ABSTRACT

   Tuberculosis (TB), caused by Mycobacterium tuberculosis (M.tb), remains
   a serious infectious disease posing significant global health
   challenges. A critical evolutionary feature of M.tb is its genome
   encoding a set of eukaryotic-like secretory proteins, which facilitate
   intracellular survival by manipulating host immune responses. However,
   the specific eukaryotic-like secretory proteins that facilitate M.tb
   intracellular survival and their regulatory mechanisms on host immunity
   remain uncharacterized. In this study, a mutant library comprising 137
   potential eukaryotic-like secretory proteins was constructed using
   clustered regularly interspaced short palindromic repeats
   (CRISPR)-non-homologous end joining genome editing technology.
   Subsequently, macrophages were infected with the mutant library, and
   CRISPR sequencing enabled preliminary identification of virulence
   factors associated with bacterial intracellular persistence. To
   validate the screen, two genes (Rv0066c and Rv3139) exhibiting the most
   pronounced reduction in intracellular survival rates when mutated were
   selected for the construction of large-fragment knockout strains
   (ΔRv0066c and ΔRv3139). Subsequent macrophage infection assays
   reconfirmed the impaired intracellular survival of these two mutants.
   RNA-seq analysis was conducted to characterize host gene expression
   profiles during ΔRv0066c-infected macrophage interactions. RNA-seq
   analysis of macrophages infected with wild-type and ΔRv0066c strains
   identified 138 differentially expressed genes, with 75 upregulated and
   63 downregulated in ΔRv0066c. Gene ontology clustering of these
   differentially expressed genes highlighted molecular functions related
   to chemokine binding, chemokine-mediated signaling pathways, Ras
   protein signal transduction, and calcineurin-mediated signaling.
   Collectively, this work established a potential eukaryotic-like
   secretory protein mutant library and identified two novel M.tb
   effectors governing intracellular survival, providing potential new
   targets for anti-TB drug development.

IMPORTANCE

   Eukaryotic-like secretory proteins that subvert host immunity to enable
   intracellular persistence are a key evolutionary adaptation of
   Mycobacterium tuberculosis (M.tb). In this study, we established a
   mutant library targeting 137 potential eukaryotic-like secretory
   proteins through clustered regularly interspaced short palindromic
   repeats (CRISPR)-non-homologous end joining genome editing technology.
   The library was subjected to macrophage infection assays, and CRISPR
   sequencing enabled identification of M.tb persistence-associated
   virulence determinants. Validation screens highlighted two genes
   (Rv0066c and Rv3139) that displayed the most significant intracellular
   survival defects to generate large-fragment knockout strains (ΔRv0066c
   and ΔRv3139). Macrophage infection experiments reconfirmed the
   compromised intracellular viability of both mutants. RNA-seq profiling
   of ΔRv0066c-infected macrophages identified 138 differentially
   expressed genes, with functional enrichment in chemokine signaling, Ras
   protein signal transduction, and calcineurin-mediated signaling. To
   conclude, this study identified two novel M.tb effectors contributing
   to intracellular survival as potential new targets for anti-TB drug
   development.

   KEYWORDS: Mycobacterium tuberculosis, macrophage, CRISPR screen,
   eukaryotic-like secretory protein, virulence effector

INTRODUCTION

   Tuberculosis (TB), caused by the intracellular pathogen Mycobacterium
   tuberculosis (M.tb), is a highly infectious airborne disease ([42]1),
   which claims millions of lives worldwide annually and remains a
   critical global public health challenge. Macrophages serve as the
   primary defense of the host immune system against M.tb and play a
   crucial role in determining the progression of TB ([43]2, [44]3). A
   notable feature of M.tb evolution is its possession of a set of
   eukaryotic-like secretory proteins ([45]4), which manipulate host
   immune mechanisms through diverse strategies ([46]5, [47]6). These
   eukaryotic-like secretory proteins have become major therapeutic
   targets for anti-TB drugs.

   Within the genome of M.tb H37Rv, 540 genes encode proteins with
   eukaryotic-like motifs or domains, among which 201 proteins are
   classified as M.tb potential secretory proteins ([48]4, [49]7).
   Previous studies have demonstrated that Rv1988, a functional
   methyltransferase secreted by M.tb, can localize to the host cell
   nucleus and directly modify host chromatin. This protein catalyzes
   methylation at the H3R42 site of histone H3, thereby suppressing the
   expression of genes associated with anti-mycobacterial immunity. M.tb
   lacking Rv1988 exhibits significantly impaired survival capacity in
   host organisms ([50]8). Additionally, Wang et al. revealed that the
   secreted tyrosine protein tyrosine phosphatase A (PtpA) promotes
   intracellular survival of M.tb by competitively binding to the RING
   domain of host TRIM27 (tripartite motif containing 27) to diminish its
   pro-apoptotic function, while concurrently targeting the JNK/p38 MAPK
   signaling cascades ([51]9). Previous work by Kim et al. demonstrated
   that M.tb enhanced intracellular survival protein promotes M.tb
   survival in macrophages by acetylating dual-specificity protein
   phosphatase 16/mitogen-activated protein kinase phosphatase-7, thereby
   inhibiting JNK-dependent autophagy, phagosome maturation, and reactive
   oxygen species (ROS) production, ultimately suppressing host immune
   responses ([52]10). However, the functions of many other
   eukaryotic-like secretory proteins remain unknown. Therefore,
   identifying virulence factors among M.tb-secreted eukaryotic-like
   proteins and elucidating their roles in TB pathogenesis will provide
   critical insights for developing improved therapeutic strategies.

   To systematically investigate critical virulence factors among these
   potential eukaryotic-like secretory proteins, we established a mutant
   library of 137 genes using the clustered regularly interspaced short
   palindromic repeats (CRISPR)-Cas9-mediated genome editing coupled with
   non-homologous end joining (NHEJ)-driven repair mechanism ([53]11). The
   library was screened via macrophage infection models, and CRISPR
   sequencing enabled the identification of virulence genes associated
   with M.tb intracellular survival. Genes exhibiting the most pronounced
   reduction in intracellular survival rates were selected for the
   construction of large-fragment knockout strains to validate the
   screens, followed by a preliminary exploration of how the encoded
   proteins promote M.tb persistence. This study further demonstrated that
   the application of CRISPR-assisted genome editing to construct a mutant
   library and the use of macrophages for in vitro screening are effective
   methods to identify M.tb virulence genes. Furthermore, the newly
   identified virulence-associated proteins provide potential targets for
   anti-TB drug development.

RESULTS

Optimized single-guide RNA (sgRNA) selection for CRISPR-mediated knockout of
genes encoding eukaryotic-like secretory proteins in M.tb

   To construct a CRISPR-based mutant library of eukaryotic-like secretory
   proteins in M.tb, we first analyzed previously reported candidate genes
   encoding potential eukaryotic-like secretory proteins. By integrating
   annotations from the Mycobrowser database
   ([54]https://mycobrowser.epfl.ch) and reviewing the literature, a total
   of 201 potential eukaryotic-like secretory proteins were classified.
   Among the 201 potential eukaryotic-like secretory proteins, 29
   essential genes were deemed as necessary for M.tb survival as reported
   in a previous study ([55]12). As the CRISPR-mediated genome editing
   approach requires protospacer adjacent motif (PAM) sites, our analysis
   of 172 non-essential genes revealed that 159 contained effective PAM
   sites while 13 lacked suitable PAM sites ([56]Fig. 1). Therefore, after
   excluding the 42 genes, a total of 159 sgRNA sequences were designed
   for the construction of mutant strains ([57]Table S1).

Fig 1.

   Pie chart presents proportions of gene categories: 79 percent
   non-essential with PAM sites, 14 percent essential, 7 percent
   non-essential lacking effective PAM sites, for CRISPR targeting
   analysis.
   [58]Open in a new tab

   Classification of the 201 potential eukaryotic-like secretory proteins
   of M.tb. The 201 genes encoding potential eukaryotic-like secretory
   proteins in M.tb were classified as essential genes, non-essential
   genes, and those with or without effective PAM sites.

Construction of the library of potential eukaryotic-like secretory protein
mutants

   To investigate the roles of these potential eukaryotic-like secretory
   proteins in maintaining M.tb intracellular survival, CRISPR-assisted
   genomic editing technology ([59]11) was used to construct a library of
   potential eukaryotic-like secretory protein mutant strains in M.tb
   H[37]Rv strain. During the process of building the mutant library,
   knockout mutants of 22 genes could not be produced. Eventually, a
   library of 137 potential eukaryotic-like secretory protein mutants was
   successfully constructed, covering 86.2% (137/159) of the target genes
   ([60]Fig. 2A), with two independent mutants generated for each gene
   (total of 274 mutants). The mutation spectrum analysis was as follows:
   small fragment deletions (1 bp–10 bp): 36%; large-fragment deletions
   (10 bp–100 bp): 17%; extra large-fragment deletions (>100 bp): 2%;
   insertion mutations: 45% ([61]Fig. 2B). All mutants were confirmed by
   colony PCR and sequencing, with mutation sites documented. Individual
   mutant strains were cultured and adjusted to an OD[600] of 0.4 before
   cryopreservation at −80°C. The 274 mutants of 137 genes and three
   control strains (empty vector, non-targeting sgRNA) were mixed in a 1:1
   ratio to form a functional screening library of M.tb potential
   eukaryotic-like secretory protein mutants for subsequent human monocyte
   leukemia cell line (THP-1)-derived macrophage infection assays.

Fig 2.

   [62]Two pie charts present gene editing outcomes. Top chart shows 86
   percent mutant, 14 percent wild type. Bottom chart shows mix of
   deletions, insertions, and indels across six mutation types.
   [63]Open in a new tab

   Statistics of the number and mutant information of M.tb potential
   eukaryotic-like secretory protein mutant strains. (A) Genome editing
   efficiency of the M.tb potential eukaryotic-like secretory protein
   mutant library. (B) Frequency of deletions or insertions detected in
   the mutant strains.

Intracellular survival analysis of eukaryotic-like secretory protein mutants
using macrophage infection models

   To explore the role of these potential eukaryotic-like secretory
   proteins toward M.tb pathogenicity, THP-1-derived macrophages were
   infected with the mutant library. Genes that affected the intracellular
   survival of M.tb were identified by analyzing the fitness of different
   mutants within the infected macrophages. At 4, 24, 48, and 72 h
   post-infection, THP-1-derived macrophages were lysed, serially diluted,
   and plated on Middlebrook 7H10 agar for colony enumeration. As shown in
   [64]Fig. 3A, 4 h after infection, each dish was colonized by
   approximately 2.5 × 10⁶ bacteria, indicating adequate coverage. At 24,
   48, and 72 h post-infection, each dish was colonized by approximately
   3.4 × 10⁶, 4.4 × 10⁶, and 3.7 × 10⁶ bacteria, respectively. The
   analysis of M.tb colony-forming unit (CFU) counts within the infected
   macrophages at different time points revealed that the M.tb library of
   potential eukaryotic-like secretory protein mutants progressively
   proliferated inside the macrophages, indicating the successful
   establishment of a cell-based screening model. The correlation analysis
   of reads from each replicate library showed that all Pearson’s
   correlation coefficients were above 0.9, indicating high
   reproducibility and good correlation of the screening results, which
   were suitable for subsequent screening and analysis ([65]Fig. 3B).

Fig 3.

   [66]Bar graph presents colony forming units from 4 to 72 hours, peaking
   at 48 hours. Nine scatter plots compare replicates for 24, 48, and 72
   hours with Pearson correlation coefficients ranging from 0.9337 to
   0.9852, indicating strong reproducibility.
   [67]Open in a new tab

   Analysis of the potential eukaryotic-like secretory protein mutant
   library colonies. (A) Statistics of the bacterial loads after THP-1
   macrophage infection with the M.tb potential eukaryotic-like secretory
   protein mutant library at 4 h, 24 h, 48 h, and 72 h post-infection. (B)
   Pearson’s correlation coefficient analysis of the potential
   eukaryotic-like secretory protein mutant library.

Screening for genes affecting M.tb intracellular survival in macrophages by
CRISPR sequencing

   At 4 h, 24 h, 48 h, and 72 h post-infection of THP-1 macrophages with
   the M.tb potential eukaryotic-like secretory protein mutant library, 2
   × 10⁶ to 4 × 10⁶ CFUs were harvested from each of three cell culture
   dishes at every time point. Genomic DNA was extracted, and the sgRNA
   regions were PCR-amplified for CRISPR sequencing and analysis ([68]Fig.
   4A). Quantifying the abundance of sgRNA for each mutant allowed us to
   assess the replication and expansion frequency of M.tb within THP-1
   macrophages. Based on the sequencing results, mutants of Rv0066c and
   Rv3139, whose sgRNA reads dropped significantly at 24 h, 48 h, and 72 h
   post-infection compared to the wild-type (WT) control strain, were
   identified. When analyzing the screening results, the Rv0153c (PtpB)
   mutant within the library was used as a positive control to confirm the
   accuracy of our library screen ([69]Fig. 4B).

Fig 4.

   [70]Dot plots present relative ratios of strains over 24, 48, 72 hours,
   highlighting ΔRv0066c and ΔRv3139 with low ratios. Bar graphs compare
   WT and mutants, showing significant reduction in ΔRv0066c and ΔRv3139
   at all time points.
   [71]Open in a new tab

   CRISPR sequencing identifies mutants with compromised survival in THP-1
   cells. (A) The M.tb potential eukaryotic-like secretory protein mutant
   library was used to infect THP-1 macrophages. At 24 h, 48 h, and 72 h
   post-infection, cells were harvested and genomic DNA extracted. CRISPR
   sequencing was used to analyze the abundance of sgRNA in different
   knockout strains. (B) Relative ratio of ΔRv0066c, ΔRv3139, and ΔPtpB
   (ΔRv0153c) strains. Data are presented as the mean ± SD of three
   independent experiments. Error bars are mean ± SD of three groups.
   Statistical significance was determined using one-way analysis of
   variance (Dunnett’s multiple comparisons test). n.s., not significant.
   *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

In vitro growth kinetics assay of potential virulence factors

   To further investigate whether proteins encoded by Rv0066c and Rv3139
   promote M.tb survival during infection, we constructed large-fragment
   knockout strains of Rv0066c (ΔRv0066c) and Rv3139 (ΔRv3139) using
   CRISPR-assisted genome editing. Sequencing analysis confirmed
   successful deletion of 2,017 bp and 1,193 bp fragments encoding Rv0066c
   and Rv3139, respectively, in the M.tb genome ([72]Fig. S1). Stable
   genetic traits and growth characteristics are conducive to guiding the
   analysis of cell infection experiments. As gene knockouts in M.tb may
   affect bacterial growth, we measured the in vitro growth kinetics of
   the WT strain, ΔRv0066c strain, and ΔRv3139 strain in 7H9 liquid
   complete medium. When compared to the WT control strain, the growth
   rates of the ΔRv0066c and ΔRv3139 strains in the medium were not
   significantly different, and the growth curves were similar ([73]Fig.
   5A). These results indicate that the ΔRv0066c and ΔRv3139 strains did
   not affect extracellular replication of M.tb or bacterial growth in
   vitro.

Fig 5.

   [74]Line graph presents strain density increases similarly across WT,
   ΔRv0066c, and ΔRv3139. Bar graphs presents at 72 hours, WT shows higher
   CFU than mutants, with complementation restoring CFU in
   ΔRv0066c::Rv0066c.
   [75]Open in a new tab

   Rv0066c and Rv3139 promote M.tb survival in macrophages. (A) In vitro
   growth kinetics of the WT, ΔRv0066c, and ΔRv3139 strains in Middlebrook
   7H9 broth media (n = 3). (B) THP-1 cells were infected with WT,
   ΔRv0066c, or ΔRv3139 strains at a multiplicity of infection (MOI) of 10
   for 4 h and 72 h. Bacterial loads in cells at the indicated time points
   after infection were determined and represented as CFU. (C) THP-1 cells
   were infected with WT, ΔRv0066c, or ΔRv0066c::Rv0066c strains at an MOI
   of 10 for 4 h and 72 h. Bacterial loads in cells at the indicated time
   points after infection were determined as CFU. Data are presented as
   the mean ± SD of three independent experiments. Error bars are mean ±
   SD of three groups. Statistical significance was determined using
   one-way analysis of variance (Dunnett’s multiple comparisons test).
   ****P < 0.0001.

Confirmation that Rv0066c and Rv3139 promote M.tb survival in macrophages

   To confirm that proteins encoded by the two genes Rv0066c and Rv3139,
   identified by sequencing, affect intracellular bacterial survival,
   THP-1 macrophages were infected with the WT, ΔRv0066c, or ΔRv3139
   strains at a multiplicity of infection (MOI) of 10 for 4 h and 72 h,
   and intracellular bacterial loads were measured. There were no
   significant differences in bacterial load among the three groups at 4
   h, indicating equivalent phagocytosis of the three strains by THP-1
   macrophages. However, at 72 h post-infection, the intracellular
   bacterial load of ΔRv0066c and ΔRv3139 strains in THP-1 macrophages was
   significantly lower than that of the WT strain (P < 0.0001). This
   suggests that secretory proteins encoded by Rv0066c and Rv3139 promote
   M.tb survival in THP-1 macrophages and that knocking out these genes
   significantly inhibits intracellular survival ([76]Fig. 5B).
   Furthermore, the ΔRv0066c mutant was complemented and used to infect
   macrophages for functional validation of its role in M.tb survival. The
   complemented strain ΔRv0066c::Rv0066c exhibited restored intracellular
   survival capacity in macrophages compared to the deletion mutant
   ΔRv0066c ([77]Fig. 5C).

Transcriptome profiling reveals macrophage gene expression modulation by
ΔRv0066c strain infection

   To elucidate how Rv0066c promotes intracellular survival of M.tb,
   phorbol-12-myristate-13-acetate (PMA)-differentiated THP-1 macrophages
   were infected with either WT or ΔRv0066c strains at an MOI of 10 for 24
   h, followed by RNA-sequencing (GENEWIZ, Inc.). Comparative
   transcriptome analysis between WT-infected and ΔRv0066c-infected groups
   revealed 138 differentially expressed genes, with 75 upregulated and 63
   downregulated genes in ΔRv0066c-infected macrophages ([78]Fig. 6A). The
   genes with significant differential expression (Top 10, |log2FC| > 1,
   P-value <0.05) and their functional annotations are shown in [79]Table
   S2. Gene ontology clustering of these differentially expressed genes
   highlighted molecular functions related to chemokine binding,
   chemokine-mediated signaling pathways, Ras protein signal transduction,
   and calcineurin-mediated signaling ([80]Fig. 6B). The Kyoto
   Encyclopedia of Genes and Genomes (KEGG) database, a comprehensive
   resource for functional annotation and integration of genomic data with
   higher-order biological networks ([81]13), was referred to for pathway
   enrichment analysis. The top 30 enriched pathways identified through
   KEGG analysis were associated with thyroid hormone signaling, cyclic
   adenosine monophosphate (cAMP) signaling, apoptosis signaling, and
   others ([82]Fig. 6C and D).

Fig 6.

   [83]Volcano plot identifies 138 differentially expressed genes with 75
   upregulated and 63 downregulated. Enrichment analysis reveals altered
   pathways in chemokine binding, Ras protein signal transduction, and
   calcineurin-mediated signaling in ΔRv0066c.
   [84]Open in a new tab

   Transcriptome profiling reveals modulation of macrophage gene
   expression following ΔRv0066c infection. (A) The volcano plot
   illustrates the expression profiles of differentially expressed genes
   in THP-1 macrophages infected with the WT strain versus the ΔRv0066c
   strain. The y-axis and x-axis represent the P-value and fold change in
   expression for each gene, respectively. Statistically significant
   upregulated and downregulated genes are highlighted in red and blue,
   respectively. (B) Bar plot of P-values for molecular function GO
   enrichment analysis of differentially expressed genes. The vertical
   axis displays enriched GO terms, while the horizontal axis represents
   the corresponding log10 (P-value) values. (C) Bar chart of
   significantly enriched KEGG annotation categories. The y-axis shows
   pathway names, and the x-axis shows the number of genes. (D) KEGG
   enrichment scatter plot of differentially expressed genes. The y-axis
   shows pathway names, and the x-axis shows the Rich factor. The size of
   the dots indicates the number of differentially expressed genes in the
   pathway, and the color corresponds to different Q-value ranges. These
   data represent three biological replicates.

DISCUSSION

   During the interaction between the host immune system and pathogenic
   microorganisms, pathogens secrete complex molecules known as effectors
   to subvert host immune responses. As a highly successful intracellular
   pathogen, M.tb has co-evolved extensively with human hosts. It has
   evolved various effector proteins to escape host innate and adaptive
   immunity, ensuring its intracellular survival and driving TB
   development and progression. A notable evolutionary feature of M.tb is
   its eukaryotic-like secretory proteins, acquired through horizontal
   gene transfer and convergent evolution ([85]4). These proteins play
   crucial roles in regulating host immune responses and facilitating M.tb
   intracellular survival ([86]4, [87]14). Understanding the functional
   roles of these secretory proteins in TB pathogenesis and developing
   therapeutic approaches targeting these proteins are of paramount
   importance ([88]15). Guided by previous reports on 201 potential
   eukaryotic-like secretory proteins in M.tb ([89]4, [90]7), we employed
   the CRISPR-assisted genome editing technology and established a mutant
   library encompassing 137 potential eukaryotic-like secretory proteins.
   Subsequently, through CRISPR sequencing, two novel virulence factors,
   Rv0066c and Rv3139, were screened and validated as influencing M.tb
   intracellular survival.

   Prior to constructing the mutant library, we found that among the 201
   potential eukaryotic-like secretory proteins, there were 29 essential
   genes including Rv0014c (PknB), Rv0015c (PknA), Rv0018c (PstP), Rv0046c
   (Ino1), Rv0285 (PE5), Rv0291 (MycP3), Rv0558 (MenH), Rv1093 (GlyA1),
   Rv1423 (WhiA), and Rv3303c (LpdA). These genes were previously verified
   as necessary for bacterial survival and mainly involved in core
   pathways such as cell wall synthesis and energy metabolism ([91]12,
   [92]16). Among the 172 non-essential genes of potential eukaryotic-like
   secretory proteins, 13 genes without effective PAM sites were excluded.
   Five of these genes are PE/PPE family genes Rv0978c (PE_PGRS17), Rv1790
   (PPE27), Rv2353c (PPE39), Rv2356c (PPE40), and Rv3872 (PE35). Their
   high sequence repetition prevents designing unique knockout sites.
   However, PE/PPE family genes have previously been reported to be
   related to virulence ([93]17, [94]18), so the omission of these genes
   from this study has limited consequences on the discovery of new
   virulence factors. Eight genes that lack PAM sites include two related
   to intermediate metabolism and respiration, Rv3117 (CysA3) ([95]19) and
   Rv3251c (rubA) ([96]20), as well as three linked to cell wall and
   cellular processes, which are Rv1081c, Rv1615, and Rv3615c (EspC)
   ([97]21). There were only three genes with unknown functions: Rv0336,
   Rv0546c, and Rv3686c. During the process of constructing the mutant
   library, knockout mutants of 22 genes, including Rv0410c (PknG),
   Rv0442c (PPE10), Rv0733 (Adk), and Rv3159c (PPE53), could not be
   generated, likely due to inefficient sgRNA design. We successfully
   constructed a mutant library of 137 potential eukaryotic-like secretory
   proteins, covering 86.2% of the target proteins. Through CRISPR
   sequencing and analysis, we identified that Rv0066c and Rv3139 mutants
   exhibited significantly reduced survival in macrophages. A significant
   reduction in the survival of the Rv0153c mutant within macrophages was
   also noted. Protein tyrosine phosphatase B (PtpB, Rv0153c) is an
   M.tb-secreted tyrosine phosphatase ([98]22[99]–[100]24). Studies have
   identified PtpB as a critical virulence factor in M.tb, where PtpB
   suppresses macrophage apoptosis by inhibiting p53 expression, thereby
   promoting M.tb survival ([101]25). A separate study revealed its
   capacity to inhibit pyroptosis through modulation of host membrane
   composition, enabling M.tb evasion of host immune responses ([102]4).
   These findings support the feasibility and reliability of our screening
   approach.

   While previous bioinformatics-based studies have shown that Rv0066c and
   Rv3139 encode M.tb secretory proteins ([103]26, [104]27), their
   functional contribution to intracellular bacterial persistence had yet
   to be established. In this study, the construction of large-fragment
   deletion mutants (ΔRv0066c and ΔRv3139) combined with macrophage
   infection assays further validated their critical association with M.tb
   intracellular survival. Key enzymes such as isocitrate dehydrogenase
   (ICD), pyruvate kinase, and succinate dehydrogenase are reported to
   play a crucial role in the growth and energy metabolism of M.tb
   ([105]28). M.tb Rv0066c (ICD-2) is an isocitrate dehydrogenase
   ([106]29), and is reported to participate in the tricarboxylic acid
   (TCA) cycle by catalyzing the oxidative decarboxylation of isocitrate,
   providing energy and metabolic intermediates for the bacterium
   ([107]30, [108]31). A recent report indicated that metabolic
   dysregulation in M.tb can significantly increase macrophage apoptosis
   or pyroptosis ([109]32). Knocking out genes closely related to
   bacterial respiration and intermediate metabolism can abolish M.tb’s
   suppression of macrophage apoptosis or pyroptosis ([110]32), reducing
   its virulence capacity. These findings highlight the importance of
   M.tb’s metabolic plasticity in its pathogenicity within the host.

   In our study, comparative transcriptome analysis demonstrated that
   Rv0066 may be functionally linked to macrophage apoptotic pathways.
   Host immune system-induced apoptosis regulates intracellular bacterial
   viability ([111]33), restricts pathogen dissemination by preventing
   microbial release ([112]34), and enhances host control of bacterial
   infections ([113]35), playing a pivotal role in antimicrobial defense.
   For instance, PtpA, a low-molecular-weight secretory tyrosine
   phosphatase essential for M.tb intracellular survival ([114]9),
   interacts with and dephosphorylates host glycogen synthase kinase 3α,
   thereby suppressing apoptosis and improving bacterial survival within
   host cells ([115]36). These findings suggest that Rv0066c may regulate
   apoptosis through similar or novel mechanisms. Rv3139, a probable
   acyl-CoA dehydrogenase (FadE24), belongs to the FadE gene family, which
   is critical for cholesterol metabolism in M.tb, with FadE gene
   expression regulated by cholesterol. This suggests that the
   FadE-encoded protein is vital for M.tb adaptation to and utilization of
   host nutrients ([116]37). Cholesterol is a key carbon source for M.tb’s
   survival and pathogenicity within the host, and M.tb can use host cell
   cholesterol as an energy source ([117]38, [118]39). As ATP levels are
   closely related to M.tb growth and replication, host cholesterol could
   serve as an energy source for its activities ([119]40). An
   activity-based protein analysis study suggested that FadE23 and FadE24
   may bind ATP ([120]41), indicating a possible catalytic function beyond
   acyl-CoA dehydrogenation.

   This study has several limitations that warrant consideration. First,
   the macrophage infection model does not adequately mimic
   pathophysiological host niches. Second, the current investigation
   focused solely on complementation verification and RNA-sequencing
   analysis of the most statistically significant candidate, Rv0066c,
   leaving the mechanistic underpinnings of its virulence regulation
   unexplored and other candidates (e.g., Rv3139) unvalidated. Future work
   will systematically evaluate the mutant library in the in vivo murine
   infection models to identify genes critical for M.tb survival under
   physiologically relevant conditions and delineate molecular pathways by
   which Rv0066c/Rv3139 interact with host defenses.

   In summary, this study employed a CRISPR-NHEJ genome editing approach
   to construct a library of M.tb potential eukaryotic-like secretory
   protein mutants. Adopting a strategy that focused on M.tb-host
   interactions, in vitro cellular screening identified critical genes
   promoting bacterial survival within host macrophages. These findings
   validate CRISPR-assisted genome editing combined with cell-based
   screens as an effective strategy to identify bacterial virulence genes,
   establishing a robust foundation for more extensive investigations on
   M.tb and other intracellular pathogens. The newly identified virulence
   genes provide novel therapeutic targets for advancing anti-TB drug
   development.

MATERIALS AND METHODS

Bacterial strain cultivation

   The Mycobacterium tuberculosis H37Rv strain (ATCC: 27294) was
   inoculated into Middlebrook 7H9 medium (BD Biosciences) with 10% OADC
   (oleic acid, albumin, dextrose, and catalase) (BD Biosciences) and
   incubated at 37°C for 14 days or when the OD[600] value was
   approximately 0.7, indicating the strain was in the logarithmic growth
   phase. The bacteria were then used for cell infection experiments. The
   bacterial strains utilized in this study are listed in [121]Table S3.

Cell culture

   THP-1 cells were passaged and cultured in Roswell Park Memorial
   Institute (RPMI) 1640 medium (Gibco) supplemented with 10% fetal bovine
   serum (Gibco), at 37°C with 5% CO[2] to the logarithmic growth phase.
   When the required cell numbers were achieved, the cells were plated,
   and PMA stock solution was added to a final concentration of 100 ng/mL.
   The THP-1 cells were incubated for ~36 h to differentiate them into
   macrophages, and PMA induction was stopped when more than 90% of the
   cells adhered to the plate. The cells were washed three times with
   serum-free RPMI 1640 medium and further incubated for about 15 h in
   complete medium without PMA.

Bacterial CFU assay

   Depending on the experimental goals, macrophages were infected with
   logarithmic-phase bacteria at an MOI of 10 and incubated at 37°C. After
   4 h, the bacteria-containing medium was carefully replaced with RPMI
   1640. Cells were washed three times to eliminate extracellular
   bacteria, after which fresh complete medium was added, and cells were
   further incubated. At each preset time point, cells were washed three
   times with RPMI 1640. After the media was removed, 500 µL of 0.5%
   Triton X-100 was added to each well. Plates were incubated at room
   temperature for 10 minutes, after which cells were disrupted by a
   pipetting action. Serial dilution of the cells was performed, and the
   bacteria were evenly spread onto 7H10 agar plates with 10% OADC and
   incubated at 37°C for 3–4 weeks to determine CFU at each dilution.

Construction of the potential eukaryotic-like secretory protein mutant
library

   M.tb competent cells were prepared as previously described ([122]11).
   The strain containing plasmids expressing NHEJ and RecX was cultured in
   roller bottles (Greiner Bio-one) with 7H9 medium at 37°C for 15 days.
   When the OD[600] value reached 0.7, glycine stock solution was added to
   a final concentration of 1.5% and the roller bottles were incubated at
   37°C for another 24 h. Bacteria were collected by centrifugation,
   washed three times with 10% glycerol, and finally resuspended in 10%
   glycerol.

   One hundred thirty-seven sgRNA expression plasmids targeting potential
   eukaryotic-like secretory protein genes were constructed based on the
   vector pYC1446. Three sgRNAs that did not target any gene were used as
   controls. The prepared competent cells were mixed with the plasmids and
   transferred to an electroporation cuvette (Bio-Rad) for electroporation
   ([123]11). After electroporation, bacteria were cultured in 7H9 medium
   for 24 h and spread onto 7H10 agar plates containing kanamycin (25
   µg/mL), zeocin (50 µg/mL), and anhydrotetracycline (50 ng/mL).
   Transformants were picked for PCR and sequence analysis to confirm
   successful genome editing. Finally, the mutant library was formed by
   mixing 137 potential eukaryotic-like secretory protein mutants and
   three control strains at a ratio of 1:1. The plasmids utilized in this
   study are detailed in [124]Table S4.

Macrophage infection, CRISPR sequencing library preparation, and data
analysis

   THP-1 monocytes were seeded in 150 mm dishes and induced with PMA (100
   ng/mL) for about 36 h, then cultured for another 15 h in PMA-free
   complete medium. Differentiated THP-1 macrophages were infected with
   the mutant library at an MOI of 10 and incubated at 37°C with 5% CO[2].
   After 4 h, the medium with bacteria was removed, and cells were washed
   three times with RPMI 1640 to eliminate extracellular bacteria, then
   cultured in fresh complete medium. At 4 h, 24 h, 48 h, and 72 h
   post-infection, THP-1 cells were lysed, serially diluted, and spread
   onto antibiotic-free 7H10 agar plates, which were then incubated
   inverted at 37°C. Monoclonal colonies from each time point were scraped
   from the 7H10 plates with a spreader and resuspended in
   phosphate-buffered saline containing 0.05% Tween-80. Each sample
   collected contained at least 6 × 10³ CFU, and genomic DNA was extracted
   from the harvested bacteria. The sgRNA coding region was PCR-amplified
   using genomic DNA as the template. The PCR fragments were sent to
   GENEWIZ, Inc. (Suzhou, China) for amplicon sequencing. Then libraries
   with different indices were multiplexed and loaded on an Illumina
   Novaseq instrument. Sequencing was carried out using a 2 × 150 bp
   paired-end (PE) configuration ([125]42). The sequencing results were
   analyzed using the MAGeCK (model-based analysis of genome-wide
   CRISPR-Cas9 knockout) robust ranking method (version 0.5.9.4)
   ([126]43). During CRISPR sequencing, the read count was normalized
   relative to the total sgRNA reads in each sample.

Construction of large-scale gene deletion knockout strains of M.tb

   The ΔRv0066c and ΔRv3139 mutants were constructed using the CRISPR-NHEJ
   gene editing strategy ([127]11). Specifically, the respective sgRNA
   sequence was amplified with specific primers and ligated into the
   pYC1876 vector. The H37Rv strain was transduce and recombinants were
   screened on 7H10 agar plates containing kanamycin, zeocin, and ATc.
   Mutants were screened via PCR and confirmed by DNA sequencing. The
   pYC874-L5-Rv0066c plasmid (synthesized by GENEWIZ) was used for
   complementation of the ΔRv0066c strain and defined as
   ΔRv0066c::Rv0066c. The primer sequences used in this study are listed
   in [128]Table S5.

In vitro strain growth kinetics assay

   To evaluate the impact of gene deletions on the in vitro growth of
   M.tb, the WT strain, ΔRv0066c strain, and ΔRv3139 strain were
   inoculated into 7H9 liquid medium to measure in vitro growth kinetics.
   Each strain was inoculated at an OD[600] of 0.04 and cultured to
   mid-log phase. Absorbance (OD[600]) was measured every 3 days for 15
   days, and growth curves were plotted based on the recorded values.

RNA-sequencing and analysis

   The WT and ΔRv0066c strains were used to infect differentiated THP-1
   macrophages at an MOI of 10 for 24 h. Total RNA was extracted using the
   Total RNA Kit (OMEGA). Subsequently, RNA-seq analysis was performed on
   these samples by GENEWIZ, Inc. (Suzhou, China). One microgram of total
   RNA was used for the library preparation. Poly(A) mRNA isolation was
   performed using Oligo(dT) beads followed by mRNA fragmentation using
   divalent cations and high temperature. Priming was performed using
   random primers. First-strand cDNA and second-strand cDNA were
   synthesized. The purified double-stranded cDNA was then subjected to
   end repair and dA tailing in a single reaction, followed by T-A
   ligation to add adaptors at both ends. Size selection of
   adaptor-ligated cDNA was then performed using DNA Clean Beads. Each
   sample was then amplified by PCR using P5 and P7 primers, and the PCR
   products were validated. Then libraries with different indices were
   multiplexed and loaded on an Illumina Novaseq 6000 instrument for
   sequencing using a 2 × 150 bp PE configuration according to
   manufacturer’s instructions.

Statistical analysis

   Data are presented as mean ±  standard deviation of at least three
   biological replicates. One-way analysis of variance (Dunnett’s multiple
   comparisons test) was used for analysis of multiple groups with a
   single independent variable. The statistical analyses were performed
   using GraphPad Prism 8.0. P < 0.05 was defined as statistically
   significant.

ACKNOWLEDGMENTS