Female Xinjiang Brown cattle (age 3–8 months) with no prior infection with Brucella were selected for analysis in this study and housed in an outdoor and restricted access isolation facility. All animal experiments were strictly performed in accordance with the Experimental Animal Regulation Ordinances (2017) formulated by the China National Science and Technology Commission. The protocol was approved by the Committee on Ethics and Welfare of Experimental Animals of Tecon biological Co., Ltd.

B. abortus A19 strain and recombinant plasmid pBBR1MCS-2E containing the E-lysis gene and thermosensitive element λpR-cI857 were constructed and preserved in Tecon Biological Co., Ltd (Urumqi, China) (15).

The method for constructing A19BG was described in detail in another work (15). In brief, the fragment containing the temperature-sensitive regulation system λpR-cI857 and E-gene lysis was amplified from plasmid pBV220::E. The polymerase chain reaction products were cloned into pBBR1MCS-2 to generate the lysis plasmid pBBR1MCS-E (16, 17). The recombinant plasmid pBBR1MCS-2E containing the E-lysis gene and the thermosensitive element λpR-cI857 was electroplated into B. abortus A19 competent cells under the conditions of 200 Ω, 25 µF, and 1800 V. The electroporated bacteria were spread on TSA plates containing 100 µg/mL of kanamycin (Sigma, USA) and incubated at 28°C for 48 h.

A single positive colony was selected and cultured using the shake-culture technique in a 10 mL Brucella broth medium (BD, USA) containing 100 µg/ml kanamycin at 28°C, 150 rpm for 48 h. Then, 2 mL of the bacterial suspension was re-inoculated into the 200 mL Brucella broth medium containing kanamycin (100 μg/mL) and incubated up to the logarithmic growth period (OD₆₀₀ = 0.6–0.8) at 28°C. Next, the culture temperature was elevated to 42°C and culturing was performed for 72h. The bacteria pellet was collected, washed three times with deionised water, and resuspended in 2.5 mL of deionised water, followed by the addition of 7.5 mL of lysis solution and autoclaving before being sent to PTM BioLab, Inc (Hangzhou, China).

Protein samples were prepared as described previously with some modifications (18). Briefly, 1% protease inhibitor (Merck Millipore, Germany) was added to samples followed by ultrasonic lysis and centrifugation. The supernatant was transferred to a new centrifuge tube, and the protein concentration was measured with a BCA kit (Thermo Scientific, USA). An equal amount of each sample was taken for enzymatic hydrolysis, and the volume was adjusted to the same with the lysis solution. One volume of pre-cooled acetone was added following mixing, and then four volumes of pre-cooled acetone were added before precipitation at -20°C for two hours. The precipitate was collected after centrifuging at 4,500 g for 5 min and washing twice with pre-cooled acetone. TEAB was added to the pellet to a final concentration of 200 mM after drying and then ultrasonically dispersed. Trypsin was added at a ratio of 1:50 (protease: protein, m/m), and hydrolysed overnight. Dithiothreitol (DTT) was added to a final concentration of 5 mM and incubated at 56°C for 30 min. Next, iodoacetamide (IAA) was added to make a final concentration of 11 mM and incubated for 15 min at room temperature in the dark. The samples before and after lysis at 42°C were used as two control groups.

The peptides were separated by Ultra-High Performance Liquid system (Thermo Scientific, USA), injected into the nano-electrospray ionisation (NSI) ion source for ionisation, and then entered into the Orbitrap Exploris™ 480 mass spectrometer (Thermo Scientific, USA) for analysis. The ion source voltage was set to 2.3 kV, the FAIMS compensation voltage (CV) was set to -45V and -65V, and then peptide precursor ions and their secondary fragments were detected and analysed by high-resolution Orbitrap. The scanning range of the primary mass spectrum was set to 400-1200 m/z, and the scanning resolution was 60000; the fixed starting point of the scanning range of the secondary mass spectrum was 110 m/z, the secondary scanning resolution was set to 15000, and TurboTMT was set to Off. The data acquisition mode was cycled time-based data-dependent scanning (DDA); that is, the peptide precursor ions were selected according to the order of signal intensity from high to low within a cycle of 1.0 s, and then entered the HCD collision cell using 27% fragmentation. The energy was fragmented, and the second-stage mass spectrometry analysis was also carried out sequentially. In order to improve the effective utilisation of the mass spectrometer, the automatic gain control (AGC) was set to 100%, the signal threshold was set to 5E4 ions/s, the maximum injection time was set to Auto, and the dynamic rejection time of the tandem mass spectrometry scan was set to 20s to avoid repetitive scan of precursor ions.

The raw data from the mass spectrometer were imported into the database search software Proteome Discoverer (v2.4.1.15) for retrieval. The dataset was Brucella_abortus_biovar_1_strain_9941_262698_Brucella_abortus_strain_2308_359391_PR_20210301_combine_20210508.fasta (6100 sequences). Anti-database was added to calculate the false positive rate (FDR) caused by random matching. The common pollution database was added to eliminate the influence of contaminating proteins in the medium. The restriction digestion method was set to Trypsin (Full). The number of missed cleavage sites was set to 2. The minimum length of the peptide was set to 6 amino acid residues. The maximum modification number of the peptide was set to 3. The mass error tolerance of the precursor ions and the secondary fragment ion was set to 10 ppm and 0.02 Da, respectively. Carbamidomethyl was specified as a fixed modification, while oxidation, acetyl (N-terminus), met-loss, and met-loss+acetyl were specified as variable modifications. The FDR for protein, peptide, and PSM identification was set to 1%.

Gene Ontology (GO) annotation of proteins was based on three categories: molecular function, biological process, and cellular component (19). GO annotation was executed via eggnog-mapper software (v2.0) based on the eggnog database. Kyoto Encyclopaedia of Genes and Genomes (KEGG) database was employed to annotate the pathways in which differentially expressed proteins (DEPs) are involved (20). Clusters of Orthologous Groups (COG) database was used to assign the distribution of DEPs (21).

The microarray was developed by using in vitro expression of a cloned recombinant expression vector and aldehyde-modified microarray with a fluorescently labelled (CY5) histidine antibody. Antibodies against IgG were labelled with CY5 and fluorescently labelled secondary antibodies were prepared; immune sera were reacted with the microarrays, followed by reactions with different concentrations of fluorescently labelled secondary antibodies, and the concentrations of primary and secondary antibodies were determined by reaction intensity analysis to establish the microarray detection method for the antibodies. Fifteen Brucella antibody-negative female Xinjiang Brown cattle (age 5–8 months) were randomly divided into three groups (n = 5 per group). In the A19 group, each animal was subcutaneously immunised with 6.0 × 10¹⁰ CFU of the A19 vaccine on the neck. In the A19BG group, the cattle were intramuscularly injected with 5.0 × 10¹⁰ BGs A19BG vaccine on the buttocks. Finally, the control group was injected with normal saline. The cattle were isolated and reared under the same conditions. The sera were selected at 28 days-post immunisation (14 days post-booster immunisation) and reacted with the proteomic microarray, while the levels of IgG antibodies in the serum was measured.

The open reading frames of BMEII0032 and BMEI0892 were amplified by PCR using the DNA from the A19 strain. The amplified DNA fragments were cloned into the pET-28a (Thermo Scientific, USA) vector and transformed in E. coli BL21 (DE3) cells (Transgen, Beijing). The transformed E. coli BL21 cells carrying the pET-28a-BMEII0032 and BMEI0892 plasmid were used for expression studies. Single colonies of transformed cells are incubated overnight at 37°C with continuous shaking at 200 rpm in a 5 ml LB broth medium containing kanamycin (100 µl/ml). 500 μl of culture material was removed and incubated in 200 ml LB broth. The cultures were grown to OD600 = 0.6–0.8at 37°C with vigorous shaking at 200 rpm. Isopropyl-β-D-thiogalactopyranoside (Invitrogen, USA) was added to a final concentration of 1 mM for the expression of BMEII0032 and BMEI0892 recombinant protein. Incubation was continued for 4 hours at 37°C with 200 rpm oscillation. The recombinant proteins were separated and analysed with SDS-PAGE (12%). The recombinant proteins, BMEII0032 and BMEI0892, was purified using an affinity chromatography Ni-NTA column (Cytiva, Sweden), protein folding with Pierce Protein Refolding Kit (Thermo scientific Number 89867, USA). Bradford method with bovine serum albumin (BSA) as a standard was used to assay protein concentration.

Clinical sera from bovine immunised with A19 and A19BG respectively were analysed by indirect ELISA using recombinant BMEII0032 or BMEI0892 as antigens. Immunoassay plates (Corning 42592, USA) were coated with purified recombinant BMEII0032 and BMEI0892 proteins at a 1 µg/ml concentration in the carbonate coating solution and incubated overnight at 4°C. The wells were emptied and washed 3 times with phosphate-buffered saline-Tween 20 (PBST) and then closed with 10% rabbit serum (SBJ Bio, Nanjing, China) at 37°C for 2 hours. Fill the plate with 1/100 dilution of serum and incubate at 37°C for 1 hour. After 5 washes with PBST, the plates are incubated with HRP coupling at 37°C for 1 hour. After washing with PBST, a substrate solution containing TMB (3,3′,5,5′-tetra methyl benzidine, KPL, USA) was loaded into the wells of the plate and the plate was incubated in the dark for 5 minutes at room temperature before the reaction was terminated by the addition of a termination solution (Sera care Life Sciences KPL TMB Microwell Peroxidase). The absorbance was measured at 650 nm in an ELISA reader (Bio-Rad, USA). Each sample was run in duplicate. In addition, as a control for each serum, wells were left uncoated.

The data were analysed by one-way ANOVA using SPSS 20 (IBM, Armonk, NY, USA). The significant differences among treatments were tested based on the least significant difference (LSD) at P ≤ 0.01.

A total of 665,382 mass spectra were generated and 21,942 specific spectra were obtained from the original data with an FDR of 1.0%. Each protein contained at least one specific peptide and the number of proteins could be further quantified to 2,014 ( Figure S1A , Supplementary Table S1 ). For comparative analysis, A19 lysed at 42°C was set as the experimental group (A19BG), and A19 cultured at 28°C was set as the control group (A19). With ratios of 1.50 and 0.67 as the cutoffs for differential up and downregulated expression, the numbers of DEPs that were significantly upregulated and down-regulated in A19BG group were 535 and 581, respectively ( Figure S1B , Supplementary Table S2 ). Quality control results showed that most peptides were 7–20 amino acids in length ( Figure S1C ), which conformed to the general rules based on trypsin enzymatic hydrolysis and HCD fragmentation. Most proteins corresponded to more than two peptides ( Supplementary Table S1 ) and were beneficial to increasing the quantitative results’ accuracy and credibility. The coverage of most proteins was below 20%, and the distribution of proteins above 10 kD was relatively uniform ( Figure S1D ), indicating that there was no obvious bias in the molecular weight of the proteins weighing more than 10 kD, and that proteins with a larger molecular weight (above 100 kD) were not lost due to poor solubility during the preparation process. The repeatability test of samples, including principal component analysis and Pearson’s correlation coefficient ( Figures S1E, F ), also showed a good repeatability.

Most upregulated DEPs were related to the transportation and metabolism of multiple substances, and transcription ( Figure 1A and Supplementary Table S3 ). The functional classification of most down-regulated DEPs included the transportation and metabolism of multiple substances, translation, structure and biosynthesis of ribosome, energy production and conversion, and cell wall/membrane/envelope biosynthesis ( Figure 1B and Supplementary Table S3 ). The number of upregulated DEPs involved in material transportation and metabolism (210) was higher than that of down-regulated ones (182), indicating that the metabolic activity was increased, while energy production was decreased in the A19BG group. In addition, the upregulated DEPs were predominant in transcription function, but the down-regulated DEPs were predominant in the translation, ribosome structure and biosynthesis functions, indicating that only the upstream transcription might be carried out efficiently, while the downstream process of ribosomal translation might be silenced or even blocked in the protein synthesis of A19BG.

The results showed that the down-regulated DEPs were mainly located in the cytoplasm (57%) followed by cytoplasmic membrane (16%) ( Figure 1D and Supplementary Table S4 ). The number of down-regulated DEPs (421) in the cytoplasmic membrane and cytoplasm was higher than that of upregulated ones (305), and DEPs located in the outer membrane were all upregulated proteins ( Figure 1C and Supplementary Table S4 ).

GO enrichment was analysed based on biological process, molecular function and cellular component ( Supplementary Table S5 ). The results showed that most upregulated DEPs after treatment were enriched in cell envelope and periplasmic space. After lysis, the ribosomal and intracytoplasmic metabolism-related proteins were down-regulated in A19BG, such as the ribosome, ribosomal subunit, structural constituent of ribosome, structural molecule activity, rRNA binding, organelle assembly and ribosome assembly, indicating that E protein cleavage inhibits A19 proliferation and leads to loss of cellular contents ( Figure 2A ). The analysis of KEGG functional enrichment showed that the DEPs were mainly concentrated in the ribosome, cell cycle-Caulobacter and galactose metabolism ( Figure 2B ).

KEGG pathway, GO and protein structural domains were then clustered to find the correlation between the functions of the differentially expressed proteins ( Supplementary Tables S6 – S8 ). Four sections (called Q1 to Q4) were grouped according to their differential expression ploidy ( Figure 3A ). For each Q group, GO classification, enrichment of KEGG and protein structural domains were performed separately, and cluster analysis was performed to find the correlations between the functions of the DEPs at different ploidy levels. KEGG enrichment analysis showed that significantly upregulated proteins (Q4) were enriched in five pathways, including galactose metabolism, ABC transporters, secondary bile acid biosynthesis, benzoate degradation, and aromatic compound degradation. The significantly down-regulate proteins (Q1) were enriched in four pathways, including ribosome, cell cycle-Caulobacter, peptidoglycan biosynthesis, and selenocompound metabolism. The upregulated and down-regulated DEPs were mainly concentrated in the ABC transporter and ribosome, respectively ( Figure 3B ). GO differential protein clustering analysis showed that the main differential proteins of A19 and A19BG were ribosome, ribosomal subunit, structural constituent of ribosome, structural molecule activity, organelle assembly and ribosome assembly. Down-regulated expression of a large number of ribosome-associated and intracellular life-activity-related proteins ( Figures 4A–C ). Clustering of differentially expressed proteins from different groups with functionally corresponding protein structural domains revealed upregulation of bacterial extracellular solute-binding proteins, MarR family, amidohydrolase family and bacterial regulatory proteins, gntR family et al. The structural domains of these proteins, such as transketolase, C-terminal domain, S4 domain, biotin-lipoyl like and HemN C-terminal domain were down-regulated in expression ( Figure 4D ).

In order to understand the antigen changes after 42°C -lysis, the differences in immunogenic antigens between A19 and A19BG were compared at the proteome level. Immunogenetic DEPs in A19BG are listed in Table 1 . Bcsp31, Omp25, Omp10, Omp19, Omp28, and Omp2a were the main outer membrane proteins (OMPs), and DnaK, GroS, and GroL were the heat shock proteins (HSPs). These proteins have been evaluated as protective or immunoreactive antigens for subunit vaccines (22–24), and were expressed in both A19 and A19BG, but enriched in A19BG ( Table 1 ). Moreover, LPS in A19BG was similar to that in A19, indicating that A19BG contained the most protective antigens of the parental strain A19. All these findings revealed that A19BG possessed enough immunogenetic antigens to elicit a protective immune response like A19.

Proteome microarray was used to analyse antibody responses after vaccine immunisation to identify and compare the differences of immunogenic proteins. Antibody responses at 28 days post the immunisation were compared with before immunisation. The results showed that after immunisation with A19, antibodies to many proteins were detected, indicating that these proteins are immunogenic in A19. Of the top 20 immunogenic proteins, most of them are membrane proteins. The highly antigenic outer membrane proteins include VirB8, COML competence lipoprotein (BamD), peptidoglycan-associated lipoprotein, peptide ABC transporter substrate-binding protein, outer membrane lipoprotein, membrane fusion protein MTRC and porin family protein etc ( Table 2 and Supplementary Table S9 ). This also indicated that membrane proteins play a major role in the immune response induction for live Brucella A19 vaccines. Then, antibody profiles of A19BG were analysed and evaluated. Compared with that before immunisation, antibodies to a number of proteins were detected 28 days after immunisation with A19BG. Highly antigenic proteins included molecular chaperone GroEL, COML competence lipoprotein (BamD), peptidoglycan-associated lipoprotein, translocation protein TolB, outer membrane lipoprotein, outer membrane protein W, immunogenic protein bp28, outer membrane protein assembly factor BamA, porin family protein and VirB8 etc. However, antibodies to most ribosomal proteins and some intracellular proteins detected in the A19 group were absent in the A19BG group ( Table 2 and Supplementary Table S10 ). The above results showed that the humoral immune responses induced by A19 and A19BG were not similar.

In order to investigate whether the immunogenic proteins of A19 and A19BG have the differential diagnostic ability, both comparative proteome and antibody profiles of the two strains were combined and analysed. The screening process of target antigens is shown in Figure S1 . Proteins that are highly expressed and have high antibody levels in A19, but not detected in A19BG are ideal differential diagnosis antigens for A19BG. With this criteria, we screened the antigenic proteins, a number of proteins were identified ( Table 3 ). To test its feasibility of use as differential diagnostic antigen, these proteins were expressed in E. coli and purified. Using these proteins as coating antigens, the antibody levels of the two clinical immune sera were detected by an indirect ELISA method. Finally, the results showed that the antibody levels of these two proteins, BMEII0032 and BMEI0892, in A19 immunization were significantly higher than those in A19BG immunization ( Figures 5A, B ), with both proteins having A19/A19BG values greater than 2. This suggests that the two antigens can distinguish A19 and A19BG immunization. In view of the high similarity between A19 and wild type strain, the two antigens have the potential to distinguish A19BG immunization from natural infection (25, 26).