Page 1 of 68

ChemotaxisChemotaxis----likelikelikelike pathway of Azorhizobium caulinodans controlscontrolscontrols ffflagellaflagellalagellalagella----drivendrivendrivendriven

motilitymotilitymotility,motility, which regulateregulateregulatesregulates biofilms biofilmsss formation, exopolysaccharide biosynthesis and

competitive nodulation

Wei Liu1, Yu Sun1, Rimin Shen1, 2, Xiaoxiao Dang1, Xiaolin Liu1, Fu Sui1, Yan Li1,

Zhenpeng Zhang1, Gladys Alexandre3, Claudine Elmerich4, and Zhihong Xie1*

1Key Laboratory of Coastal Biology and Bioresource Utilization, Yantai Institute of

Coastal Zone Research, Chinese Academy of Sciences, Yantai, China;

2 Shanxi Agricultural University, Taigu, Shanxi, China

3Biochemistry, Cellular and Molecular Biology Department, University of

Tennessee, Knoxville, USA;

4Institut Pasteur, Paris, France

*Corresponding author

E-mail: [email protected] Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

Wei Liu, 1, MPMI

Page 2 of 68

Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

Wei Liu, 2, MPMI

Page 3 of 68

1 AAABSTRACTABSTRACTBSTRACTBSTRACT

2 The genome of the Azorhizobium caulinodans ORS571 contains a unique

3 chemotaxis gene cluster (che) including 5 chemotaxis genes: cheAWY1BR.

4 Analysis of the role of the chemotaxis cluster of A. caulinodans, using deletion

5 mutant strains revealed that CheA or Che signaling pathway controls chemotaxis

6 behavior and flagella-driven motility, and plays important roles in biofilms

7 formation and production of extracellular polysaccharides. Furthermore, the

8 deletion mutants (ΔcheA and ΔcheA-R) were defective in competitive adsorption

9 and colonization on the root surface of host plants. In addition, a functional CheA

10 or Che pathway promoted competitive nodulation on roots and stems.

11 Interestingly, a non-flagellated mutant ΔfliM, displayed a phenotype highly similar

12 to that of the ΔcheA or ΔcheA-R mutant strains. These findings suggest that

13 through controlling flagella-driven motility behavior, the chemotaxis signaling Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

14 pathway in A. caulinodans coordinates biofilms formation, extracellular This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

15 polysaccharides, and competitive colonization and nodulation.

16 IIINTINTNTNTRODUCTIONRODUCTION

Wei Liu, 3, MPMI

Page 4 of 68

17 Nitrogen-fixing bacteria are generally divided into two major classes: free-living

18 and symbiotic. Symbiosis between rhizobia and hosts generally leads to formation

19 of nodules on roots. Some members of legumes are also able to develop stem

20 nodules (Tsien et al. 1983; James et al. 2001). Azorhizobium caulinodans

21 ORS571 has the capability of fixing nitrogen both in the free-living and under

22 symbiotic conditions with its host Sesbania rostrata (Dreyfus et al. 1983). In

23 addition, it retains a capacity to induce nodule formation on roots as well as on

24 stems of S. rostrata under symbiotic conditions (Dreyfus et al. 1983, 1988; Tsien

25 et al. 1983).

26 The formation of biofilms on root surface promotes colonization in many

27 plant-associated bacteria, including genera such as Rhizobium,

28 Gluconacetobacter and Azospirillum (Rinaudi and Giordano 2010; Siuti et al.

29 2011; Yaryura et al. 2008). Furthermore, rhizobial extracellular polysaccharides Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

30 (EPS) have been proposed to play important roles in adhesion and colonization This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

31 of root surfaces (Burdman et al. 2000; Matthysse et al. 2005; Michiels et al. 1991;

32 Santaella et al. 2008), biofilms formation on roots during nodulation (Bianciotto et

Wei Liu, 4, MPMI

Page 5 of 68

33 al. 2001; Fraysse et al. 2003; Fujishige et al. 2006; Meneses et al. 2011; Ramey

34 et al. 2004; Rinaudi and Giordano 2010; Skorupska et al. 2006). Mutants with

35 altered EPSs are defective in symbiosis with host plants (Cheng and Walker

36 1998; Doherty et al. 1988; Wells et al. 2007). In soil bacteria, chemotaxis and

37 motility favor the chemical or physical interaction with host roots and occupation

38 of infection sites resulting in more competitive colonization and nodulation ability

39 (Caetano-Anollés et al. 1988; Greer-Phillips et al. 2004; Miller et al. 2007; Vande

40 Broek and Vanderleyden 1995).

41 Bacterial chemotaxis is a motility-based response initiated by perception of

42 environmental chemical signals. The chemotaxis signal transduction pathway has

43 been best studied in Escherichia coli (Wadhams and Armitage 2004). This

44 pathway is comprised of conserved proteins that including a histidine kinase

45 CheA, an adaptor protein CheW, a response regulator CheY, a methylesterase Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

46 CheB, a methyltransferase CheR and multiple chemoreceptors. CheA acts as the This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

47 central processing unit of this system. Environmental signals sensed by

48 chemoreceptors regulate the phosphorylated state of CheA via CheW. The

Wei Liu, 5, MPMI

Page 6 of 68

49 phosphoryl group is then transferred from CheA to CheY resulting in

50 phosphorylated CheY. Phospho-CheY then binds proteins in the flagellar motor

51 switch complex (comprised of FliM, FliN and FliG) to affect flagellar rotational

52 direction (Stock and Surette 1996).

53 Generally, the basic molecular mechanism of chemotaxis is conserved in

54 bacteria (Wuichet and Zhulin 2010). There is only a single chemotaxis operon in E.

55 coli genome. However, many motile soil bacteria possess two or more chemotaxis

56 operons encoding homologues of the E. coli chemotaxis proteins in their genomes

57 (Scharf et al. 2016; Szurmant and Ordal 2004). Furthermore, gene order and

58 composition of operons vary within members of the alpha-proteobacteria.

59 Chemotaxis operons in some proteobacterial species were reported to regulate

60 chemotaxis response, flagellar synthesis, cell size, cyst formation, biofilms

61 formation or other cellular functions (Berleman and Bauer 2005; Bible et al. 2008; Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

62 D’Argenio et al. 2002; Ferrandez et al. 2002; Kirby and Zusman 2003; Vlamakis et This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

63 al. 2004). There are two chemotaxis operons in the genome of Sinorhizobium

64 meliloti (which is a synonym of the accepted name Ensifer meliloti) and the

Wei Liu, 6, MPMI

Page 7 of 68

65 chemotactic response is controlled by the Che1 pathway (Meier et al. 2007;

66 Sourjik and Schmitt 1996). In Azospirillum brasilense, there are four main sets of

67 chemotaxis gene clusters encoded within the genome (Wisniewski-Dyé et al.

68 2011). Che1 signaling pathway controls swimming speed and has a minor role in

69 chemotaxis (Bible et al. 2012) while Che4 controls changes in swimming direction

70 and competitive colonization on plant root surfaces (Mukherjee et al. 2016).

71 A. caulinodans ORS571 are motile soil bacteria that have versatile rhizobial

72 lifestyles. Analysis of available genome sequence (Lee et al. 2008) indicates that

73 A. caulinodans relies on a single chemotaxis system in the whole genome (Jiang

74 et al. 2006a). The chemotaxis system (che) contains 5 chemotaxis genes:

75 cheAWY1BR, transcribed in same direction, from a putative σ54-dependent

76 promoter upstream of cheA (Jiang et al. 2016a). While the function of some

77 chemoreceptors has been investigated (Jiang et al. 2016b; Liu et al. 2017a), the Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

78 role of signaling mediated by the products of the chemotaxis cluster (che) in A. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

79 caulinodans ORS571 has not yet been established. Our results show that the che

80 cluster of A. caulinodans ORS571 regulates directional motility and other cellular

Wei Liu, 7, MPMI

Page 8 of 68

81 functions in a free-living state or in symbiosis with its host S. rostrata. We provide

82 evidence that che-mediated regulation of other cellular functions depends on

83 functional flagella-driven motility.

84 RRRESULTSRESULTSESULTSESULTS

85 CCChemotaxisChemotaxis system of A. caulinodans ORS571ORS571ORS571

86 Analysis of the A. caulinodans ORS571 genome (Lee et al. 2008) revealed a

87 large DNA region (AZC_0620 to 0666), spanning about 45-kb, that contains ORFs

88 whose translation products shared similarity with proteins involved in synthesis of

89 flagella and chemotaxis response (Fig. 1A). A preliminary comparative analysis

90 identified a group of 5 che genes (AZC_0661 to 0665): cheAWY1BR, transcribed

91 in the same direction (Fig. 1A). In this gene cluster, translational overlapping is

92 manifest between the cheA and cheW, between cheY1 and cheB and between

93 cheB and cheR and given that chemotaxis proteins are typically organized as Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

94 operon, it is assumed that the five genes here are also organized as a single This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

95 transcription unit. A gene (AZC_0660), named tlpA1, encoding a transmembrane

96 MCP product (Liu et al. 2017a), is located 200 nucleotides upstream of cheA (Fig.

Wei Liu, 8, MPMI

Page 9 of 68

97 1A). Another mcp gene (AZC_0666) also encoding a transmembrane MCP

98 product is located downstream of the che cluster (Fig. 1A). The icpB gene

99 (AZC_3718) encoding a soluble chemotaxis receptor previously characterized is

100 not linked to the cluster (Jiang et al. 2016b). Two other chemotaxis genes cheY2

101 (AZC_0620) and cheZ (AZC_0621), are located in the distal region of the che

102 gene cluster and are predicted to be transcribed divergently. A flagella-related

103 gene cluster (AZC_0625-0655) containing fliM (AZC_0643) is encoded between

104 the two chemotaxis genes (cheY2 and cheZ) and the che gene cluster (Fig. 1A).

105 This favors the hypothesis that the che gene cluster encodes a chemotaxis

106 pathway that controls flagella-mediated motility. Furthermore, using transmission

107 electron microscopy it was found that A. caulinodans ORS571 cells possess one

108 to three flagella when grown in liquid medium (Fig. 1B).

109 ccchechehehe clusterclustercluster controls chemotaxis responseresponseresponse of A. caulinodans ORS571ORS571ORS571 Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

110 To determine the role of the che cluster in the chemotactic response of A. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

111 caulinodans ORS571, we constructed two mutant strains (see methods). The

112 ΔcheA mutant is carrying a complete deletion of the cheA gene while a second

Wei Liu, 9, MPMI

Page 10 of 68

113 mutant carries a deletion of the che cluster spanning cheAWY1BR, and referred to

114 as ΔcheA-R. Both mutants displayed a wild-type flagellation pattern (Fig. 1C and

115 D) while, as expected a mutant strain carrying a deletion of the fliM gene (ΔfliM )

116 encoding one of three components of the flagella motor's switch complex was

117 devoid of flagella (Fig. 1E). Chemotaxis behavior of these mutant strains was

118 compared to that of the wild type in soft agar plate assay with carbon sources

119 previously identified as chemoeffectors for A. caulinodans ORS571 (Jiang et al.

120 2016b). In presence of succinate, lactate or proline as the sole carbon source and

121 chemoeffector, the ΔcheA and ΔcheA-R mutant strains expanded from the point

122 of inoculation and formed chemotactic rings that were significantly reduced in size,

123 compared to the wild type, regardless of whether they grew under nitrogen fixation

124 conditions or with combined nitrogen available (Fig. 2A and B). The chemotaxis

125 defect of the ΔcheA mutant strain was rescued by introduction of a plasmid Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

126 carrying the wild type cheA gene into the ΔcheA mutant (ΔcheA-com) (Fig. 2A This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

127 and B). Moreover, the chemotaxis pattern displayed by the ΔcheA and ΔcheA-R

128 mutants in this assay was similar to that of the ΔfliM, which is non-motile mutant

Wei Liu, 10, MPMI

Page 11 of 68

129 strain. These results suggest that the che cluster plays an essential role in

130 chemotaxis of A. caulinodans ORS571.

131 To rule out the possibility that the reduced chemotactic ring diameter of the

132 mutants were linked to a defect in bacterial growth, growth rates of each strain

133 was determined in L3 medium (10 mM succinate as carbon source and 10 mM

134 NH4Cl as nitrogen source). The results show that the growth rate of mutants is

135 similar to that of the wild type (Fig. 3A). Furthermore, since the cheA gene is

136 located directly upstream of cheWY1BR and transcribed in the same direction (Fig.

137 1A), we determined whether deletion of the cheA affected cheWY1BR genes

138 expression by comparing the expression of these genes in the wild type, ΔcheA

139 and ΔcheA-R strains and as expected, found similar expression profiles of

140 cheWY1BR genes in the wild type and ΔcheA (Fig. 3B). These data suggests that

141 the diminished diameters observed on swim plate assays are related to defects in Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

142 chemotaxis response. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

143 To further assess chemotaxis behavior of these mutants, competitive

144 quantitative capillary assays were conducted using bacterial suspension

Wei Liu, 11, MPMI

Page 12 of 68

145 containing equal ratio of the wild type and each of these mutants respectively. Fig.

146 4 shows data obtained with capillary tubes filled with buffer, succinate or sodium

147 lactate as the attractants. Determination of the number of bacterial cell in capillary

148 tubes filled only with buffer revealed an excess of the wild type as compared to

149 the ΔcheA, ΔcheA-R or ΔfliM (Fig. 4), suggesting that the swimming motility bias

150 of the mutant cells was impaired. Furthermore, when the capillary tubes were

151 filled with succinate or sodium lactate, the number of the ΔcheA, ΔcheA-R or

152 ΔfliM mutant cells was also significantly reduced compared to the wild type under

153 similar conditions (Fig. 4). These results confirmed that the ΔcheA and ΔcheA-R

154 mutants are impaired in cell chemotaxis response and motility behavior.

155 Involvement of the cccheAcheAheAheA andandand checheche clusterclustercluster in the controlthe control ofofof flagellaflagellaflagella-flagella---drivdrivdrivdriveeeennnn motility

156 behavior

157 Chemotaxis ultimately controls the probability of change in the swimming Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

158 direction or swimming motility bias. Chemotaxis mutants are thus expected to This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

159 display change in the swimming motility bias compared to parent strain. The

160 free-swimming behavior of the wild type and the mutant strains were analyzed by

Wei Liu, 12, MPMI

Page 13 of 68

161 motion tracking. The swimming paths of the wild type cells showed that their

162 swimming as long runs alternated with sudden direction changes (turns) (Movie 1).

163 In contrast, the ΔcheA (Movie 2) and ΔcheA-R (Movie 3) mutants swam with

164 incessant changes in the swimming direction (tumbles). The non-motile mutant

165 ΔfliM was used as a control (Movie 4). The motility patterns indicate that

166 chemotaxis signaling in A. caulinodans ORS571 regulates the probability of

167 change in swimming direction and that the unique Che pathway, through CheA, is

168 the major regulator of this behavior.

169 TheTheThe ΔΔΔchechecheAcheAAA and ΔΔΔcheAcheAcheA-cheA-R--RRR mutantmutantmutantsmutantsss affectaffectaffect bbbiofilmbiofilmiofilmiofilmssss ffformationformationormationormation

170 Motility and chemotaxis play important roles in attachment and biofilms

171 formation under static conditions, with the swimming bias being important traits

172 controlling this behavior (Merritt et al. 2007). The relative biofilms production

173 ability of the wild type and che mutant was compared using crystal violet staining Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

174 (Fig. 5A). Quantitative data further confirmed that the decrease in staining This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

175 observed with the mutant strains correlated with a decrease in biofilm mass

176 compared to the wild type (Fig. 5B). In order to further distinguish the possible

Wei Liu, 13, MPMI

Page 14 of 68

177 roles of motility or chemotaxis in biofilms formation, the non-motility mutant ΔfliM

178 was also analyzed for its ability to form biofilms and found to be defective

179 compared to the wild type. These data indicate that the Che signaling pathway

180 modulates biofilms formation probably through its effect on cell flagella-driven

181 motility.

182 TheTheThe ΔΔΔchechecheAcheAAA and ΔΔΔcheAcheAcheA-cheA-R--RRR mutantsmutantsmutants impairedimpairedimpaired in biosynthesis of exopolysaccharides

183 Exopolysaccharides (EPS) are known as key matrix components involved in the

184 process of biofilms formation by bacteria (Ryder et al. 2007; Watnick and Kolter

185 2000). Thus, EPS production by the ΔcheA and ΔcheA-R mutants was compared

186 with that of the wild type. The quantitative results showed that the ΔcheA and

187 ΔcheA-R mutants produced less total EPS than the wild type under all tested

188 conditions. Interestingly, the ΔfliM mutant was also defective in EPS production

189 (Fig. 6). Taken together, these data indicate that chemotaxis signaling via che and Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

190 flagellar motility modulates EPS production, with this phenotype likely also This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

191 contributing to the reduced biofilms formation.

192 TheTheThe checheche clusterclustercluster is essential for adsorption and colonization ononon roots.roots.roots.

Wei Liu, 14, MPMI

Page 15 of 68

193 Quantitative assays were used to analyze the adsorption and colonization

194 ability of the different strains on the root surfaces to document the role of the che

195 cluster in the symbiotic association with the host plant S. rostrata. We used a

196 competitive assay to compare the adsorption ability of the mutants and wild type.

197 Surface-sterilized seedlings were co-inoculated with the wild type and mutants at

198 an approximate 1:1 ratio for 4h. Results indicate that the ΔcheA and ΔcheA-R

199 mutants were significantly impaired in their absorption to S. rostrata roots

200 compared to the wild type, whereas the complementation strain (ΔcheA-com)

201 could restore the phenotype of ΔcheA. In addition, we found that the absorption

202 ability of the ΔfliM mutant was also severely impaired (Fig. 7A).

203 We further hypothesized that the lack of adsorption to roots should result in a

204 defective colonization by the ΔcheA and ΔcheA-R mutants. First, to establish

205 overall colonization pattern, quantitative assay was used to compare the Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

206 competitive colonization ability of the wild type and the mutants. Three-day-old This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

207 seedlings were co-inoculated with the wild type and mutants at 1:1 ratio for seven

208 days. Colonization efficiency was determined by counting the number of cells

Wei Liu, 15, MPMI

Page 16 of 68

209 re-isolated from the root surfaces at the end of the incubation period. As expected,

210 the number cells of the ΔcheA and ΔcheA-R mutants recovered from the root

211 surface was significantly reduced compared to wild type cells. However, the wild

212 type phenotype could be restored in the complementation strain ΔcheA-com.

213 Similarly, the ΔfliM mutant deficient in flagella was shown to be less competitive in

214 colonization of the host (Fig. 7B). Give that both the ΔcheA and ΔcheA-R mutants

215 as well as the non-flagellated ΔfliM mutant are severely impaired in adsorption

216 and surface colonization of S. rostrata roots, this suggests that the Che signaling

217 pathway contributes to root symbiosis through regulating flagella-based motility.

218 The Che pathwaypathwaypathway modulatesmodulatesmodulates competitive nodule formation

219 Surface colonization is essential for effective symbiotic nodulation of Sesbania

220 (Mitra et al. 2016).To further test whether the chemotaxis mutants have any

221 symbiotic nodulation defects, the nodule formation ability was used to compare Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

222 between the wild type and mutants. When inoculated alone on stems or roots of S. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

223 rostrata, the ΔcheA and ΔcheA-R mutants formed normal stem (Fig. 8A) or root

224 nodules (Fig. 8B), with no obvious difference in the morphology compared to the

Wei Liu, 16, MPMI

Page 17 of 68

225 wild type. Further, the competitive nodulation ability on stems and roots was used

226 to compare between the wild type and ΔcheA or ΔcheA-R or ΔfliM mutants. All

227 these mutants were severely defective in the ability to compete for nodulation of

228 root and stem compared to the wild type (Fig. 8C). Together, these results

229 demonstrate that Che pathway mediated flagella motility is important for

230 promoting competitive nodule formation.

231 DISCUSSION

232 Many motile soil bacteria possess multiple chemotaxis operons playing

233 pleiotropic roles in regulating flagella-mediated motility and other functions

234 (Scharf et al. 2016). A. caulinodans is a motile, nitrogen-fixing soil bacterium with

235 a preference for organic acids as carbon source (Dreyfus et al. 1988). There is a

236 single chemotaxis gene cluster (cheAWY1BR) (Fig. 1A) predicted from the

237 genome sequence, but the DNA region containing this cluster contains other Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

238 genes involved also in the chemotactic response, such as tlpA1 (AZC_0660) (Liu This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

239 et al. 2017a), and cheZ (AZC_0621) (Liu et al. 2017b). Here, we confirmed that

240 the Che pathway (CheAWY1BR) directly controls the cellular chemotaxis

Wei Liu, 17, MPMI

Page 18 of 68

241 response and flagella-mediated motility behavior. A mutant lacking partial cheA

242 (Liu et al. 2017a) does not yield a null chemotaxis phenotype in soft agar plates

243 (data not shown), but there is no obvious difference in capillary tube chemotaxis

244 assays and flagella motility behavior (Liu et al. 2017a) compared to ΔcheA-R

245 mutants. This is why, in the present work, the role of CheA was reevaluated using

246 a new construction carrying a complete deletion of the cheA gene (ΔcheA).

247 Indeed, the chemotaxis phenotype and motility behavior of the newly constructed

248 ΔcheA mutant is similar to that of the mutant strain carrying the complete deletion

249 of the che cluster ΔcheA-R (Fig. 2). This indicates that chemotaxis signaling in A.

250 caulinodans ORS571 controls the chemotaxis response and flagella-driven

251 motility and that the unique che cluster, through the unique CheA, is the major

252 regulator of these behaviors.

253 Che operons appear to regulate different cell functions in different bacteria Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

254 (Berleman and Bauer 2005; D’Argenio et al. 2002; Ferrandez et al. 2002; Kirby This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

255 and Zusman 2003; Vlamakis et al. 2004). In the Azospirillum genus, the Che1

256 pathway of Azospirillum brasilense also encodes a conserved set of CheAWYBR

Wei Liu, 18, MPMI

Page 19 of 68

257 homologues controlling swimming speed during chemotaxis and regulating the

258 cell size (Bible et al. 2008; Gullett et al. 2017). In addition, several che operon

259 mutants of A. brasilense had an altered production of total EPS that affected

260 colony morphology as well as clumping and flocculation behaviors (Bible et al.

261 2008). In the present work, the flocculation behavior showed no significant

262 difference between wild type and mutants (Data not shown). However, our data

263 showed that defect in chemotaxis caused by mutation within che genes affected

264 chemotaxis response as well as the total amount of EPS produced.

265 Interestingly, decreased EPS biosynthesis in the ΔfliM mutant is similar to that

266 observed in the che mutants. Given that a non-flagellated mutant had a similar

267 phenotype to non-chemotactic mutants, this suggests that the ΔcheA and

268 ΔcheA-R reduced EPS production is likely indirectly related to chemotaxis

269 behavior and rather, is associated with flagella-driven motility. Several other Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

270 genes involved in the chemotactic response in A. caulinodans were also shown to This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

271 be involved in EPS production. In particular, a mutation of the icpB gene,

272 encoding a soluble chemoreceptor, and mutation of the cheZ-like gene, led to

Wei Liu, 19, MPMI

Page 20 of 68

273 increased EPS production (Jiang et al. 2016b, Liu et al. 2017b) suggesting their

274 involvement in a negative regulation of EPS production while Che pathway and

275 FliM would be involved in a positive regulation of EPS.

276 The decreased in EPS production of the ΔcheA and ΔcheA-R and

277 non-flagellated (ΔfliM) mutant is correlated with decreased biofilm formation (Fig.

278 5), as also reported for a mutant of the transmembrane chemoreceptor tlpA1 (Liu

279 et al. 2017a), while in contrast the cheZ-like mutant strain displayed increased

280 ability of biofilm formation. This suggests a complex regulatory network which

281 needs to be further investigated. In E. coli, biofilm formation in the ΔcheA-Z

282 (non-chemotactic) was similar to that of the wild type, and only the non-flagellated

283 or paralyzed flagellated mutants were defective in biofilms formation. This finding

284 indicates bacterial motility, but not chemotaxis is essential for biofilms formation in

285 E. coli (Pratt and Kolter 1998), but the chemotaxis and its effect on the swimming Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

286 motility pattern appear to be important in biofilms formation by several soil This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

287 bacteria such as Azorhizobium caulinodans and Azospirillum brasilense but an

288 exact mechanism is yet to be identified.

Wei Liu, 20, MPMI

Page 21 of 68

289 Previous research indicates that chemotaxis and motility can confer a

290 significant advantage of bacterial competitiveness in the rhizosphere (Brencic and

291 Winans 2005), colonization on plant roots (Bernabéu-Roda et al. 2015; de Weger

292 et al. 1997; Dekkers et al. 1998; Miller et al. 2007; Simons et al. 1996; Vande

293 Broek and Vanderleyden 1995; Vande Broek et al. 1998), and the efficiency of

294 nodule initiation (Caetano-Anollés et al. 1988). In S. meliloti, several non-motile or

295 non-chemotatic mutants (che-, mot-, and fla-) were all less effective than parent as

296 competitors in root adsorption and nodulation, and the motility and chemotaxis

297 appears are quantitatively important in root adsorption and nodule initiation

298 (Caetano-Anollés et al. 1988). In Rhizobium leguminosarum, only the che1 cluster

299 controls the swimming motility bias and chemotaxis as well as competitive nodule

300 formation, while che2 cluster that has a minor effect on chemotaxis is dispensable

301 for competitive nodulation. In A. brasilense, Che1 chemotaxis pathway has a Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

302 minor role in chemotaxis and root surface colonization. However, Che4 pathway This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

303 regulates the swimming pattern of motile A. brasilense cells and wheat root

304 surface colonization. The Che4 pathway is orthologous to the major pathway

Wei Liu, 21, MPMI

Page 22 of 68

305 controlling all chemotaxis responses in S. meliloti (Armitage and Schmitt 1997)

306 and R. leguminosarum (Caetano-Anollés et al. 1988), where this chemotaxis

307 system is also essential for plant root colonization. In this study, although the

308 ΔcheA and ΔcheA-R mutants are not impaired in their absorption and nodulation

309 ability after inoculation alone, all these mutants appeared to be severely impaired

310 in their competitive colonization and nodulation behavior, suggesting that these

311 genes positively control efficient nodulation, as also reported for the IcpB and

312 TlpA1. In contrast, CheZ, was found to have a negative role (Liu et al. 2017b). In

313 addition, the non-motility (ΔfliM) mutants were also severely defective in

314 competitive nodule formation on roots and stems of host plant. These results

315 suggest that chemotaxis signal transduction pathway controls flagella-driven

316 motility, which regulates competitiveness in the rhizosphere.

317 In establishment of a successful symbiosis between rhizobia and legumes, in Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

318 addition to chemotaxis and motility, bacterial surface lipopolysaccharides (LPS) This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

319 and extracellular polysaccharide (EPS) are essential for colonization and nodule

320 development (Gao et al. 2001; Mathis et al. 2005; Mitra et al. 2016). Our results

Wei Liu, 22, MPMI

Page 23 of 68

321 reported here support the conclusion that the Che signaling pathway in A.

322 caulinodans not only provide a competitive advantage in root colonization by

323 regulating chemotaxis and mediating flagella-driven but also indirectly affected

324 competitive colonization through affection biofilm formation or EPS biosynthesis.

325 MMMATERIALSMATERIALS AND METHODS

326 StrainsStrainsStrains and growth condition

327 All strains and plasmids used in this study are listed in Table 1. A. caulinodans

328 ORS571 wild type strain and its derivatives were grown in L3 or TY liquid medium

329 at 37°C with shaking (180 rpm). The L3 medium is supplemented or not 10 mM

330 NH4Cl. L3 and TY medium containing 100 μg/ml ampicillin and 25 μg/ml nalidixic

331 acid antibiotics were prepared as described previously (Jiang et al. 2016b). E. coli

332 was grown in LB medium or agar plates at 37°C.

333 Construction of mutantmutantmutantsmutantsss and cccomplementcomplementomplementomplementationationationation ssstrainstraintraintrain Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

334 To construct a marker less che gene cluster deletion mutant (deletion of cheA, This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

335 cheW, cheY1, cheB and cheR genes), a 825-bp upstream fragment (UF) was

336 amplified by PCR from the genomic DNA of A. caulinodans ORS571 using

Wei Liu, 23, MPMI

Page 24 of 68

337 primers CheUF and CheUR (Table 2), and a 801-bp downstream fragment (DF)

338 was amplified by PCR using primers CheDF and CheDR (Table 2). The PCR

339 product corresponding to the upstream DNA fragment was digested with

340 EcoRI-NdeI and then inserted into the pCM351 plasmid (Marx and Lidstrom 2002)

341 digested with EcoRI-NdeI. The resulting plasmid was designated as pCM351::UF.

342 The PCR product corresponding to the downstream fragment was digested with

343 SacII-SacI and was cloned into the pCM351::UF. The final plasmid

344 pCM351::UF::DF was transformed into E. coli DH5α and the DNA sequence was

345 verified by sequencing prior to mating. The recombinant plasmid was transferred

346 into A. caulinodans ORS571 by tri-parental conjugation with the helper plasmid

347 pRK2013 (Figurski and Helinski 1979). Recombinants from double homologous

348 recombination were obtained and screened from on TY plates by selecting for

349 gentamicin-resistance and tetracycline-sensitivity (Marx and Lidstrom, 2002). The Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

350 potential mutant was generated by in-frame deletion of an internal 5756-bp This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

351 fragment and insertion of a gentamicin resistance gene. Then the gentamicin

352 gene was then deleted by introduction of the Cre expression plasmid pCM157

Wei Liu, 24, MPMI

Page 25 of 68

353 (Marx and Lidstrom 2002). The correct mutation was verified by PCR with primer

354 pair CheAF and CheAR (Table 2), and a confirmed mutation was named

355 ΔcheA-R.

356 To construct a complete deletion mutant of cheA gene, a 501-bp upstream

357 fragment (UF) was amplified using primer pair CheAUF and CheAUR (Table 2),

358 and a 576-bp downstream fragment (DF) was amplified by PCR using primers

359 CheADF and CheADR (Table 2). The PCR product of the upstream DNA

360 fragment was digested with NsiI-NdeI and then inserted into the pCM351 plasmid

361 digested with NsiI-NdeI. The resulting plasmid was termed as pCM351::UF. The

362 PCR product corresponding to the downstream fragment was digested with

363 ApaI-AgeI and was cloned into the pCM351::UF. The plasmid pCM351::UF::DF

364 was transformed into E. coli DH5α and the DNA sequence was verified by

365 sequencing. The recombinant plasmid was transferred into A. caulinodans Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

366 ORS571 by tri-parental conjugation. The potential mutant was generated by This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

367 in-frame deletion of an internal 2793-bp fragment and insertion of a gentamicin

368 resistance gene. Then the gentamicin gene was then deleted by introduction of

Wei Liu, 25, MPMI

Page 26 of 68

369 plasmid pCM157 (Marx and Lidstrom 2002). The introduction of the correct

370 mutation was verified by PCR with primer pair CheAF and CheAR (Table 2), and a

371 confirmed mutation was named ΔcheA.

372 For complementation of ΔcheA, the cheA ORF with 601-bp immediately

373 upstream non-coding sequence was amplified by PCR using primer pair CheACF

374 and CheACR (Table 2). The amplified 3394-bp fragment was inserted into SpeI

375 and XbaI sites of the broad-host-range cloning vector pBBR1MCS-2 (Kovach et al.

376 1995). The ligation mixture was transformed into E. coli DH5α. DNA sequences

377 from individual clones were sequenced and one verified clone was designated as

378 pBBR1MCS-2-cheA-com. This recombinant plasmid, named pBBR-cheA was

379 then introduced into the ΔcheA mutant via tri-parental mating and the

380 transformants were recovered by selection for kanamycin resistance. One of the

381 resulting strains was designated as ΔcheA-com. Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

382 To construct a fliM gene deletion mutant, a 695-bp upstream fragment (UF) was This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

383 amplified by PCR from the genomic DNA of A. caulinodans ORS571 using primer

384 pair FliMUF and FliMUR (Table 2), and a 579-bp downstream fragment (DF) was

Wei Liu, 26, MPMI

Page 27 of 68

385 amplified by PCR using primers FliMDF and FliMDR (Table 2). The PCR product

386 corresponding to the upstream DNA fragment was digested with KpnI-NdeI and

387 then inserted into the pCM351 plasmid digested with KpnI-NdeI. The resulting

388 plasmid was termed as pCM351::UF. The PCR product corresponding to the

389 downstream fragment was digested with AgeI-SacI and was cloned into the

390 pCM351::UF. The final plasmid pCM351::UF::DF was transformed into E. coli

391 DH5α and the DNA sequence was verified by sequencing prior to mating. The

392 recombinant plasmid was transferred into A. caulinodans ORS571 by tri-parental

393 conjugation. The potential mutant was generated by in-frame deletion of an

394 internal 930-bp fragment and insertion of a gentamicin resistance gene. Then the

395 gentamicin gene was then deleted (Marx and Lidstrom 2002). The mutation was

396 verified by PCR with primer pair FliMF and FliMR (Table 2), and a confirmed

397 mutation was named ΔfliM. Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

398 Electron microscopy

This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. Electron microscopy

399 Strains were grown overnight with shaking at 37°C in TY to mid-logarithmic

400 phase. Cultures were inspected to ensure motility, and cells were taken directly

Wei Liu, 27, MPMI

Page 28 of 68

401 from the cultures and adsorbed to Formvar-coated copper grids for 5 min. Excess

402 culture was blotted with filter paper and the grid was placed face down on a drop

403 of 1% phosphotungstic acid for 1 min. Excess stain was blotted with filter paper,

404 and the grid was dried. The images were obtained using JEM-1400 (Japan)

405 electron microscope.

406 BehavioBehavioBehavioralBehavioral assays

407 Soft agar plates were used to detect chemotactic responses of the wild type

408 and mutants to different carbon sources. L3 plates (0.35% agar) supplemented

409 with 10 mM carbon source (succinate, sodium lactate or proline, respectively) as

410 attractant and containing or not 10 mM NH4Cl as nitrogen source. Cells were

411 grown up to mid-exponential phase in TY medium and were harvested by

412 centrifugation. The cultures were washed twice and then resuspended in the

413 chemotaxis buffer (10 mM K2HPO4, 10 mM KH2PO4, 0.1 mM EDTA, pH 7.0) to Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

414 give an OD600 of 1.0. Aliquots of 5 μl suspension were used to inoculate on L3 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

415 plates. The plates were incubated at 37° for 48-72 h and then swimming

416 diameters were recorded.

Wei Liu, 28, MPMI

Page 29 of 68

417 Quantitative capillary chemotaxis assays were performed as previously

418 described in Reyes-Darias et al. (2016) with some minor modifications. The wild

419 type and mutants were grown to early logarithmic phase (OD600 of 0.3-0.4) in TY

420 medium. The cultures were washed and resuspended with chemotaxis buffer to a

421 density of OD600 = 0.05. Then, 300 μl aliquots of cell suspensions (1:1 ratio of

422 wild-type to mutant) were added into each well of a 96-well plate. The open end of

423 capillary tubes (containing 10 mM succinate or sodium lactate as attractant) was

424 placed into the wells. Capillary tubes filled with chemotaxis buffer were used as

425 controls. Subsequently, the sealed end was broken and the content emptied into 1

426 ml sterile water after incubation for 1 h. Serial 10-fold dilutions were placed onto

427 TY plates. Colonies (480) randomly selected on TY plates were analyzed by PCR.

428 To observe the swimming behavior of cells, cultures were grown in TY medium

429 to mid-exponential phase and the swimming behavior was observed and recorded Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

430 using Olympus DP73 digital microscope camera. Videos were analyzed to This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

431 observe the cell paths for 2-3 s. For each strain, at least 5 independent

432 experiments were observed.

Wei Liu, 29, MPMI

Page 30 of 68

433 ReverseReverseReverse transcription PCR

434 Total RNA was isolated from cultured free-living cells of wild type and mutants.

435 cDNA was generated using the TransScript® All-in-One First-Strand cDNA

436 Synthesis SuperMix for qPCR (One-Step gDNA Removal) Kit (Transgen) and

437 diluted 5-fold as template for subsequent PCR amplification. PCR was performed

438 using with gene-specific primer pairs (Table 2). PCR program included an initial

439 denaturation step at 95°C for 3 min, followed by 30 cycles of 95°C for 30 s, 62°C

440 for 30 s and 72°C for 30 s.

441 Biofilm assays

442 To form biofilm on abiotic surfaces (glass tubes), cell cultures (OD600 of 2.5)

443 were resuspended in L3 medium. 150 μl cultures were inoculated into 1.5 ml glass

444 tubes at 37°C for 3-5 days. For crystal violet staining, cultures were removed and

445 washed with ddH2O for 3-5 times. Then 0.5% crystal violet was added into each Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

446 tube. After incubation for 20 min, the crystal violet was removed and washed with This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

447 ddH2O for 5 times, the water was removed completely in the final wash. 1.5 ml of

448 30% acetic acid was added to dissolve the “crystal violet ring” off and the OD570

Wei Liu, 30, MPMI

Page 31 of 68

449 was determined.

450 Exopolysaccharides production assay

451 For the examination of EPS production, overnight cultures were adjusted to

452 OD600 of 1.0, and 15 μl bacterial suspension was inoculated onto solid L3 plates

453 (0.8% agar) and then, incubated at 37°C for 1-3 days. L3 plates contained 10 mM

454 of different carbon sources and are supplemented or not 10 mM NH4Cl as the

455 nitrogen source. Quantification of EPS production was performed based on the

456 method of Nakajima et al. (2012) with some minor modifications. Bacteria on the

457 L3 plates were collected and resuspended in λ-buffer (10 mM Tris-HCl, pH 7.0, 10

458 mM MgSO4). Supernatant containing the EPS soluble fraction were first treated

459 with 1 ml of concentrated sulfuric acid containing 0.2% anthrone, mixed and

460 incubated for 7 min at 100°C, and then quickly chilled on ice. The OD620 of the

461 mixture was measured on a Nanodrop 2000C. D-glucose was used for the Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

462 standard curve. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

463 Adsorption and colonization assay of roots

Wei Liu, 31, MPMI

Page 32 of 68

464 The number of root-adsorbed rhizobia was determined as described in Zamudio

465 and Bastarrachea (1994) with some minor modifications. S. rostrata seeds were

466 germinated and grown under sterile conditions for three days. Cultures of each

467 strain (OD600 nm of 0.2) were used to inoculate seedlings for 4 h at 28°C with slow

468 shaking at 50 rpm. The slow shaking was included to reduce the likelihood that

469 motility differences account for differences in reaching the root surfaces. Then,

470 the seedlings were washed by gentle suspension with sterile water shaking in the

471 rotary bath at 100 rpm for 1 min. Five consecutive washes were performed to

472 remove the excess bacteria. The terminal 2 cm of the inoculated roots of five

473 plants were dispersed in sterile water by homogenizer and dilution spread on TY

474 agar plates. The adsorption pattern of the root surfaces was examined by

475 counting the colonies forming units on TY agar plates. In competitive adsorption

476 experiment, surface-sterilized seedlings were inoculated with mixtures of wild type Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

477 and mutants at approximate 1:1 for 4 h. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

478 Colonization assays were carried out as described in Greer-Phillips et al. (2004)

479 with a few modifications. Wild type and mutants were mixed in 1:1 and added to

Wei Liu, 32, MPMI

Page 33 of 68

480 glass bottles containing 200 ml molten agar (0.35% agar). Ten surface-sterilized

481 seedlings were embedded into each bottle and incubated at 30℃. After seven

482 days, the roots were washed and homogenized by homogenizer. Serial dilutions

483 of the supernatant were spread on TY agar plates to count colony forming units.

484 Colonies were further assigned as detected by PCR. A non-inoculated control

485 flask was included for all experiments.

486 Nodulation assays

487 Nodulation and competition assays were carried out as described in Yost et al.

488 (1998) with the following modifications. For nodulation assays on roots,

489 surface-sterilized seedlings were co-inoculated with each strain in approximate

490 OD600 of 0.5, respectively. For stem nodulation assays, stems root primordial

491 were inoculated with the bacterial suspension by painting using sterile cotton wool.

492 For competitive nodulation assays on roots and stems, surface-sterilized Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

493 seedlings and plant stems were co-inoculated with wild type and mutants in This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

494 approximate 1:1 ratio. All plants were grown at 27°C in the greenhouse with a

495 daylight illumination period of 12 h. Stem nodules were harvested 20-30 days post

Wei Liu, 33, MPMI

Page 34 of 68

496 inoculation. Bacteria were re-isolated from surface-sterilized nodules and colonies

497 identified by PCR. Controls included un-inoculated plants.

498 ACKNOWLEDGMENTS

499 We thank Professors Toshihiro Aono, Shunpeng Li and Zhentao Zhong for

500 kindly providing A. caulinodans ORS571, S. rostrata seeds. We thank Mary E.

501 Lidstrom for kindly providing pCM351 and pCM157 plasmids. This work is

502 financed by the National Natural Science Foundation of China (31370108, and

503 31570063), One Hundred-Talent Plan of Chinese Academy of Sciences (CAS),

504 the High-tech Industrialization Cooperation Funds of Jilin province and the

505 Chinese Academy of Science (2017SYHZ0007), Agriculture Scientific and

506 Technological Innovation Project of Shandong Academy of Agriculture Sciences

507 (+CXGC2016B10), Shandong Province Science Foundation for Youths

508 (ZR2016CB44). Work in the Alexandre laboratory is supported by NSF-MCB Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

509 1330344. This study was conducted with the support of the Institut Pasteur, Paris, This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

510 France.

511

Wei Liu, 34, MPMI

Page 35 of 68

512 LITERATURE CITED

513 Armitage, J. P., and Schmitt, R. 1997. Bacterial chemotaxis: Rhodobacter

514 sphaeroides and Sinorhizobium meliloti-variations on a theme? Microbiol.

515 143:3671-3682.

516 Berleman, J. E., and Bauer, C. E. 2005. Involvement of a Che-like signal

517 transduction cascade in regulating cyst cell development in Rhodospirillum

518 centenum. Mol. Microbiol. 56:1457-1466.

519 Bernabéu-Roda, L., Calatrava-Morales, N., Cuéllar, V., and Soto, M. J. 2015.

520 Characterization of surface motility in Sinorhizobium meliloti : regulation and

521 role in symbiosis. Symbiosis. 67:79-90.

522 Bianciotto, V., Andreotti, S., Balestrini, R., Bonfante, P., and Perotto, S. 2001.

523 Mucoid mutants of the biocontrol strain Pseudomonas fluorescens CHA0 show

524 increased ability in biofilm formation on mycorrhizal and nonmycorrhizal carrot Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

525 roots. Mol. Plant-Microbe Interact. 14:255-260. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

526 Bible, A. N., Stephens, B. B., Ortega, D. R., Xie, Z. H., and Alexandre, G. 2008.

527 Function of a chemotaxis-like signal transduction pathway in modulating motility,

Wei Liu, 35, MPMI

Page 36 of 68

528 cell clumping, and cell length in the alphaproteobacterium Azospirillum

529 brasilense. J. Bacteriol. 190:6365-6375.

530 Bible, A., Russell, M. H., and Alexandre, G. 2012. The Azospirillum brasilense

531 Che1 chemotaxis pathway controls swimming velocity, which affects transient

532 cell-to-cell clumping. J. Bacteriol. 194:3343-3355.

533 Brencic, A., and Winans, S. C. 2005. Detection of and response to signals

534 involved in host-microbe interactions by plant-associated bacteria. Microbiol.

535 Mol. Biol. Rev. 69:155-194.

536 Burdman, S., Okon, Y., and Jurkevitch, E. 2000. Surface characteristics of

537 Azospirillum brasilense in relation to cell aggregation and attachment to plant

538 roots. Critical Rev. Microbiol. 26:91-110.

539 Caetano-Anollés, G., Wall, L. G., De Micheli, A. T., Macchi, E. M., Bauer, W. D.,

540 and Favelukes, G. 1988. Role of motility and chemotaxis in efficiency of Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

541 nodulation by Rhizobium meliloti. Plant Physiol. 86:1228-1235. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

542 Cheng, H. P., and Walker, G. C. 1998. Succinoglycan is required for initiation and

543 elongation of infection threads during nodulation of alfalfa by Rhizobium meliloti.

Wei Liu, 36, MPMI

Page 37 of 68

544 J. Bacteriol. 180:5183-5191.

545 D'Argenio, D. A., Calfee, M. W., Rainey, P. B., and Pesci, E. C. 2002. Autolysis

546 and autoaggregation in Pseudomonas aeruginosa colony morphology mutants.

547 J. Bacteriol. 18:6481-6489.

548 de Weger, L. A., Kuiper, I., van der Bij, A. J., and Lugtenberg, B. J. J. 1997. Use of

549 a lux-based procedure to rapidly visualize root colonization by Pseudomonas

550 fluorescens in the wheat rhizosphere. Antonie Van Leeuwenhoek 72:365-372.

551 Dekkers, L. C., Bloemendaal, C. J., de Weger, L. A., Wijffelman, C. A., Spaink, H.

552 P., and Lugtenberg, B. J. 1998. A two-component system plays an important

553 role in the root-colonizing ability of Pseudomonas fluorescens strain WCS365.

554 Mol. Plant-Microbe Interact. 11:45-56.

555 Doherty, D., Leigh, J. A., Glazebrook, J., and Walker, G. C. 1988. Rhizobium

556 meliloti mutants that overproduce the R. meliloti acidic calcofluor-binding Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

557 exopolysaccharide. J. Bacteriol. 170:4249-4256. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

558 Dreyfus, B. L., Elmerich, C., and Dommergues, Y. R. 1983. Free-living Rhizobium

559 strains able to grow on N2 as sole nitrogen source. Appl. Environ. Microbiol.

Wei Liu, 37, MPMI

Page 38 of 68

560 45:711-713.

561 Dreyfus, B., Garcia, J. L., and Gillis, M. 1988. Characterization of Azorhizobium

562 caulinodans gen. nov., sp.nov., a stem-nodulating nitrogen-fixing bacterium

563 isolated from Sesbania rostrata. Int. J. Syst. Bacteriol. 38:633-639.

564 Ferrandez, A., Hawkins, A. C., Summerfield, D. T., and Harwood, C. S. 2002.

565 Cluster II che genes from Pseudomonas aeruginosa are required for an optimal

566 chemotactic response. J. Bacteriol. 184:4374-4383.

567 Figurski, D. H., and Helinski, D. R. 1979. Replication of an origin-containing

568 derivative of plasmid RK2 dependent on a plasmid function provided in trans.

569 Proc. Natl. Acad. Sci. USA 76:1648-1652.

570 Fraysse, N., Couderc, F., and Poinsot, V. 2003. Surface polysaccharide

571 involvement in establishing the Rhizobium-legume symbiosis. Eur. J. Biochem.

572 270:1365-1380. Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

573 Fujishige, N. A., Kapadia, N. N., De Hoff, P. L., and Hirsch, A. M. 2006. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

574 Investigations of Rhizobium biofilm formation. FEMS Microbiol. Ecol.

575 56:195-206.

Wei Liu, 38, MPMI

Page 39 of 68

576 Gao, M. S., D’Haeze, W., Rycke, R. D., Wolucka, B., and Holsters, M. 2001.

577 Knockout of an Azorhizobial dTDP-L-rhamnose synthase affects

578 lipopolysaccharide and extracellular polysaccharide production and disables

579 symbiosis with Sesbania rostrata. Mol. Plant-Microbe Interact. 14:857-866.

580 Greer-Phillips, S. E., Stephens, B. B., and Alexandre, G. 2004. An energy taxis

581 transducer promote root colonization by Azospirillum brasilense. J. Bacteriol.

582 186:6595-6604.

583 Gullett, J. M., Bible, A. N., and A., Alexandre, G. 2017. Distinct domains confer

584 CheA with unique functions in chemotaxis and cell length in Azospirillum

585 brasilense Sp7. J. Bacteriol. 199(13):JB.00189-17.

586 James, E. K., Loureiro, M. D. F., Pott, A., Pott, V. J., Martins, C. M., Franco, A. A.,

587 and Sprent, J. I. 2001. Flooding-tolerant legume symbioses from the Brazilian

588 Pantanal. New Phytol. 150:723-738. Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

589 Jiang, N., Liu, W., Li, Y. and Xie, Z. H. 2016a. Comparative genomic and protein This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

590 sequence analyses of the chemotaxis system of Azorhizobium caulinodans.

591 Acta Microbiologica Sinica 56:1256-1265.

Wei Liu, 39, MPMI

Page 40 of 68

592 Jiang, N., Liu, W., Li, Y., Wu, H. L., Zhang, Z. H., Alexandre, G., Elmerich, C., and

593 Xie, Z. H. 2016b. A chemotaxis receptor modulates nodulation during the

594 Azorhizobium caulinodans-Sesbania rostrata symbiosis. Appl. Environ.

595 Microbiol. 82:3174-3184.

596 Kirby, J. R., and Zusman, D. R. 2003. Chemosensory regulation of developmental

597 gene expression in Myxococcus xanthus. Proc. Natl. Acad. Sci. USA

598 100:2008-2013.

599 Kovach, M. E., Elzer, P. H., Hill, D. S., Robertson, G. T., Farris, M. A., Roop, R. M,

600 and Peterson, K. M. 1995. Four new derivatives of the broad-host-range cloning

601 vector pBBR1MCS, carrying different antibiotic-resistance cassettes. Gene.

602 166:175-176.

603 Lee, K. B., De Backer, P., Aono, T., Liu, C. T., Suzuki, S., Suzuki, T., Kaneko, T.,

604 Yamada, M., Tabata, S., Kupfer, D. M., Najar, F. Z., Wiley, G. B., Roe, B., Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

605 Binnewies, T. T., Ussery, D. W., D'Haeze, W., Den Herder, J., Gevers, D., This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

606 Vereecke, D., Holsters, M., and Oyaizu, H. 2008. The genome of the versatile

607 nitrogen fixer Azorhizobium caulinodans ORS571. BMC Genomics 9:271-284.

Wei Liu, 40, MPMI

Page 41 of 68

608 Liu, W., Yang, J. B., Sun, Y., Liu, X. L., Li, Y., Zhang, Z. P., and Xie, Z. H. 2017a.

609 Azorhizobium caulinodans transmembrane chemoreceptor TlpA1 involved in

610 host colonization and nodulation on roots and stems. Front. Microbiol. 8:1327.

611 doi: 10.3389/fmicb.2017.01327

612 Liu, X. L., Liu, W., Sun, Y., Xia, C. L., Elmerich, C., and Xie, Z. H. 2017b.

613 Identification of a cheZ-like gene in Azorhizobium caulinodans, a key gene in

614 the control of chemotaxis and colonization of the host plant. Appl. Environ.

615 Microbiol. doi:10.1128/AEM.01827-17.

616 Marx, C. J., and Lidstrom, M. E. 2002. Broad-host-range cre-lox system for

617 antibiotic marker recycling in gram-negative bacterium. BioTechniques

618 33:1062-1067.

619 Mathis, R., Van Gijsegem, F., de Rycke, R., D’Haeze, W., Van Maelsaeke, E.,

620 Anthonio, E., Van Montagu, M., Holsters, M., and Vereecke, D. 2005. Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

621 Lipopolysaccharides as a communication signal for progression of legume This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

622 endosymbiosis. Proc. Natl. Acad. Sci. USA 102:2655-2660.

623 Matthysse, A. G., Marry, M., Krall, L., Kaye, M., Ramey, B. E., Fuqua, C., and

Wei Liu, 41, MPMI

Page 42 of 68

624 White, A. R. 2005. The effect of cellulose overproduction on binding and biofilm

625 formation on roots by Agrobacterium tumefaciens. Mol. Plant-Microbe Interact.

626 18:1002-1010.

627 Meier, V. M., Muschler, P., and Scharf, B. E. 2007. Functional analysis of nine

628 putative chemoreceptor proteins in Sinorhizobium meliloti. J. Bacteriol.

629 189:1816-1826.

630 Meneses, C. H., Rouws, L. F., Simões-Araújo, J. L., Vidal, M. S., Baldani, J. I.

631 2011. Exopolysaccharide production is required for biofilm formation and plant

632 colonization by the nitrogen-fixing endophyte Gluconacetobacter diazotrophicus.

633 Mol. Plant-Microbe Interact. 24:1448-1458.

634 Merritt, P. M., Danhorn, T. and Fuqua, C. 2007. Motility and chemotaxis in

635 Agrobacterium tumefaciens surface attachment and biofilm formation. J.

636 Bacteriol. 189:8005-8014. Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

637 Michiels, K. W., Croes, C. L., and Vanderleyden, J. 1991. Two different modes of This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

638 attachment of Azospirillum brasilense Sp7 to wheat roots. J. Gen. Microbiol.

639 137:2241-2246.

Wei Liu, 42, MPMI

Page 43 of 68

640 Miller, L. D., Yost, C. K., Hynes, M. F., and Alexandre, G. 2007. The major

641 chemotaxis gene cluster of Rhizobium leguminosarum bv. viciae is essential for

642 competitive nodulation. Mol. Microbiol. 63:348-362.

643 Mitra, S., Mukherjee, A., Wiley-Kalil, A., Das, S., Owen, H., Reddy, P. M., Ané, J.

644 M., James, E. K., and Gyaneshwar, P. 2016. A rhamnose-deficient

645 lipopolysaccharide mutant of Rhizobium sp. IRBG74 is defective in root

646 colonization and beneficial interactions with its flooding-tolerant hosts Sesbania

647 cannabina and wetland rice. J. Exp. Bot. 67:5869-5884.

648 Mukherjee, T., Kumar, D., Burriss, N., Xie, Z. H., and Alexandre, G. 2016.

649 Azospirillum brasilense chemotaxis depends on two signaling pathways

650 regulating distinct motility parameters. J. Bacteriol. 198:1764-1772.

651 Nakajima, A., Aono, T., Tsukada, S., Siarot, L., Ogawa, T., and Oyaizua, H. 2012.

652 Lon protease of Azorhizobium caulinodans ORS571 is required for suppression Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

653 of reb gene expression. Appl. Environ. Microbiol. 78:6251-6261. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

654 Pratt, L. A., and Kolter, R. 1998. Genetic analysis of Escherichia coli biofilm

655 formation: roles of flagella, motility, chemotaxis and type I pili. Mol. Microbiol.

Wei Liu, 43, MPMI

Page 44 of 68

656 30:285-293.

657 Ramey, B. E., Koutsoudis, M., von Bodman, S. B., and Fuqua, C. 2004. Biofilm

658 formation in plant-microbe associations. Curr. Opin. Microbiol. 7:602-609.

659 Reyes-Darias, J. A., García, V., Rico-Jiménez, M., Corral-Lugo, A., and Krell, T.

660 2016. Identification and characterization of bacterial chemoreceptors using

661 quantitative capillary and gradient plate chemotaxis assays. Bio-protocol. 6(8).

662 Rinaudi, L. V., and Giordano, W. 2010. An integrated view of biofilm formation in

663 rhizobia. FEMS Microbiol. Lett. 304:1-11.

664 Ryder, C., Byrd, M., and Wozniak, D. J. 2007. Role of polysaccharides in

665 Pseudomonas aeruginosa biofilm development. Curr. Opin. Microbiol.

666 10:644-648.

667 Santaella, C., Schue, M., Berge, O., Heulin, T., and Achouak, W. 2008. The

668 exopolysaccharide of Rhizobium sp. YAS34 is not necessary for biofilm Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

669 formation on Arabidopsis thaliana and Brassica napus roots but contributes to This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

670 root colonization. Environ. Microbiol. 10:2150-2163.

671 Scharf, B. E., Hynes, M. F., and Alexandre, G. M. 2016. Chemotaxis signaling

Wei Liu, 44, MPMI

Page 45 of 68

672 systems in model beneficial plant-bacteria associations. Plant Mol. Biol.

673 90:549-559.

674 Simons, M., van der Bij, A. J., Brand, I., de Weger, L. A., Wijffelman, C. A., and

675 Lugtenberg, B. J. J. 1996. Gnotobiotic system for studying rhizosphere

676 colonization by plant growth-promoting Pseudomonas bacteria. Mol.

677 Plant-Microbe Interact. 9:600-607.

678 Siuti, P., Green, C., Edwards, A. N., Doktycz, M. J., and Alexandre G. 2011. The

679 chemotaxis-like Che1 pathway has an indirect role in adhesive cell properties of

680 Azospirillum brasilense. FEMS Microbiol. Lett. 323:105-112.

681 Skorupska, A., Janczarek, M., Marczak M, Mazur, A., and Krol, J. 2006. Rhizobial

682 exopolysaccharides: genetic control and symbiotic functions. Microb. Cell Fact.

683 16:5-17.

684 Sourjik, V., and Schmitt, R. 1996. Different roles of CheY1 and CheY2 in the Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

685 chemotaxis of Rhizobium meliloti. Mol. Microbiol. 22:427-436. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

686 Stock, J. B., and Surette, M. G. 1996. Chemotaxis. In: Escherichia

687 coli and Salmonella typhimurium: Molecular and Cellular Biology (Neidhardt,

Wei Liu, 45, MPMI

Page 46 of 68

688 F.C., Ed.), Vol. 1, pp. 1103–1129. American Society for Microbiology,

689 Washington, DC.

690 Szurmant, H., and Ordal, G. W. 2004. Diversity in chemotaxis mechanisms

691 among the bacteria and archaea. Microbiol. Mol. Biol. Rev. 68:301-319.

692 Trinick, M. J. 1979. Structure of nitrogen-fixing nodules formed by Rhizobium on

693 roots of Parasponia andersonii Planch. Can. J. Microbiol. 25:565-578.

694 Tsien, H. C., Dreyfus, B. L., and Schmidt, E. L. 1983. Initial stages in the

695 morphogenesis of nitrogen-fixing stem nodules of Sesbania rostrata. J.

696 Bacteriol. 156:888-897.

697 Vande Broek, A., and Vanderleyden, J. 1995. The role of bacterial motility,

698 chemotaxis, and attachment in bacteria-plant interactions. Mol. Plant-Microbe.

699 Interact. 8:800-810.

700 Vande Broek, A., Lambrecht, M., and Vanderleyden, J. 1998. Bacterial Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

701 chemotactic motility is important for the initiation of wheat root colonization by This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

702 Azospirillum brasilense. Microbiol. 144:2599-2606.

703 Vlamakis, H. C., Kirby, J. R., and Zusman, D. R. 2004. The Che4 pathway of

Wei Liu, 46, MPMI

Page 47 of 68

704 Myxococcus xanthus regulates type IV pilus-mediated motility. Mol. Microbiol.

705 52:1799-1811.

706 Wadhams, G. H., and Armitage, J. P. 2004. Making sense of it all: bacterial

707 chemotaxis. Nat. Rev. Mol. Cell Biol. 5:1024-1037.

708 Watnick, P., and Kolter, R. 2000. Biofilm, city of microbes. J. Bacteriol.

709 182:2675-2679.

710 Wells, D. H., Chen, E. J., Fisher, R. F., Long, S. R. 2007. ExoR is genetically

711 coupled to the ExoS-ChvI two-component system and located in the periplasm

712 of Sinorhizobium meliloti. Mol. Microbiol. 64:647-664.

713 Wisniewski-Dyé, F., Borziak, K., Khalsa-Moyers, G., Alexandre, G., Sukharnikov,

714 L. O., Wuichet, K., Hurst, G. B., McDonald, W. H., Robertson, J. S., Barbe, V.,

715 Calteau, A., Rouy, Z., Mangenot, S., Prigent-Combaret, C., Normand, P., Boyer,

716 M., Siguier, P., Dessaux, Y., Elmerich, C., Condemine, G., Krishnen, G., Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

717 Kennedy, I., Paterson, A. H., González, V., Mavingui, P., and Zhulin, I. B. 2011. This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

718 Azospirillum genomes reveal transition of bacteria from aquatic to terrestrial

719 environments. PLoS Genet. 7:1-13.

Wei Liu, 47, MPMI

Page 48 of 68

720 Wuichet, K., and Zhulin, I. B. 2010. Origins and diversification of a complex signal

721 transduction system in prokaryotes. Sci. Signal. 3(128):ra50.

722 Yaryura, P. M., León, M., Correa, O. S., Kerber, N. L., Pucheu, N. L., and Garcı´a,

723 A. F. 2008. Assessment of the role of chemotaxis and biofilm formation as

724 requirements for colonization of roots and seeds of soybean plants by Bacillus

725 amyloliquefaciens BNM339. Curr. Microbiol. 56:625-632.

726 Yost, C. K., Rochepeau, P., and Hynes, M F. 1998. Rhizobium leguminosarum

727 contains a group of genes that appear to code for methyl-accepting chemotaxis

728 proteins. Microbiol. 144:1945-1956.

729 Zamudio, M., and Bastarrachea, F. 1994. Adhesiveness and root hair deformation

730 capacity of Azospirillum strains for wheat seedlings. Soil Biol. Biochem.

731 26:791-797.

732 Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

Wei Liu, 48, MPMI

Page 49 of 68

733 Table 1. Strains and plasmids used in this study.

Strain or plasmid Relevant properties Source or

reference

Strain:Strain:Strain:

A. caulinodans Wild type strain, Ampr, Nalr (Dreyfus et al.

ORS571 1998 )

ΔcheA-R ORS571 derivative, che deletion This study

mutant, Ampr, Nalr,

ΔcheA ORS571 derivative, cheA complete This study

deletion mutant, Ampr, Nalr,

ΔfliM ORS571 derivative, fliM deletion This study

mutant, Ampr, Nalr,

WT-pBBR Wild type strain carrying This study Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

r r r

This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. pBBR1MCS-2, Amp , Nal , Kan

ΔcheA-com ORS571 derivative, cheA deletion This study

Wei Liu, 49, MPMI

Page 50 of 68

mutant carrying pBBR1MCS-2-cheA,

Ampr, Nalr, Kanr

E. coli DH5a F- SupE44 ΔlacU169 (φ80 lacZΔM15) Transgen

hsdR17 recA1 endA1 gyrA96 thi-1

relA1

plasmid :plasmid :

pCM351 Mobilizable allelic exchange vector, (Marx and

Ampr, Genr Lidstrom 2002)

pCM157 IncP plasmid that expresses Cre (Marx and

recombinase, Tetr Lidstrom 2002)

pBBR1MCS-2 Broad host range plasmid, Kanr (Kovach et al.

1995)

pBBR-cheA pBBR1MCS-2 carrying the native This study Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. promoter of the cheA and cheA gene,

Kanr

Wei Liu, 50, MPMI

Page 51 of 68

pRK2013 Helper plasmid, ColE1 replicon, Tra+, (Figurski and

Kanr Helinski 1979)

734 Ampr, ampicillin resistance; Nalr, Nalidixic acid; Genr, gentamicin resistance; Kanr,

735 kanamycin resistance, Tetr, tetracycline resistance.

736

737 Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

Wei Liu, 51, MPMI

Page 52 of 68

738 TTTableTable 222.2. Primers used in this study.

PrimerPrimerPrimer Sequence (5’(5’----3’)*3’)*3’)*3’)* PurposePurposePurpose

CheUF-EcoRⅠ GGAATTCCGAGATCGGTACCCAG ΔcheA-R mutant

GTG construction

CheUR-NdeⅠ GGAATTCCATATGCTGTCACTCCA ΔcheA-R mutant

GCGGGCTG construction

CheDF-SacII TCCCCGCGGCGGGCACATAAACC ΔcheA-R mutant

CAAGGATAG construction

CheDR-SacⅠ CGAGCTCCGTCACGATGCTGCGG ΔcheA-R mutant

AAG construction

CheAF CAGGAAGACTCCGAATACAAGGT Validation of cheA or

che

CheAR GATCATCTCGATGTTCGAGCG Validation of cheA or Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. che

CheAUF-NsiⅠ ATGCATGGACACCATCCGCAAGAT ΔcheA construction

Wei Liu, 52, MPMI

Page 53 of 68

TGAC

CheAUR-NdeⅠ CATATGCTCGCGCAGCAGATCATC ΔcheA construction

CAT

CheADF-ApaⅠ GGGCCCGCCCATGGAGAGGCGG ΔcheA construction

CATGA

CheADR-AgeⅠ ACCGGTCACCACGCCTGAATCATC ΔcheA construction

CAC

CheACF-SpeⅠ ACTAGTTGGCGAGCGAAGTGAAG ΔcheA-com

TC construction

CheACR-XbaⅠ TCTAGAGTCGAACTTGGCGATGTA ΔcheA-com

GTC construction

FliMUF-KpnⅠ GGTACCCGACCGAGCTTCATGAG ΔfliM construction

AT Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. FliMUR-NdeⅠ CATATGGCGGGTCGAAGAGCAGA ΔfliM construction

TC

Wei Liu, 53, MPMI

Page 54 of 68

FliMDF-AgeⅠ ACCGGTCATCGAGCAGCATGTCT ΔfliM construction

FliMDR-SacⅠ GAGCTCGACGCAGTATATCTGCG ΔfliM construction

AC

FliMF ACATAAGCGCCTTCGGCCT Validation of fliM

FliMR GTCATCTTCGACCGCCTGG Validation of fliM

CheAQF GTCTCGTGGTGGACGAGATT RT-PCR for cheA

CheAQR TGAGGTTGCGGTAGAAGAGG RT-PCR for cheA

CheWQF GTCTTCGTACCGGACCACAT RT-PCR for cheW

CheWQR ACCTCGTCGATCATCAGTCC RT-PCR for cheW

CheY1QF CCGACTGGAACATGGAGCCGAT RT-PCR for cheY1

CheY1QR TCAGGCCTCGAACACGGTGTC RT-PCR for cheY1

CheBQF ACATGGTGCTGGAGAAGTCC RT-PCR for cheB

CheBQR AACGACGCTTGTTTCTTCGT RT-PCR for cheB Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. CheRQF CGCTGGCGATGACGCTGAAG RT-PCR for cheR

CheRQR AGGCCCACGAACGGGTTGAG RT-PCR for cheR

Wei Liu, 54, MPMI

Page 55 of 68

16SQF ACGGATTTCTTCCAGCAATG RT-PCR for 16S rRNA

16SQR ACCGGCAGTCCCTTTAGAGT RT-PCR for 16S rRNA

739 ***Engineered* restriction sites are underlined

740

741 Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

Wei Liu, 55, MPMI

Page 56 of 68

742 FigureFigureFigure captionscaptionscaptions

743 Fig.Fig.Fig. 1. Chemotaxis-related genes encoded within the A. caulinodans ORS571

744 genome and flagellation pattern. A,A,A, Organization of chemotaxis-related genes of

745 A. caulinodans ORS571. The arrows indicate the direction of transcription of open

746 reading frames and are drawn relative to scale. B,B,B, Transmission electron

747 micrographs of A. caulinodans ORS571 showing one to three flagella. Cells were

748 negatively stained with phosphotungstic acid. Bar, 2 μm. CCC-C---EEEE,,,, Transmission

749 electron micrographs showing presence of flagella for the ΔcheA-R (CCCC) and

750 ΔcheA (DDDD) mutant strains, and absence of flagella for the ΔfliM (EEEE) mutant.

751 Fig. 2.Fig. 2. Chemotaxis behavior of the ORS571 wild type and mutant strains ΔcheA-R,

752 ΔcheA, the complementation strain ΔcheA-com and the ΔfliM mutant toward

753 attractants on soft agar plates. Carbon sources and nitrogen source were

754 included at a final concentration of 10 mM. A,A,A, Representative L3 plates with Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

755 succinate as the sole carbon source. B, The average diameters of mutants are

This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. B,B,

756 expressed relative to that of wild type (taken as 100%). Error bars represent

757 standard deviations of mean from three repetitions.

Wei Liu, 56, MPMI

Page 57 of 68

758 Fig. 3.Fig. 3. Growth of ORS571 and mutant strains and gene expression. AAA,A,,, Growth

759 curves of the wild type (WT), ΔcheA-R, ΔcheA and ΔfliM in L3 medium. BBB,B,,,

760 Expression of cheWY1BR genes in the wild-type (ORS571) and chemotaxis

761 mutants. The cheWY1BR genes expression was detected by PCR using

762 appropriate primers (Table 2). 16S rRNA gene was amplified as a positive gene

763 control; cheA gene was amplified as a negative gene control. “+” indicates

764 samples incubated with reverse transcriptase; “-” indicates samples with no

765 reverse transcriptase.

766 Fig. 4.Fig. 4. Competitive quantitative capillary chemotaxis assays of the ORS571 wild

767 type, ΔcheA, ΔcheA-R and ΔfliM mutants.... The capillary tubes filled with

768 chemotaxis buffer (left), 10 mM succinate (middle) or sodium lactate (right) as the

769 carbon source were then inserted into the 96-well plate containing the bacterial

770 suspension (mutants and wild type 1:1 ratio). Each data point represents the Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

771 mean of three independent experiments and the error bars represent the standard This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

772 error of the mean.

773 Fig. 5.Fig. 5. Biofilms generated by ORS571 wild type, ΔcheA, ΔcheA-R and ΔfliM

Wei Liu, 57, MPMI

Page 58 of 68

774 mutants. A,A,A, Bacterial strains were grown in glass tubes containing L3 (10 mM

775 succinate as carbon source) medium for 3-5 days without shaking. Biofilms were

776 visualized by staining with crystal violet. B,B,B, Quantification of crystal violet staining

777 was determined the OD value at 570 nm wavelength. The error bars represent

778 standard errors from the mean of five repetitions. Asterisks indicate statistically

779 significant differences (P≤0.05) between the wild type and the mutants.

780 Fig. 6.Fig. 6. Quantitative analysis of the EPS production of the ORS571 wild type,

781 ΔcheA, ΔcheA-R and ΔfliM on different carbon sources plates. The error bars

782 indicate the standard deviations from the mean of three independent experiments.

783 Fig. 7.Fig. 7. Adsorption and colonization ability of the A. caulinodans ORS571 wild type

784 and chemotaxis mutants strains on S. rostrata root surface. AAA,A, Competitive

785 adsorption on S. rostrata roots with the wild type and either the ΔcheA or

786 ΔcheA-R or ΔfliM or complementation strain ΔcheA-com in approximate 1:1. BBB,B,,, Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

787 Competitive colonization levels of S. rostrata roots with the wild type and either This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

788 the ΔcheA or ΔcheA-R or ΔfliM or complementation strain ΔcheA-com in

789 approximate 1:1. The adsorption and colonization ratios were analyzed by PCR.

Wei Liu, 58, MPMI

Page 59 of 68

790 Error bars represent standard errors of the mean calculated from three

791 independent experiments.

792 Fig. 8.Fig. 8. Nodulation tests between the A. caulinodans ORS571 wild type and

793 mutants. A,A,A, Stem nodules were induced by the wild type, ΔcheA and ΔcheA-R.

794 Leghemoglobin of stem nodules shows characteristic orange-brown color. B, Root

795 nodules induced by the wild type, ΔcheA and ΔcheA-R mutants. C,C,C, Analysis the

796 number of the WT, ΔcheA, ΔcheA-R, ΔfliM and complementation strain

797 (ΔcheA-com) in competitive nodulation assay on stems and roots. Error bars

798 represent standard errors of the mean calculated from three independent

799 experiments.

800 MoMoMovMovvviiiieeee 111. 1... The swimming paths of ORS571 wild type cells showed that their

801 swimming as long runs alternated with sudden direction changes.

802 MoMoMovMovvviiiieeee 222.2... The swimming paths of ΔcheA mutant cells showed that their swimming Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018

803 with incessant changes in swimming direction (tumble). This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

804 MoMoMovMovvviiiieeee 333. 3... The swimming paths of ΔcheA-R mutant cells showed that their

805 swimming with incessant changes in swimming direction (tumble).

Wei Liu, 59, MPMI

Page 60 of 68

806 MovMovMoviMoviiieeee 4. The swimming paths of ΔfliM mutant cells. Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.

Wei Liu, 60, MPMI

Page 61 of 68

Fig. 1. Chemotaxisrelated genes encoded within the A. caulinodans ORS571 genome and flagellation pattern. A, Organization of chemotaxisrelated genes of A. caulinodans ORS571. The arrows indicate the direction of transcription of open reading frames and are drawn relative to scale. B, Transmission electron micrographs of A. caulinodans ORS571 showing one to three flagella. Cells were negatively stained with phosphotungstic acid. Bar, 2 m. CE, Transmission electron micrographs showing presence of flagella for the ∆cheA-R (C) and ∆cheA (D) mutant strains, and absence of flagella for the ∆fliM (E) mutant

Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 189x203mm (300 x 300 DPI)

This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. Page 62 of 68

Fig. 2. Chemotaxis behavior of the ORS571 wild type and mutant strains ∆cheA-R, ∆cheA, the complementation strain ∆cheAcom and the ∆fliM mutant toward attractants on soft agar plates. Carbon Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 sources and nitrogen source were included at a final concentration of 10 mM. A, Representative L3 plates with succinate as the sole carbon source. B, The average diameters of mutants are expressed relative to that of wild type (taken as 100%). Error bars represent standard deviations of mean from three repetitions.

This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. 219x273mm (300 x 300 DPI)

Page 63 of 68

Fig. 3. Growth of ORS571 and mutant strains and gene expression. A, Growth curves of the wild type (WT), ∆cheA-R, ∆cheA and ∆fliM in L3 medium. B, Expression of cheWY1BR genes in the wildtype (ORS571) and chemotaxis mutants. The cheWY1BR genes expression was detected by PCR using appropriate primers (Table 2). 16S rRNA gene was amplified as a positive gene control; cheA gene was amplified as a negative gene control. “+” indicates samples incubated with reverse transcriptase; “” indicates samples with no reverse transcriptase.

90x45mm (300 x 300 DPI)

Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. Page 64 of 68

Fig. 4. Competitive quantitative capillary chemotaxis assays of the ORS571 wild type, ∆cheA, ∆cheA-R and ∆fliM mutants. The capillary tubes filled with chemotaxis buffer (left), 10 mM succinate (middle) or sodium lactate (right) as the carbon source were then inserted into the 96well plate containing the bacterial suspension (mutants and wild type 1:1 ratio). Each data point represents the mean of three independent experiments and the error bars represent the standard error of the mean.

80x36mm (300 x 300 DPI)

Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. Page 65 of 68

Fig. 5. Biofilms generated by ORS571 wild type, ∆cheA, ∆cheA-R and ∆fliM mutants. A, Bacterial strains were grown in glass tubes containing L3 (10 mM succinate as carbon source) medium for 35 days without shaking. Biofilms were visualized by staining with crystal violet. B, Quantification of crystal violet staining was determined the OD value at 570 nm wavelength. The error bars represent standard errors from the mean of five repetitions. Asterisks indicate statistically significant differences (P≤0.05) between the wild type and the mutants.

149x127mm (300 x 300 DPI)

Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. Page 66 of 68

Fig. 6. Quantitative analysis of the EPS production of the ORS571 wild type, ∆cheA, ∆cheA-R and ∆fliM on different carbon sources plates. The error bars indicate the standard deviations from the mean of three independent experiments.

90x45mm (300 x 300 DPI)

Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. Page 67 of 68

Fig. 7. Adsorption and colonization ability of the A. caulinodans ORS571 wild type and chemotaxis mutants strains on S. rostrata root surface. A, Competitive adsorption on S. rostrata roots with the wild type and either the ∆cheA or ∆cheA-R or ∆fliM or complementation strain ∆cheAcom in approximate 1:1. B, Competitive colonization levels of S. rostrata roots with the wild type and either the ∆cheA or ∆cheA-R or ∆fliM or complementation strain ∆cheAcom in approximate 1:1. The adsorption and colonization ratios were analyzed by PCR. Error bars represent standard errors of the mean calculated from three independent experiments.

90x45mm (300 x 300 DPI)

Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ. Page 68 of 68

Fig. 8. Nodulation tests between the A. caulinodans ORS571 wild type and mutants. A, Stem nodules were induced by the wild type, ∆cheA and ∆cheA-R. Leghemoglobin of stem nodules shows characteristic orange brown color. B, Root nodules induced by the wild type, ∆cheA and ∆cheA-R mutants. C, Analysis the number of the WT, ∆cheA, ∆cheA-R, ∆fliM and complementation strain (∆cheAcom) in competitive nodulation assay on stems and roots. Error bars represent standard errors of the mean calculated from three independent experiments.

172x169mm (300 x 300 DPI)

Molecular Plant-Microbe Interactions "First Look" paper • http://dx.doi.org/10.1094/MPMI-12-17-0290-R posted 02/09/2018 This paper has been peer reviewed and accepted for publication but not yet copyedited or proofread. The final published version may differ.