Rough brucella strain RM57 is attenuated and confers protection against Brucella melitensis

Accepted Manuscript Rough brucella strain RM57 is attenuated and confers protection against Brucella melitensis Yu Feng, Xiaowei Peng, Hui Jiang, Yong Peng, Liangquan Zhu, Jiabo Ding PII:

S0882-4010(17)30136-5

DOI:

10.1016/j.micpath.2017.03.045

Reference:

YMPAT 2207

To appear in:

Microbial Pathogenesis

Received Date: 14 February 2017 Revised Date:

30 March 2017

Accepted Date: 30 March 2017

Please cite this article as: Feng Y, Peng X, Jiang H, Peng Y, Zhu L, Ding J, Rough brucella strain RM57 is attenuated and confers protection against Brucella melitensis, Microbial Pathogenesis (2017), doi: 10.1016/j.micpath.2017.03.045. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

Rough brucella strain RM57 is attenuated and confers protection against

2

Brucella melitensis

3

Yu Fenga,b*, Xiaowei Penga*, Hui Jianga, Yong Penga, Liangquan Zhua#, Jiabo Dinga#

4

a

5

Beijing 100081,China.

6

Agriculture University, Taian, Shandong 270018, China;

7

* Co-first author; # Correspondence author

b

RI PT

Department of Inspection Technology Research, China Institute of Veterinary Drug Control,

AC C

EP

TE D

M AN U

SC

College of Animal Science And Veterinary Medicine, Shandong

ACCEPTED MANUSCRIPT Abstract： ：Brucella melitensis (B. melitensis) is a facultative intracellular pathogen, which is the main

9

epidemic strain in China. To overcome disadvantages of traditional live attenuated vaccines, in this

10

study a rough mutant RM57 was induced from a B. melitensis isolate M1981. In order to uncover the

11

reason of changes in the LPS of RM57, the nucleotide sequences and transcription levels of all known

12

genes related to LPS synthesis were detected. As LPS plays an important role in outer membrane

13

integrity, the sensitivities of RM57 to SDS and polymyxin B were examined. The results showed that

14

the expression level of LPS genes of RM57 was not significantly changed, and RM57 was sensitive to

15

polymyxin B, compared to its parent strain. In further study, the virulence and protective efficacy of

16

RM57 in mice and guinea pigs were determined, and our data indicated that RM57 was attenuated and

17

had good protection effects, especially in guinea pigs model. Overall, these results demonstrated that

18

the artificially induced rough mutant strain RM57 was an efficacious vaccine candidate against the

19

challenge of virulent B. melitensis. Our data presented here provided additional insight into the

20

mechanism of LPS synthesis of Brucella spp.

21

Keywords: Brucella; rough strain; vaccine candidate; LPS

AC C

EP

TE D

M AN U

SC

RI PT

8

ACCEPTED MANUSCRIPT Introduction

23

Brucella spp. are intracellular bacterial pathogens that cause infection in domestic and wild animals,

24

and these Gram-negative pathogens usually infect their hosts through exposure to the infected animals

25

and animal products [1]. Among brucella spp., B.abortus, B.melitensis and B.suis are of particular

26

importance to human health, with an estimated annual incidence of 500,000 human cases worldwide,

27

especially in developing countries [1].

28

Similar to many LPSs of Gram-negative pathogens, the Brucella LPS plays an important role in the

29

interaction with the corresponding host. The Brucella LPS consists of three covalently bound canonical

30

motifs anchored in the OM: lipid A, the most hydrophobic part that is embedded in the OM; a core

31

oligosaccharide that creates a bridge to the O-antigen, and O-polysaccharide (O-PS), the third and most

32

external surface moiety [2]. Depending on the inclusion or lack of O-polysaccharide (O-PS) moiety,

33

Brucella strains can be divided into smooth type (S) or rough type (R). Through preventing deposition

34

of complement factors at the cell surface and impairing the binding of antimicrobial peptides to the

35

membrane, Brucella O-PS confers resistance to host innate bactericidal response [3]. In addition, the

36

O-PS could hinder the production of proinflammatory cytokines [4], and along with the core provides

37

receptor moieties for brucella phages [5]. All these studies have demonstrated that Brucella LPS is

38

critical to counteract an early host immune response, thus allowing a successful intracellular infection.

39

Vaccination against brucellosis is an important control strategy to prevent the disease in

40

high-prevalence regions [1]. At present, live Brucella vaccines (B. abortus S19, B. melitensis Rev.1, B.

41

suis S2) induce effective immune protection against brucellosis, and are widely used in many countries

42

around the world [6, 7]. However, as all three vaccines carry a bacterial surface antigen with an

43

immunodominant region (O-PS), they could persistently induce antibodies that interfere with the

AC C

EP

TE D

M AN U

SC

RI PT

22

ACCEPTED MANUSCRIPT diagnosis of brucellosis. In addition, vaccination with these vaccines may cause abortion in pregnant

45

animals [8, 9]. Thus, a safer vaccine, which could not interfere with diagnosis of brucellosis, is

46

urgently needed for brucellosis eradication campaigns worldwide. To overcome the disadvantages of

47

these vaccine strains mentioned above, scientists have endeavored to improve current vaccine strains or

48

to design novel vaccines with the deletion of O-PS (rough LPS) while remain satisfactory

49

immunogenic properties [10].

50

In the present study, we obtained a rough mutant RM57, and assessed the phenotype of this rough

51

mutant. To explain the changes of LPS in RB57, sequences of LPS related genes and their transcription

52

levels were detected , and evaluate the mutant RM57 as a potential candidate vaccine against B.

53

melitensis infections.

54

Materials and Methods

55

Bacterial strains, primers and media

56

Strains used in this study are listed in Table S1. The primers used in this study are listed in Table S2.

57

Brucella strains were cultured on tryptic soy agar (TSA, Bacto) or in tryptic soy broth (TSB) at 37 ℃

58

on a rotary shaker. All bacterial strains were stored at -80 ℃ in medium supplemented with 25 %

59

(vol/vol) glycerol. All experiments described in this study was performed in a biosafety level 3

60

laboratory in the China Institute of Veterinary Drug Control.

61

Animals

62

Four- to 6-week-old female BALB/c mice and 350-400 g female Hartley guinea pigs were purchased

63

from Wei Tong Li Hua Laboratory Animal Services Centre (Beijing, China), bred in individually

64

ventilated cage rack systems, and subsequently transferred to the biosafety level 3 facilities of China

AC C

EP

TE D

M AN U

SC

RI PT

44

ACCEPTED MANUSCRIPT Institute of Veterinary Drug Control at the beginning of the experiments. All animal experiments were

66

conducted following the regulations enacted by the Beijing Administration Office of Laboratory

67

Animals.

68

Construction of RM57 rough mutant

69

Method of inducing RM57 rough mutant was similar to that of the M111B vaccine strain used in China

70

[11]

71

in TSB and then anti-sera of rough Brucella was added to a final concentration of 1U/ml. This mixture

72

was placed for 24h at 4℃, and the upper bacterial suspension was discarded. The rest of mixture was

73

diluted with TSB medium with different dilutions and then overlaid onto TSA plates. The plates were

74

incubated at 37°C for 72 h, and then tested by the crystal violet exclusion assay for screening of rough

75

clones. Subsequently the rough clones (F1) were selected and the processes described above were

76

repeated 57 times. More than 97% of clones exhibited the rough genotype and this rough brucella

77

strain was designated RM57.

78

To determine the genetic stability of RM57 in host, guinea pigs infection model was used. In brief,

79

guinea pigs were infected intraperitoneally with 1× 105 CFU, and the guinea pigs were sacrificed at 2

80

weeks postinfection. Serial dilutions of spleen homogenates were plated on TSA, and the clones on the

81

TSA were stained using crystal violet. Rough clones were cultured and infected guinea pigs

82

intraperitoneally as mentioned above. According the The People's Republic of China beast uses the

83

biological preparations, these steps were repeated 14 times..

84

Phenotypic characterization of the RM57 mutant

RI PT

65

AC C

EP

TE D

M AN U

SC

. In brief, a single clone of B. melitensis M1981 (F0), isolated in a farm, was grown for ~12h at 37℃

ACCEPTED MANUSCRIPT The phenotypes of the RM57 rough mutant was characterized by coagglutination of the killed bacterial

86

suspensions with the acriflavine solution and by colony staining with crystal violet solution [12].The

87

parent strain B. melitensis M1981, B. melitensis M28 (smooth) and B. canis RM6/66 (rough) were used

88

as phenotype controls.

89

Detergent sensitivity and polymyxin B sensitivity tests

90

Detergent sensitivity and polymyxin B sensitivity tests were performed as previously described [13].

91

Sequencing of known LPS related genes

92

One microliter of genomic DNA of B.melitensis M1981 or RM57 was used as the template for the PCR.

93

Primers were designed according to the sequences of the known LPS related genes (Table S2). The

94

PCR products were analyzed with 1% agarose gel electrophoresis, purified and sequenced by the

95

Beijing Genomics Institute (Shenzhen, China).

96

RNA isolation and real-time quantitative PCR

97

Bacteria were grown in TSB at 37 °C until the log phase was reached. Total RNA was isolated and

98

reverse-transcribed into cDNA using random primers as previously described [14]. Primers of LPS

99

related genes were designed and were listed in Table S2. Reactions performed without reverse

100

transcriptase were used as negative controls to test for DNA contamination. Real-time qPCRs were

101

performed with SYBR green mix in 96-well optical reaction plates (Applied Biosystems). The results

102

for each target mRNA were normalized to 16S rRNA transcript levels and averaged.

103

Virulence studies

AC C

EP

TE D

M AN U

SC

RI PT

85

ACCEPTED MANUSCRIPT Mice were inoculated intraperitoneally with 100 µl (105 CFU) of RM57, the parental strain M1981,

105

vaccine strain S2, or M28, respectively. Five mice of one group were euthanized via carbon dioxide

106

asphyxiation at 1, 2, 3, 4, 5, or 6 weeks postinfection. Hartley guinea pigs were inoculated

107

intraperitoneally with 1 ml (109 CFU) of RM57, the parental strain M1981, vaccine strain S2, or M28,

108

respectively. Five mice in one group were euthanized via carbon dioxide asphyxiation at 1, 2, 4, 6, or 8

109

weeks post-infection.

110

At each time point, the spleens were harvested, weighed, and homogenized in 1 ml of peptone saline.

111

Serial dilutions were prepared, and 100-µl aliquots of each dilution (including the undiluted organ)

112

were plated in duplicate onto TSA plates.

113

Protection test in mice and guinea pigs

114

Five mice were intraperitoneally inoculated at a dose of 105 CFU/mouse for RM57 or the vaccine

115

strain B. suis S2, respectively. Another five mice were intraperitoneally inoculated with 0.1 ml PBS as

116

a control. Each mouse was challenged with 1 × 105 CFU of the wild-type strain B. melitensis M28 at 8

117

weeks after vaccination. Four weeks later, the challenged mice were euthanized as described above.

118

Nine guinea pigs were intraperitoneally inoculated at a dose of 3 × 109 CFU/mouse for RM57 or the

119

vaccine strain B. suis S2, respectively. Another nine guinea pigs were intraperitoneally inoculated with

120

1 ml PBS as a control. Each mouse was challenged with 30 CFU of the wild-type strain B.

121

melitensis M28 at 60 days after vaccination. Thirty days later, the challenged mice were euthanized as

122

described above.

123

Spleens were collected and homogenized in 1 ml of PBS, serially diluted, and plated onto TSA. The

124

challenged bacterial burden of the spleen was used to measure the protective immunity index.

AC C

EP

TE D

M AN U

SC

RI PT

104

ACCEPTED MANUSCRIPT Statistical analysis.

126

A Student's t test was performed to analyze the data from the mouse virulence and protection

127

experiments, and a P value of <0.05 was considered significant.

128

Result

129

RM57 was a rough mutant induced from smooth-type parent strain

130

After inducing with anti-sera of rough Brucella for 57 generations, smooth B. melitensis M1981

131

became a stable rough mutant RM57. As shown in Fig. 1, clones of the RM57 and 6/66 strains (a

132

naturally rough strain, brucella canis ) were obviously stained by crystal violet (Fig. 1A) and cultures

133

of RM57 could be agglutinated by acriflavine solution (Fig. 1B). These results demonstrated that

134

RM57 was a rough strain.

135

The sequences and transcription levels of LPS related genes were not changed in RM57

136

To determine if some genes involved in LPS synthesis were mutated, all known LPS related genes

137

were amplified and sequenced. The results indicated that the nucleotide sequences of these genes in

138

RM57 were completely the same as the parent strain M1981. The transcription levels of these genes

139

in both RM57 and M1981 were detected, and we found that only the transcription level of the pgm

140

gene of RM57 was changed more than 2 fold, compared to that of the wild type M1981(Fig 2).

141

RM57 was sensitive to polymyxin B

142

Bacterial outer membrane plays important roles in resistance to various stresses, such as detergents and

143

antimicrobial peptides. As RM57 had defects in LPS, the sensitivities of RM57 to polymyxin B and

144

SDS were detected. RM57 exhibited a significantly decreased survival ratio as compared with the

AC C

EP

TE D

M AN U

SC

RI PT

125

ACCEPTED MANUSCRIPT parent strain in survival assays in the presence of polymyxin B (Fig. 3A), while no significant

146

difference was observed when it was exposed to SDS (Fig. 3B). These data indicated that RM57 was

147

sensitive to polymyxin B.

148

Attenuation of RM57 in mice and guinea pigs

149

Many genes related to Brucella LPS were previously reported to be important for smooth-type Brucella

150

strains [15]. To determine if the virulence of RM57 altered, BALB/c mice and guinea pigs were

151

infected i.p. with RM57, M1981, M28 and S2 respectively. The CFU, an indicative of concentration of

152

Brucella in each spleen, was determined at different time points post inoculation.

153

In mice infection model, the numbers of viable bacteria recovered from the spleens of RM57 and

154

S2-inoculated mice were found to be much lower than those from mice infected with M1981 or M28.

155

As shown in Fig. 4A, RM57 was not detectable in the inoculated mice at 5 weeks post-inoculation, and

156

the vaccine strain S2 was completely cleared at 4 weeks post-inoculation. Similarily, in guinea pigs

157

infection model, the numbers of viable bacteria recovered from the spleens of RM57 and S2-inoculated

158

guinea pigs were also found to be significantly lower than those from mice infected with M1981 or

159

M28. As shown in Fig. 4B, RM57 and S2 were not detected from most inoculated guinea pigs at 3

160

weeks post-inoculation.

161

Protective efficacy of RM57 in mice and guinea pigs

162

To evaluate the potential protective immunity induced by RM57 against the M28 virulent-strain

163

challenge, the number of challenge strain recovered from the spleens of all vaccinated mice or

164

guinea pigs were compared to the numbers recovered from the control animals. Protective immunity

165

was expressed as log10 units of protection [16]. As shown in Table 1, S2 (5.47 protection units)

AC C

EP

TE D

M AN U

SC

RI PT

145

ACCEPTED MANUSCRIPT conferred greater protection than RM57 (2.62 protection units) (P < 0.05). Meanwhile, the challenge

167

strain M28 was recovered in all of the challenge control mice, and there was a statistically significant

168

difference (P < 0.05) between the immunization groups (S2 and RM57) and the challenge control

169

groups in protective immunity. As shown in Table 2, there was a statistically significant decrease in the

170

splenic bacterial loads from the guinea pigs vaccinated with RM57 or S2 relative to those of the naive

171

guinea pigs regardless of the vaccination dose (1.35 and 1.55 protection units, respectively; P<0.05).

172

Discussion

173

In this study, we induced a rough mutant RM57 from M1981, which was a smooth type B. melitensis

174

strain isolated in China. The phenotypes of RM57 were characterized, and the virulence and protective

175

efficacy of RM57 in mice and guinea pigs were also evulatted. Our results demonstrated that RM57

176

was a rough-type Brucella, while the sequences and transcription levels of LPS related genes were

177

almost unchanged in RM57, compared to its parent strain. In addition, we showed that RM57 was

178

sensitive to ploymyxin B, which might explain the decrease of virulence of RM57 in mice and

179

guinea pigs infection models. Most importantly, RM57 provided good protection against challenge

180

with B. melitensis M28 in guinea pigs model.

181

Brucella LPS are exposed on the bacteria surface, and could be divided into three parts: the lipid A, the

182

core oligosaccharides, and the O-specific polysaccharide (O-PS). Brucella LPS is considered as one of

183

the virulence factors that allow the pathogen to escape early detection by the host immune system [4,

184

17]. Various studies have demonstrated that mutants lacking the O-PS (rough mutants) are attenuated

185

[15]. In this study, the rough mutant RM57 was also proved to be attenuated in both mice and

186

guinea pigs models. In addition, consist to other rough strains reported, RM57 was very sensitive to

AC C

EP

TE D

M AN U

SC

RI PT

166

ACCEPTED MANUSCRIPT polymyxin B [18]. To uncover the reason why RM57 was a rough strain, the sequences and

188

transcription levels of all known genes involved in LPS synthesis were tested. Among these genes, 11

189

genes (gmd, per, wbkC, wbkA, wbkE, wbkD, wbkF, wzm, wzt, wboA, and wboB) were involved in

190

synthesis of O-PS, and no change of these O-PS related genes was found at both gene and transcription

191

level. The other genes related to synthesis of lipid A or core oligosaccharide were also detected, but

192

still no change of gene sequences was observed. Compared to the parent strain M1981, pgm was the

193

only gene with altered transcription level. Previous studies have demonstrated that the pgm gene

194

deleted mutant was rough and attenuated in both cell and mice infection models [18]. Therefore, the

195

increased level of pgm expression likely contributed to the generation of the rough phenotype of RM57.

196

Nevertheless, as the transcription level of pgm only increased a little more than two fold, we considered

197

that it might not be the sole or principal reason for the phenotype change of RM57. Thus, most likely

198

the change of LPS in RM57 was caused by some unknown mechanism. To address this question,

199

genomics and transcriptomics approaches will be used in the following study.

200

Although many widely used live attenuated vaccines targeting brucellosis (such as S19, Rev.1 and S2)

201

are effective, there are a number of drawbacks in practice. At present, all live attenuated vaccines

202

mentioned above are unable to differentiate between the natural and vaccinated forms of the infection,

203

in addition, these vaccines could cause abortion in pregnant animals [19]. To overcome these problems,

204

some researchers have tried to develop novel vaccines, such as subunit, recombinant proteins and DNA

205

vaccines. Although these vaccines are safety, the protection efficacy of these vaccines is lower than the

206

live attenuated vaccines in animal models [20].

207

In addition, some researchers have also attempted to develop rough-type live attenuated vaccines as a

208

solution. The B. abortus RB51 vaccine is a good example, as it does not interfere with diagnosis and

AC C

EP

TE D

M AN U

SC

RI PT

187

ACCEPTED MANUSCRIPT retains the capacity to induce protection [21-23]. In addition to RM51, B. melitensis B115 is also an

210

attenuated rough strain, and is considered to have significant protective immunity in mice against the

211

challenge of B. melitensis 16M, B.ovis, and B.abortus 2308, equivalent to what is provided by B.

212

melitensis Rev.1 [24, 25]. A different attenuated live rough vaccine strain, B. abortus 45/20, confers

213

protection in cattle, but the vaccine strain easily reverts to smooth pathogenic forms in vivo [23, 26].

214

Besides, through deleting the wboA gene involved in synthesis of LPS in B. melitensis, some

215

researchers constructed rough vaccine candidate, and found this candidate could induce great

216

protection against the challenge of B. melitensis 16M, B. abortus 2308, and B. melitensis NI [27].

217

However, it was reported that the protective immunity induced by rough Brucella mutants was inferior

218

to that induced by the smooth vaccine strains in sheep and goats, so the feasibility of developing

219

rough Brucella vaccine strains was questioned gradually [28, 29].

220

In China, B.melitensis is the main epidemic strain, and our rough mutant RM57 was also induced from

221

a B. melitensis isolate M1981. Unlike many induced rough vaccines (such as RB51), our rough strain

222

did not induce exogenous antibiotic resistance, since a new method of changing Brucella LPS was used

223

in inducing RM57. In this study, our data have demonstrated this rough mutant was attenuated in

224

different animal models, and showing good protective efficacy against B.melitensis challenge,

225

particularly in guinea pigs model. Overall, these data indicated that RM57 is a potential vaccine

226

candidate without interference in serological diagnosis. Further study using ruminant models (cattle

227

and sheep) to evaluate the protective efficacy of RM57 is now under way.

228

Acknowledgments

229

YF performed the experiments, analyzed the data (including statistical analysis), and drafted the

AC C

EP

TE D

M AN U

SC

RI PT

209

ACCEPTED MANUSCRIPT manuscript. XWP analyzed the data, drafted the manuscript, HJ and YP treated the samples and

231

performed the experiments. LQZ prepared and provided the reagents exclusively for this study.

232

LQZ and JBD performed the experiments, analyzed the data, and edited and finalized the

233

manuscript. All authors read and approved the final manuscript.

234

Funding

235

This study was supported by a grant award from Chinese National Natural Science Foundation (NO.

236

31602017) and National Key Research and Development Program of China (NO.2016YFD0500902).

237

Salaries and research support were also provided by funds appropriated to the China Institute of

238

Veterinary Drug Control, and College of Science and Veterinary Medicine, Shandong Agriculture

239

University.

240

Conflict of interest

241

The authors declare that they have no conflict of interest.

242

References

243

[1] Pappas G, Akritidis N, Bosilkovski M, Tsianos E. Brucellosis. British Medical Journal.

244

2005;352:2325-36.

245

[2] Mancilla M. Smooth to Rough Dissociation in Brucella: The Missing Link to Virulence. Frontiers in

246

Cellular & Infection Microbiology. 2015;5:98.

247

[3] Eisenschenk FC, Houle JJ, Hoffmann EM. Mechanism of serum resistance among Brucella abortus

248

isolates. Veterinary Microbiology. 1999;68:235-44.

249

[4] Barquero-Calvo E, Chaves-Olarte E, Weiss DS, Guzmán-Verri C, Chacón-Díaz C, Rucavado A, et al.

250

Brucella abortus uses a stealthy strategy to avoid activation of the innate immune system during the

AC C

EP

TE D

M AN U

SC

RI PT

230

ACCEPTED MANUSCRIPT onset of infection. Plos One. 2007;2:e631.

252

[5] Monreal D, Grilló MJ, González D, Marín CM, De Miguel MJ, López-Goñi I, et al. Characterization of

253

Brucella abortus O-polysaccharide and core lipopolysaccharide mutants and demonstration that a

254

complete core is required for rough vaccines to be efficient against Brucella abortus and Brucella ovis

255

in the mouse model. Infection & Immunity. 2003;71:3261-71.

256

[6] Alton GG. Further studies on the duration of the immunity produced in goats by the Rev. 1 Brucella

257

melitensis vaccine. Journal of Comparative Pathology. 1968;78:173-8.

258

[7] Nielsen, Klaus. Animal brucellosis: CRC Press; 1990.

259

[8] Alton GG, Elberg SS, Crouch D. Rev. 1 Brucella melitensis vaccine. Journal of Comparative Pathology.

260

1967;77:293-300.

261

[9] Blasco JM. A review of the use of B. melitensis Rev 1 vaccine in adult sheep and goats. Preventive

262

Veterinary Medicine. 1997;31:275-83.

263

[10] Shumilov KV, Sklyarov O, Klimanov A. Designing vaccines against cattle brucellosis. Vaccine.

264

2010;28 Suppl 5:F31.

265

[11] Ding J, Pan Y, Jiang H, Cheng J, Liu T, Qin N, et al. Whole genome sequences of four Brucella strains.

266

Journal of Bacteriology. 2011;193:3674.

267

[12] Wang Z, Niu J, Wang S, Lv Y, Wu Q. In vivo differences in the virulence, pathogenicity, and induced

268

protective immunity of wboA mutants from genetically different parent Brucella spp. Clinical &

269

Vaccine Immunology Cvi. 2013;20:174-80.

270

[13] † WJL, Dong H, Peng XW, Wu QM. The Cyclic AMP-Binding Protein CbpB in Brucella melitensis and

271

its role in cell envelope integrity, resistance to detergent and virulence. Fems Microbiology Letters.

272

2014;356:79-88.

AC C

EP

TE D

M AN U

SC

RI PT

251

ACCEPTED MANUSCRIPT [14] Liu W, Dong H, Liu W, Gao X, Zhang C, Wu Q. OtpR regulated the growth, cell morphology of B.

274

melitensis and tolerance to β-lactam agents. Veterinary Microbiology. 2012;159:90-8.

275

[15] D G, MJ G, MJ DM, T A, V A-G, RM D, et al. Brucellosis Vaccines: Assessment of Brucella melitensis

276

Lipopolysaccharide Rough Mutants Defective in Core and O-Polysaccharide Synthesis and Export. Plos

277

One. 2008;3:e2760.

278

[16] Juan Esteban Ugalde DJC, M. Susana Leguizamón, Rodolfo Augusto Ugalde. Evaluation of Brucella

279

abortus Phosphoglucomutase (pgm) Mutant as a New Live Rough-Phenotype Vaccine. Infection &

280

Immunity. 2003;71:6264.

281

[17] Lapaque N, Moriyon I, E, Gorvel J. Brucella lipopolysaccharide acts as a virulence factor. Current

282

Opinion in Microbiology. 2005;8:60-6.

283

[18] Ugalde JE, Czibener C, Feldman MF, Ugalde RA. Identification and characterization of the Brucella

284

abortus phosphoglucomutase gene: role of lipopolysaccharide in virulence and intracellular

285

multiplication. Infection & Immunity. 2000;68:5716-23.

286

[19] Zhiqiang LI, Zhang J, Zhang KE, Qiang FU, Wang Z, Tiansen LI, et al. Brucella melitensis 16MΔTcfSR

287

as a potential live vaccine allows for the differentiation between natural and vaccinated infection.

288

Experimental & Therapeutic Medicine. 2015;10:1182.

289

[20] Arenasgamboa AM, Riceficht AC, Kahlmcdonagh MM, Ficht TA. Protective Efficacy and Safety of

290

Brucella melitensis 16MΔmucR against Intraperitoneal and Aerosol Challenge in BALB/c Mice.

291

Infection & Immunity. 2011;79:3653.

292

[21] Cheville NF. Development, testing and commercialization of a new brucellosis vaccine for cattle.

293

Annals of the New York Academy of Sciences. 2000;916:147-53.

294

[22] Schurig GG, Nd RR, Bagchi T, Boyle S, Buhrman D, Sriranganathan N. Biological properties of RB51;

AC C

EP

TE D

M AN U

SC

RI PT

273

ACCEPTED MANUSCRIPT a stable rough strain of Brucella abortus. Veterinary microbiology. 1991;28:171-88.

296

[23] Schurig GG, Sriranganathan N, Corbel MJ. Brucellosis vaccines: past, present and future.

297

Veterinary Microbiology. 2002;90:479.

298

[24] Adone R, Francia M, Ciuchini F. Evaluation of Brucella melitensis B115 as rough-phenotype

299

vaccine against B. melitensis and B. ovis infections. Vaccine. 2008;26:4913-7.

300

[25] Adone R, Francia M, Pistoia C, Pesciaroli M, Pasquali P. B. melitensis rough strain B115 is

301

protective against heterologous Brucella spp. infections. Vaccine. 2011;29:2523-9.

302

[26] Moriyón I, Grilló MJ, Monreal D, González D, Marín C, Lópezgoñi I, et al. Rough vaccines in animal

303

brucellosis: structural and genetic basis and present status. Veterinary Research. 2004;35:1-38.

304

[27] Wang Z, Niu J, Wang S, Lv Y, Wu Q. In vivo differences in the virulence, pathogenicity, and induced

305

protective immunity of wboA mutants from genetically different parent Brucella spp. Clinical &

306

Vaccine Immunology Cvi. 2012;20:174-80.

307

[28] Barrio MB, Grilló MJ, Muñoz PM, Jacques I, González D, de Miguel MJ, et al. Rough mutants

308

defective in core and O-polysaccharide synthesis and export induce antibodies reacting in an indirect

309

ELISA with smooth lipopolysaccharide and are less effective than Rev 1 vaccine against Brucella

310

melitensis infection of sheep. 2009;27:1741-9.

311

[29] Elzer PH, Enright FM, Mcquiston JR, Boyle SM, Schurig GG. Evaluation of a rough mutant of

312

Brucella melitensis in pregnant goats. Research in Veterinary Science. 1998;64:259.

SC

M AN U

TE D

EP

AC C

313

RI PT

295

ACCEPTED MANUSCRIPT Figure legends

315

Figure 1. RM57 was a rough mutant strain. (1A) The results of crystal violet staining for RM57

316

(A),6/66(B), M1981(C) and M28 (D). (B) Results of acriflavine agglutination test for RM57

317

(A),6/66(B), M1981(C) and M28 (D).

318

Figure 2. The transcription levels of LPS synthesis related genes on RM57, brucella M1981 as the

319

reference strain.

320

Figure 3. RM57 exhibited a cell envelope defect in vitro. (A) Sensitivity of RM57 and its parent strain

321

to polymyxin B. (B) Sensitivity of RM57 and its parent strain to SDS. Values represent the means of

322

one of three experiments performed in duplicate, and error bars indicate the SD. *P < 0.05, **P < 0.01

323

(unpaired t-test).

324

Figure 4. RM57 was significantly attenuated in mice model (A) and guinea pigs (B). Values represent

325

the means of one of three experiments performed in duplicate, and error bars indicate the SD.

AC C

EP

TE D

M AN U

SC

RI PT

314

ACCEPTED MANUSCRIPT Table 1 Protection against challenge with B. melitensis M28 in mice.

Log10 CFU in spleen (±SD)a

UPb

RM57

4.18±0.46 c

2.62

S2

1.33±2.64c

5.47

PBS control

6.80±0.33

SC

a Mean and SD of the log10 CFU per spleen

RI PT

Treatment group(n=5)

b UP, unit of protection. Average of log10 CFU in the spleens of PBS-inoculated mice minus average of

M AN U

log10 CFU in the spleens of vaccinated mice.

AC C

EP

TE D

c P<0.05 (significant) compared with the value for PBS control in each challenge group.

ACCEPTED MANUSCRIPT Table 2 Protection against challenge with B. melitensis M28 in guinea pigs.

Log10 CFU in spleen (±SD) a

UPb

RM57

2.32±0.49c

1.35

S2

2.12±0.50c

1.55

PBS control

3.67±0.68

SC

a Mean and SD of the log10 CFU per spleen

RI PT

Treatment group(n=9)

b UP, unit of protection. Average of log10 CFU in the spleens of PBS-inoculated guinea pigs minus

M AN U

average of log10 CFU in the spleens of vaccinated guinea pigs.

AC C

EP

TE D

c P<0.05 (significant) compared with the value for PBS control in each challenge group.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Research highlight 1. A rough mutant RM57 was induced from a B. melitensis isolate M1981 2. RM57 was attenuated and had good protection effects, especially in guinea pigs model.

AC C

EP

TE D

M AN U

SC

RI PT

3. RM57 was an efficacious vaccine candidate against the challenge of virulent B. melitensis.