and pg/pi-tp disrupted strains occurred, respectively, 2 days and 1 day later, compared to the wild type during fermentation on ground WK. During surface growth, microscopic analysis revealed that the hyphal growth unit length (L (hgu)) of the pclA and pg/pi-tp disrupted strains was, on average, 50 and 74% of that of the wild-type strain. This implies that in both mutant strains, a higher branching frequency occurs than in the wild-type strain. Compared to the wild-type strain, the pclA and pg/pi-tp disrupted strains produced at least 50% more amylase, at least 100% more glucoamylase and at least 90% more protease activity levels after growth on WK. These results support the hypothesis that branching mutants with an increased branching frequency can improve the solid state fermentation process.

Different control mechanisms regulate glucoamylase and protease gene transcription in Aspergillus oryzae in solid-state and submerged fermentation.

Solid-state fermentation (SSF) with Aspergillus oryzae results in high levels of secreted protein. However, control mechanisms of gene expression in SSF have been only poorly studied. In this study we show that both glucoamylase (glaB) and protease (alpA, nptB) genes are highly expressed during surface cultivation on wheat-based solid medium, and even higher during cultivation on wheat kernels. In wheat-based liquid medium, low levels of gene expression are observed. Typical SSF cultivation conditions, such as low water activity and the formation of aerial hyphae, did not contribute to the high-level gene expression on wheat-based solid medium. Analysis of wheat-based solid and liquid cultivations showed differences in carbon and nitrogen utilisation and external pH. The results presented show that the difference in regulation of transcription of the alpA and nptB genes in wheat-based liquid and solid medium could be pH dependent, involving a pH-dependent transcription regulator. The results obtained suggest that the difference in regulation of transcription of the glaB gene in wheat-based liquid and solid medium is caused by a difference in carbohydrate degradation and consumption under the different culture conditions.

Identification of growth phenotype-related genes in Aspergillus oryzae by heterologous macroarray and suppression subtractive hybridization.

Aspergillus oryzae requires polarized growth for colonization of solid substrates, and this growth phenotype differs from that seen in liquid medium. Various experimental approaches were used to identify genes that are differentially expressed when A. oryzae is grown on wheat kernels and in a wheat-based liquid medium. Hybridization of A. oryzae RNAs to a macroarray bearing cDNAs isolated from a library representing at least 16% of the total number of A. niger genes identified 14 differentially expressed cDNA clones, showing that heterologous macroarray analysis with an A. niger cDNA library can be used to identify regulated gene transcripts in the related species A. oryzae. Moreover, Northern analysis with a selection of eight probes for A. niger genes encoding proteins involved in morphological development and cell wall biosynthesis identified five more differentially expressed genes. A suppression subtractive hybridization procedure revealed another 12 differentially expressed genes. The results presented show that, of the 29 identified genes which are expressed at higher levels during growth on wheat kernels, six encode proteins that are functionally related to polarized growth, four encode products known to be involved in morphogenesis, three code for proteins related to cell wall composition, and nine of the cDNA clones encode novel proteins. These findings pinpoint genes associated with the changes in cellular morphogenesis seen in A. oryzae grown on wheat kernels as opposed to wheat-based liquid medium.

Elongation factor Ts can act as a steric chaperone by increasing the solubility of nucleotide binding-impaired elongation factor-Tu.

Several elongation factor (EF) Tu mutants (T25A, H22Y/T25S, D80N, D138N) that have impaired nucleotide binding show decreased solubility on overexpression in the E. coli cell, an indication that they do not fold correctly. Moreover, EF-Tu[T25A] and EF-Tu[D80N] were shown to inhibit cell growth on expression, an effect attributed to their sequestration of EF-Ts [Krab, I. M., and Parmeggiani, A. (1999) J. Biol. Chem. 274, 11132--11138; Krab, I. M., and Parmeggiani, A. (1999) Biochemistry 38, 13035--13041]. We present here results showing that the co-overexpression of EF-Ts at a 1:1 ratio dramatically improves the solubility of mutant EF-Tu, although in the case of EF-Tu[D138N]--which cannot at all bind the nucleotides available in the cell--this is a slow process. Moreover, with co-overexpression of EF-Ts, the mentioned growth inhibition is relieved. We conclude that for the formation of a correct EF-Tu structure the nucleotide plays an important role as a "folding nucleus", and also that in its absence EF-Ts can act as a folding template or steric chaperone for the correct folding of EF-Tu.

The arginine finger loop of yeast and human GAP is a determinant for the specificity toward Ras GTPase.

In this work, we have studied the role of the arginine finger region in determining the specificity of the GTPase activating proteins (GAPs) Saccharomyces cerevisiae Ira2p and human p120-GAP toward yeast Ras2p and human Ha-Ras p21. It is known that p120-GAP can enhance both Ras2p and Ha-Ras GTPase activities, whereas Ira2p is strictly specific for Ras2p and fails to activate Ha-Ras GTPase. Substitution in Ira2p of the arginine following the arginine finger with alanine, the residue found in the corresponding position of p120-GAP, or by glycine as found in neurofibromin, evokes a low but significant stimulation of Ha-Ras GTPase. The stimulatory activity of Ira2p on Ha-Ras increased by substituting segments of the finger loop region with p120-GAP residues, especially with the six residues forming the tip of the arginine loop. In p120-GAP, substitution of the entire finger loop with the corresponding region of Ira2p led to a construct completely inactive on Ha-Ras GTPase but active on yeast Ras2p GTPase. Analysis of these results and modeling of Ira2p.Ras complexes emphasize the importance of the finger loop region not only for the catalytic activity but also as a structural determinant involved in the specificity of GAPs toward Ras proteins from different organisms.

Comparison of cell wall proteins of Saccharomyces cerevisiae as anchors for cell surface expression of heterologous proteins.

The carboxyl-terminal regions of five cell wall proteins (Cwp1p, Cwp2p, Ag alpha 1p, Tip1p, and Flo1p) and three potential cell wall proteins (Sed1p, YCR89w, and Tir1p) all proved capable of immobilizing alpha-galactosidase in the cell wall of Saccharomyces cerevisiae. The fraction of the total amount of fusion protein that was localized to the cell wall varied depending on the anchor domain used. The highest proportion of cell wall incorporation was achieved with Cwp2p, Ag alpha 1p, or Sed1p as an anchor. Although 80% of these fusion proteins were incorporated in the cell wall, the total production of alpha-galactosidase-Ag alpha 1p was sixfold lower than that of alpha-galactosidase-Cwp2p and eightfold lower than that of alpha-galactosidase-Sed1p. Differences in mRNA levels were not responsible for this discrepancy, nor was an intracellular accumulation of alpha-galactosidase-Ag alpha 1p detectable. A lower translation efficiency of the alpha-galactosidase-AG alpha 1 fusion construct is most likely to be responsible for the low level of protein production. alpha-Galactosidase immobilized by the carboxyl-terminal 67 amino acids of Cwp2p was most effective in the hydrolysis of the high-molecular-weight substrate guar gum from Cyamopsis tetragonoloba. This indicates that the use of a large anchoring domain does not necessarily result in a better exposure of the immobilized enzyme to the exterior of the yeast cell.

The beta-1, 6-glucan containing side-chain of cell wall proteins of Saccharomyces cerevisiae is bound to the glycan core of the GPI moiety.

Cell wall proteins of Saccharomyces cerevisiae are anchored by means of a beta-1, 6-glucan-containing side-chain. It is not known whether this chain is linked to the protein part (e.g. through carbohydrate side-chains) or to the glycosylphosphatidylinositol (GPI) moiety of cell wall proteins. An IgA protease recognition site was introduced in Cwp2p, a beta-1, 6-glucosylated cell wall protein, immediately N-terminal from the omega amino acid (the attachment site of the GPI moiety). Proteolytic cleavage of this site revealed that the beta-1, 6-glucan epitope was not linked to the protein part. We conclude that neither N-or O-glycosylation is involved in beta-glucosylation of cell wall proteins. This confirms that the glycan core of the GPI moiety is the probable beta-1, 6-glucan attachment site.