ABSTRACT

   Mycobacterium phlei (M. phlei) is an understudied microbe with medical
   values as an immunomodulating agent. Here, we establish an industrial
   strain of M. phlei, CUD, and characterize its genomic, metabolic, and
   immunological profiles. The established strain has been stably passed
   for more than a decade, indicated by next-generation sequencing of its
   5.3 Mb genome. We show that the intramuscular inoculation of
   heat-inactivated CUD in immunocompetent mice is well tolerated, and can
   mount immunological responses. Immunophenotyping demonstrates induced
   innate and adaptive immune responses in peripheral blood, spleen, and
   inguinal lymph nodes of CUD-treated mice. Using GC-TOF-MS, we find that
   the metabolomic profiles of different batches are highly concordant.
   These results demonstrate a highly reproducible production of M. phlei
   under GMP conditions.

   IMPORTANCE Heat-inactivated M. phlei demonstrates promising efficacy to
   treat BCG-unresponsive non-muscle-invasive bladder cancer patients in
   clinical trials. However, lack of GMP-grade heat-inactivated M. phlei
   hampers further clinical investigations. Here, we described a
   GMP-grade, heat-inactivated M. phlei product, and presented initial
   characterization of its safety and immunomodulating properties. This
   product will serve as a starting point for further preclinical studies
   as well as clinical trials such as in combination with immune
   checkpoint inhibitors to treat bladder cancer.

   KEYWORDS: Mycobacterium phlei, Mycobacterium bovis BCG, genomics,
   metabolomics, immunity, BCG

INTRODUCTION

   Mycobacterium is intertwined with human life. Among approximately 190
   species in this genus, Mycobacterium bovis bacillus Calmette-Guérin
   (BCG) and Mycobacterium tuberculosis (Mtb) are intensively studied
   ([32]1). Mtb has co-evolved with human host ([33]2), and causes
   infectious disease tuberculosis (TB), resulting in global health burden
   ([34]3). To counteract Mtb infection, BCG is widely used as vaccine
   ([35]4). Nevertheless, the efficacy of BCG vaccine is about 70% to 80%
   against childhood tuberculosis ([36]5). BCG is also used as
   immunomodulatory agent in other contexts. For example, BCG remains the
   gold-standard treatment for non-muscle-invasive bladder cancer (NMIBC)
   for more than 40 years ([37]6). However, despite initial response in
   many patients at 1 year after treatment, 75% of patients become
   BCG-unresponsive within 5 years resulting in high mortality ([38]7).
   BCG is also used to reduce recurrent respiratory tract infections as
   well as acute attacks of chronic bronchitis. Other species are less
   studied, and medical values of mycobacterial space remain to be
   explored.

   Mycobacterium phlei (M. phlei) is an understudied microbe, first
   isolated by Alfred Moëller in 1898 ([39]8, [40]9). M. phlei is a
   nonpathogenic, nontuberculous mycobacterial species ([41]10). Previous
   studies suggested that heat-inactivated M. phlei has antitumor
   activities in vitro, similar to BCG ([42]11). The partial genome
   sequence of several strains of M. phlei has recently been published
   ([43]12). Nevertheless, GMP-grade, industrial scale production of M.
   phlei has not been reported, limiting further exploration of its
   medical value.

   Heat-inactivated M. phlei was approved about 20 years ago by the China
   National Medical Products Administration (NMPA, formerly known as the
   China Food and Drug Administration) (S20040067-70) as an agent to
   potentially boost human immunity. It should be noted that
   heat-inactivated M. phlei was approved as a general immunomodulating
   agent, but not as a drug to treat specific disease. The safety of
   heat-inactivated M. phlei in humans has been ascertained in the past
   20 years. However, its mechanism-of-action as well as its efficacy on
   specific disease are still poorly understood. Interestingly, recent
   clinical studies have demonstrated the efficacy of heat-inactivated M.
   phlei to treat BCG-unresponsive NMIBC patients ([44]11, [45]13). Novel
   treatment options for NMIBC present a major unmet medical need
   worldwide ([46]14), promoting us to produce GMP-grade heat-inactivated
   M. phlei for further clinical testing. The original production and
   quality control processes were suboptimal due to technical limitations
   20 years ago, and new guidelines from NMPA have been upgraded since
   then. Therefore, in the present study, we aimed to use advanced genomic
   and metabolic techniques to characterize an industrial strain of M.
   phlei (named CUD) as well as a GMP-grade product containing
   heat-inactivated CUD. It is our hope that this study will serve as a
   starting point to provide GMP-grade heat-inactivated M. phlei for
   ongoing clinical trials to examine its efficacy on cancer and
   respiratory diseases.

RESULTS

Genomic features of Mycobacterium phlei CUD.

   We analyzed the genome of Mycobacterium phlei CUD (CUD) by
   next-generation sequencing using Illumina platform ([47]Fig. 1a).
   Forty-three overlapping scaffolds were generated. We found that the
   genome of CUD contained approximately 5.3 million base pairs (Mb) (GC
   content, 69.47%), close to those of M. phlei CCUG 21000 (5.35 Mb, NCBI
   Assembly: GCA_001583415.1), M. phlei NCTC8156 (5.31 Mb, NCBI Assembly:
   GCA_900453675.1) and M. phlei RIVM700384 (5.68 Mb, NCBI Assembly:
   GCA_900453675.1) ([48]Fig. 1b).

FIG 1.

   [49]FIG 1
   [50]Open in a new tab

   Genomic features of Mycobacterium phlei strain, CUD. (a) A circos plot
   of CUD genome assembly. From inner to outer circles, blue blocks denote
   43 scaffolds of CUD. Purple histogram represents gene densities of each
   scaffold. The green histogram track shows single nucleotide
   polymorphisms (SNPs) compared with the whole-genome sequence of M.
   phlei CCUG21000 (NCBI reference sequence, [51]NZ_ANBP01000001.1),
   containing 5,454,064 nucleotides. Of note, the assembly of CCUG21000
   genome is complete, thus CUD genome assembly is ~98.4% complete
   (5,366,584 nucleotides). (b) Phylogenetic tree of six M. phlei strains
   as well M. tuberculosis and M. bovis BCG based on comparison of 2,000
   core conserved genes calculated by core-pan analysis. Branch length
   represents the size of evolutionary distance. (c) A circos plot of
   detected variants between two batches of GMP-grade CUD. Out of 58
   variants detected by second generation sequencing, 19 variants were
   located in putative protein-coding regions, denoted on corresponding
   scaffolds. Of note, these variants were localized in four genes. (d)
   KEGG pathway analysis of predicted protein-coding genes of CUD. (e)
   Analysis of homologous and divergent genes between mycobacterial
   species CUD, M. bovis BCG and M. tuberculosis. (f) Resistance Gene
   Identifier (RGI) analysis of CUD using AMR gene family, drug class and
   resistance mechanism based on antibiotic resistance database (CARD).

   To determine the genomic stability of CUD, different batches produced
   under GMP standard were subjected to next-generation sequencing
   ([52]Fig. 1c). CUD has been maintained for more than 2 decades, and we
   compared two batches produced in 2010 (batch A) and 2020 (batch B). We
   found that batch A and batch B shared >99.99% identity, and only
   differed in 57 nucleotides out of its 5.3 Mb genome ([53]Fig. 1c).
   Because > 99.99% identity exceeds the accuracy of Illumina sequencing
   platform (Q20), this suggests that the nuance in nucleotide composition
   between two batches might arise from sequencing error of Illumina
   platform. Thus, the genome of CUD is stable under GMP condition.

   GeneMarkS was used to predict protein-coding genes of M. phlei CUD
   ([54]15). Overall, 5,126 genes were predicted, covering 92.04% of M.
   phlei CUD genome. Among them, 3,780 genes were predicted to encode
   proteins. Microbial pathway enrichment analysis using the KEGG database
   (hsa00001) revealed that most genes were involved in metabolic
   pathways, including amino acid metabolism, carbohydrate metabolism as
   well as xenobiotics biodegradation and metabolism ([55]Fig. 1d).

   Core-pan analysis was performed to predict homologous genes between M.
   phlei CUD and M. bovis BCG and M. tuberculosis ([56]Fig. 1e). We found
   that CUD, BCG and M. tuberculosis had 2,005 homologous genes. For
   pairwise comparison, we found that 1,397 genes of CUD had homologues in
   BCG but not in M. tuberculosis, and 46 genes in CUD had homologues in
   M. tuberculosis but not in BCG ([57]Fig. 1e). This result suggests that
   CUD may impact human immune system in a similar fashion as BCG.
   Finally, we used RGI software in the comprehensive antibiotic
   resistance database (CARD) to predict potential resistance of CUD to
   antibiotics. Consistent with its high homology to M. bovis, BCG, and M.
   tuberculosis, M. phlei CUD was predicted to resist macrolide and
   rifamycin ([58]Fig. 1f).

Safety profile of heat-inactivated M. phlei CUD in immunocompetent mice.

   Six-week-old female BALB/c mice were used to examine whether CUD had
   immunostimulatory properties. M. bovis BCG was used as a positive
   control. Two doses of heat-inactivated M. phlei CUD or M. bovis BCG
   were injected at day 0 and day 3, followed by measurement of animal
   body weight, spleen/kidney weight, and immune cells at 1-week interval
   during a 4-week period ([59]Fig. 2a). Saline was used as an injection
   control. Two independent experiments were carried out with three mice
   for each treatment group. No overt toxicity was found in mice injected
   with CUD or BCG, indicated by comparable body weight gain with mice
   injected with saline ([60]Fig. 2b). Spleen was enlarged in mice
   injected with CUD at week 1 and week 2, and at week 3 recovered to a
   similar size as mice injected with saline. As expected, BCG also
   induced the enlargement of spleen, albeit at slower kinetics and with
   lower amplitude ([61]Fig. 2c). Consistently, the number of lymphocytes
   in spleen ([62]Fig. 2d) and immune cells in peripheral blood
   ([63]Fig. 2e) were increased in CUD- or BCG-treated mice. These results
   demonstrate that heat-inactivated CUD is generally safe, and can
   stimulate immune responses in mice.

FIG 2.

   [64]FIG 2
   [65]Open in a new tab

   Safety profile of heat-inactivated M. phlei CUD in immunocompetent
   mice. (a) Experimental scheme. Mice were injected twice with M. phlei
   CUD and M. bovis bacillus Calmette-Guérin (BCG), individually, at day 0
   and day 3, followed by measurements of body weight, spleen weight, and
   immune cells each week for 4 weeks. D, day; W, week. (b) Dot plot
   representation of body weight of mice treated with saline, M. phlei CUD
   (phlei) or M. bovis BCG (BCG). Data are represented as mean ± S.D.
   (n = 6). (c) Box plot representation of spleen weight of mice treated
   with saline, M. phlei CUD (phlei) or M. bovis BCG (BCG). Spleen weight
   was normalized using the weight of kidney because kidney weight was
   comparable between each group. Data are represented as mean ± S.D.
   (n = 6). (d) Box plot representation of total number of cells in spleen
   of mice treated with saline, M. phlei CUD (phlei), or M. bovis BCG
   (BCG). The quantity of immune cells in control mice was arbitrarily
   denoted as 1, and those of treatment groups were normalized to the
   control group. Data are represented as mean ± S.D. (n = 6). (e) Box
   plot representation of total number of immune cells in peripheral blood
   of mice treated with saline, M. phlei CUD (phlei) or M. bovis BCG
   (BCG). The quantity of immune cells in control mice was arbitrarily
   denoted as 1, and those of treatment groups were normalized to the
   control group. Data are represented as mean ± S.D. (n = 6). Statistical
   analysis (one-way ANOVA): *, P-value < 0.05; **, P-value < 0.01; ***,
   P-value < 0.001.

Analyses of immunological reactions induced by M. phlei CUD.

   We further used fluorescence-activated cell sorting (FACS) to analyze
   subtypes of immune cells in peripheral blood, spleen, and inguinal
   lymph nodes. Gating strategy for peripheral blood was shown in
   [66]Fig. 3a. CD3^+ T cells were increased in CUD- and BCG-treated mice
   and lasted for 2 weeks after the final injection ([67]Fig. 3b).
   CD69-positive CD4^+ and CD8^+ T cell subsets were also increased during
   the same time window, indicating substantial T cells were activated
   ([68]Fig. 3c and [69]d). CD69-positive B220^+ B cells were increased
   for 2 weeks ([70]Fig. 3e). Interestingly, dendritic cells, natural
   killer cells, and macrophages were increased in CUD-treated mice only
   at week 1 ([71]Fig. 3f, g, h). These results demonstrated that CUD and
   BCG could stimulate quantitively similar immune reactions.

FIG 3.

   [72]FIG 3
   [73]Open in a new tab

   Analyses of immunological reactions in peripheral blood induced by M.
   phlei CUD. (a) Gating strategy of fluorescence-activated cell sorting
   (FACS) analyses to quantitate indicated cell types in peripheral blood.
   Lymphocytes were gated based on side scatter area (SSC-A) and forward
   scatter area (FSC-A) (panel 1). Time parameter was used as a quality
   control to exclude events bursts (panel 2), followed by gating of
   singlet cell (panel 3). CD3^+ T cells (panel 4), B220^+ B cells (panel
   5), CD11c^+ dendritic cells (panel 6), F4/80^+ macrophage (panel 7) and
   NK1.1^+ natural killer (NK) cells (panel 8) were analyzed in peripheral
   blood. CD3^+ T cells were further analyzed to identify CD4^+CD69^+ T
   cells (panel 9) and CD8^+CD69^+T cells (panel 10). B220^+ B cells were
   further analyzed to identify B220^+CD69^+ B cells (panel 11). (b to e)
   Box plot presentation of (b) total CD3^+ T cells, (c) CD4+CD69^+ T
   cells, (d) CD8^+CD69^+ T and (e) B220^+CD69^+ B cells in peripheral
   blood. The quantity of immune cells in control mice was arbitrarily
   denoted as 1, and those of treatment groups were normalized to the
   control group. Data are represented as mean ± S.D. (n = 6). (f to g)
   Box plot presentation of dendritic cells (f), natural killer cells (g),
   and macrophage (h) in peripheral blood. The quantity of immune cells in
   control mice was arbitrarily denoted as 1, and those of treatment
   groups were normalized to the control group. Data are represented as
   mean ± S.D. (n = 6). Statistical analysis (one-way ANOVA): *,
   P-value < 0.05; **, P-value < 0.01; ***, P-value < 0.001.

   FACS analyses were carried to immunophenotype cells in spleen
   ([74]Fig. 4). Gating strategy was shown in [75]Fig. 4a. The size of
   spleen in CUD-treated mice were increased at week 1 and 2. Accordingly,
   FACS analyses revealed increased T and B cells in these mice
   ([76]Fig. 4b, c). However, unlike peripheral blood, only CD4^+ but not
   CD8^+ T cells were significantly activated ([77]Fig. 4d, e).

FIG 4.

   [78]FIG 4
   [79]Open in a new tab

   Analyses of immunological reactions in spleen induced by M. phlei CUD.
   (a) Gating strategy of fluorescence-activated cell sorting (FACS)
   analyses to quantitate indicated cell types in spleen. Lymphocytes were
   gated based on side scatter area (SSC-A) and forward scatter area
   (FSC-A) (panel 1). Time parameter was used as a quality control to
   exclude events bursts (panel 2), followed by gating of singlet cell
   (panel 3), CD3^+ T cells (panel 4), B220^+ B cells (panel 5) in spleen.
   CD3^+ T cells were further analyzed to identify CD4^+CD69^+ T cells
   (panel 6) and CD8^+CD69^+T cells (panel 7). B220^+ B cells were further
   analyzed to identify B220^+CD69^+ B cells (panel 8). (b to e) Box plot
   presentation of (b) total CD3^+ T cells, (c) B220^+CD69^+ B cells, (d)
   CD8^+CD69^+ T, and (e) CD4^+CD69^+ T cells in spleen. The quantity of
   immune cells in control mice was arbitrarily denoted as 1, and those of
   treatment groups were normalized to the control group. Data are
   represented as mean ± S.D. (n = 6). Statistical analysis (one-way
   ANOVA): *, P-value < 0.05; **, P-value < 0.01; ***, P-value < 0.001.

   Finally, we used FACS analyses to quantitate immune cells in inguinal
   lymph nodes. Gating strategy was shown in [80]Fig. 5a. Similar to
   peripheral blood ([81]Fig. 3), numbers of total as well as activated T
   and B cells were increased in both CUD- and BCG-treated mice
   ([82]Fig. 5b to [83]e).

FIG 5.

   [84]FIG 5
   [85]Open in a new tab

   Analyses of immunological reactions in inguinal lymph node induced by
   M. phlei CUD. (a) Gating strategy of fluorescence-activated cell
   sorting (FACS) analyses to quantitate indicated cell types in inguinal
   lymph node. Lymphocytes were gated based on side scatter area (SSC-A)
   and forward scatter area (FSC-A) (panel 1). Time parameter was used as
   a quality control to exclude events bursts (panel 2), followed by
   gating of singlet cell (panel 3), CD3^+ T cells (panel 4), B220^+ B
   cells (panel 5) in spleen. CD3^+ T cells were further analyzed to
   identify CD4^+CD69^+ T cells (panel 6) and CD8^+CD69^+T cells (panel
   7). B220^+ B cells were further analyzed to identify B220^+CD69^+ B
   cells (panel 8). (b to e) Box plot presentation of (b) total CD3^+ T
   cells, (c) B220^+CD69^+ B cells, (d) CD8^+CD69^+ T, and (e) CD4^+CD69^+
   T cells in inguinal lymph node. The quantity of immune cells in control
   mice was arbitrarily denoted as 1, and those of treatment groups were
   normalized to the control group. Data are represented as mean ± S.D.
   (n = 6). Statistical analysis (one-way ANOVA): *, P-value < 0.05; **,
   P-value < 0.01; ***, P-value < 0.001.

Metabolomics analysis of six batches of GMP-grade M. phlei CUD.

   To understand the immunogenic properties of M. phlei CUD, we further
   used gas chromatography time-of-flight mass spectrometry
   (GC-TOF-MS)-based metabolomics approach to identify metabolites
   generated by CUD. Six batches of CUD produced under GMP conditions were
   analyzed. Metabolites reproducibly identified in six batches (t test P
   value < 0.05, OPLS-DA VIP score >1) were calculated and aligned again
   KEGG ([86]16) and HMDB databases ([87]17). This analysis identified 281
   known metabolites. Normalized peak area for each metabolite was used to
   estimate its abundance, and their abundance was ranked in a descending
   order ([88]Fig. 6a). We also carried out metabolic pathway enrichment
   analysis against KEGG metabolomics database. Top 25 pathways
   (normalized P value < 0.05) are shown in [89]Fig. 6b. These results
   demonstrate highly reproducible production of metabolites among
   different GMP-grade batches of CUD, reflecting a stable manufacturing
   process. Several metabolites are also of interest due to their
   immunomodulatory properties (see Discussion).

FIG 6.

   [90]FIG 6
   [91]Open in a new tab

   Metabolomic analysis of six batches of GMP-grade M. phlei CUD. (a)
   Heatmap presentation of 281 metabolites reproducibly identified from
   six batches of GMP-grade CUD and color represents relatively high and
   low expression, respectively. Highly expressed metabolites were listed
   in the dotted box. (b) Bubble chart of metabolic pathway enrichment
   analysis based on KEGG metabolomics database. Dot size reflected
   identified metabolites matched in the pathway.

DISCUSSION

   Mycobacteria are important microorganisms due to their relevance to
   human health. Prominent examples include human pathogens Mycobacterium
   tuberculosis-causing TB and Mycobacterium leprae-causing leprosy
   ([92]18). On the other hand, Mycobacterium bovis BCG is widely used in
   medical practice as anti-TB vaccine, immunostimulatory adjuvant,
   anticancer reagent, and immune modulator to reduce recurrent
   respiratory tract infections as well as acute attacks of chronic
   bronchitis ([93]19).

   Mycobacterium phlei is an understudied mycobacterial species of
   potential medical value. Here, we present genomic, metabolic, and
   immunological characterization of an industrial strain, namely, CUD,
   manufactured under GMP conditions. We show that different batches of
   CUD are genetically stable via next-generation sequencing, and produce
   highly reproducible metabolites via metabolic analysis. CUD product is
   safe and can induce immunological reactions in immunocompetent mice. A
   standardized, GMP-grade microbial product is a valuable tool for
   further clinical investigations. This is exemplified by the fact that
   BCG products from various vendors can exhibit differential clinical
   efficacy ([94]20[95]–[96]22). Of note, recent clinical studies
   demonstrated that heat-inactivated M. phlei may be effective to treat
   BCG-unresponsive bladder cancer patients ([97]11, [98]13). It will be
   of great interest to examine the effect of CUD in this patient
   population. Currently, the effect of CUD on bladder cancer is being
   evaluated in animal models such as the syngeneic mouse model with mouse
   bladder cancer MB49 cells.

   In summary, we present a GMP-grade product of Mycobacterium phlei
   strain, CUD, which is safe and can modulate host immunity. Further
   studies are warranted to investigate its efficacy in the context such
   as cancer immunotherapy. These applications are under investigation.

MATERIALS AND METHODS

Genomic DNA extraction, sequencing, and assembly.

   BCG was obtained from Chengdu Institute of Biological Products Co.,
   Ltd. (S20123007, NMPA-approved agent to treat bladder cancer). M. phlei
   strain, CUD (DSM 43471), was cultured in fermentation tanks.
   Heat-inactivated CUD was manufactured by proprietary procedures
   (Chengdu Jinxing Sanum-Kehlbeck Medcine Co., Ltd.) in compliance with
   GMP guidelines from the China National Medical Products Administration.
   Genomic DNA of CUD was extracted using GeneJET Genomic DNA purification
   kit, analyzed by agarose gel electrophoresis, and quantified by Qubit
   2.0 Fluorometer (Thermo Scientific). Sequencing libraries were prepared
   using NEBNext Ultra DNA Library Prep Kit for Illumina (NEB, USA),
   following the manufacturer’s instructions. Index codes were added to
   attribute sequences to each sample. Genomic DNA was fragmented by
   sonication to ~350 bp, and ligated with the full-length adaptor for
   Illumina sequencing with further PCR amplification. PCR products were
   purified (AMPure XP system), analyzed for size distribution by
   Agilent2100 Bioanalyzer, and quantified using real-time PCR. Clean data
   were assembled by SOAP denovo ([99]23), SPAdes ([100]24), ABySS
   ([101]25), and Gapclose ([102]26) to generate scaffolds with minimal
   gaps (Novogene, China).

Gene prediction and annotation.

   tRNA, rRNA, and sRNA were separately predicted by tRNAscan-SE, rRNAmmer
   and BLAST against Rfam database ([103]27[104]–[105]29). Tandem Repeat
   Finder was used for tandem repeats analysis ([106]30). IslandPath-DIOMB
   was applied in genomic islands analysis ([107]30[108]–[109]34). A
   whole-genome BLAST search (E-value ≤ 1e-5, minimal alignment length
   percentage ≥ 40%) was performed against five databases: KEGG (Kyoto
   Encyclopedia of Genes and Genomes) ([110]35[111]–[112]37), COG
   (Clusters of Orthologous Groups) ([113]38, [114]39), NR (Non-Redundant
   Protein Database databases), Swiss-Prot ([115]40), and GO (Gene
   Ontology) ([116]41).

Phylogenetic analysis.

   Genomic alignment between the six M. phlei genomes and M. tuberculosis
   and M. bovis BCG in NCBI database was performed using the MUMmer
   ([117]42) and LASTZ tools. Core genes and specific genes were analyzed
   by the CD-HIT rapid clustering of similar proteins software with a
   threshold of 50% pairwise identity and 0.7-length difference cutoff in
   amino acid ([118]43). The phylogenetic tree was constructed by the
   PhyML and the setting of bootstraps was 1,000 with the orthologous
   genes ([119]44). SNP, indel, and SV were found by the genomic alignment
   results among samples by the MUMmer and LASTZ.

Antibiotic resistance analysis.

   The Resistance Gene Identifier (RGI) was used to predict resistomes
   according to the genome of M. phlei CUD via the web portal
   ([120]https://card.mcmaster.ca/analyze/rgi).

GC-TOF-MS metabolomic analysis.

   Metabolites from six batches of GMP-grade product containing
   heat-inactivated CUD (from > 1 × 10^7 cells) in saline were analyzed by
   a commercial vendor (Biotree, China), using Agilent 7890B gas
   chromatography (Agilent Technologies, USA) coupled with LECO Pegasus
   HT/BT time-of-flight mass spectrometer (LecoCorp., USA) (GC-TOF-MS).
   Data, including peak numbers, sample names, and normalized peak areas,
   were fed to SIMCA14 software package (V14, Umetrics AB, Umea, Sweden)
   for principal-component analysis (PCA) and orthogonal projections to
   latent structures-discriminate analysis (OPLS-DA). High-resolution
   GC-TOF-MS spectral peak data were applied to perform metabolic pathway
   enrichment analysis based on KEGG metabolomics database, and a heat map
   was drawn using R (v3.6.3) software.

Immunological analysis.

   Animal studies were approved by the Institutional Animal Care and Use
   Committee at West China Second University Hospital of Sichuan
   University. Mice were housed under the Program of Laboratory Animal
   Center at West China Second University Hospital of Sichuan University
   with the principles and procedures of the Guide for the Care and Use of
   Laboratory Animals. Six-week-old female BALB/c mice were injected
   intramuscularly with 172 μg of heat-inactivated CUD or BCG resuspended
   in 500 μL saline (0.9% NaCl) per mouse, respectively. Saline was used
   as an injection control. The number of mice for each group was
   determined according to recently published guidelines (n = 3)
   ([121]45). Two independent experiments were carried out (n = 6 in total
   for each group). Each mouse was injected on day 0 and day 3, and
   corresponding samples were collected on day 7, day 14, day 21, and day
   28 for subsequent analyses. For fluorescence-activated cell sorting
   (FACS) analysis, CD3^+ T cells, B220^+ B cells, CD11c^+ dendritic
   cells, F4/80^+ macrophage and NK1.1^+ natural killer (NK) cells were
   analyzed in peripheral blood, spleen, and inguinal lymph node. CD3^+ T
   cells were further analyzed to identify CD4^+CD69^+ T cells and
   CD8^+CD69^+T cells. B220^+ B cells were further analyzed to identify
   B220^+CD69^+ B cells.

Statistical analysis.

   R (v3.6.3) and GraphPad Prim 8 software were used for all statistical
   analyses. Statistical comparisons were made by using one-way ANOVA and
   Student’s t test. The statistical significance of difference was set as
   P < 0.05. In all figures, * denotes P value < 0.05; ** denotes P
   value < 0.01; *** denotes P value < 0.001.

Data availability.

   The assembled genome sequence of CUD has been submitted to the GenBank
   database under the accession number [122]PRJNA843359.

ACKNOWLEDGMENTS