Essential roles of Lon protease in the morpho-physiological traits of the rice pathogen Burkholderia glumae

The highly conserved ATP-dependent Lon protease plays important roles in diverse biological processes. The lon gene is usually nonessential for viability; however, lon mutants of several bacterial species, although viable, exhibit cellular defects. Here, we show that a lack of Lon protease causes pleiotropic effects in the rice pathogen Burkholderia glumae. The null mutation of lon produced three colony types, big (BLONB), normal (BLONN), and small (BLONS), in Luria–Bertani (LB) medium. Colonies of the BLONB and BLONN types were re-segregated upon subculture, while those of the BLONS type were too small to manipulate. The BLONN type was chosen for further studies, as only this type was fully genetically complemented. BLONN-type cells did not reach the maximum growth capacity, and their population decreased drastically after the stationary phase in LB medium. BLONN-type cells were defective in the biosynthesis of quorum sensing (QS) signals and exhibited reduced oxalate biosynthetic activity, causing environmental alkaline toxicity and population collapse. Addition of excessive N-octanoyl-homoserine lactone (C8-HSL) to BLONN-type cell cultures did not fully restore oxalate biosynthesis, suggesting that the decrease in oxalate biosynthesis in BLONN-type cells was not due to insufficient C8-HSL. Co-expression of lon and tofR in Escherichia coli suggested that Lon negatively affects the TofR level in a C8-HSL-dependent manner. Lon protease interacted with the oxalate biosynthetic enzymes, ObcA and ObcB, indicating potential roles for the oxalate biosynthetic activity. These results suggest that Lon protease influences colony morphology, growth, QS system, and oxalate biosynthesis in B. glumae.


Introduction
The ATP-dependent protease Lon is highly conserved in prokaryotes and eukaryotes and is involved in diverse biological processes [1]. Lon was first identified in Escherichia coli and derived its name from the long, non-septate filamentation phenotype of the lon mutant [2]. Since its discovery, studies have shown that Lon plays an important role in the degradation of abnormal or unstable regulatory proteins as a quality control protease [1]. In E. coli, the short- adjusted to an optical density (OD) of 0.05 at 600 nm using a BioSpectrometer (Eppendorf) followed by 2 mL subculture in glass test tubes (PYREX) for all assays. To compare initial growth rate of wild type and mutant, an OD 600 of 0.05 was diluted 20 times followed by subculture (Figs 2A and S1A). Aliquots of 100 μL from each sample were serially diluted and 10 μL each of three repeats was spotted on LB agar medium to monitor colony-forming units (CFUs) at the designated time point. LB agar plates were incubated at 37˚C for 24 h to allow colonies to grow. CFUs were counted under a dissecting microscope and multiplied by the appropriate dilution factor to calculate CFU/mL. In an extracellular pH assay, the culture supernatant was sampled from each vial at the designated time point and the pH was measured using a pH meter (Lab 860, SCOTT Instruments).
Construction of the B. glumae lon::Gm, lon::Gm/lon, tofI::Sp/lon::Gm double mutant, and obcA::Tp The lon (bglu_1g13520) gene was deleted via insertion of the gentamycin cassette ScaI and marker exchange as described previously [21]. The pLa2 is a 9.0-kb EcoRI-HindIII fragment cloned into pLAFR3, including the lon gene from cosmid clone pLa1. A gentamycin cassette was generated by PCR using ScaI-Gm-F and Gm-ScaI-R primers, cloned into pBluescript II SK(+), and termed "pBS_Gm4". A 0.8-kb DNA fragment obtained from pBS_Gm4 via digestion performed with ScaI was ligated into pLa2 to produce pLa2_Gm. pLa2_Gm was mobilized from DH5α into B. glumae BGR1 or BGS2 by conjugation using pRK2013 as a helper plasmid and marker exchange, as described previously [21]. For genetic complementation of the lon mutant, a 2.876-kb DNA fragment containing the lon gene and its own promoter was generated by PCR using HindIII-La and La-EcoRI primers, cloned into pBluescript II SK(+), and termed "pLa3". A 2.876-kb DNA fragment obtained by pLa3 by digestion performed with HindIII and EcoRI was ligated into pLAFR3 to produce pLa4. pLa4 was mobilized from DH5α into B. glumae BLONN through conjugation using pRK2013 as a helper plasmid. The obcA (bglu_2g18790) gene was deleted via insertion of the trimethoprim resistance gene KpnI-XhoI and marker exchange as described previously [21]. A 2.716-kb DNA fragment including the obcA and obcB genes and their promoter was generated via PCR using BamHI-obcA and obcB-EcoRI primers, cloned into pLAFR3, and termed "pOBC4". pTP2 contained a trimethoprim-resistant gene produced by PCR using the Tp(BamHI)-F and Tp-R primers and then cloned into pBluescript II SK(+). A 0.887-kb KpnI-XhoI fragment from pTP2 was cloned into pOBC4 to produce pOBC4::Tp. The BOBCA strain was generated via triparental mating of B. glumae BGR1, DH5α (pRK2013), and DH5α (pOBC4::Tp). Southern blot analysis was performed to confirm the BLONN, S2LON, and BOBCA strains, and the BLONC was assessed via plasmid extraction.

Measurements of oxalate levels and oxalate biosynthetic activity assay
Oxalate levels and oxalate biosynthetic activity were measured as described previously [20,22]. Oxalate was measured using an oxalate assay kit (Trinity Biotech) according to the manufacturer's instructions. Briefly, oxalate was converted into carbon dioxide and hydrogen peroxide using oxalate oxidase, and the production of hydrogen peroxide was measured via reaction with 3-(dimethylamino) benzoic acid, during which it formed a blue compound catalyzed by peroxidase. Absorbance at 590 nm was measured using a microplate reader (PerkinElmer); 0.5 mM oxalate was used as a standard. The absorbance of the sample was divided by the absorbance of standard and multiplied by the appropriate dilution factor. The reaction buffer used in the assay measuring oxalate biosynthetic activity consisted of 200 mM Tris-Cl (pH 8.0), 10 mM EDTA, 20 mM CoCl 2 , 2 mM acetyl-CoA, and 200 mM oxaloacetate. The total cell lysate was added to the reaction mixture, followed by incubation at 37˚C for 10 min. The level of biosynthesized oxalate from the reaction was measured using an oxalate assay kit. Units were calculated as the concentration of biosynthesized oxalate divided by the amount of total protein and reaction time.

RNA extraction and quantitative reverse-transcription PCR (qRT-PCR)
Total RNAs were extracted from B. glumae BGR1, BLONN (BGR1 lon::Gm), and BLONC (BGR1 lon::Gm/lon), grown in LB medium at 37˚C for 10 h after subculture, were extracted using RNeasy mini kits (Qiagen) following the manufacturer's protocols. The extracted total RNA was treated with RNase-free DNase I (Ambion) to remove DNA. Total RNA (1 μg) was reverse-transcribed into cDNA using M-MLV Reverse Transcriptase (Promega, Madison, WI, USA) and incubated for 1 h at 42˚C. The primer pairs used for qRT-PCR are listed in S2 Table. The 16S rRNA gene was used as the positive control. Transcription levels were determined using SsoFast EvaGreen Supermix (Bio-Rad, Hercules, CA, USA) and the CFX96 Real-Time PCR System (Bio-Rad). The thermal cycling parameters were as follows: 95˚C for 30 s, followed by 40 cycles of 95˚C for 5 s and 60˚C for 5 s. All reactions were performed in triplicate, and all data were normalized to the expression levels of the 16S rRNA gene expression using Bio-Rad CFX Manager software.

Overexpression of Lon and TofR and co-expression of Lon and TofR
The lon and tofR genes of B. glumae were amplified using the primers listed in S2 Table. The amplification products were cloned into the NheI and HindIII restriction sites of the pET-28b expression vector and into the NdeI and XhoI restriction sites of the pACYC-Duet expression vector (Novagen), resulting in pLa8 and pTOFR5, respectively. His-Lon-His and TofR-His were overexpressed in E. coli BL21(DE3), which was induced by addition of 1 mM isopropyl β-o-thiogalactopyranoside (IPTG), followed by additional growth for 1 h at 37˚C. To test for protease resistance against Lon protease, 4 μM N-octanoyl-homoserine lactone (C8-HSL) was added to the BL21(DE3) cultures. The BL21(DE3) carrying both pLa8 and pTOFR5 was used for co-expression of Lon and TofR.

Coimmunoprecipitation (Co-IP) and western blot analysis
According to the manufacturer's instructions, Co-IP was performed using the Pierce Direct IP kit (Thermo Fisher Scientific). The anti-ObcA (10 μg), anti-ObcB (10 μg), and anti-Lon (10 μg) antibodies were coupled to AminoLink Plus Coupling Resin (Thermo Fisher Scientific). Cell lysates from B. glumae BGR1 and S2HA containing 1 mg of protein were added to the antibodycoupled resin in a spin column and mixed at 4˚C for 24 h. The immunoprecipitates were eluted, followed by the addition of a 5× non-reducing lane marker sample buffer. Samples were boiled for 5 min, cooled to room temperature, and separated via SDS-PAGE. The samples were then transferred to PVDF membranes. Anti-Lon, anti-ObcA, anti-ObcB, and anti-HA (Invitrogen) antibodies were used to detect interactions of ObcA with ObcB and Lon, of ObcB with Lon, and of TofI-HA with Lon, respectively. The immunoreactive bands were detected using electrochemiluminescence (ECL) reagents (Bio-Rad) and captured using the ChemiDoc™ XRS+ imager (Bio-Rad). All experiments were performed using at least three independent replicates.

Autoinducer assay
An autoinducer assay was performed as described previously [18] with the following modifications; B. glumae BGR1, BLONN (BGR1 lon::Gm), and BLONC (BGR1 lon::Gm/lon) were grown in LB supplemented with HEPES (pH 7.0) for 24 h. Autoinducer was extracted from the 500 μL culture-free supernatant using the same amount of ethyl acetate and dried. The dried pellet was dissolved in 10 μL dimethyl sulfoxide. Autoinducer was developed via the thin layer chromatography (TLC), using 70% methanol in the TLC tank. LB agar with biosensor strain Chromobacterium violaceum CV026 overlaid was used to develop the TLC plate to visualize C8-HSL and C6-HSL.

Statistical analysis
Statistical analyses were performed using SPSS version 23 (IBM Corp.). Comparisons were performed using one-way analysis of variance (ANOVA) followed by Tukey's post-hoc analysis (set at 5% significance level).

Continuous segregation of the lon mutant into morphologically distinct colonies
To explore the role of Lon protease in B. glumae BGR1, we constructed a lon null mutant by inserting a gentamicin resistance cassette into the lon gene via double-crossover homologous recombination. The genome of B. glumae BGR1 contains three homologs of ATP-dependent Lon protease: bglu_1g13520, bglu_1g31260, and bglu_1g33380. The bglu_1g13520 homolog had typical domains of Lon protease, including the Lon substrate binding domain (accession num-ber_pfam02190), P-loop_NTPase domain (accession number_ cl38936), and C-terminal proteolytic domain (accession number_pfam05362) (S1 Fig). On the other hand, bglu_1g31260 and bglu_1g33380 contained only N-terminal domain of Lon protease (accession number_cl19481) and P-loop_NTPase domain, respectively (S1 Fig). The bglu_1g13520 was selected for investigation of its biological roles in B. glumae BGR1. The marker-exchanged lon null mutant was segregated into three morphologically distinct colony types: big (BLONB), normal (BLONN), and small (BLONS) (Fig 1A). The BLONS-type cells grew very slowly and were too small to manipulate. The largest-sized colonies, called BLONB, grew faster than the other two types but segregated into BLONB and BLONS only, not BLONN, after subculture in LB medium ( Fig 1B). BLONB was bigger than BLONN in all tested media, and was darker, than BLONN under a dissecting microscope (Fig 1A, 1B and 1E-1H). The occurrence frequency of BLONB and BLONS segregated from BLONB was 67.5% and 32.5%, respectively (Fig 1C). When a colony of the BLONN type was subcultured in LB medium, cells were segregated into the BLONB, BLONN, and BLONS types at frequencies of 5.35%, 84.52%, and 10.13%, respectively ( Fig 1D). Such segregation of the BLONB and BLONN types did not change in M9 minimal medium supplemented with 0.2% glucose (Fig 1E), nutrient broth medium (Fig 1F), King's B medium (Fig 1G), and LB supplemented with 100 mM HEPES (pH 7.0) ( Fig 1H). We selected the BLONN type for further analysis, as this colony type exhibited fully complemented phenotypes with pLa4 carrying a full length of lon gene along with its own promoter in pLAFR3 (Fig 2A).

Impaired growth of the lon mutant
To determine the biological significance of Lon protease, we measured the growth and extracellular pH of the lon mutant. The population of the wild type BGR1 reached a maximum capacity at 3.20 × 10 9 colony forming unit (CFU)/mL after 12 hours of subculture in LB medium ( Fig 2B). On the other hand, the lon mutant increased up to 2.26 × 10 8 CFU/ml after the stationary phase and decreased gradually for 1 to 3 days after subculture, followed by complete population death (Fig 2B). During the initial growth stage, the lon mutant had a significantly lower growth rate than that of the wild type (Fig 2A). The lon mutant showed alkalization of the LB culture medium, whereas the pH of wild-type BGR1 slightly increased to 7.33 and then decreased to 4.97, and finally returned to 6.67 ( Fig 2C). Environmental pH increased during growth of the lon mutant, which resulted in a population crash (Fig 2). Addition of 100 mM HEPES buffer to the growth medium spared the lon mutant from cell death after the stationary phase but did not affect its initial growth rate (S2 Fig).

Failure of QS signal biosynthesis in the lon mutant
To determine whether impaired growth of the lon mutant affects QS systems, we performed an autoinducer (AI) assay of the wild type and the lon mutant. We performed AI assays with  wild type BGR1 at two different population densities, approximately at 1.28 × 10 8 and 1.33 × 10 9 CFU/mL, as the lon mutant did not reach the maximum population density. AI signals were detected at 1.28 × 10 8 CFU/mL but much higher levels were observed at 1.33 × 10 9 CFU/mL of wild type BGR1 (Fig 3A). However, the lon mutant produced no detectable QS signals, and the genetic complementation resulted in recovery of QS signal biosynthesis to the wild type level ( Fig 3A). As the failure of the lon mutant to reach maximum growth capacity might be one reason that the mutant did not produce detectable amounts of QS signals, we harvested cells from 10 culture tubes, each containing 3.57 × 10 8 CFU/mL, and pooled them to No QS signal was detected at 3.57 × 10 9 CFU/mL of the lon mutant (S3 Fig). To determine whether the lack of QS signals in the lon mutant was due to low expression of QS signal synthase gene, tofI, we assessed tofI expression at the transcription level during the mid-log phase using qRT-PCR. Expression of tofI was lower in the lon mutant than in the wild type at the transcription level ( Fig 3B). To detect the level of TofI-HA in the wild type and lon mutant via western blot using an anti-HA antibody, stationary-phase strains carrying the TofI-HA clone, pTOFI6, were used. The level of TofI-HA in the lon mutant was approximately 71.1% of the wild type level (Fig 3C).

Low levels of oxalate biosynthetic enzymes produced by the lon mutant
Because the lon mutant produced no detectable QS signal, we measured the capacity for oxalate biosynthesis, which is known to be QS-dependent, in the lon mutant. We found that the oxalate level was significantly lower in the lon mutant than in the wild type ( Fig 4A). Reduced production of oxalate by the lon mutant was in accordance with the observed increase in pH of the lon mutant (Figs 2C and 4A). To determine whether the cause of the low oxalate level in the lon mutant was due to the low expression of obcA and obcB, which are responsible for oxalate biosynthesis, we measured the expression levels of both genes in the wild type and the lon mutant. The expression of the obcA and obcB genes was significantly lower in the lon mutant than in the wild type and the complemented strains (Fig 4B). We also analyzed the levels of ObcA and ObcB by western blotting using ObcA and ObcB antibodies at the time points corresponding to the mid-log, early stationary, and mid-stationary growth phases after subculture (6, 12, and 24 hours, respectively). ObcA and ObcB were not detected in the lon mutant at 6 hours but were detected at 12 hours after subculture (Fig 4C). Oxalate production was significantly lower in the lon mutant than in the wild type, even though the ObcA and ObcB were accumulated in the lon mutant during growth (Fig 4A and 4C). To investigate whether Lon affects the activities of ObcA and ObcB, we measured specific oxalate biosynthetic activity in the wild type and lon mutant. Specific oxalate biosynthetic activity was reduced by 68.86% in the lon mutant compared with the wild type ( Fig 4D). When specific oxalate biosynthetic activity was converted into the units per CFU, the oxalate biosynthetic activity was significantly lower in the lon mutant than in the wild type (S4 Fig).

Reduced oxalate biosynthetic enzyme activity in the lon mutant was not due to low production of C8-HSL
As obcA and obcB are expressed in a QS-dependent manner in B. glumae, and the lon mutant produced no detectable amounts of QS signal, we evaluated whether oxalate biosynthetic capacity can be recovered after exogenous addition of C8-HSL to the lon mutant culture medium. When levels of ObcA and ObcB and oxalate biosynthetic activity were measured in the lon mutant cells grown in LB with the addition of 10 μM C8-HSL, levels of ObcA and ObcB and oxalate biosynthetic activity in the lon mutant did not recover to wildtype levels (Fig 5A and 5B). Oxalate production was monitored during growth in LB buffered with 100 mM HEPES (pH 7.0) and supplemented with up to 10 μM C8-HSL for 2 days after subculture, but the lon mutant did not produce oxalate as much as the wild type did (Fig 5C).

Influence of Lon protease on the TofR level
As addition of excessive QS signal to the lon mutant did not recover the QS-dependent oxalate production to wild-type levels, we hypothesized that Lon protease might influence the activity of the QS signal receptor, TofR. To determine whether Lon protease affects TofR levels, we measured total TofR levels in the lon mutant. TofR levels were 1.81-and 1.6-fold higher in the lon mutant and the tofI/lon double mutant, respectively, than in the wild type ( Fig 6A). Although no QS signals were detected in the tofI mutant or lon mutant, TofR levels in the tofI mutant were 12.21% of its wild type and 6.72% of its lon mutant (Fig 6A). When lon and tofR were individually expressed with or without C8-HSL in E. coli, C8-HSL did not influence total levels of TofR ( Fig 6B). However, TofR levels were reduced by 4.9-fold when the two genes were co-expressed without C8-HSL compared to the individual expression of the tofR gene ( Fig 6B). Addition of C8-HSL to E. coli cultures co-expressing the two genes resulted in a 2.2-fold increase in TofR compared with the same cultures grown without C8-HSL (Fig 6B).

Interaction of Lon protease with ObcA and ObcB but not with TofI
As the lon mutant exhibited low oxalate biosynthetic activity and produced significantly less QS signals compared with the wild type, we hypothesized that Lon protease interacts with ObcA, ObcB, and TofI, affecting their activities. Immunoprecipitation was performed using anti-ObcB and anti-Lon antibodies, ObcA protein was detected, which indicated that ObcA interacts with ObcB and Lon in the wild type strain (Fig 7A). In addition to the 59.11 kDa band of ObcA, another positive band was detected with a size of approximately 52 kDa (Fig 7A). ObcB protein was pulled down using an anti-Lon antibody in an affinity pulldown experiment, which showed that ObcB and Lon interacted in the wild type strain (Fig 7B). To determine whether Lon interacts with TofI, we performed immunoprecipitation using an anti-Lon antibody and detected TofI-HA using an anti-HA antibody by western blot analysis (S5 Fig). TofI-HA was not detected in the immunoprecipitated sample using the anti-Lon antibody (S5 Fig), indicating that Lon and TofI did not interact.

Discussion
Continuous colony differentiation and abnormal growth due to the loss of Lon protease allowed us to determine that Lon protease is vital to the integrity of cellular functions. It has been reported that the null mutant of lon formed large and small colonies in A. tumefaciens, but its large colonies were homogeneous after subculture [14]. Although the lon mutant has a longer doubling time than that of the wild type A. tumefaciens, they reached similar maximum population densities [14]. This study is the first to report that the lon mutant colonies are continuously segregated morphologically, and that this mutant cannot grow to its maximum capacity. We explored whether several physiological issues caused by a deficiency of Lon protease threaten bacterial survival in vitro.
In general, heterogeneity of the bacterial population is induced by nutrient exhaustion or stress, and it occurs to allow adaptation of the bacteria to environmental changes [23]. Heterogeneous populations may exhibit phenotypic differences within a genetically identical population, or genetic heterogeneity may be present [23]. Persistent colony differentiation was observed in the lon mutant of B. glumae under all tested culture conditions; therefore, heterogeneity of the lon mutant did not appear to be caused by the external environments. The BLONB, generated from BLONN, did not return to the BLONN and was not complemented with pLa4 containing the lon gene; other genetic changes appear to have occurred in the BLONB, although we did not identify any mutation in the BLONB genome. The BLONN maintained a homogeneous population only via genetic complementation, indicating that Lon protease plays an essential role in normal growth and cellular integrity.
The primary biological function of Lon protease is control of protein quality by aiding degradation of unfolded proteins and unstable proteins such as global regulators, thereby preventing the accumulation of abnormal proteins [1]. Protein quality control is also important for stress responses; in particular, Lon plays a role in environmental stress responses in E. coli [1]. Expression of the lon gene in E. coli is dependent on the σ 32 transcription factor (RpoH) and is activated by heat shock [24]. These stress-induced proteins are essential for removing damaged polypeptides from stressed cells. High demand for Lon, which suggests that the lon gene is upregulated under certain stress conditions, explains the reduced growth of the lon mutant under thermal stress [24]. As the promoter region of the lon gene in B. glumae had no recognition sequence for sigma factors, Lon in B. glumae is unlikely to belong to the heat regulon, as observed in E. coli.
In addition to proteolytic activity, chaperone-like functions of Lon have also been reported in eukaryotic cells. The human Lon protease LONP1 is a mitochondrial matrix protein that, in conjunction with mitochondrial HSP70, promotes protein folding [25]. Lon plays a vital role in enhancing cell survival by maintaining the stability of the Hsp60-mtHsp70 complex [26]. The interactions of Lon with ObcA and ObcB in B. glumae might be due to Lon acting as a chaperone for ObcA and ObcB. The ObcA and ObcB proteins were detected in samples obtained by affinity pulldown using an anti-Lon antibody. Lon clearly interacts with ObcA and ObcB; however, the biochemical mechanism by which Lon functions as a chaperone is not understood and should be addressed in future works. The appearance of additional unidentified positive signals of approximately 52 kDa, as shown in Fig 7A, suggests that ObcA might undergo posttranslational processing. However, the purpose of posttranslational processing of ObcA remains unknown.
The B. glumae Lon protease has significant homology with the Lon proteases of other bacteria, and the organization of the lon locus is conserved. The locus encoded by clpX and hupB in the flanking region of lon is consistent with those of P. aeruginosa, P. putida, E. coli, and A. tumefaciens. The Lon protease of these four different bacteria had 81% (accession number_AAG05192.1), 67% (AAN67915.1), 70% (AAC73542.1), and 65% (WP_080866091.1) identity with B. glumae Lon (bglu_1g13520), respectively (S6 Fig). However, despite these similarities, the function of B. glumae Lon is quite different from the functions of Lon in these bacteria. The P. aeruginosa Lon negatively controls the LasR/LasI system via degradation of LasI, which directs the synthesis of 3-oxo-dodecanoyl-HSL [15]. Lon negatively regulates the expression of rhlR/rhlI via the degradation of LasI [15]. The lon mutant of P. putida produces significantly higher QS signals than those produced by the wild type [16]. However, in B. glumae, no QS signals were detected from the lon mutant, even though the TofI level in the lon mutant was approximately 71.1% of the wild type level. Lon is unlikely to function as a chaperon for TofI activity, as TofI does not interact with Lon in B. glumae. However, Lon might affect other proteins that are involved in AI biosynthesis, such as acyl carrier protein (ACP) or enzymes associated with S-adenosylmethionine (SAM). Whether Lon affects the biosynthesis of two substrates of TofI at any level is an interesting topic for future research.
No recovery of the QS-dependent phenotype in the lon mutant with excessive addition of C8-HSL QS signal suggests that TofR might not function properly as a receptor protein of C8-HSL in the lon mutant. It has been reported that ClpB and Lon protease of E. coli are responsible for degradation of TraR, a receptor protein for the QS signal, in A. tumefaciens [27]. However, there is no evidence that ClpB or Lon protease of A. tumefaciens directly affects the TraR level dependent on the presence of the QS signal. In contrast to TraR of A. tumefaciens, we showed that TofR appears to be degraded by the Lon protease of B. glumae in E. coli when C8-HSL is unavailable. As overexpressed TofR in the absence of C8-HSL was completely insoluble, degradation of TofR by Lon protease in vitro was not demonstrable.
Our results revealed previously unidentified roles of Lon protease in bacteria. Continuous segregation and growth defects of the lon mutant of B. glumae are unique among the phenotypes reported in bacteria. The negative role of Lon protease in bacterial QS is contradictory to the previously known activities of this enzyme in other bacteria. Our findings raise fundamental questions about the functions of bacterial Lon protease, including the mechanism underlying continuous segregation of colony morphology. Our findings suggest that Lon protease plays essential roles in QS and morpho-physiological traits, and that it is a potential target for the development of control agents for the rice pathogen B. glumae.