CsgD, a regulator of curli and cellulose synthesis, also regulates serine hydroxymethyltransferase synthesis in Escherichia coli K-12

Neema T. Chirwa and Muriel B. Herrington

Biology Department, Concordia University, 1455 Maisonneuve Boulevard West, Montreal, Quebec, Canada H3G 1M8

Correspondence
Muriel B. Herrington
muriel{at}vax2.concordia.ca


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The homologous CsgD and AgfD proteins are members of the FixJ/UhpA/LuxR family and are proposed to regulate curli (thin aggregative fibres) and cellulose production by Escherichia coli and Salmonella enterica serovar Typhimurium, respectively. A plasmid containing part of the csgD gene was isolated during a screen for multicopy suppressors of glycine auxotrophy caused by deleting the folA gene in E. coli. The sequence of the plasmid suggests it encodes a chimaeric protein. Plasmids containing the intact csgD or agfD gene also caused suppression. Cells transformed with the recombinant plasmids contained higher serine hydroxymethyltransferase (SHMT) activity than controls. The increase could also be monitored by assaying {beta}-galactosidase activity from a reporter strain with part of the SHMT gene, glyA, fused to lacZ. The increase in SHMT activity was sufficient to correct the glycine auxotrophy of strains lacking folA. The recombinant plasmids also enabled K-12 strains that are not curli-proficient to make curli. Curlin, the major component of curli, contains more glycine than normal E. coli proteins. It is proposed that CsgD upregulates glyA to facilitate synthesis of curli. It is suggested that recombinant plasmids produce enough CsgD or chimaeric protein to titrate out a ligand that switches CsgD into its inactive form. As a result, sufficient active CsgD is present to activate genes in its regulon. It is concluded that CsgD increases expression of the glyA gene either directly or indirectly.


Abbreviations: DHFR, dihydrofolate reductase; SHMT, serine hydroxymethyltransferase; FEP, folate end-product(s) (thymine, glycine, histidine, methionine, adenine and pantothenic acid); THF, tetrahydrofolate


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Commensal strains of Escherichia coli are found in the lower intestine of many different animal species (Souza et al., 1999). Pathogenic strains of E. coli can cause diarrhoea of varying severity, urinary tract infections and sepsis in humans, and a variety of diseases in domestic animals. Because of their wide geographical distribution and primary habitat, they are also common in soil and water contaminated with faeces. Their numbers can increase in these habitats given a favourable environment and relatively low predation (Solo-Gabriele et al., 2000).

Thus, E. coli switches between colonizing an animal host and water or soil. Colonization of a new habitat by bacterial cells is enhanced by their ability to attach to surfaces; such attachment is facilitated by fimbriae. E. coli can produce fimbriae called curli (Olsén et al., 1989). Salmonella enterica produces homologous fimbriae called thin aggregative fibres or SEF17 (Collinson et al., 1991; Römling et al., 1998a). We will use curli to refer to these fimbriae in both E. coli and S. enterica. Curli promote binding to a variety of host proteins, increase internalization of bacterial cells by eukaryotic cells in culture and trigger events leading to septic shock (Olsén et al., 1989, 1998; Sjöbring et al., 1994; Ben Nasr et al., 1996; Bian et al., 2000; Gophna et al., 2001; Uhlich et al., 2002). Curli also promote clumping of bacterial cells in culture and binding to abiotic surfaces such as glass and polystyrene, making them important for biofilm formation (Vidal et al., 1998; Römling et al., 1998b; Austin et al., 1998; Prigent-Combaret et al., 2000).

Although E. coli, S. enterica serovars Typhimurium (Salty) and Enteritidis, and Shigella species have the genes required for curli synthesis, the expression of curli is not common to all strains. Many strains contain insertion sequence disruptions or deletions in the genes encoding curli components and assembly factors, or mutations elsewhere that block curli formation (Olsén et al., 1993b; Arnqvist et al., 1994; La Ragione et al., 1999; Uhlich et al., 2001; Sakellaris et al., 2000). Natural isolates of enterohaemorrhagic, enterotoxigenic and sepsis E. coli strains from humans, and isolates from avian and bovine infections, are frequently curliated, whereas those from enteroinvasive and enteropathogenic strains are not (Olsén et al., 1989, 1993a; Maurer, 1998; Bian et al., 2000; Uhlich et al., 2001). E. coli K-12 strains generally are not curli-proficient but can produce curli as a result of mutations (Arnqvist et al., 1992, 1994; Vidal et al., 1998). Expression of curli in E. coli and Salty normally occurs during stationary phase and requires growth at low temperature and low osmolarity, but some strains and S. enterica serovar Enteritidis can express curli at 37 °C (Olsén et al., 1989; Collinson et al., 1991; Römling et al., 1998b; Uhlich et al., 2001).

The csg (agf in S. enterica) genes are required for curli synthesis. They are organized in two adjacent divergently transcribed operons. The csgBA (agfBA) operon contains the gene (agfB) for a nucleator protein and the gene (agfA) for curlin, the major protein component of the curli fibre. The csgDEFG (agfDEFG) operon contains a putative regulatory gene (csgD), a gene (csgG) encoding a lipoprotein involved in secretion of curlin and CsgB, a gene (csgE) encoding a chaperone-like protein and a gene (csgF) encoding a second nucleator protein (Collinson et al., 1996, 1997; Hammar et al., 1995, 1996; Loferer et al., 1997; Römling et al., 1998a; Chapman et al., 2002).

Changes in the csgD (agfD) promoter region can influence curli formation. Some isolates of the enteroinvasive E. coli O157 : H7 show a phase shift between curliated and non-curliated cells that results from a single base pair change in the csgD promoter (Uhlich et al., 2002). A change in the agfD promoter in Salty enables the cells to make curli at 37 °C (Römling et al., 2000).

The csgD-encoded protein is homologous to proteins of the FixJ/UhpA/LuxR family, suggesting that it is regulatory protein (Hammar et al., 1995). This role is supported by the observations that inactivation of csgD prevents the transcription of the csgBA operon (Hammar et al., 1995) and of adrA, a putative regulatory gene required for cellulose synthesis (Römling et al., 2000; Zogaj et al., 2001). Cellulose is found in the matrix of E. coli and Salty biofilms. The FixJ/UhpA/LuxR family is characterized by a DNA-binding C-terminal domain and an N-terminal receiver domain. Depending on the regulator, the receiver domain is either phosphorylated or binds homoserine lactone or other small molecules (Kahn & Ditta, 1991; Sitnikov et al., 1995; Baikalov et al., 1996; Ducros et al., 2001). As yet there is no evidence for DNA binding by CsgD, so it is not known whether it controls the csgBA operon and the adrA gene directly or indirectly. Similarly, it is not known what ligand modulates its function. Expression of csgD or agfD increases in late exponential phase and is influenced by temperature, osmolarity and the availability of nutrients, oxygen and iron (Römling et al., 1998b; Prigent-Combaret et al., 2001; Gerstel & Römling, 2001).

In this study, we show that csgD is a multicopy suppressor of the glycine auxotrophy of a strain lacking dihydrofolate reductase (DHFR) activity. DHFR is required for the de novo synthesis of tetrahydrofolate (THF) and for the recycling of dihydrofolate produced during thymidylate synthesis. THF is converted to methylene THF during glycine sythesis or degradation. Methylene THF is required for thymidylate and pantothenate biosynthesis and is the precursor for methyl-THF and formyl-THF. Methyl-THF is required for methionine biosynthesis. Formyl-THF is used for purine (and, indirectly, histidine) synthesis and for the formylation of Met-tRNAfmet (reviewed by Matthews, 1996).

Strains with the DHFR gene (folA) deleted grow slowly on rich or supplemented minimal medium. The supplements required depend on the strain background, but most strains will grow with thymidine or a combination of folate end-products (FEP: thymine, glycine, histidine, methionine, adenine and pantothenic acid) (Ahrweiler & Frieden, 1988; Howell et al., 1988; Hamm-Alvarez et al., 1990; Krishnan & Berg, 1993; Herrington & Chirwa, 1999). When glycine or glycine and histidine are omitted from the mixture the cells grow extremely slowly, and when methionine or methionine and histidine are omitted the cells do not grow. Under these conditions the purine pathway is repressed and this could reduce turnover of formyl-THF (Herrington & Chirwa, 1999). We show here that plasmids containing csgD sequences enabled faster growth in the absence of glycine. This results from an increased expression of glyA, the gene encoding serine hydroxymethyltransferase (SHMT, EC 2 . 1 . 2 . 1), suggesting that csgD also regulates glyA.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and plasmids.
E. coli K-12 strains and plasmids are listed in Table 1.


View this table:
[in this window]
[in a new window]
 
Table 1. E. coli K-12 strains and plasmids

 
Media and growth conditions.
Minimal medium A with glucose (GM), and Luria–Bertani (LB) medium containing 50 µg thymidine ml-1 (LB-Thy) (Miller, 1992) were used routinely. Media were solidified with 15 g agar l-1. GM was supplemented with amino acids and thymidine at 50 µg ml-1, adenine at 30 µg ml-1 and pantothenate at 1 µg ml-1. Media containing subsets of FEP were identified by indicating the missing FEP. For example, FEP-His-Met is GM with glycine, adenine, pantothenate and thymidine, but not histidine or methionine. Any compounds required by strains were also supplied. Ampicillin (100 µg ml-1) was always added when growing ampicillin-resistant transformants. Chloramphenicol (25 µg ml-1), spectinomycin (100 µg ml-1), streptomycin (20 µg ml-1) and tetracycline (25 µg ml-1) were added when required. Congo red plates (Hammar et al., 1995) with added thymidine were used to test curli proficiency. Liquid cultures were grown at the indicated temperatures with shaking. Growth on solid media was tested by spotting 10 µl of dilutions of overnight cultures grown in LB-Thy and by monitoring colony formation (Herrington & Chirwa, 1999).

Genetic and molecular techniques.
P1 transductions (Miller, 1992) were performed with either P1CM or P1vir. P1CM lysates prepared from JC1089 were used to lysogenize other strains as needed. P1vir was obtained from C. G. Cupples, Concordia University, Montreal, Canada.

Normally, cells were rendered competent for transformation using CaCl2 (Sambrook, 1989). To avoid heat shock, the one-step PEG method (Chung et al., 1989) was used to make strain GS162{lambda}glyAlacZ and its derivatives competent.

Strain construction.
P1vir-mediated transduction was used to transfer the purR6 : : Tn10 mutation from strain SØ5052 to strain MH829. We constructed strains MH910 and MH911 by, respectively, transducing csgA2 : : Tn105 and csgG1 : : Tn105 mutations into strain MH829 with P1vir lysates made from strains MHR204 and MHR210. Strains MH901 and MH902 were obtained by transferring these mutations to strain GS162{lambda}glyAlacZ. P1CM lysates prepared on strain FB10186 were used to transduce the csgD : : Tn5<KAN-I-SceI> mutation to strains MH917 and GS162{lambda}glyAlacZ. Construction of the spectinomycin-resistant folA null mutant MH917 was similar to that of strain MH829 (Herrington & Chirwa, 1999).

Isolation and identification of multicopy suppressors.
Strain {chi}PH43 is a double lysogen for Mu cts and the mini-Mu replicon MudII4042 which confers chloramphenicol resistance on the host strain (Groisman & Casadaban, 1987). Lysates from strain {chi}PH43 were used to prepare mini-Mu lysogens of MG1655. Mini-Mu lysates prepared on strain MG1655 were then used to transduce strain MH829/Mu cts at 31 °C. Chloramphenicol-resistant colonies were selected and then were replica-plated to screen for growth on various media.

Suppressing plasmids were restricted and the resulting fragments were ligated into pUC18. DH5{alpha} was transformed with the ligation mixture. Recombinant plasmids were tested for their effect on growth of strain MH829. Inserts in suppressor plasmids were sequenced by the York University Core Molecular Facility, Toronto, Canada.

Enzyme assays.
Cells were grown at 37 °C to an OD600 between 0·4 and 0·8 for all enzyme assays. Threonine dehydrogenase assays were measured in toluene-permeabilized cells (Ravnikar & Somerville, 1987). The SHMT- (Taylor & Weissbach, 1965) and glycine-cleavage activities were assayed in crude extracts prepared by sonicating cells. Glycine cleavage was assayed by a modification of the SHMT assay. The reaction mix contained 50 mM potassium phosphate, pH 7·4, 5 mM DTT, 1·14 mM THF, 1 mM pyridoxal phosphate and 2-[14C]glycine (0·16 mCi mM-1). Reactions were incubated at 32 °C. Reactions were stopped and analysed as for SHMT. Protein was measured by the modified bicinchoninic acid procedure (Brown et al., 1989). {beta}-Galactosidase activity in reporter strains for glyA (Lorenz & Stauffer, 1995) and gcvT (Jourdan & Stauffer, 1998) expression was assayed as described by Miller (1992) using the SDS/chloroform lysis method.

Curli formation.
Curli proficiency was assayed on Congo red plates (Hammar et al., 1995), by the ability of cells to bind to polystyrene (Vidal et al., 1998) and by the binding of Congo red in solution (Gophna et al., 2001). Congo red binding units were 1000 ({Delta}A500/OD600) where {Delta}A500 was the difference between the A500 value of the Congo red solution without cells and the A500 value measured after cells had been incubated in the solution and then removed. OD600 measurements for these assays were made on Milton Roy Spectronic 1001 Plus apparatus, using a 1 cm path length. Cells were diluted in 0·9 % saline.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
A plasmid containing part of the csgD gene is a multicopy suppressor
A mini-Mu lysate prepared on the wild E. coli strain MG1655 was used to transduce strain MH829/Mu cts to chloramphenicol resistance. A total of 59 195 chloramphenicol-resistant colonies were screened on FEP-Gly, FEP-Met, FEP-His-Met and FEP-His-Gly. The majority of colonies did not grow on FEP-Gly or FEP-Met but grew slowly on FEP-His-Gly and FEP-His-Met. Thirty colonies grew faster on FEP-His-Gly and FEP-His-Met. Five of these grew on FEP-Gly and on FEP-Met and 16 grew on FEP-Gly but not on FEP-Met. This suggested that different suppressors were obtained in this screen. Plasmids were isolated and transformed into strain MH829/Mu cts to confirm the phenotype.

Mini-Mu plasmids can carry insertions of up to 22·3 kb (Groisman & Casadaban, 1987). To reduce the size of the inserts, HindIII fragments of suppressing plasmids were subcloned into the high-copy-number vector pUC18. The resultant recombinant plasmids were tested for their ability to correct the growth defect of strain MH829.

One recombinant plasmid, pSD6P, allowed growth on FEP-Gly but not on FEP-Met. It contained a 210 bp insert. The sequence of the insert was determined and a BLAST (Altschul et al., 1997) search of the E. coli genome sequence indicated that it matched part of the csgD gene (GenBank accession no. AE000205). When pSD6P was analysed for ORFs, two were observed. One corresponded to the bla gene of pUC18 (GenBank accession no. L08752). The other, which is in the opposite orientation to the sequence encoding the lacZ {alpha} fragment, initiated at the complement of nt 499 of pUC18, spanned the insert and terminated at complement of nt 143 of pUC18. The predicted chimaeric protein contains 70 aa corresponding to the central region of CsgD (Fig. 1). The presence of a weak promoter located at the complement of nucleotide 746 of pUC18 suggests the chimaeric ORF could be transcribed.



View larger version (10K):
[in this window]
[in a new window]
 
Fig. 1. Comparison of CsgD and the putative protein encoded by pSD6P. The amino acid sequences of CsgD and chimaeric protein (SD6P) encoded by pSD6P were aligned with MultAlin (Corpet, 1988). The cross-hatched areas indicate the 70 aa encoded by the insert in pSD6P and the corresponding CsgD region. The black segment in CsgD is the helix–turn–helix characteristic of DNA-binding domains. Dashed lines in the regions of SD6P encoded by the vector indicate positions in the chimaeric protein where the amino acid is identical to that in CsgD.

 
If suppression results from expression of a chimaeric protein, changing the orientation of the insert in pSD6P should eliminate suppression. When pSD6P was cut with HindIII, ligated, and transformed into strain MH829, approximately 50 % of the transformants did not grow on FEP-Gly. One non-suppressing plasmid was sequenced and shown to have the insert in the opposite orientation to pSD6P. These results suggest that a chimaeric protein made by pSD6P was responsible for suppression.

Plasmids containing intact csgD genes suppress
Colony formation on LB-Thy and FEP-Gly was monitored in strain MH829 transformed with different plasmids containing csgD or agfD sequences and with control plasmids. All transformants produced small colonies within 24 h on LB-Thy (Fig. 2a and Table 2). Transformants with pSD6P (partial csgD), pCP994 (intact csgD coding region) or pCSGD (agfD from S. enterica serovar Typhimurium) formed colonies on FEP-Gly within 4 days. In contrast, the pUC18 control made no colonies (pUC18) and the pKK233-2 control made extremely small colonies after 5 days (Fig. 2b and Table 2). The efficiency of plating on LB-Thy and on FEP-Gly was similar. These results indicate that multiple copies of the intact csgD gene also suppress the glycine auxotrophy.



View larger version (51K):
[in this window]
[in a new window]
 
Fig. 2. Growth and Congo red binding of transformants of strain MH829. Transformants were grown overnight in LB-Thy ampicillin, and aliquots (10 µl) were spotted on plates that were incubated for 5 days. 1, MH829 transformed with pCP994; 2, MH829 transformed with pKK233-2. (a, b) Upper row, samples were diluted 10-3; lower row, samples were diluted 10-5. (a) LB-Thy at 37 °C; (b) FEP-Gly at 37 °C; (c) undiluted cultures spotted on Congo red plates and incubated at 23 °C. The spot stained red appears darker than the unstained spot in the black and white image.

 

View this table:
[in this window]
[in a new window]
 
Table 2. Growth of strain MH829 and derivatives on FEP-Gly

 
Transformed cells were grown in liquid FEP-Gly containing ampicillin with varying amounts of glycine. In the absence of glycine or at low concentration, the yield of cells transformed with pCP994 was higher than those transformed with PKK233-2 (Fig. 3). Similar results were obtained with pSD6P transformants (data not shown).



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 3. Growth with varying amounts of glycine. Strain MH829 transformed with pKK233-2 (white bars) and pCP994 (black bars) were grown in minimal medium containing FEP with the glycine concentration ranging from 0 to 50 µg ml-1. One millilitre cultures were inoculated with 106 cells and the OD600 was measured after 5 days incubation at 37 °C. Results represent the mean of two assays in which each sample was done in duplicate. Values for pKK233-2 and pCP994 transformants were compared using Student's t-test; *, conditions in which the pCP994 plasmid significantly improved growth.

 
SHMT activity is increased by expression of CsgD from a plasmid
The csgD gene could reduce the requirement for exogenous glycine by increasing synthesis or decreasing degradation. Most of the glycine made by E. coli is produced from serine by SHMT (Stauffer, 1996) with the concomitant production of 5,10-methylene THF from THF. To test whether csgD sequences modified the expression of SHMT, we assayed SHMT activity in transformants of strains MH828 (folA+ thyATs), MH829 ({Delta}folA : : kan3 thyATs) and GS162{lambda}glyAlacZ (folA+ thyA+). SHMT activity was significantly and reproducibly 1·5- to 3-fold higher in pSD6P and pCP194 transformants compared to control transformants when MH829 and GS162{lambda}glyAlacZ transformants were grown in GM+Thy and when MH828 and GS162{lambda}glyAlacZ transformants were grown in FEP-Gly (Table 3). Although pSD6P transformants of strain MH828 grown in GM+Thy had higher activity than control transformants in the experiment shown (Table 3), this difference was not observed in other experiments.


View this table:
[in this window]
[in a new window]
 
Table 3. SHMT activity in transformants

 
SHMT activity was higher in the {Delta}folA strain MH829 than in the folA+ strains. Similar results were observed in a wild-type strain when DHFR was inhibited by trimethoprim (Stauffer, 1996). This could be a response to limited glycine, methionine and purine synthesis when DHFR is not available.

Growth in the presence of purines and methionine normally represses SHMT (Mansouri et al., 1972; Miller & Newman, 1974; Greene & Radovich, 1975). We observed repression on FEP-Gly with strain GS162{lambda}glyAlacZ transformants, but not with the strain MH828 transformants. This suggests that the thyA mutation in strain MH828 affected the regulation of SHMT. Under the conditions used, strains with the thyA(ts) allele normally express 5 % of wild-type thymidylate synthase activity (Herrington & Chirwa, 1999). The reduced demand for methylene THF for thymidylate synthesis could be perceived by the cell as a signal to provide more formyl-THF for protein synthesis; hence the upregulation of SHMT.

Expression of {beta}-galactosidase from a glyAlacZ fusion is also increased
Strain GS162{lambda}glyAlacZ is lysogenized with the glyA reporter, {lambda}glyAlacZ. This phage contains the entire hmpglyA intergenic region and the sequence encoding the first 50 aa of SHMT fused in-frame with the lacZ gene (Lorenz & Stauffer, 1995). {beta}-Galactosidase activity (Table 4) in strain GS162{lambda}glyAlacZ transformed with pKK233-2 and pCP994 paralleled SHMT activity (Table 3). In all media tested, the presence of pCP994 elicited 1·5- to 3-fold more {beta}-galactosidase activity than the control plasmid pKK233-2. A similar but smaller increase (1·3–1·4-fold) in {beta}-galactosidase activity was observed with pSD6P transformants compared to pUC18 transformants (data not shown).


View this table:
[in this window]
[in a new window]
 
Table 4. Expression of {beta}-galactosidase from the {lambda}glyAlacZ fusion

Transformants were grown in the media indicated. {beta}-Galactosidase activities were normalized against the activity of control plasmid pKK233-2 transformants that were grown without FEP. One hundred per cent activity for strain GS162{lambda}glyAlacZ was 7100 Miller units (Miller, 1992); for strain MH901, 8750 Miller units; and for strain MH902, 12 000 Miller units. The SD was less than 15 % in all samples. Data are from two representative experiments in which two cultures of each transformant were assayed in duplicate. Strain GS162{lambda}glyAlacZ was assayed under all growth conditions in one experiment, and strains MH901 and MH902 were assayed in the other. Experiments were reproduced at least twice.

 
Both the SHMT activity assay and the assay of {beta}-galactosidase from the glyAlacZ fusion indicated that SHMT activity was increased by the presence of csgD or the chimaeric protein. This was not simply caused by overproduction of any protein. We observe suppression only with some recombinant plasmids although many would be expected to overexpress protein. Moreover, deliberate high-level expression of luciferase in E. coli actually reduces glyA expression (Oh & Liao, 2000).

Increased SHMT activity was sufficient for growth of strain MH829 on FEP-Gly
Several lines of evidence indicate that increasing the SHMT activity above a threshold level is sufficient to enable strain MH829 to make enough glycine and methylene THF to grow on FEP-Gly. SHMT activity could be increased by transforming cells with the glyA-containing plasmid, pGS29 (Stauffer et al., 1981), by inactivating the PurR repressor by mutation of purR (Steiert et al., 1990), or by growth in FEP-His-Gly (Table 4). Strain MH829 grows on FEP-His-Gly (Herrington & Chirwa, 1999) and grew on FEP-Gly when transformed with pGS29 or made purR- (Table 2), indicating that the increase in SHMT activity was sufficient to compensate for the lack of glycine in the medium.

Are there alternative sources of glycine in cells transformed with a csgD-containing plasmid?
Glycine can also be produced from threonine via threonine dehydrogenase (Fraser & Newman, 1975; Ravnikar & Somerville, 1987) or can be spared by reducing glycine cleavage (Stauffer, 1996). We assayed threonine dehydrogenase activity and glycine-cleavage activity in transformants of MH829 and {beta}-galactosidase activity in transformants of GS162{lambda}glyTlacZ. These activities were very low, and there were no reproducible differences between control and csgD transformants when transformants were grown in FEP-Gly (data not shown). Therefore, glycine was not provided by either of these routes.

Are curli produced in transformants?
Curli are not expressed by most laboratory strains of E. coli, but expression of csgD from a plasmid induces their formation (Prigent-Combaret et al., 1999). We tested for curli formation by spotting cells on Congo red plates (Hammar et al., 1995), by measuring Congo red binding (Gophna et al., 2001) and by monitoring binding to polystyrene (Prigent-Combaret et al., 1999). On Congo red plates, spots of strains MH829 ({Delta}folA : : kan3) and MH828 (folA+) were white, indicating that they do not produce curli (data not shown). Both strains transformed with the control plasmids pUC18 and pCP994 pKK233-2 were white, whereas pSD6P and pCP994 transformants were red (Fig. 2c and data not shown). Strain MH828 transformed with control plasmids pUC18 and pKK233-2 bound 49±6·5 and 47±0·1 Congo red units, respectively, whereas pSD6P and pCP994 transformants bound 67±7·0 and 96±32 units, respectively. The csgD plasmids also promoted binding to plastic whereas the control plasmids did not (data not shown). These results indicate that our strains made curli when the csgD plasmids were present.

Does strain MH829 have to produce curli to grow on FEP-Gly?
To determine whether the ability to make curli was necessary for growth on FEP-Gly or increasing glyA expression we constructed csgA- and csgG- derivatives of strains MH829 and GS162{lambda}glyAlacZ. Both genes are essential for curli biogenesis. Transformants of strains MH910({Delta}folA csgA) and MH911({Delta}folA csgG) were tested for growth on FEP-Gly (Table 2). The csg mutations did not affect growth. {beta}-Galactosidase activity was higher in pCP994 transformants of the glyAlacZ fusion strains MH901 and MH902 than in control transformants (Table 4). We conclude that formation of curli fibres was not necessary for the CsgD-mediated suppression of glycine auxotrophy or elevation of glyA expression.

Is chromosomally encoded CsgD involved?
To determine if the chromosomal csgD gene was required for suppression by the plasmids pSD6P and pCP994, we constructed csgD- derivatives of strains MH917 and GS162{lambda}glyAlacZ, and transformed the resultant strains with pSD6P, pCP94 and control plasmids. The pCP994 transformants of strain MH937 ({Delta}folA : : aadA csgD-) grew as well as transformants of MH829 (Table 2) and were red on Congo red plates (data not shown). In contrast, pSD6P transformants grew poorly and did not stain as strongly on the Congo red plates. The activity of {beta}-galactosidase in pCP994 transformants of the csgD- reporter strain MH938 was higher than in the controls (Table 5) and the magnitude of the increase was similar to that observed in the csgD+ strain GS162{lambda}glyAlacZ (Table 4). In contrast, pSD6P enhanced the activity slightly in two of the three growth conditions tested (Table 5). Taken together, these results suggest that the chimaeric protein coded by pSD6P is not sufficient for optimum suppression, curli formation and glyA expression, and that it influences the expression or activity of the chromosomally encoded CsgD protein.


View this table:
[in this window]
[in a new window]
 
Table 5. {beta}-Galactosidase activities in strain MH938

Transformants were grown in the media indicated. {beta}-Galactosidase activities were normalized against the activity of control plasmid transformants grown without FEP. One hundred per cent activity for strain MH938(pUC18) was 3460 Miller units (Miller, 1992); for strain MH938(pKK233-2), 3750 Miller units. The SD was less than 15 % in all samples. Data are from a representative experiment in which two cultures of each transformant were assayed in duplicate. The experiment was reproduced.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Multicopy plasmids expressing either the intact csgD gene, a chimaeric form of csgD or the glyA gene suppressed the glycine auxotrophy of strain MH829. These plasmids all increase the amount of SHMT in the cell. Strain MH829 lacks DHFR activity and is auxotrophic for glycine when histidine, adenine, thymidine, methionine and pantothenate are present (Herrington & Chirwa, 1999). In spite of the lack of DHFR, folA mutants can still produce some THF (Hamm-Alvarez et al., 1990; Herrington & Chirwa, 1999). The distribution of reduced folates in the cell presumably depends on the growth conditions. In FEP-Gly, methyl-THF and formyl-THF pools could be elevated relative to the THF pool because these cofactors would not be turned over via methionine and purine synthesis. This could limit synthesis of glycine, and restrict protein synthesis, further limiting turnover of formyl-THF. An increase in the amount of SHMT could increase glycine synthesis, thereby enhancing protein synthesis and turnover of formyl-THF.

The glycine auxotrophy of strain MH829 can also be alleviated by mutations resulting in curli proficiency. Curli-proficient mutants of strain MH829 grew on FEP-Gly, and some mutants selected for growth on FEP-Gly were curli-proficient (T. MacRae, C. Zambrana, N. T. Chirwa and M. B. Herrington, unpublished results). We have not yet tested whether glyA is upregulated in these mutants, but it is reasonable to expect that some could express higher levels of SHMT.

We propose that upregulation of glyA is an integral response to signals eliciting curli formation. Cells make curli in stationary phase, suggesting that the demand for amino acids is high at a time when their availability could be limited. Glycine makes up 11·5 % of the amino acid residues in total E. coli protein (Neidhardt & Umbarger, 1996). Curlin, the major protein of curli, contains 1·7 times as much glycine, and cells make a lot of curli, suggesting the demand for glycine is particularly high. Thus, increasing the cell's ability to make glycine could facilitate its production of curli.

Expression of csgD is highly regulated. A variety of environmental factors including nutrient limitation, low osmotic strength, low temperature, microaerophilic conditions and iron limitation increase expression of csgD or its Salty homologue, agfD (Römling et al., 1998b; Prigent-Combaret et al., 2001; Gerstel & Römling, 2001). At least four regulatory proteins (RpoS, OmpR, Cpx and Mlr) control CsgD expression (Arnqvist et al., 1994; Brown et al., 2001; Prigent-Combaret et al., 2001) (Fig. 4). AgfD is detectable by Western blotting in late exponential and stationary phases, but transcript measurements and reporter constructs suggest low levels of AgfD and CsgD are made even during exponential phase (Prigent-Combaret et al., 2001; Gerstel & Römling, 2001).



View larger version (22K):
[in this window]
[in a new window]
 
Fig. 4. Regulatory roles of CsgD. CsgD[L] represents the protein with bound ligand and CsgD[L] the protein with no ligand. Gene names in the dashed boxes indicate genes partially or completely dependent on RpoS for transcription (Olsén et al., 1993b; De Wulf et al., 1999; Römling et al., 2000; Brown et al., 2001). Solid arrows are used to indicate gene–protein relationships and the two states of CsgD. Patterned arrows indicate activation and patterned lines with a bar represent repression. Patterns are used to differentiate between known regulatory circuits that have been identified by mutants (Hammar et al., 1995; Römling et al., 2000; Brown et al., 2001; Zogaj et al., 2001) or by mutants and DNA binding studies (Stauffer, 1996; Dorel et al., 1999; Prigent-Combaret et al., 2001) and possible circuits proposed in this study. Dotted lines, mutants; dash–dot lines, hypothetical; dashed lines, mutants and DNA binding.

 
CsgD is a homologue of the two-component FixJ/UhpA/LuxR family of proteins characterized by a DNA binding (helix–turn–helix) domain in the C-terminal region and a receiver domain in the N-terminal region of the proteins. Depending on the protein, the receiver domain contains a phosphorylation site or binds a small molecule. The helix–turn–helix domain of CsgD is a good fit to the consensus (obtained from the Conserved Domain Database; Marchler-Bauer et al., 2002) for the family. The chimaeric protein made by pSD6P does not contain the domain (Fig. 1), indicating that glyA activation and curli formation do not involve the binding of the chimaeric protein to DNA. This suggests CsgD activation could involve both protein–DNA and protein–protein interactions. Comparison of AgfD with homologues suggests that it does not contain a phosphorylation site similar to that of the homologues (Römling et al., 2000).

In the cell, CsgD presumably can have ligand bound to it (CsgD[L]) or not (CsgD[L]). We propose that CsgD[ ] activates transcription of target genes whereas CsgD[L] is inactive (Fig. 4). Normally, growing cells contain very little CsgD (Prigent-Combaret, 2001) so CsgD[L] predominates. In curli-proficient cells, activation of csgDEFG transcription by OmpR, MlrA or other activators results in increased amounts of CsgD so that the active form predominates. CsgD[L] then activates transcription of csgBA, adrA and possibly other genes. In curli non-proficient cells, CsgD[L] predominates because either ligand levels are high or csgD expression remains very low.

When curli non-proficient cells are transformed with pCP994, the amount of CsgD made is higher than the amount of ligand so CsgD[L] will activate curli synthesis and glyA expression. In cells transformed with pSD6P, the chimaeric protein will bind the ligand and there will be a mixture of chimaeric protein and chromosomally encoded CsgD in the activator form. The net activation will be weaker than in cells transformed with pCP994 because the chimaeric protein is a poor activator. This is demonstrated by eliminating CsgD. In the csgD null mutant, the response to pSD6P was weaker than in the csgD+ strain.

How does CsgD regulate glyA expression? CsgD[L] could activate glyA directly. Alternatively, it could alter known regulation either by binding one of the ligands or by altering expression of one of the regulatory proteins. MetR and its co-activator, homocysteine, activates transcription of glyA and PurR and its co-repressor, guanine or inosine, represses transcription. Other, as yet unidentified, factors also influence glyA transcription ((Stauffer, 1996). A search of the Conserved Domain Database (Marchler-Bauer, 2002) using the N-terminal region of CsgD returned no hits, suggesting that the receiver domain does not resemble any proteins known to bind homocysteine, guanine, inosine or a related molecule. This suggests that CsgD does not titrate out a known ligand needed for glyA expression. CsgD[L] could enhance expression of glyA by activating metR or repressing purR (Fig. 4).

Our results are the first to demonstrate that CsgD alters the expression of a gene not directly involved in multicellular behaviour. Our observation that a chimaeric protein containing only 70 aa of CsgD is sufficient to stimulate curli formation is of interest because it defines a functionally important region of the protein.


   ACKNOWLEDGEMENTS
 
The authors are grateful for financial support from the Natural Sciences and Engineering Council of Canada (grant no. OGP0006727 to M. B. H.) and from Concordia University.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Ahrweiler, P. & Frieden, C. (1988). Construction of a fol mutant strain of Escherichia coli for use in dihydrofolate reductase mutagenesis experiments. J Bacteriol 170, 3301–3304.[Medline]

Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 3389–3402.[Abstract/Free Full Text]

Amann, E. & Brosius, J. (1985). ‘ATG vectors' for regulated high-level expression of cloned genes in Escherichia coli. Gene 40, 183–190.[CrossRef][Medline]

Arnqvist, A., Olsén, A., Pfeifer, J., Russell, D. G. & Normark, S. (1992). The Crl protein activates cryptic genes for curli formation and fibronectin binding in Escherichia coli HB101. Mol Microbiol 6, 2443–2452.[Medline]

Arnqvist, A., Olsén, A. & Normark, S. (1994). Sigma S-dependent growth-phase induction of the csgBA promoter in Escherichia coli can be achieved in vivo by sigma 70 in the absence of the nucleoid-associated protein H-NS. Mol Microbiol 13, 1021–1032.[Medline]

Austin, J. W., Sanders, G., Kay, W. W. & Collinson, S. K. (1998). Thin aggregative fimbriae enhance Salmonella enteritidis biofilm formation. FEMS Microbiol Lett 162, 295–301.[CrossRef][Medline]

Baikalov, I., Schroder, I., Kaczor-Grzeskowiak, M., Grzeskowiak, K., Gunsalus, R. P. & Dickerson, R. E. (1996). Structure of the Escherichia coli response regulator NarL. Biochemistry 35, 11053–11061.[CrossRef][Medline]

Ben Nasr, A., Olsén, A., Sjobring, U., Muller-Esterl, W. & Bjorck, L. (1996). Assembly of human contact phase proteins and release of bradykinin at the surface of curli-expressing Escherichia coli. Mol Microbiol 20, 927–935.[Medline]

Bian, Z., Brauner, A., Li, Y. & Normark, S. (2000). Expression of and cytokine activation by Escherichia coli curli fibers in human sepsis. J Infect Dis 181, 602–612.[CrossRef][Medline]

Brown, P. K., Dozois, C. M., Nickerson, C. A., Zuppardo, A., Terlonge, J. & Curtiss, R., 3rd (2001). MlrA, a novel regulator of curli (AgF) and extracellular matrix synthesis by Escherichia coli and Salmonella enterica serovar Typhimurium. Mol Microbiol 41, 349–363.[CrossRef][Medline]

Brown, R. E., Jarvis, K. L. & Hyland, K. J. (1989). Protein measurement using bicinchoninic acid: elimination of interfering substances. Anal Biochem 180, 136–139.[Medline]

Chapman, M. R., Robinson, L. S., Pinkner, J. S., Roth, R., Heuser, J., Hammar, M., Normark, S. & Hultgren, S. J. (2002). Role of Escherichia coli curli operons in directing amyloid fiber formation. Science 295, 851–855.[Abstract/Free Full Text]

Chung, C. T., Niemela, S. L. & Miller, R. H. (1989). One-step preparation of competent Escherichia coli: transformation and storage of bacterial cells in the same solution. Proc Natl Acad Sci U S A 86, 2172–2175.[Abstract]

Collinson, S. K., Emody, L., Muller, K. H., Trust, T. J. & Kay, W. W. (1991). Purification and characterization of thin, aggregative fimbriae from Salmonella enteritidis. J Bacteriol 173, 4773–4781.[Medline]

Collinson, S. K., Clouthier, S. C., Doran, J. L., Banser, P. A. & Kay, W. W. (1996). Salmonella enteritidis agfBAC operon encoding thin, aggregative fimbriae. J Bacteriol 178, 662–667.[Abstract]

Collinson, S. K., Clouthier, S. C., Doran, J. L., Banser, P. A. & Kay, W. W. (1997). Characterization of the agfBA fimbrial operon encoding thin aggregative fimbriae of Salmonella enteritidis. Adv Exp Med Biol 412, 247–248.[Medline]

Corpet, F. (1988). Multiple sequence alignment with hierarchical clustering. Nucleic Acids Res 16, 10881–10890.[Abstract]

Csonka, L. N. & Clark, A. J. (1979). Deletions generated by the transposon Tn10 in the srl recA region of the Escherichia coli K-12 chromosome. Genetics 93, 321–343.[Abstract/Free Full Text]

De Wulf, P., Kwon, O. & Lin, E. C. (1999). The CpxRA signal transduction system of Escherichia coli: growth-related autoactivation and control of unanticipated target operons. J Bacteriol 181, 6772–6778.[Abstract/Free Full Text]

Dorel, C., Vidal, O., Prigent-Combaret, C., Vallet, I. & Lejeune, P. (1999). Involvement of the Cpx signal transduction pathway of E. coli in biofilm formation. FEMS Microbiol Lett 178, 169–175.[CrossRef][Medline]

Ducros, V. M., Lewis, R. J., Verma, C. S., Dodson, E. J., Leonard, G., Turkenburg, J. P., Murshudov, G. N., Wilkinson, A. J. & Brannigan, J. A. (2001). Crystal structure of GerE, the ultimate transcriptional regulator of spore formation in Bacillus subtilis. J Mol Biol 306, 759–771.[CrossRef][Medline]

Fraser, J. & Newman, E. B. (1975). Derivation of glycine from threonine in Escherichia coli K-12 mutants. J Bacteriol 122, 810–817.[Medline]

Gerstel, U. & Römling, U. (2001). Oxygen tension and nutrient starvation are major signals that regulate agfD promoter activity and expression of the multicellular morphotype in Salmonella typhimurium. Environ Microbiol 3, 638–648.[CrossRef][Medline]

Gophna, U., Barlev, M., Seijffers, R., Oelschlager, T. A., Hacker, J. & Ron, E. Z. (2001). Curli fibers mediate internalization of Escherichia coli by eukaryotic cells. Infect Immun 69, 2659–2665.[Abstract/Free Full Text]

Greene, R. C. & Radovich, C. (1975). Role of methionine in the regulation of serine hydroxymethyltransferase in Escherichia coli. J Bacteriol 124, 269–278.[Medline]

Groisman, E. A. & Casadaban, M. J. (1987). Cloning of genes from members of the family Enterobacteriaceae with mini-Mu bacteriophage containing plasmid replicons. J Bacteriol 169, 687–693.[Medline]

Hamm-Alvarez, S. F., Sancar, A. & Rajagopalan, K. V. (1990). The presence and distribution of reduced folates in Escherichia coli dihydrofolate reductase mutants. J Biol Chem 265, 9850–9856.[Abstract/Free Full Text]

Hammar, M., Arnqvist, A., Bian, Z., Olsén, A. & Normark, S. (1995). Expression of two csg operons is required for production of fibronectin- and congo red-binding curli polymers in Escherichia coli K-12. Mol Microbiol 18, 661–670.[Medline]

Hammar, M., Bian, Z. & Normark, S. (1996). Nucleator-dependent intercellular assembly of adhesive curli organelles in Escherichia coli. Proc Natl Acad Sci U S A 93, 6562–6566.[Abstract/Free Full Text]

Herrington, M. B. & Chirwa, N. T. (1999). Growth properties of a folA null mutant of Escherichia coli K12. Can J Microbiol 45, 191–200.[CrossRef][Medline]

Howell, E. E., Foster, P. G. & Foster, L. M. (1988). Construction of a dihydrofolate reductase-deficient mutant of Escherichia coli by gene replacement. J Bacteriol 170, 3040–3045.[Medline]

Jourdan, A. D. & Stauffer, G. V. (1998). Mutational analysis of the transcriptional regulator GcvA: amino acids important for activation, repression, and DNA binding. J Bacteriol 180, 4865–4871.[Abstract/Free Full Text]

Kahn, D. & Ditta, G. (1991). Modular structure of FixJ: homology of the transcriptional activator domain with the -35 binding domain of sigma factors. Mol Microbiol 5, 987–997.[Medline]

Kilstrup, M., Meng, L. M., Neuhard, J. & Nygaard, P. (1989). Genetic evidence for a repressor of synthesis of cytosine deaminase and purine biosynthesis enzymes in Escherichia coli. J Bacteriol 171, 2124–2127.[Medline]

Krishnan, B. R. & Berg, D. E. (1993). Viability of folA-null derivatives of wild-type (thyA+) Escherichia coli K-12. J Bacteriol 175, 909–911.[Abstract]

La Ragione, R. M., Collighan, R. J. & Woodward, M. J. (1999). Non-curliation of Escherichia coli O78 : K80 isolates associated with IS1 insertion in csgB and reduced persistence in poultry infection. FEMS Microbiol Lett 175, 247–253.[CrossRef][Medline]

Loferer, H., Hammar, M. & Normark, S. (1997). Availability of the fibre subunit CsgA and the nucleator protein CsgB during assembly of fibronectin-binding curli is limited by the intracellular concentration of the novel lipoprotein CsgG. Mol Microbiol 26, 11–23.[Medline]

Lorenz, E. & Stauffer, G. V. (1995). Characterization of the MetR binding sites for the glyA gene of Escherichia coli. J Bacteriol 177, 4113–4120.[Abstract]

Mansouri, A., Decter, J. B. & Silber, R. (1972). Studies on the regulation of one-carbon metabolism. II. Repression-derepression of serine hydroxymethyltransferase by methionine in Escherichia coli 113-3. J Biol Chem 247, 348–352.[Abstract/Free Full Text]

Marchler-Bauer, A., Panchenko, A. R., Shoemaker, B. A., Thiessen, P. A., Geer, L. Y. & Bryant, S. H. (2002). CDD: a database of conserved domain alignments with links to domain three-dimensional structure. Nucleic Acids Res 30, 281–283.[Abstract/Free Full Text]

Matthews, R. G. (1996). One-carbon metabolism. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 600–611. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.

Maurer, J. J., Brown, T. P., Steffens, W. L. & Thayer, S. G. (1998). The occurrence of ambient temperature-regulated adhesins, curli, and the temperature-sensitive hemagglutinin tsh among avian Escherichia coli. Avian Dis 42, 106–118.[Medline]

Messing, J. (1983). New M13 vectors for cloning. Methods Enzymol 101, 20–78.[Medline]

Miller, B. A. & Newman, E. B. (1974). Control of serine transhydroxymethylase synthesis in Escherichia coli K12. Can J Microbiol 20, 41–47.[Medline]

Miller, J. H. (1992). A Short Course in Bacterial Genetics: a Laboratory Manual and Handbook for Escherichia coli and Related Bacteria. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Neidhardt, F. C. & Umbarger, H. E. (1996). Chemical composition of Escherichia coli. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 13–16. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.

Oh, M. K. & Liao, J. C. (2000). DNA microarray detection of metabolic responses to protein overproduction in Escherichia coli. Metab Eng 2, 201–209.[CrossRef][Medline]

Olsén, A., Jonsson, A. & Normark, S. (1989). Fibronectin binding mediated by a novel class of surface organelles on Escherichia coli. Nature 338, 652–655.[CrossRef][Medline]

Olsén, A., Arnqvist, A., Hammar, M. & Normark, S. (1993a). Environmental regulation of curli production in Escherichia coli. Infect Agents Dis 2, 272–274.[Medline]

Olsén, A., Arnqvist, A., Hammar, M., Sukupolvi, S. & Normark, S. (1993b). The RpoS sigma factor relieves H-NS-mediated transcriptional repression of csgA, the subunit gene of fibronectin-binding curli in Escherichia coli. Mol Microbiol 7, 523–536.[Medline]

Olsén, A., Wick, M. J., Morgelin, M. & Bjorck, L. (1998). Curli, fibrous surface proteins of Escherichia coli, interact with major histocompatibility complex class I molecules. Infect Immun 66, 944–949.[Abstract/Free Full Text]

Prigent-Combaret, C., Vidal, O., Dorel, C. & Lejeune, P. (1999). Abiotic surface sensing and biofilm-dependent regulation of gene expression in Escherichia coli. J Bacteriol 181, 5993–6002.[Abstract/Free Full Text]

Prigent-Combaret, C., Prensier, G., Le Thi, T. T., Vidal, O., Lejeune, P. & Dorel, C. (2000). Developmental pathway for biofilm formation in curli-producing Escherichia coli strains: role of flagella, curli and colanic acid. Environ Microbiol 2, 450–464.[CrossRef][Medline]

Prigent-Combaret, C., Brombacher, E., Vidal, O., Ambert, A., Lejeune, P., Landini, P. & Dorel, C. (2001). Complex regulatory network controls initial adhesion and biofilm formation in Escherichia coli via regulation of the csgD gene. J Bacteriol 183, 7213–7223.[Abstract/Free Full Text]

Ravnikar, P. D. & Somerville, R. L. (1987). Genetic characterization of a highly efficient alternate pathway of serine biosynthesis in Escherichia coli. J Bacteriol 169, 2611–2617.[Medline]

Römling, U., Bian, Z., Hammar, M., Sierralta, W. D. & Normark, S. (1998a). Curli fibers are highly conserved between Salmonella typhimurium and Escherichia coli with respect to operon structure and regulation. J Bacteriol 180, 722–731.[Abstract/Free Full Text]

Römling, U., Sierralta, W. D., Eriksson, K. & Normark, S. (1998b). Multicellular and aggregative behaviour of Salmonella typhimurium strains is controlled by mutations in the agfD promoter. Mol Microbiol 28, 249–264.[CrossRef][Medline]

Römling, U., Rohde, M., Olsén, A., Normark, S. & Reinkoster, J. (2000). AgfD, the checkpoint of multicellular and aggregative behaviour in Salmonella typhimurium regulates at least two independent pathways. Mol Microbiol 36, 10–23.[CrossRef][Medline]

Sakellaris, H., Hannink, N. K., Rajakumar, K., Bulach, D., Hunt, M., Sasakawa, C. & Adler, B. (2000). Curli loci of Shigella spp. Infect Immun 68, 3780–3783.[Abstract/Free Full Text]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sitnikov, D. M., Schineller, J. B. & Baldwin, T. O. (1995). Transcriptional regulation of bioluminesence genes from Vibrio fischeri. Mol Microbiol 17, 801–812.[Medline]

Sjobring, U., Pohl, G., & Olsén, A. (1994). Plasminogen, absorbed by Escherichia coli expressing curli or by Salmonella enteritidis expressing thin aggregative fimbriae, can be activated by simultaneously captured tissue-type plasminogen activator (t-PA). Mol Microbiol 14, 443–452.[Medline]

Solo-Gabriele, H. M., Wolfert, M. A., Desmarais, T. R. & Palmer, C. J. (2000). Sources of Escherichia coli in a coastal subtropical environment. Appl Environ Microbiol 66, 230–237.[Abstract/Free Full Text]

Souza, V., Rocha, M., Valera, A. & Eguiarte, L. E. (1999). Genetic structure of natural populations of Escherichia coli in wild hosts on different continents. Appl Environ Microbiol 65, 3373–3385.[Abstract/Free Full Text]

Stauffer, G. V., Plamann, M. D. & Stauffer, L. T. (1981). Construction and expression of hybrid plasmids containing the Escherichia coli glyA genes. Gene 14, 63–72.[CrossRef][Medline]

Stauffer, G. V. (1996). Biosynthesis of serine, glycine and one-carbon units. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 506–513. Edited by F. C. Neidhardt. Washington, DC: American Society for Microbiology.

Steiert, J. G., Rolfes, R. J., Zalkin, H. & Stauffer, G. V. (1990). Regulation of the Escherichia coli glyA gene by the purR gene product. J Bacteriol 172, 3799–3803.[Medline]

Sutcliffe, J. G. (1978). Nucleotide sequence of the ampicillin resistance gene of Escherichia coli plasmid pBR322. Proc Natl Acad Sci U S A 75, 3737–3741.[Abstract]

Taylor, R. T. & Weissbach, H. (1965). Radioactive assay for serine transhydroxymethylase assay. Anal Biochem 13, 80–84.

Uhlich, G. A., Keen, J. E. & Elder, R. O. (2001). Mutations in the csgD promoter associated with variations in curli expression in certain strains of Escherichia coli O157 : H7. Appl Environ Microbiol 67, 2367–2370.[Abstract/Free Full Text]

Uhlich, G. A., Keen, J. E. & Elder, R. O. (2002). Variations in the csgD promoter of Escherichia coli O157 : H7 associated with increased virulence in mice and increased invasion of HEp-2 cells. Infect Immun 70, 395–399.[Abstract/Free Full Text]

Vidal, O., Longin, R., Prigent-Combaret, C., Dorel, C., Hooreman, M. & Lejeune, P. (1998). Isolation of an Escherichia coli K-12 mutant strain able to form biofilms on inert surfaces: involvement of a new ompR allele that increases curli expression. J Bacteriol 180, 2442–2449.[Abstract/Free Full Text]

Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103–119.[CrossRef][Medline]

Zogaj, X., Nimtz, M., Rohde, M., Bokranz, W. & Römling, U. (2001). The multicellular morphotypes of Salmonella typhimurium and Escherichia coli produce cellulose as the second component of the extracellular matrix. Mol Microbiol 39, 1452–1463.[CrossRef][Medline]

Received 28 June 2002; revised 25 October 2002; accepted 30 October 2002.