Identication of a Two-Component System Important for Cell Division of the Rice Pathogen Burkholderia glumae in Response to Nutrient Conditions

Bacterial two-component regulatory systems control the expression of sets of genes to coordinate physiological functions in response to environmental cues. Here, we report that a genetically linked but functionally unpaired two-component system comprising the sensor kinase GluS (BGLU_1G13350) and the response regulator GluR (BGLU_1G13360) is critical for cell division in the rice pathogen Burkholderia glumae BGR1. The gluR null mutant, unlike the gluS mutant, formed lamentous cells in Luria–Bertani medium and was sensitive to exposure to 42°C. Expression of genes responsible for cell division and cell-wall (dcw) biosynthesis in the gluR mutant was elevated compared to the wild type, resulting in an imbalance between FtsZ and FtsA. GluR-His bound to the putative promoter regions of ftsA and ftsZ, indicating that repression of genes in the dcw cluster by GluR is critical for cell division in B. glumae. The gluR mutant did not form lamentous cells in M9 minimal medium, whereas exogenous addition of glutamine or glutamate to the medium induced lamentous cell formation. These results implicate glutamine and glutamate as external stimuli that modulate GluR-mediated cell division in B. glumae. Therefore, GluR controls cell division of B. glumae in a nutrition-dependent manner.


Introduction
Two-component systems (TCS) consisting of a sensor kinase and a cognate response regulator are common in bacteria 1 . They are essential for the responses of bacteria to changes in environmental factors such as pH, osmotic pressure, antibiotics, and quorum-sensing (QS) signals 1 . The sensor kinases are autophosphorylated after sensing an environmental stimulus, which is followed by phosphotransfer from the phosphorylated sensor kinases to the response regulators 1 . The phosphorylated response regulators then undergo conformational changes to become active, thereby controlling the expression of target genes 1 . The genes encoding sensor kinases and response regulators are often genetically linked in bacterial genomes and functionally paired 2,3 . In addition to paired TCSs, sensor kinases and transcriptional regulators can crosstalk, thus modulating multiple biological processes in response to environmental signals irrespective of their genetic linkage 2,4−6 .
We study the social behavior and host interactions of the rice bacterial pathogen Burkholderia glumae, the cause of rice panicle blight 7,8 . A phytotoxin, toxo avin, is the major virulence factor of B. glumae and exerts a toxic effect on photosynthetic organisms by generating radicals under light 8,9 . The virulencefactor biosynthesis and motility of B. glumae are dependent on QS 8,10 . As well as QS, we are interested in TCSs in B. glumae BGR1 because they coordinate and regulate the expression of genes critical for adaptation to stress, survival, tness in the host, and virulence [11][12][13][14][15][16][17] . For instance, CpxAR of Actinobacillus pleuropneumoniae 14 , ArcAB of Escherichia coli 15 , and KdpDE 16 and PhoPQ 17 in a variety of bacterial taxa reportedly promote growth, tness, and survival in the host. In addition, AgrAC, SsrAB, SaeRS, and ArlRS of Staphylococcus aureus and BvgAS of Bordetella pertussis are necessary for virulence [11][12][13] . Few studies have focused on TCSs in B. glumae, probably because of concern over repeating works on other pathogens. However, Karki et al. reported that the PidS/PidR TCS is essential for the pigmentation and virulence of B. glumae 411gr-6 18 .
In this study, we identi ed a TCS composed of the sensor kinase GluS and the response regulator GluR, which was critical for normal cell division in B. glumae BGR1. gluR and gluS were cotranscribed, but GluR functioned independently of GluS in normal cell division. We report that GluR regulates the gene cluster involved in cell division and cell wall (dcw) biosynthesis to maintain balanced expression of FtsZ and FtsA for normal septum formation 19 . We conclude that external nutritional conditions modulate cell division in a TCS-dependent manner in B. glumae. These ndings provide insight into how the recognition of external signals by TCS affects the sophisticated molecular mechanisms involved in controlling bacterial cell division.

Results
Identi cation of a TCS critical for normal cell division of B. glumae BGR1. To identify a key TCS important for normal cell division of B. glumae BGR1, we rst mutagenized it with mini-Tn5 and examined the morphology of the mutants. The mutant RT271 formed lamentous cells when grown in Luria-Bertani medium (LB) (Fig. 1a). To determine the insertion site of mini-Tn5 in the RT271 mutant, a mini-Tn5 insertion along with anking sequences was rescued by digestion of its genomic DNA with EcoRI, self-ligation, and transformation into E. coli DH5α. Flanking sequences of mini-Tn5 from the rescued plasmid pRT271E revealed that an annotated gene BGLU_1G13360 had an insertional mutation (Fig. 1b). This gene, gluR, encoded a 27.7 kDa protein that exhibited 99.6% similarity to known OmpR-type response regulators such as BURPS305_7006 in B. pseudomallei 305, RisA (BMA10247_1253) in B. mallei NCTC 10247, and BCENMCO3_1962 of B. cenocepacia MCO-3 ( Supplementary Fig. S1a). Downstream of gluR was a putative sensor kinase, gluS (BGLU_1G13350), (Fig. 1b) S1b).
Aberrant cell division due to a mutation in gluR. To determine whether the insertion of Tn3-gusA in gluR or gluS conferred a similar cell morphology to the RT217 mutant, we observed the morphology of the gluR and gluS mutants under a light microscope. The gluR mutant BGLUR133 showed extensive lamentous cells in LB medium (Fig. 2), consistent with the initial phenotype of the gluR::min-Tn5 mutant RT271 in LB ( Fig. 1a). However, the gluS mutant BGLUS35 formed normal cells in LB medium (Fig. 2). The gluR mutant BGLUR133 maintained a normal rod-shaped cell morphology similar to that of the gluS mutant BGLUS35 in M9 minimal medium (Fig. 2). Transmission electron microscopy (TEM) of ultrathin sections of the gluR mutant BGLUR133 revealed characteristic features of lamentous cells with multiple nuclei and indents along the cell membrane at points where the septum would have formed to separate dividing cells (Fig. 3a). The genetically complemented strain of the gluR mutant BGLUR133 with pBGH13, BGLUR133C, had morphologically uniform rod-shaped cells (Fig. 3a). The growth of the gluR mutant BGLUR133 and the wild-type BGR1 for 30 hours in LB medium was similar ( Supplementary Fig. S2,   Fig. 3b). Although lamentous cells of the gluR mutant BGLUR133 remained viable for 30 hours, their abundance decreased after 18 hours (Fig. 3b).
Direct control of genes involved in cell division by GluR. Because TEM suggested the involvement of GluR in cell division, we determined whether GluR in uences the expression of genes in the dcw cluster involved in cell division. In B. glumae, there were 15 annotated genes; e.g., ftsA, ftsI, ftsL, ftsQ, ftsW, and ftsZ in the dcw cluster and ftsB and ftsK in other regions (Fig. 4a, b). The expression levels of ftsA, ftsB, ftsI, ftsK, ftsL, ftsQ, ftsW, and ftsZ in the gluR mutant BGLUR133 were signi cantly increased compared to those in the wild-type BGR1 (Fig. 4c). The expression levels of the eight genes in BGLUR133C were similar to those in the wild type (Fig. 4c). To determine whether GluR directly controls their expression, we performed electrophoretic mobility shift assays (EMSA) on the putative promoter regions of ftsA and ftsZ and puri ed His-tagged GluR (GluR-His). The binding of GluR-His to the putative promoter regions of ftsA and ftsZ con rmed that GluR-His directly represses the expression of cell division genes in B. glumae (Fig. 4d).
Alleviation of aberrant cell morphology by constitutive expression of ftsA in the gluR mutant BGLUR133. Because the FtsA to FtsZ ratio is critical for normal bacterial cell division, we evaluated the role of GluR in its maintenance. Taking the expression levels of ftsA and ftsZ in the wild type as 1.00, the expression levels of these two genes were 1.21 and 1.67, respectively, in the gluR mutant BGLUR133 (Fig. 4c). To con rm that imbalanced expression of ftsA and ftsZ causes abnormal cell division, we constitutively expressed ftsA under the control of the trc promoter in the wild type, the gluR mutant BGLUR133, and the complemented strain BGLUR133C; the resulting strains were designated BGR1(pFtsA), BGLUR133(pFtsA), and BGLUR133C(pFtsA), respectively. Cells of BGLUR133(pFtsA) showed normal cell division as well as a 5.2fold increase in ftsA expression (Fig. 5). Moreover, ftsA expression was increased more than 100fold in BGR1(pFtsA) and BGLUR133C(pFtsA), whose cells underwent abnormal division (Fig. 5).
In uence of glutamate and glutamine on GluR-mediated control of cell division. Because the gluR mutant BGLUR133 formed lamentous cells in LB medium but not in M9 minimal medium, we reasoned that the amino acids in LB medium might be the cause of lamentous cell formation. Therefore, we added 10% casamino acids to M9 minimal medium to evaluate their in uence on the morphology of the gluR mutant BGLUR133. Adding casamino acids to M9 minimal medium transformed the morphologically normal cells of the gluR mutant BGLUR133 into lamentous cells (Fig. 6). To identify the amino acid(s) responsible for triggering lamentous cells in the gluR mutant BGLUR133, 20 amino acids were individually added to M9 minimal medium. Of the 20 amino acids, only glutamine and glutamate individually or in combination triggered cells of the gluR mutant BGLUR133 to become lamentous in M9 minimal medium (Fig. 7a, b, c). When 13 amino acids excluding glutamine and its six amino-acid metabolites (glutamate, serine, alanine, proline, aspartate, and asparagine) were added to M9 minimal medium, the gluR mutant BGLUR133 maintained a normal morphology (Fig. 7d). These results con rmed the role of glutamine in the gluR-mediated control of cell division in B. glumae.
Because environmental glutamine and glutamate affected the cell morphology of the gluR mutant BGLUR133 in M9 minimal medium, we examined the expression levels of seven fts genes in M9 minimal medium with or without glutamine and glutamate. The expression levels of the seven fts genes were signi cantly lower in the gluR mutant BGLUR133 than in the wild type or the BGLUR133C complemented strain (Fig. 7e). However, addition of glutamine to M9 minimal medium increased the expression levels of the seven fts genes in the gluR mutant BGLUR133 (Fig. 7f).
Heat sensitivity due to an imbalance of FtsA and FtsZ in the gluR mutant. Because fts genes were identi ed in a temperature-sensitive lamenting mutant, we assessed whether the lamenting gluR mutant BGLUR133 is heat sensitive. The number of cells of the gluR mutant BGLUR133 decreased signi cantly after 6 hours at 42°C and they were entirely nonviable after 12 hours in LB medium (Fig. 8a).
The wild-type BGR1, the gluS mutant BGLUS35, and the complemented strain BGLUR133C showed no growth but prolonged survival at 42°C (Fig. 8a). In M9 minimal medium at 42°C, the gluR mutant BGLUR133 mutant retained viability for 18 hours and subsequently lost viability (Fig. 8b). By contrast, the wild-type BGR1 and the complemented strain BGLUR133C increased in cell number during the static period of gluR mutant BGLUR133 in M9 medium (Fig. 8b).

Discussion
In addition to QS systems, pathogens likely manipulate environmental factors. Here we investigated the response regulator, GluR, which is crucial for normal cell division in B. glumae. Although gluR and gluS were co-transcribed, GluS was not the counterpart of GluR because a mutation in gluS did not affect normal cell division of B. glumae. Such a genetically linked but functionally independent TCS system was reported for risS and risA, which encode a sensor kinase and a response regulator, respectively, in B. pertussisi 3 . risS and risA were genetically linked but functionally independent 3 . Phosphorylation of RisA was mediated by crosstalk with a distant histidine kinase, RisK 4 . Therefore, an as-yet-unidenti ed sensor kinase may be responsible for phosphorylation of GluR in B. glumae.
Cell division involves ingrowth of the cell wall and membrane and septum formation in the chromosome of rod-shaped bacteria such as B. glumae 20 . To ensure equal partitioning of chromosomes into daughter cells, the expression of genes involved in cell division must be properly regulated 19,20 . In most bacteria, cell division and cell-wall synthesis are regulated by a series of genes in the dcw cluster 19 . Within bacterial groups of the same taxon and cell shape, the order and regulation of genes in the dcw cluster are highly conserved 21 . Therefore, it was not surprising that in B. glumae, the dcw cluster displayed signi cant similarities to that of E. coli 22 . Pioneer studies of the dcw cluster genes in E. coli spotlighted ftsZ as the key element in cell division 23,24 . It was later noted that FtsZ is not su cient to drive septation, leading to discovery of, for instance, ftsA, ftsQ, and ftsI 22 .
The mechanisms of regulation of the dcw cluster are unclear, despite the presence therein of several regulatory elements, e.g., internal promoters, transcript stabilizers, and protein ratios 22,25 . Studies on the control of cell division have concentrated on FtsZ. Multiple promoter regions have been reported upstream of ftsZ in the dcw cluster, indicating regulation at the transcriptional level 22,25,26 . We found that GluR binds to the upstream promoter regions of ftsZ and ftsA located in the ftsA and ftsQ coding regions, respectively. Unlike positive regulators in E. coli, such as SdiA 27 , the phase-speci c sigma factor 28 , and RcsB 29 , GluR negatively regulates cell division in B. glumae.
In the dcw cluster, biased expression of genes resulting from a mutation in gluR induced aberrant cell division. The GluR-controlled expression of dcw cluster genes was essential for normal cell division in B. glumae. FtsZ may be tethered to the membrane by two cytoplasmic membrane-associated proteins, ZipA and FtsA 30 , thus mediating cell division. Each E. coli cell is estimated to contain 3000-5000 molecules of FtsZ, 50-200 of FtsA, and 1500 of ZipA 31 . In addition, an imbalanced FtsZ and FtsA ratio is detrimental to E. coli 19,32 . In the lamentous cell-forming gluR mutant, expression of ftsZ was higher than that of ftsA, but the constitutive expression of ftsA restored a lamentous morphology. Although the optimal FtsZ and FtsA ratio is unknown, a 5.2fold increase in ftsA expression in the gluR mutant restored normal cell division to lamentous cells.
Because LB medium is rich in amino acids, and lamentation of the gluR mutant was facilitated by supplementation of extracellular glutamine or glutamate in M9, the glutamine-and glutamate-dependent lamentous cell formation at an early stage of growth in LB was explicable. However, the number of lamentous cells of the gluR mutant BGLUR133 decreased over time, possibly as a result of depletion of amino acids, including glutamine and glutamate, 12 hours after incubation ( Supplementary Fig. S3). Because the gluR mutant formed lamentous cells in a glutamine or glutamate-dependent manner, we hypothesized that GluR phosphorylation is caused by extracellular glutamine or glutamate, which promotes proper cell division by repressing dcw cluster genes. Extracellular glutamine and glutamate reportedly alter the expression of genes involved in cell division and cell-wall synthesis of B. subtilis 33 . Beuria et al. reported that an increased FtsZ polymerization rate and extent in E. coli resulted from extracellular glutamine 34 . It was noted that FtsZ showed optimal polymerization as large, bundled lamentous structures in E. coli in the presence of 1 M glutamine 34 . Interestingly, FtsZ polymers formed in the absence of glutamine were ninefold less stable than those in its presence, emphasizing the roles of these amino acids in the stability of FtsZ polymers 34 .
Connections between TCS and glutamine metabolism have been reported in other bacterial taxa. For example, GlnK-GlnL of Bacillus subtlis 35 , GluR-GluK of Streptomyces coelicolor 36 , and AauR-AauS of Pseudomonas putida 37 reportedly sense and control glutamate uptake. In other bacteria, the TCSs involved in glutamine sensing and uptake are located close to the glutamine ABC transporter [35][36][37] . GluR is not likely to be involved in glutamine uptake because we reported that GltI is responsible for glutamine uptake in B. glumae 38 . A bona de sensor kinase responsible for glutamine sensing and GluR phosphorylation is yet to be identi ed in B. glumae.
The fact that the gluR mutant BGLUR133 was sensitive to heat treatment at 42°C was somewhat expected because the name fts was coined from filamentous temperature sensitive mutants in E. coli 39 . Mutations in septation genes conferred an elongated morphology on E. coli; similarly, lamentous B. glumae caused by a gluR mutation were heat sensitive. This supports the hypothesis that GluR is crucial for cell division and an optimum gene expression pro le. Taken together, our ndings indicate that GluR is key for maintaining the gene expression pro le required for glutamine-or glutamate-dependent control of cell division in B. glumae BGR1.
DNA manipulation and sequencing. Basic DNA manipulation was conducted following standard protocols 40 . Plasmid DNA from E. coli was isolated using the Biomedic® Plasmid DNA Miniprep Kit (Ibiomedic, Korea) following the manufacturer's instructions. DNA sequencing was performed by Macrogen, Inc. (Seoul, Korea). The genetic information and gene IDs for DNA construction were obtained from the B. glumae BGR1 genome database (GenBank accession numbers: CP001503-CP001508; kropbase.snu.as.kr/cgi_bg.cg). A cosmid library of B. glumae BGR1 was constructed as described previously 8 .
Rescue mini-Tn5, Tn3-gusA, and marker-exchange mutagenesis. Using E. coli S17-1 (pRescue mini-Tn5), random mutations were created in B. glumae BGR1 as described previously 41 . Successful mutants were isolated by selection on LB agar containing kanamycin. The rescued mini-Tn5 mutants were screened for phenotypic changes. Following a previous method 42 , the anking regions were sequenced using the Oend primer (5′-GGTTTTCACCGTCATCACCG-3′), and the TCS genes were disrupted using the identi ed rescue mini-Tn5 insertions.
The pLAFR3 derivatives of pBGH1 carrying gluR (BGLU_1G13360) and gluS (BGLU_1G13350) were mutagenized using Tn3-gusA as described previously 43 . The Tn3-gusA insertion site and orientation in each mutant were mapped by restriction enzyme digestion analysis, and the plasmid sequenced using the Tn3gus primer (5′-CCGGTCATCTGAGACCATTAAAAGA-3′). The plasmids carrying Tn3-gusA insertions were marker-exchanged into B. glumae BGR1 via tri-parental mating 44 to generate BGLUR133 and BGLUS35. All marker-exchange mutants were con rmed by southern hybridization analysis.
Bacterial growth and viability assay. Overnight liquid cultures of the B. glumae strains were adjusted to an OD 600 of 0.05 and subcultured into fresh LB medium. The cultures were incubated for 30 hours at 37°C with shaking at 250 rpm. At 6hour intervals, bacterial growth was assayed by spotting 10 µL of serial dilutions in triplicate on LB agar plates. Bacterial growth was expressed as log CFU/mL after 2 days of incubation at 37°C.
Transmission electron microscopy. Bacterial cells were harvested from overnight cultures and prepared for observation by transmission electron microscopy (TEM) as reported previously 38 . Electron micrographs were acquired using a JEM 1010 microscope (JEOL, Tokyo, Japan) with acceleration voltages of 180 and 100 kV from a LIBRA 120 energy-ltration microscope (Carl Zeiss, Oberkochen, Germany).
PCR was performed in triplicate and gene expression values were normalized to that of 16S rRNA using Bio-Rad CFX Manager software.
Constitutive expression of ftsA. To express ftsA under the control of the trc promoter in pKK38, we ampli ed the ftsA-coding region from the B. glumae strains BGR1, BGLUR133, and BGLUR133C using the primers FtsA_Nco1 (5'-GGCCATGGAGCAAAGACTACAAAGATCT-3') and FtsA_HindIII (5'-CCAAGCTTTCAGAAATTGCTCAGGAACC-3') and a TaKaRa PCR Kit (TaKaRa Bio Inc., Kusatsu, Shiga, Japan) following the manufacturer's instructions. The PCR fragments were rst cloned into the Sma1 sites of pBluescript II SK (+) and transferred to the Nco1-HindIII sites of pKK38 as described previously 38 .
Growth and viability of B. glumae strains at 42°C. The B. glumae BGR1, TCS null mutants, and BGLUR133C strains were cultured overnight at 37°C, and the optical density at 600 nm (OD 600 ) was adjusted to 0.05. The strains were incubated at 42°C with shaking at 250 rpm for 24 hours in LB and M9 minimal media, and the cell density was measured at 6hour intervals.
Environmental stimuli driving GluR responses. We cultured the wild type, gluR mutant BGLUR133, and BGLUR133C in M9 minimal medium (6 g of Na 2 HPO 4 , 3 g of KH 2 PO 4 , 0.5 g of NaCl, and 1 g of NH 4  Scanning electron microscopy. B. glumae strains cultured overnight in LB or M9 minimal medium with/without amino acids were harvested, xed with Karnovsky's xative [2% glutaraldehyde, 2% paraformaldehyde in 0.05 M sodium cacodylate buffer (pH 7.4)], and post-xed with 1% sodium tetroxide in 0.1 M sodium cacodylate buffer for 1 hour at 4°C as described previously 46 . Before imaging, the samples were coated with platinum at 10 mA for 270 seconds using a G20 Ion Sputter Coater (GSEM Co., Suwon, Korea) and electron micrographs were acquired using a Carl Zeiss microscope (Auriga, Zeiss Germany).
Electrophoretic mobility shift assay. GluR-His was puri ed using an established method 10 . Using the primer set gluR_Nde1/gluR_BamH1 (Supplementary Table S2), we ampli ed the promoter regions of the putative GluR targets, ftsAp and ftsZp. The resulting PCR products were labeled with biotin using Lightshift Chemiluminescent Electrophoretic Mobility Shift Assay Kits, as described by the manufacturer (Pierce, Appleton, Wisconsin). We used 329 bp upstream of katE1 as a nonspeci c competitor DNA ampli ed using KatE1-F and KatE1-R primers (Supplementary Table S2). Puri ed GluR-His (0.75 µM) was incubated in binding buffer (10 mM Tris-HCl [pH 7.5], 100 mM NaCl, and 5% [v/v] glycerol) containing 1 nM biotin-labeled DNA as described previously 10 . For competition assays, unlabeled target DNA at 20fold molar excess was added to each reaction with the labeled DNA. Using 4% (w/v) polyacrylamide gels, the reactions were separated and transferred to nitrocellulose membranes. The bands were detected using streptavidin/horseradish peroxidase-derived chemiluminescence kits, as described by the manufacturer (Pierce) and visualized using ChemiDoc XRS + and Image Lab Software (Bio-Rad).
Statistical analysis. All experiments were conducted in triplicate with the appropriate controls. One-way analysis of variance (ANOVA) followed by Tukey's honestly signi cant difference post hoc analysis in SPSS software (ver. 25 x86-x64; IBM Corp., Armonk, NY) were conducted to detect signi cant differences. A value of p < 0.05 was considered indicative of statistical signi cance. Figure 1 The analysis showing that gluS and gluR genes are co-transcribed in BGR1. Primers were designed to amplify a 237 bp (*) product encompassing the gluR and gluS genes in the wild type. Lane G, PCR product using genomic DNA as a template; Lane R, PCR product using RNA as a template; Lane C, PCR product using cDNA as a template. (d) No polar effect resulted from Tn3-gusA insertion. Lane bp, marker; Lane G1, PCR product from gluR chromosomal DNA as a template; Lane G2, PCR product from gluS chromosomal DNA as a template; Lane R, PCR product from total RNA as a template; Lane C1, PCR product from gluR cDNA as a template; Lane C2, PCR product from gluS cDNA as a template. Full gel images are presented in Supplementary Fig. S4 and S5.

Figure 2
Tn3-gusA mutations in gluR resulted in nutrient-dependent cell lamentation. In LB medium, the gluR mutant BGLUR133 formed lamentous cells, but a normal rod-shaped cell morphology was observed in M9 minimal medium. No morphological defects were observed in the gluS mutant BGLUS35 in the different culture media. The gluR mutant forms a heterogeneous population of viable lamentous and normal rod-shaped cells.
(a) The indicated bacterial strains were grown to early stationary phase, and the morphological phenotypes of ultrathin sections were observed by TEM. BGLUR133-M shows that the lamentous cells formed by the gluR mutant contained multiple nuclei (arrows) with indents (arrowhead) along the cell wall, symbolizing failed septum formation. (b) Cell viability of the wild type, BGLUR133, and complemented BGLUR133C strains assessed by combination staining with propidium iodide (PI) and SYTO 9 green. Fluorescence images were obtained by confocal laser scanning microscopy. Dead cells stained with PI are red, and SYTO 9-stained viable cells are green. GluR maintains the molar ratio of ftsZ to ftsA to ensure normal cell division. FtsA was constitutively expressed to counteract ftsZ in the wild type (BGR1), BGLUR133, and BGLUR133C strains. (a) mRNA levels were quanti ed by qRT-PCR and are shown as normalized fold expression values. Data are means ± standard error (SE) of triplicates. (b) Bacterial strains cultured to early stationary phase were visualized using a Carl Zeiss GmbH Auriga microscope. pFtsA represents constitutive expression of ftsA in the indicated bacterial strains.

Figure 6
Extracellular amino acids promote lamentation in response to GluR mutation. The indicated bacterial strains were cultured overnight in M9 minimal medium with or without 10% casamino acids (CA), processed for SEM analysis, and their morphology observed using a Carl Zeiss GmbH Auriga microscope. Exponential population decline at 42oC as a result of mutations in GluR. At 6 hour intervals, the indicated strains' population densities in LB medium (a) and M9 medium (b), were quanti ed by CFU counting and the results expressed as log CFU/mL. Data are means ± standard error (SE) of triplicates.