Whole document restricted, see Access Instructions file below for details of how to access the print copy. / 1. The most appropriate methods were investigated for producing Rous sarcoma virus of suitable quantity and quality for use in the study of the viral RNA by electron microscopy. The roller bottle method of Smith and Bernstein (1973) which was adopted, produced virus yields of up to 5mg per litre of transformed cell culture supernatant after 24 hour incubations, and 0.2mg per litre of culture supernatant after 4 hour incubations. 2. The method of purifying RNA tumour viruses which resulted in the least damage to the virions was found to be isopyncic and velocity sedimentation in Ficoll density gradients containing 5mM tris-HCl and 1mM EDTA pH 8.5. The use of solutions of sucrose or >0.1M salt resulted in both osmotic changes in the virus and viral aggregation. 3. The lipoprotein coat of the Rous sarcoma virus was shown by freeze-fracturing and electron microscopy to have properties similar to those of plasma membranes, except that the number of intramembranous particles was smaller. The hydrated diameter of Rous sarcoma virus was estimated from freeze-fracture replicas of purified virions to be 140nm. 4. Vesicular contaminants, derived from serum, were present in Rous sarcoma virus preparations that had been purified from transformed cell culture supernatants. The isolated contaminants resembled virus when examined by both freeze-fracturing and negative-staining, but were readily distinguished from virus in thin sections. The virus-like serum vesicles were present in sera from several different sources. When treated with detergent and subjected to polyacylamide gel electrophoresis, the vesicles were found to contain polypeptides that possessed similar electrophoretic mobilities to those of Rous sarcoma virus polypeptides. It is probable that extraneous nucleic acid molecules, observed in preparations of Rous sarcoma virus RNA were the result of VLSV contamination of virus suspensions. 5. Contamination of purified virus suspensions by virus-like material derived from serum was reduced by centrifugation of the serum prior to its addition to cell culture medium. Virus suspensions, purified from cell supernatants from which the contaminating vesicles had been removed, were resolved in sharp bands at p = 1.07 g/ml in Ficoll density gradients; in the analytical ultracentrifuge they sedimented as homogenous populations with a sedimentation value of 740s20,w and were observed by electron microscopy to be relatively free of contaminants. 6. The maximum length of molecules from preparations of both 60-70s and 30-40s viral RNA prepared in 80% and 50% formamide respectively was 2.5μm, but both preparations were not homogeneous since they contained other, smaller molecules. 7. A model is proposed in which the difference in physical properties between the native (60-70S) form and the denatured (30-40S) form of the viral RNA is suggested to be the result of two possible conformations of a single RNA molecule. This model is an alternative to the prevailing model in which the RNA tumour virus genome is proposed to contain a number of RNA molecules of equivalent size.

Whole document restricted, see Access Instructions file below for details of how to access the print copy. / 1. The most appropriate methods were investigated for producing Rous sarcoma virus of suitable quantity and quality for use in the study of the viral RNA by electron microscopy. The roller bottle method of Smith and Bernstein (1973) which was adopted, produced virus yields of up to 5mg per litre of transformed cell culture supernatant after 24 hour incubations, and 0.2mg per litre of culture supernatant after 4 hour incubations. 2. The method of purifying RNA tumour viruses which resulted in the least damage to the virions was found to be isopyncic and velocity sedimentation in Ficoll density gradients containing 5mM tris-HCl and 1mM EDTA pH 8.5. The use of solutions of sucrose or >0.1M salt resulted in both osmotic changes in the virus and viral aggregation. 3. The lipoprotein coat of the Rous sarcoma virus was shown by freeze-fracturing and electron microscopy to have properties similar to those of plasma membranes, except that the number of intramembranous particles was smaller. The hydrated diameter of Rous sarcoma virus was estimated from freeze-fracture replicas of purified virions to be 140nm. 4. Vesicular contaminants, derived from serum, were present in Rous sarcoma virus preparations that had been purified from transformed cell culture supernatants. The isolated contaminants resembled virus when examined by both freeze-fracturing and negative-staining, but were readily distinguished from virus in thin sections. The virus-like serum vesicles were present in sera from several different sources. When treated with detergent and subjected to polyacylamide gel electrophoresis, the vesicles were found to contain polypeptides that possessed similar electrophoretic mobilities to those of Rous sarcoma virus polypeptides. It is probable that extraneous nucleic acid molecules, observed in preparations of Rous sarcoma virus RNA were the result of VLSV contamination of virus suspensions. 5. Contamination of purified virus suspensions by virus-like material derived from serum was reduced by centrifugation of the serum prior to its addition to cell culture medium. Virus suspensions, purified from cell supernatants from which the contaminating vesicles had been removed, were resolved in sharp bands at p = 1.07 g/ml in Ficoll density gradients; in the analytical ultracentrifuge they sedimented as homogenous populations with a sedimentation value of 740s20,w and were observed by electron microscopy to be relatively free of contaminants. 6. The maximum length of molecules from preparations of both 60-70s and 30-40s viral RNA prepared in 80% and 50% formamide respectively was 2.5μm, but both preparations were not homogeneous since they contained other, smaller molecules. 7. A model is proposed in which the difference in physical properties between the native (60-70S) form and the denatured (30-40S) form of the viral RNA is suggested to be the result of two possible conformations of a single RNA molecule. This model is an alternative to the prevailing model in which the RNA tumour virus genome is proposed to contain a number of RNA molecules of equivalent size.

Whole document restricted, see Access Instructions file below for details of how to access the print copy. / 1. The most appropriate methods were investigated for producing Rous sarcoma virus of suitable quantity and quality for use in the study of the viral RNA by electron microscopy. The roller bottle method of Smith and Bernstein (1973) which was adopted, produced virus yields of up to 5mg per litre of transformed cell culture supernatant after 24 hour incubations, and 0.2mg per litre of culture supernatant after 4 hour incubations. 2. The method of purifying RNA tumour viruses which resulted in the least damage to the virions was found to be isopyncic and velocity sedimentation in Ficoll density gradients containing 5mM tris-HCl and 1mM EDTA pH 8.5. The use of solutions of sucrose or >0.1M salt resulted in both osmotic changes in the virus and viral aggregation. 3. The lipoprotein coat of the Rous sarcoma virus was shown by freeze-fracturing and electron microscopy to have properties similar to those of plasma membranes, except that the number of intramembranous particles was smaller. The hydrated diameter of Rous sarcoma virus was estimated from freeze-fracture replicas of purified virions to be 140nm. 4. Vesicular contaminants, derived from serum, were present in Rous sarcoma virus preparations that had been purified from transformed cell culture supernatants. The isolated contaminants resembled virus when examined by both freeze-fracturing and negative-staining, but were readily distinguished from virus in thin sections. The virus-like serum vesicles were present in sera from several different sources. When treated with detergent and subjected to polyacylamide gel electrophoresis, the vesicles were found to contain polypeptides that possessed similar electrophoretic mobilities to those of Rous sarcoma virus polypeptides. It is probable that extraneous nucleic acid molecules, observed in preparations of Rous sarcoma virus RNA were the result of VLSV contamination of virus suspensions. 5. Contamination of purified virus suspensions by virus-like material derived from serum was reduced by centrifugation of the serum prior to its addition to cell culture medium. Virus suspensions, purified from cell supernatants from which the contaminating vesicles had been removed, were resolved in sharp bands at p = 1.07 g/ml in Ficoll density gradients; in the analytical ultracentrifuge they sedimented as homogenous populations with a sedimentation value of 740s20,w and were observed by electron microscopy to be relatively free of contaminants. 6. The maximum length of molecules from preparations of both 60-70s and 30-40s viral RNA prepared in 80% and 50% formamide respectively was 2.5μm, but both preparations were not homogeneous since they contained other, smaller molecules. 7. A model is proposed in which the difference in physical properties between the native (60-70S) form and the denatured (30-40S) form of the viral RNA is suggested to be the result of two possible conformations of a single RNA molecule. This model is an alternative to the prevailing model in which the RNA tumour virus genome is proposed to contain a number of RNA molecules of equivalent size.

Alginate biosynthesis is not only a major contributor to pathogenicity of P. aeruginosa but also an important factor in colonization of adverse environmental habitats by biofilm formation. The requirement of proteins Alg8 and Alg44, encoded by their respective genes in the alginate biosynthesis gene cluster, for alginate biosynthesis of P. aeruginosa was demonstrated, since deletion mutants were unable to produce or polymerise alginate. AlgX deletion mutants failed to produce the alginate characteristic mucoid phenotype, but showed low concentrations of uronic acid monomers in the culture supernatants. Complementation experiments using PCR based approaches were used to determine the complementing ORF and all deletion mutants could be complemented to at least wildtype levels by introducing a plasmid harbouring the respective gene. Increased copy numbers of Alg44 did not impact on the amount of alginate produced, whereas increased copy numbers of the alg8 gene led to an at least 10 fold stronger alginate production impacting on biofilm structure and stability. Topological analysis using reporter protein fusions and subsequent subcellular fractionation experiments revealed that Alg8 is located in the cytoplasmic membrane and contains at least 4 transmembrane helices, 3 of them at its C terminus. Its large cytosolic loop showed similarities to inverting glycosyltransferases and the similarities were used to generate a threading model using SpsA, a glycosyltransferase involved in spore coat formation of B. subtilis, as a template. Site-directed mutagenesis confirmed the importance of identified motifs commonly detected in glycosyltransferases. Inactivation of the DXD motif, which has been shown to be involved in nucleotide sugar binding, led to loss-offunction mutants of Alg8 and further replacements revealed putative candidates for the catalytic residue(s). Contradicting the commonly reported prediction of being a transmembrane protein, Alg44 was shown to be a periplasmic protein. The highest specific alkaline phosphatase activity of its fusion protein could be detected in the periplasmic fraction and not in the insoluble membrane fraction. Bioinformatical analysis of Alg44 revealed structural similarities of its N terminus to PilZ domains, shown to bind cyclic-di-GMP, and of its C terminus to MexA, a membrane fusion protein involved in multi-drug efflux systems. Thus, it was suggested that Alg44 has a regulatory role for alginate biosynthesis in bridging the periplasm and connecting outer and cytoplasmic membrane components. AlgX was shown to interact with MucD, a periplasmic serine protease or chaperone homologue, and is suggested to exert its impact on alginate production via MucD interaction. In vitro alginate polymerisation assays revealed that alginate production requires protein components of the outer and cytoplasmic membrane as well as the periplasm, and these data were used to construct a model describing a multi-enzyme, membrane and periplasm spanning complex for alginate polymerisation, modification and export.

Bacteria communicate with their hosts in part via surface, secreted and transmembrane proteins (collectively the secretome) resulting in probiotic (beneficial) or pathogenic (harmful) outcomes to the host. Therapeutic benefits of probiotic bacteria have been shown previously, but the molecular mechanisms and the health-promoting effector components involved are still being elucidated. Some evidence suggests that probiotic bacteria can competitively adhere to intestinal mucus and displace pathogens. The adherence of probiotic bacteria to human intestinal mucus and cells appears to be mediated, at least in part, by secretome proteins. Secretome proteins-encoding open reading frames can be identified in bacterial genome sequences using bioinformatics. However, functional analysis of the translated secretome is possible only if many secretome proteins are expressed and purified individually. Phage display technology offers a very efficient way to purify and functionally characterise proteins by displaying them on the surface of the bacteriophage. While a phage display system for cloning secretome proteins has been previously reported it is not efficient for enrichment and display of Gram-positive secretome proteins. In this study a new phage display system has been developed and applied in direct selection, identification, expression and purification of Gram-positive Lactobacillus rhamnosus strain HN001 secretome proteins. The new phage display system is based on the requirement of a signal sequence for assembly of sarcosyl-resistant filamentous phage virions. Using this system 89 secretome open reading frames were identified from a library of only 106 clones, performing at least 20-fold more efficiently than the previously reported enrichment method. Seven of the identified secretome proteins are unique for L. rhamnosus HN001. A L. rhamnosus HN001 shot-gun phage display library was also constructed to capture proteins that mediate adhesion or aggregation, initial steps in establishing host-microbe contact or forming multicellular aggregates, both of which may lead to beneficial effects – colonisation of the gastro-intestinal tract and exclusion of pathogens. In search for proteins involved in adhesion, a L. rhamnosus HN001 shot-gun phage display library was screened against the human extracellular matrix component fibronectin commonly used as binding target by bacteria that colonise diverse tissues. This screen selected, instead of a fibronectin-binding protein, a protein that binds to avidin, used to immobilise biotinylated fibronectin. Affinity screening of the shot-gun library for binding to L. rhamnosus HN001 cells identified a secretome protein, Lrh33, as an HN001-cell surface binding protein. This protein contains two bacterial immunoglobulin-like domains type 3. Analysis of phage-displayed nested deletions of Lrh33 determined that the proximal (N-terminal) immunoglobulin-like domain is not sufficient for binding; only the constructs displaying both domains demonstrated binding to HN001. Lrh33 does not have any similarity to previously identified Lactobacillus-binding proteins and no match in the NCBI database (at a cutoff value of > e-13), hence it represents potentially a new type of bacterial auto-aggregation protein.

Alginate biosynthesis is not only a major contributor to pathogenicity of P. aeruginosa but also an important factor in colonization of adverse environmental habitats by biofilm formation. The requirement of proteins Alg8 and Alg44, encoded by their respective genes in the alginate biosynthesis gene cluster, for alginate biosynthesis of P. aeruginosa was demonstrated, since deletion mutants were unable to produce or polymerise alginate. AlgX deletion mutants failed to produce the alginate characteristic mucoid phenotype, but showed low concentrations of uronic acid monomers in the culture supernatants. Complementation experiments using PCR based approaches were used to determine the complementing ORF and all deletion mutants could be complemented to at least wildtype levels by introducing a plasmid harbouring the respective gene. Increased copy numbers of Alg44 did not impact on the amount of alginate produced, whereas increased copy numbers of the alg8 gene led to an at least 10 fold stronger alginate production impacting on biofilm structure and stability. Topological analysis using reporter protein fusions and subsequent subcellular fractionation experiments revealed that Alg8 is located in the cytoplasmic membrane and contains at least 4 transmembrane helices, 3 of them at its C terminus. Its large cytosolic loop showed similarities to inverting glycosyltransferases and the similarities were used to generate a threading model using SpsA, a glycosyltransferase involved in spore coat formation of B. subtilis, as a template. Site-directed mutagenesis confirmed the importance of identified motifs commonly detected in glycosyltransferases. Inactivation of the DXD motif, which has been shown to be involved in nucleotide sugar binding, led to loss-offunction mutants of Alg8 and further replacements revealed putative candidates for the catalytic residue(s). Contradicting the commonly reported prediction of being a transmembrane protein, Alg44 was shown to be a periplasmic protein. The highest specific alkaline phosphatase activity of its fusion protein could be detected in the periplasmic fraction and not in the insoluble membrane fraction. Bioinformatical analysis of Alg44 revealed structural similarities of its N terminus to PilZ domains, shown to bind cyclic-di-GMP, and of its C terminus to MexA, a membrane fusion protein involved in multi-drug efflux systems. Thus, it was suggested that Alg44 has a regulatory role for alginate biosynthesis in bridging the periplasm and connecting outer and cytoplasmic membrane components. AlgX was shown to interact with MucD, a periplasmic serine protease or chaperone homologue, and is suggested to exert its impact on alginate production via MucD interaction. In vitro alginate polymerisation assays revealed that alginate production requires protein components of the outer and cytoplasmic membrane as well as the periplasm, and these data were used to construct a model describing a multi-enzyme, membrane and periplasm spanning complex for alginate polymerisation, modification and export.

Alginate biosynthesis is not only a major contributor to pathogenicity of P. aeruginosa but also an important factor in colonization of adverse environmental habitats by biofilm formation. The requirement of proteins Alg8 and Alg44, encoded by their respective genes in the alginate biosynthesis gene cluster, for alginate biosynthesis of P. aeruginosa was demonstrated, since deletion mutants were unable to produce or polymerise alginate. AlgX deletion mutants failed to produce the alginate characteristic mucoid phenotype, but showed low concentrations of uronic acid monomers in the culture supernatants. Complementation experiments using PCR based approaches were used to determine the complementing ORF and all deletion mutants could be complemented to at least wildtype levels by introducing a plasmid harbouring the respective gene. Increased copy numbers of Alg44 did not impact on the amount of alginate produced, whereas increased copy numbers of the alg8 gene led to an at least 10 fold stronger alginate production impacting on biofilm structure and stability. Topological analysis using reporter protein fusions and subsequent subcellular fractionation experiments revealed that Alg8 is located in the cytoplasmic membrane and contains at least 4 transmembrane helices, 3 of them at its C terminus. Its large cytosolic loop showed similarities to inverting glycosyltransferases and the similarities were used to generate a threading model using SpsA, a glycosyltransferase involved in spore coat formation of B. subtilis, as a template. Site-directed mutagenesis confirmed the importance of identified motifs commonly detected in glycosyltransferases. Inactivation of the DXD motif, which has been shown to be involved in nucleotide sugar binding, led to loss-offunction mutants of Alg8 and further replacements revealed putative candidates for the catalytic residue(s). Contradicting the commonly reported prediction of being a transmembrane protein, Alg44 was shown to be a periplasmic protein. The highest specific alkaline phosphatase activity of its fusion protein could be detected in the periplasmic fraction and not in the insoluble membrane fraction. Bioinformatical analysis of Alg44 revealed structural similarities of its N terminus to PilZ domains, shown to bind cyclic-di-GMP, and of its C terminus to MexA, a membrane fusion protein involved in multi-drug efflux systems. Thus, it was suggested that Alg44 has a regulatory role for alginate biosynthesis in bridging the periplasm and connecting outer and cytoplasmic membrane components. AlgX was shown to interact with MucD, a periplasmic serine protease or chaperone homologue, and is suggested to exert its impact on alginate production via MucD interaction. In vitro alginate polymerisation assays revealed that alginate production requires protein components of the outer and cytoplasmic membrane as well as the periplasm, and these data were used to construct a model describing a multi-enzyme, membrane and periplasm spanning complex for alginate polymerisation, modification and export.

Alginate biosynthesis is not only a major contributor to pathogenicity of P. aeruginosa but also an important factor in colonization of adverse environmental habitats by biofilm formation. The requirement of proteins Alg8 and Alg44, encoded by their respective genes in the alginate biosynthesis gene cluster, for alginate biosynthesis of P. aeruginosa was demonstrated, since deletion mutants were unable to produce or polymerise alginate. AlgX deletion mutants failed to produce the alginate characteristic mucoid phenotype, but showed low concentrations of uronic acid monomers in the culture supernatants. Complementation experiments using PCR based approaches were used to determine the complementing ORF and all deletion mutants could be complemented to at least wildtype levels by introducing a plasmid harbouring the respective gene. Increased copy numbers of Alg44 did not impact on the amount of alginate produced, whereas increased copy numbers of the alg8 gene led to an at least 10 fold stronger alginate production impacting on biofilm structure and stability. Topological analysis using reporter protein fusions and subsequent subcellular fractionation experiments revealed that Alg8 is located in the cytoplasmic membrane and contains at least 4 transmembrane helices, 3 of them at its C terminus. Its large cytosolic loop showed similarities to inverting glycosyltransferases and the similarities were used to generate a threading model using SpsA, a glycosyltransferase involved in spore coat formation of B. subtilis, as a template. Site-directed mutagenesis confirmed the importance of identified motifs commonly detected in glycosyltransferases. Inactivation of the DXD motif, which has been shown to be involved in nucleotide sugar binding, led to loss-offunction mutants of Alg8 and further replacements revealed putative candidates for the catalytic residue(s). Contradicting the commonly reported prediction of being a transmembrane protein, Alg44 was shown to be a periplasmic protein. The highest specific alkaline phosphatase activity of its fusion protein could be detected in the periplasmic fraction and not in the insoluble membrane fraction. Bioinformatical analysis of Alg44 revealed structural similarities of its N terminus to PilZ domains, shown to bind cyclic-di-GMP, and of its C terminus to MexA, a membrane fusion protein involved in multi-drug efflux systems. Thus, it was suggested that Alg44 has a regulatory role for alginate biosynthesis in bridging the periplasm and connecting outer and cytoplasmic membrane components. AlgX was shown to interact with MucD, a periplasmic serine protease or chaperone homologue, and is suggested to exert its impact on alginate production via MucD interaction. In vitro alginate polymerisation assays revealed that alginate production requires protein components of the outer and cytoplasmic membrane as well as the periplasm, and these data were used to construct a model describing a multi-enzyme, membrane and periplasm spanning complex for alginate polymerisation, modification and export.

Alginate biosynthesis is not only a major contributor to pathogenicity of P. aeruginosa but also an important factor in colonization of adverse environmental habitats by biofilm formation. The requirement of proteins Alg8 and Alg44, encoded by their respective genes in the alginate biosynthesis gene cluster, for alginate biosynthesis of P. aeruginosa was demonstrated, since deletion mutants were unable to produce or polymerise alginate. AlgX deletion mutants failed to produce the alginate characteristic mucoid phenotype, but showed low concentrations of uronic acid monomers in the culture supernatants. Complementation experiments using PCR based approaches were used to determine the complementing ORF and all deletion mutants could be complemented to at least wildtype levels by introducing a plasmid harbouring the respective gene. Increased copy numbers of Alg44 did not impact on the amount of alginate produced, whereas increased copy numbers of the alg8 gene led to an at least 10 fold stronger alginate production impacting on biofilm structure and stability. Topological analysis using reporter protein fusions and subsequent subcellular fractionation experiments revealed that Alg8 is located in the cytoplasmic membrane and contains at least 4 transmembrane helices, 3 of them at its C terminus. Its large cytosolic loop showed similarities to inverting glycosyltransferases and the similarities were used to generate a threading model using SpsA, a glycosyltransferase involved in spore coat formation of B. subtilis, as a template. Site-directed mutagenesis confirmed the importance of identified motifs commonly detected in glycosyltransferases. Inactivation of the DXD motif, which has been shown to be involved in nucleotide sugar binding, led to loss-offunction mutants of Alg8 and further replacements revealed putative candidates for the catalytic residue(s). Contradicting the commonly reported prediction of being a transmembrane protein, Alg44 was shown to be a periplasmic protein. The highest specific alkaline phosphatase activity of its fusion protein could be detected in the periplasmic fraction and not in the insoluble membrane fraction. Bioinformatical analysis of Alg44 revealed structural similarities of its N terminus to PilZ domains, shown to bind cyclic-di-GMP, and of its C terminus to MexA, a membrane fusion protein involved in multi-drug efflux systems. Thus, it was suggested that Alg44 has a regulatory role for alginate biosynthesis in bridging the periplasm and connecting outer and cytoplasmic membrane components. AlgX was shown to interact with MucD, a periplasmic serine protease or chaperone homologue, and is suggested to exert its impact on alginate production via MucD interaction. In vitro alginate polymerisation assays revealed that alginate production requires protein components of the outer and cytoplasmic membrane as well as the periplasm, and these data were used to construct a model describing a multi-enzyme, membrane and periplasm spanning complex for alginate polymerisation, modification and export.

Alginate biosynthesis is not only a major contributor to pathogenicity of P. aeruginosa but also an important factor in colonization of adverse environmental habitats by biofilm formation. The requirement of proteins Alg8 and Alg44, encoded by their respective genes in the alginate biosynthesis gene cluster, for alginate biosynthesis of P. aeruginosa was demonstrated, since deletion mutants were unable to produce or polymerise alginate. AlgX deletion mutants failed to produce the alginate characteristic mucoid phenotype, but showed low concentrations of uronic acid monomers in the culture supernatants. Complementation experiments using PCR based approaches were used to determine the complementing ORF and all deletion mutants could be complemented to at least wildtype levels by introducing a plasmid harbouring the respective gene. Increased copy numbers of Alg44 did not impact on the amount of alginate produced, whereas increased copy numbers of the alg8 gene led to an at least 10 fold stronger alginate production impacting on biofilm structure and stability. Topological analysis using reporter protein fusions and subsequent subcellular fractionation experiments revealed that Alg8 is located in the cytoplasmic membrane and contains at least 4 transmembrane helices, 3 of them at its C terminus. Its large cytosolic loop showed similarities to inverting glycosyltransferases and the similarities were used to generate a threading model using SpsA, a glycosyltransferase involved in spore coat formation of B. subtilis, as a template. Site-directed mutagenesis confirmed the importance of identified motifs commonly detected in glycosyltransferases. Inactivation of the DXD motif, which has been shown to be involved in nucleotide sugar binding, led to loss-offunction mutants of Alg8 and further replacements revealed putative candidates for the catalytic residue(s). Contradicting the commonly reported prediction of being a transmembrane protein, Alg44 was shown to be a periplasmic protein. The highest specific alkaline phosphatase activity of its fusion protein could be detected in the periplasmic fraction and not in the insoluble membrane fraction. Bioinformatical analysis of Alg44 revealed structural similarities of its N terminus to PilZ domains, shown to bind cyclic-di-GMP, and of its C terminus to MexA, a membrane fusion protein involved in multi-drug efflux systems. Thus, it was suggested that Alg44 has a regulatory role for alginate biosynthesis in bridging the periplasm and connecting outer and cytoplasmic membrane components. AlgX was shown to interact with MucD, a periplasmic serine protease or chaperone homologue, and is suggested to exert its impact on alginate production via MucD interaction. In vitro alginate polymerisation assays revealed that alginate production requires protein components of the outer and cytoplasmic membrane as well as the periplasm, and these data were used to construct a model describing a multi-enzyme, membrane and periplasm spanning complex for alginate polymerisation, modification and export.