Effects of Mycobacterium vaccae vaccine in a mouse model of tuberculosis: protective action and differentially expressed genes

Background Tuberculosis is a leading cause of death worldwide. BCG is an effective vaccine, but not widely used in many parts of the world due to a variety of issues. Mycobacterium vaccae (M. vaccae) is another vaccine used in human subjects to prevent tuberculosis. In the current study, we investigated the potential mechanisms of M. vaccae vaccination by determining differentially expressed genes in mice infected with M. tuberculosis before and after M. vaccae vaccination. Methods Three days after exposure to M. tuberculosis H37Rv strain (5 × 105 CFU), adult BALB/c mice randomly received either M. vaccae vaccine (22.5 μg) or vehicle via intramuscular injection (n = 8). Booster immunization was conducted 14 and 28 days after the primary immunization. Differentially expressed genes were identified by microarray followed by standard bioinformatics analysis. Results M. vaccae vaccination provided protection against M. tuberculosis infection (most prominent in the lungs). We identified 2326 upregulated and 2221 downregulated genes in vaccinated mice. These changes could be mapped to a total of 123 signaling pathways (68 upregulated and 55 downregulated). Further analysis pinpointed to the MyD88-dependent TLR signaling pathway and PI3K-Akt signaling pathway as most likely to be functional. Conclusions M. vaccae vaccine provided good protection in mice against M. tuberculosis infection, via a highly complex set of molecular changes. Our findings may provide clue to guide development of more effective vaccine against tuberculosis.


Background
Since the discovery of Mycobacterium tuberculosis by Robert Koch over a century ago [1], human beings have made significant achievements in the fight against tuberculosis (TB). However, with the increase of multidrugresistant (MDR) strains, human immunodeficiency virus (HIV) co-infection, and lack of effective TB vaccines, TB remains a major threat to human health [2].
Bacillus Calmette-Guérin (BCG), the first vaccine used against TB, is prepared from a strain of the attenuated live Mycobacterium bovis. A major limitation of BCG is the variable efficacy across ethnicity and population [2][3][4]. Vaccae™ vaccine is one of the most promising vaccines against TB. It is a non-cell Mycobacterium vaccae vaccine produced by Anhui Zhifei Longcom [5]. Our previous study suggested that it played an important role in improving immunity, promoting phagocytosis, regulating bidirectional immunoreaction, and reducing pathological damage [2]. At present, this vaccine has been given a Chinese new drug certificate and approved by the China Food and Drug Administration (CFDA) for the adjuvant treatment of TB. Currently, a large doubleblind Phase III trial has been completed to evaluate the efficacy and safety of the Vaccae™ vaccine in 10,000 cases whose skin tests of PPD (purified protein derivative) were strongly positive in Guangxi province in China [6], and the results have not yet been published.
Although the potential mechanisms of M. vaccae vaccine immunotherapy have been studied from the immunological and proteomic levels [19][20][21], the molecular mechanism of the immunotherapeutic effect of this vaccine is still unclear. Herein, we assessed the immunotherapeutic effect of the M. vaccae vaccine in mouse animal model and identified the differential expression (DE) genes of mice before and after M. vaccae vaccine treatment for the first time by using DNA (deoxyribonucleic acid) microarray. Based on these data, we hope to identify possible target molecules and signaling pathways of M. vaccae vaccine, which will give a new perspective for the molecular mechanism of its immunotherapy.

Mice and ethics statement
Female BALB/c mice (6-8 weeks of age) were purchased from the Institute of Military Medicine, Academy of Military Sciences of Chinese PLA (People's Liberation Army) (Beijing, China). Experimental protocol was approved by the Animal Ethical Committee of the 8th Medical Center of Chinese PLA General Hospital, and conducted in compliance with the Experimental Animal Regulation Ordinances of the China National Science and Technology Commission.

Immunization and challenge
General experimental design is shown in Fig. 1. The schedule of immunization and challenge is shown in Fig. 2. Mice received 5 × 10 5 colony formation units (CFUs) of M. tuberculosis H37Rv strain via the caudal vein. Three days later, mice were randomly divided to receive intramuscular injection of either M. vaccae vaccine (22.5 μg in 100-μl distilled water) or vehicle (n = 8). Booster immunization was conducted 14 and 28 days after the primary immunization.

Infection severity assessment
Mouse body weight was measured once per week. Eight weeks after the last immunization, the mice per group were killed and the lungs, liver, and spleen were collected for gross pathological observation, histopathological examination, and CFU counting. Firstly, the organ coefficients were evaluated by the ratio of organ weight to body weight, the average areas of lesions in the liver, the number of the tubercular nodules in the lung, and the size of spleen were observed following the standards listed in Table 1. The spleen and the left lobe of lung were homogenized in 3-ml saline, serially diluted (10-fold in each step), inoculated in duplicate on Lowenstein-Jensen medium plate (100 μl) and cultured at 37°C for 4 weeks. Colonies on the medium were counted and the results are showed as CFUs per organ. The right lung was fixed in 10% (vol/vol) formalin overnight and embedded in paraffin. Sections (3 μm) thickness were stained with hematoxylin and eosin (H&E) for histopathological examination as previously described [22][23][24][25][26][27][28].

PBMCs isolation and total RNA extraction
On days 87 after challenge, 3 mice of each group were sacrificed. PBMCs (peripheral blood mononuclear cells) were prepared using a Mouse PBMCs Isolation Kit (TBDscience, Tianjin, China). Total RNA was extracted using a kit from Solarbio Life Science (Beijing, China). The integrity of RNA was assessed by electrophoresis on a denaturing agarose gel. Sharp 28S and 18S rRNA bands at a ratio of 2:1 are used as the hallmark for intact RNA.

DNA microarray
DNA microarray experiment was conducted using a Mouse 4x44K Gene Expression Array (Agilent) with 39, 000+ mouse genes and transcripts, all with public domain annotations.

RNA labeling and array hybridization
Sample labeling and array hybridization were conducted according to the Agilent One-Color Microarray-Based Gene Expression Analysis protocol (Agilent Technology). Briefly, total RNA from each sample was amplified and labeled with Cy3-UTP. Labeled cRNAs were purified by RNeasy Mini Kit (Qiagen), and NanoDrop ND-1000 was used to measure the concentration and specific activity of the labeled cRNAs (pmol Cy3/μg cRNA). One microgram of each labeled cRNA was fragmented by adding 2.2-μl 25 × fragmentation buffer and 11-μl 10 × blocking agents, heated at 60°C for 30 min, and diluted by adding 55μl 2 × GE hybridization buffer. Then, 100 μl of hybridization solution was added into the gasket slide and assembled to the gene expression microarray slide. The slides were incubated for 17 h at 65°C in an Agilent Hybridization Oven. The hybridized arrays Fig. 1 The flow chart of all experiments. Eight mice of M. vaccae group or control group were challenged with M. tuberculosis H37Rv strain. After 3 times immunization, mice were sacrificed and their lungs, livers, and spleens were collected to efficacy evaluation and pathological observation. The PBMCs of 3 mice of each group were isolated to extract total RNA. Hierarchical Clustering was performed to show the distinguishable gene expression profiling between samples. DE genes with statistical significance were identified through Volcano Plot filtering. Finally, pathway analysis and GO Analysis were applied to determine the roles of these DE genes played in these biological pathways or GO terms. GO, Gene Ontology; PBMCs, peripheral blood mononuclear cells; CFUs, colony formation units were washed, fixed and scanned using the Agilent DNA Microarray Scanner (part number G2505C).

Data analysis
Microarray images were analyzed using Agilent Feature Extraction software (version 11.0.1.1). Quantile normalization and subsequent data processing were performed using GeneSpring GX v11.5.1 software package (Agilent Technologies, USA). DE genes were identified through volcano plot filtering. Hierarchical clustering was performed using the Agilent Gene-Spring GX software (version 11.5.1). GO analysis and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis were performed using a standard enrichment computation method.

Statistical analysis
Statistical analyses were conducted using SAS (version 9.1, SAS Institute, Cary, NC). The sample size was estimated according to our previous studies [24][25][26][27]. The results of M. vaccae protective experiments, gross pathological observation, histopathological examination, and CFU count were compared with Student's t-test or Wilcoxon Two-Sample test according to data Fig. 2 The schedule of immunotherapy evaluation. Each mouse in the control group (showed as an orange belt) and M. vaccae group (showed as a green belt) was challenged with M. tuberculosis H37Rv strain at day 0. Three days after challenge, mice were immunized intramuscularly 3 times at 2-week intervals with water (orange belt) and M. vaccae vaccine (green belt), respectively. Eighty-seven days after challenge, each mouse was sacrificed and their organs were collected for protective evaluation and histopathological analysis. Additionally, the weight of each mouse was measured per week (showed as a red notebook) Note: Relative to normal liver and spleen, a swell was less than 20%; b swell was more than 20% and less than 40%; c swell was more than 40% Differences were considered statistically significant at P < 0.05. *, P < 0.05; ***, P < 0.001; ns. Not significant normality and homogeneity of variances. Differential expression was defined as fold-change ≥2. P < 0.05 was considered statistically significant.

Efficacy of the vaccine
With the exception of temporary reduction in the first week, body weight in mice receiving the vehicle increased over the entire experimental period as expected (Fig. 3a). In mice receiving the M. vaccae vaccine, body weight started to decrease on day 14 day, reached a nadir on day 28, and then increased gradually back to the control level on day 77 (Fig. 3a).
In comparison to the control, there was a statistically non-significant trend for decreased CFUs in the lungs in the M. vaccae group (Fig. 3b). The CFUs in the spleen was lower in the M. vaccae group (P = 0.041 vs. control, Fig. 3c). In comparison to the control, mice in the M. vaccae group had similar organ coefficient of the liver (Fig. 3d) and spleen (Fig. 3e), but significantly lower organ coefficient of the lungs (P = 0.026, Fig. 3f).

Histopathological and gross pathological analyses
The structure of alveoli was damaged severely in the control group (Fig. 4a). Inflammatory cell infiltration of the lungs was apparent in the M. vaccae group, but the alveolar wall was intact, with no thickening. Gross pathological analysis showed fewer tubercular nodules in the lungs (P = 0.0002, Fig. 4b) and decreased spleen size (P = 0.0196, Fig. 4c) in the M. vaccae group. The average area of the lesions in the liver did not differ significantly between the two groups ( Fig. 4d).

GO analysis
GO analysis showed that, in comparison with the control group, the upregulated genes involve 1672 terms in biological process (BP, P < 0.05, Table S3 BP sheet), 137 terms in cellular component (CC, P < 0.05, Table S3 CC sheet), and 231 terms in molecular function (MF, P < 0.05, Table S3 MF sheet). The downregulated genes involved 1080 terms in BP (P < 0.05, Table S4 BP sheet), 134 terms in CC (P < 0.05, Table S4 CC sheet), and 195 terms in MF (P < 0.05, Table S4 MF sheet).
The top 10 GO terms of the upregulated genes sorted by enrichment score (left lane in Fig. 6), fold enrichment (middle lane in Fig. 6), and classification (right lane in Fig. 6) in BP, CC, and MF are shown in Fig. 6a, Fig. 6b, and Fig. 6c, respectively. The top 10 GO terms of the downregulated genes are showed in Fig. 6d/E/F. Briefly, the upregulated genes in the M. vaccae group are mainly related to metabolic process, cellular metabolic process, primary metabolic process, intracellular, and binding. The downregulated genes are mainly associated with localization, cellular component organization, metabolic process, cell part, cell periphery, and binding.
The relationship among DE genes associated with upregulated pathways (Fig. 8a) and downregulated pathways (Fig. 8b) were determined by using Gehpi software.

Discussion
The current study showed that the M. vaccae vaccine could decrease the CFUs of M. tuberculosis in mice. Such effect was most robust in the spleen, and statistically significant in the lungs. The organ coefficient of the lungs was decreased. The vaccination attenuated the pulmonary lesion and splenomegaly. These results are generally consistent with the lower M. tuberculosis CFUs, pathological change index, and organ weight index in previous studies [20,29], and indicated that M. vaccae vaccine had a significant immunotherapeutic effect on TB.
Previous studies suggested that the effects of M. vaccae vaccine immunotherapy mainly depend on enhanced recall IFN-γ responses [8], CD3 + CD4 + T cells, IFNγ + CD4 + T cells, natural killer (NK) cells, and reduced IL-4 + CD4 + T cells [29]. However, a systematic review of clinical trials conducted suggested no benefit of M. vaccae vaccine immunotherapy [30]. One study showed that smooth type of M. vaccae could interfere with the production of helper T lymphocytes-1 (Th-1) cytokines, and rough type of M. vaccae could induce the production of Th-1 cytokines [31] by splenocytes, suggesting that the different colonial morphology (smooth type or rough  We speculated that there are significant differences in gene expression profiles before and after M. vaccae vaccine treatment, and identifying these changes could help to understand the regulatory mechanism of the M. vaccae vaccine. We identified 2326 upregulated genes and 2221 downregulated genes in M. vaccae group, suggesting that M. vaccae vaccine induce more complex and specific gene regulation activities in individuals infected with M. tuberculosis. The top 1 upregulated gene was Retnlg (also known as Xcp1, Fizz3, and Relmg), which encodes the resistin-like gamma protein (alternative names, RELM-γ or XCP1). This protein was first identified as a novel member of the resistin-like molecule/ found in inflammatory zone (RELM/FIZZ) family in mice and rats [32]. A subsequent study showed marked increase of Retnlg expression in spontaneously hypertensive hyperlipidemic rats [33], suggesting that RELM-γ has cytokine-like effect and plays a role in promyelocytic differentiation [32,34]. In addition to RELM-γ, several other upregulated genes identified in our study have been previously reported to be associated with TB. For example, decreased transcription of PTGS2 has been shown to confer a survival benefit to M. tuberculosis [35]. GBP2 is one of the prominent hubs in a highly active common core in TB [36]. Blocking CXCL2 could reduce the M. tuberculosis-induced IL-1β production [37,38]. Exposure of murine peritoneal macrophages to M. tuberculosis increases SLPI secretion and accelerates both the phagocytosis and killing of the pathogen [39][40][41], possibly by interacting with S100A8/A9 proteins to decrease lung tissue damage without affecting protective immunity against TB [42]. TNF-α has a prominent role in defense and pathological responses to TB and its production in TB patients has been shown to be increased by the M. vaccae vaccine [43][44][45]. Expression of the IL1A gene is increased in both the TB-infected and the healthy cattle to M. bovis stimulation [46].
The top 1 downregulated gene was Afp encoding alpha-fetoprotein (AFP). AFP is a shuttle protein that transports nutrients to embryonic cells through receptor-mediated endocytosis and converts drugs into AFP-positive bone marrow-derived inhibitors in adults. Previous studies have implicated AFP in the regulation of cell growth, differentiation, apoptosis, angiogenesis, and immune regulation [47]. A few previous studies reported normal AFP in TB patients, but increased AFP in TB patients with hepatocellular carcinoma [48,49]. Monocytes can undergo homotypic fusion to produce different types of multinucleated giant cells in response to M. tuberculosis infection. In comparison to CD9 Low classical monocytes, CD9 High classical monocytes expressed higher levels of tetraspanin CD151, but the role of these cells in immunity remains unknown [50]. Taken together, we identified a number of new downregulated genes in the current study, including Pcdhga9, Cdc42ep5, Hrsp12, Rnasek, Nprl3, 4932443I19Rik, Ly6g6c, Kdr, 2810416G20Rik, Tubb2a, Triqk, Slc6a16, Cxx1c, Fez2, 1810058I24Rik, Egfbp2, Efna5, and 2210013O21Rik. Whether these genes participate in the immune response needs to be investigated in the future.
The upregulated and downregulated genes in the current study are associated with 1672 and 1080 terms in the biological process, 137 or 134 terms in the cellular component, and 231 or 195 terms in the molecular function, indicating the importance of biological process in the regulatory mechanisms of M. vaccae vaccine. Interestingly, GO analysis demonstrated that the most significant GO term of upregulated genes in the biological process is the metabolic process. In contrast, the most significant GO term for downregulated genes in the biological process is localization. It is well known that the immune responses depend on energy. Maintaining adequate energy supply is the basis for immunocytes to attack M. tuberculosis. It has also been shown that M. tuberculosis can adhere to and taken up by alveolar epithelial cells [51,52]. The interactions between M. tuberculosis and host molecules within the alveolar certainly play a key role in determining whether M. tuberculosis could successfully invade the host [53]. These findings suggested that M. vaccae vaccine activate more immunocytes to participate in the elimination of M. tuberculosis by enhancing metabolism, and antagonize the invasion of M. tuberculosis by downregulating the molecules involved in recognition, adhesion, and invasion. KEGG pathway analysis in the current study identified 68 upregulated and 55 downregulated pathways by M. vaccae vaccination. The upregulated pathways most associated with M. vaccae vaccine treatment were TNF signaling pathway, NOD-like receptor signaling pathway, TLR signaling pathway, and mitogen-activated protein kinase (MAPK) signaling pathway. M. tuberculosis is primarily recognized by macrophages via TLR2/4 signaling pathways, but the TLR2 and TLR4 signal can be inhibited by the antigens secreted by bacteria [54], which makes it possible to inhibit autophagy, and allow the long-term presence of M. tuberculosis in macrophages [55]. After M. vaccae vaccination, the expression of TLR2 was significantly enhanced to induce upregulation of inflammatory cytokines (TNF-α, IL-6, IL-12, IL-18, and IL-1) and chemokines (CXCL, MCP-1) via MyD88dependent TLR signaling pathway, NOD-like receptor signaling pathway, and subsequently activating two downstream pathways NF-κB and MAPK to accelerate the killing and elimination of M. tuberculosis [56][57][58]. MyD88 is one of the most extensively investigated adaptor proteins in the TLR signaling cascade, and plays a critical role in immune response to M. tuberculosis infection [59]. Our study determined that the expression of MyD88 is significantly upregulated in response to M. vaccae vaccine. However, MyD88-independent pathway also participates in the host defense against mycobacterial infection [57]. We speculate that M. vaccae vaccination could induce the transition of the TLR signaling pathway from MyD88independent to MyD88-dependent.
Downregulated pathways associated with M. vaccae vaccination in the current study included focal adhesion, ECM-receptor interaction, Rap1 signaling pathway, and PI3K-Akt signaling pathway. Focal adhesions are Fig. 8 Network map of DE genes associated with upregulated or downregulated pathways. The network map of DE genes associated with upregulated a or downregulated b pathways was made by Gehpi software. The pathways were showed as pathway ID such as mmu05168. The color and diameter of each circle represent the degree of enrichment of the genes or pathways, the larger diameter, and redder or pinker color indicate the more upregulated or downregulated pathways the gene involved. In addition, the top 10 genes of upregulated c or downregulated d pathways were also identified integrin-containing, multi-protein structures that form mechanical links between intracellular actin bundles and the extracellular substrate in many cell types [60]. ECM is a highly dynamic structure that provides structural and biochemical support of surrounding cells [61,62]. Both play a dominant role in the control of cell-cell and cell-matrix interactions by regulating the function of integrins and other adhesion molecules in various cell types. In addition, growth factor (GF) is a naturally occurring substance capable of stimulating cellular growth, proliferation, healing, and cellular differentiation [63]. In the present study, we found reduced expression of GF and ECM by the M. vaccae vaccine. Recognition of both molecules and their receptors on cell membrane could induce the activation of PI3K and FAK, thus triggering the downstream signaling events, including PI3K-Akt signaling pathway, Wnt signaling pathway, and Rap1 signaling pathway. These pathways have been implicated in macrophage invasion, M. tuberculosis survival, and impaired immune response [64,65].
There are several limitations to this study. Firstly, the number of mice used to identify DE genes is relatively small (n = 3/group), and therefore must be considered preliminary. Secondly, the study was conducted in BALB/c mice; extrapolation to other animal species, and particularly human beings, must be cautious. Third, the changes induced by the M. vaccae vaccine were not compared to the BCG vaccine. Finally, the upregulated and downregulated signaling pathways were identified by bioinformatics based on microarray data; validation with more quantitative measures and at the protein levels is required.

Conclusions
M. vaccae vaccine produces fairly robust protection against M. tuberculosis. The vaccination resulted in 2326 upregulated and 2221 downregulated genes and 68 upregulated and 55 downregulated pathways. Enhanced release of pro-inflammatory factors via MyD88-dependent TLR signaling pathway might be a key component of the action. Accelerated apoptosis of host cells due to downregulated PI3K-Akt signaling pathway could be another important mechanism.