Comparison analysis of Bacillus Calmette-Guerin induced Autophagy mechanisms in Macrophages Derived from Human Induced Pluripotent Stem Cells and THP-1

Background Tuberculosis (TB) remains a major global public health problem and the leading cause of mortality by a single infectious agent. TB is a chronic infectious disease that is primarily caused by Mycobacterium tuberculosis (Mtb). Macrophage (Mφ) are the main hosts of Mtb, the interaction between Mtb and Mφ plays an important role in the pathogenesis of TB. Summary The macrophages used in the current study are mostly derived from tumor cell lines or peripheral blood mononuclear cells (PBMC), but the application of such cells still have many problems needed to be sloved, such as the loss of function due to changes in genetic structure and the diculty in cell acquisition. Human induced pluripotent stem cells (hiPS) represent an innovative source for the standardized in vitro generation of Mφ, and show novel promise in exploring disease pathogenesis, particularly TB. Current studies have revealed that autophagy plays a central role in the interaction between Mtb and Mφ, but the molecular mechanism involoved remains unclear and the exact role of hiPS-derived macrophages (hiPS-Mφ) in regulating autophagy induced by Mtb also remains unclear. To investigate the similarities and differences in hiPS-Mφ and THP-1-Mφ in anti-tuberculosis immunity, this study successfully obtained macrophages derived from hiPS and THP-1, then explored the mechanism behind Bacillus Calmette-Guerin (BCG)-induced autophagy through transcriptome sequencing analysis, qPCR, Western Blot Analysis and cell submicroscopic structure observation etc.. Our ndings revealed that BCG infection of hiPS-Mφ and THP-1-Mφ would promote autophagy by regulating the expression of autophagy-related genes, which also indicated that the BCG-induced autophagy in hiPS-Mφ and THP-1-Mφ may be associated with PI3K/AKT/mTOR signaling pathway. However, there are some differences in the mechanism by which BCG infects macrophages from different sources and induces autophagy. Considering the above ndings, we have provided novel insights into the role of macrophages along with autophagy in the anti-tuberculosis immune mechanism and the possibility of establishing an in vitro hiPS-Mφ-TB disease model. as mean ± SD. ns, non-signicant,*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.


Introduction
Macrophages (Mφ) are present in all tissues of mammals. As a key component of the innate immune system, macrophages exhibit extremely signi cant plasticity and play an important role [1][2][3][4] , for example, eliminating invading pathogens, remodeling tissues and clearing dead cells 5 . Macrophages are associated with various types of cancer in terms of tumors [6][7][8][9] , and in the aspect of cardiovascular diseases, atherosclerosis and metabolic regulatory abnormalities are also closely related to it 10,11 . In summary, it is not di cult to see the signi cance of macrophages in human health issues. Unfortunately, normal human macrophages (e.g. blood monocyte derived macrophages, bone-marrow derived macrophages) are too di cult to be acquired and they cannot self-renew, hindering the development of relevant researches. More importantly, each study requires large amounts of cells, which means we will require large amounts of blood from the donors and it may cause ethical issues. The differences of donors' physiological state and genes are also a tough problem, making the test results to be not representative. To solve all these problems, the macrophages currently used for immunological researches are RAW264.7 along with cells derived from tumor cell lines (e.g. U937, THP-1) and primary cells (e.g. tissue cells and PBMC). Among them, RAW264.7 is a mouse-speci c macrophage, while U937-Mφ and THP-1-Mφ are both derived from tumor cell lines, the main differences between these two sources are the origin and the maturity stage. U937 cells are of tissue origin and therefore at a more mature stage while THP-1 cells originate from hematological leukemia cells and belong to a less mature stage 12 . Although they are all immortalized cell lines and can be cultured in vitro to the 25th generation, maintaining cell sensitivity and activity, THP-1 cells are more suitable for further researches as the growth rate of THP-1 increases faster, more importantly, there are no reports related to the infectious viruses or toxic products in these cells. Such a cell line is relatively easy and safe. Due to the homologous genetic background, the degree of change in cell phenotype is minimized, which is conducive to experimental reproducibility 13,14 . But cells derived from these sources still have various defects, such as being di cult to obtain and manipulate, furthermore they are also prone to mutations. of individual-speci c pluripotent stem cells have received wide public attention. Therefore, the emergence of iPS-Mφ provides a genotype-speci c, scalable and reproducible source of human macrophages 11,16−22 , it can also be genetically manipulated by CRISPR/Cas9-mediated gene editing and used in genetic engineering 23,24 . Compared with macrophages from other sources, iPS-Mφ are less di cult to operate, they have signi cant advantages over other cells and have been studied in many researches related to mechansims underlying diseases, especially the pathogenesis of immune diseases such as tuberculosis and tumors. For all these grounds, iPS-Mφ have broad application prospects.
To date, infectious disease like TB remains a very important public health issue worldwide, and Mycobacterium tuberculosis(Mtb) is known to be the main pathogen of TB 25 . Mtb belongs to intracellular parasites, tuberculosis can be simply understood as a series of continuous and slow interactions between Mtb and the immune system, which gives the body a sustained immune response. However, it is di cult for antibodies present in serum to enter the interior of the cells and kill Mtb.
Therefore, the cellular immune response plays an important role in the body's resistance to Mtb infection.
Current studies had found that the main host cell of Mtb is macrophage. When the infection begins, Mtb is engulfed by lung macrophages that have important immune effects against Mtb 26 , when lung macrophages are infected with Mtb, they produce a series of immune responses to prevent the spread of them, and nally kill them. But Mtb can also survive in cells by evading the immune surveillance along with the attack of macrophages, and even lysing macrophages. Therefore, the interaction between Mφ and Mtb plays an important role in the study of the occurrence and development of tuberculosis.
The Bacillus Calmette-Guerin (BCG) is currently the only TB vaccine approved for marketing by the FDA. It was developed by two scientists: Guerin and Calmette, by culturing a poisonous Bovine M. tuberculosis into a vaccine-free attenuated strain. Since 1923, more than 4 billion people have been vaccinated by BCG worldwide 27 . But BCG has very little effect on adult TB and Mtb latent infection, more importantly, problems such as inability to be applied to immunode ciency and prone to strain variation are also gradually exposed 28 . A variety of improved vaccines based on BCG, such as recombinant BCG, subunit vaccine, live vector vaccine, whole-cell inactivated vaccine and live attenuated vaccine, have been gradually developed due to these defects of BCG. Despite this, BCG is still favored by many scientists due to the low price and high safety as well as the e ciency in preventing TB infection in children.
Autophagy is a relatively conservative process in the evolution of organisms prevalent in eukaryotes 29 and an intracellular degradation pathway for damaged organelles and aggregation-prone proteins 30,31 .
Through the process of autophagy, misfolded proteins, senescent organelles and invading pathogens are delivered to the lysosome by double-membrane vesicles for degradation to maintain cell homeostasis [32][33][34][35] . Autophagy is also an indispensable part of human health which is involved in a variety of diseases 36 , including cancer 37 , neurodegenerative diseases 38 and microbial infections 39,40 . Thereby, autophagy is a crucial cellular machinery conserved from yeast to higher eukaryotes that maintains organ metabolism, genome stability, and cell survival, and functions as either tumor suppressor at early stage or promotor at late stage 41 . Studies have con rmed that autophagy plays a unique role in the struggle between Mtb and Mφ. Mtb can activate autophagy after infection with Mφ, while autophagy is involved in the clearance of Mtb by Mφ 42 . Autophagy ubiquitin-like protein has been revealed to be involved in the promotion of antituberculosis function in macrophages 42,43 , the stimulation of autophagy in macrophages leads to the maturation of Mtb phagosomes into phagolysosomes, moreover, the induction of autophagy inhibits the survival of Mtb in macrophages. On the other hand, Mtb can escape the killing effect of autophagy [44][45][46] , Mtb persists in macrophage phagocytosis by interfering with macrophage phagocytic lysosomal production, dysfunctional lysosomes are more susceptible to Mtb by inhibiting autophagy and directed migration of macrophages 47 .
In conclusion, autophagy plays an important role in the involvement of macrophages in immune responses 48,49 . However, in the process of infection by Mtb, there are many factors remains unclear, for instance, the autophagy-related genes and proteome involved in the formation of autophagy and the signaling pathways involved in the entire autophagy process. Furthermore, when macrophages from different sources are infected with Mtb, the signaling pathways for activation or inhibition may differ. In our previous studies, we have successfully established macrophages derived from hiPS by using the EB differentiation method and macrophages derived from THP-1 by using PMA (PKC activator). Therefore, this study is aimed to investigate the differences and similarities of the role of autophagy and the mechanism underlying in Mtb-infected hiPS-Mφ and THP-1-Mφ. This will further elucidate the mechanism behind the action of macrophages in anti-tuberculosis immune response and lay the foundation for revealing the pathogenesis of Mtb infection.

Macrophages differentiation from hiPS
hiPS-Mφ were generated by using a previously established protocol in our previous work 50  The medium was changed every 3 days. Continuous monocyte production was cultured for 17-19 days. Nonadherent monocytes were collected every 4 days and other cells were continuously cultured, the adherent cells were further cultured for 2-3 weeks to collect suspended cells. Non-adherent monocytes were cultured in RPMI 1640 medium supplemented with 10% FBS, 100ng/mL M-CSF, 50 ng/mL interleukin (IL)-3, and 50 μM β-ME, and then used for identi cation experiments after 10 days.

Macrophages differentiation from THP-1
THP-1 cells were centrifuged at 1000 rpm for 5 min, the supernatant were discarded, and then THP-1 cells were resuspended in medium supplemented with 100 ng/mL PMA (Beyotime, SH, China) and incubated for 24 hours in 5% CO 2 with 95% humidity at 37℃. After that, cells were cultured with the original medium and the medium was changed every two days.

Giemsa stain assay.
The slides were stained for 30 min with a working solution of Giemsa stain prepared from a commercially available stock solution (Beijing Solarbio Science & Technology Co., Ltd., BJ, China) according to recommendations of the manufacturer. The slides were then washed 2×1 min in ddH 2 O and air dried, then detected with microscope.
2.5 Phagocytosis assay. The medium of hiPS-Mφ and THP-1-Mφ was aspirated, and then washed cells once with PBS. The Indian ink was diluted to the medium at a ratio of 1:1000 and mixed, the reagent was added to the cell culture dish and incubated in 5% CO 2 with 95% humidity at 37℃ for 1 h , and cells were observed under the microscope.

BCG infection
Fully dissolve BCG lyophilized powder with sterile physiological saline and the concentration of BCG strain after constant volume is 1×10 7 CFU/mL. hiPS-Mφ and THP-1-Mφ were infected with 10 MOI of BCG indicated condition priorto were harvested for analysis.

Preparation and analysis of transcriptome sequencing samples
After BCG-infected cells were cultured for 24 h, the cells were digested by 0.25% trypsin (Gibco, NY, USA) and harvested. The total RNA of the cells was extracted according to the Trizol product speci cation, and the purity and concentration of the RNA samples were determined by an Agilent 2100 Analyzer (USA). After the total RNA of the sample had passed the test, the sequencing library was established according to the standard procedure and was sequenced using Illumina Hiseq4000 with a sequencing read length of 150 bp (PE150). According to the standard transcriptome sequencing analysis process, sequencing data output statistics, gene expression level analysis, differential gene expression analysis, differential gene KEGG enrichment analysis and differential gene GO enrichment analysis were performed.

Scanning electron microscope
The sample was immersed in Gluta xative at 4℃ overnight and then post-xed for1-2 h in1% citric acid solution, dehydrated in ethanol. And the sample was dried using a Hitachi HCP-2 critical point dryer to reach the dry critical point. The above-prepared sample was observed by a scanning electron microscope (SU-8010 type, Hitachi), and photographed.

Transmission electron microscope
The sample was immersed in Gluta xative at 4℃ overnight, post-xed for 1-2 h in1% citric acid solution, dehydrated in ethanol and embedded with a mixture of Spurr embedding agent and acetone (V/V = 1/1).The sample was sliced using a LEICA EM UC7 ultrathin slicer with a thickness of 70-90 nm; the obtained sections were stained with lead citrate solution, uranyl acetate and 50% ethanol saturated solution for 5-10 min. Finally, the sections were observed and photographed in a transmission electron microscope (H-7650,Hitachi).

Real-time Quantitative PCR Detecting System (qPCR)
RNA was extracted by using TRIzol TM Reagent, and cDNA was obtained by reverse transcription using the PrimeScript RT Reagent Kit according to the instructions. The qPCR analysis was set up in duplicate with SYBR Premix Ex Taq (Takara Bio, Inc., BJ, China) and performed using the 7500 Real-Time PCR system (Applied Biosystems; Thermo Fisher Scienti c, Inc.). The cDNA was used as a template for qPCR, using βactin as the internal reference gene, the relative expression of mRNA was calculated by 2 -ΔΔCT method, and the results were statistically plotted. The mRNA uorescent quantitative PCR primer information is shown in Table3. All primers were synthesized by Bio-Bioengineering Co., Ltd (Shanghai, China) .

Laser confocal detection of autophagy protein expression
The Ad-mCherry-GFP-LC3B was purchased from Beyotime (Shanghai, China) .The cells were seeded at 1 × 10 5 cells/well into 24-well plates. The Ad-mCherry-GFP-LC3Bwas added at an MOI of 20, and we observed the expression of uorescent proteins after 24 h and 48 h of infection by laser scanning confocal microscope. Red puncta represented autolysosome, and yellow puncta overlaid by green and red appearing in the images indicated autophagosomes formation. they were counted to determine the autophagy level.

Western blot analysis
The cells were homogenized in lysis buffer (IP Buffer, 10 μg/μL Aprotinin,10 μg/μL Leupeptin, 100 mM PMSF,200 mM Vanadate (pH 10.0)overnight at 4℃, the lysates were centrifuged at 13,000 × g for 30 min at 4℃, and the supernatants were collected.The soluble protein concentration was determined via the

Statistical analysis
Statistical analysis was conducted with GraphPad Prism 7 software. One-way ANOVA with uncorrected Fisher's least signi cant differences test was applied for the analysis of two independent variables.
Comparison between two samples was done with the unpaired t test. More details are described in the gure legends. Mean values of three independent experiments are shown. Error bars are shown as mean ± SD. ns, non-signi cant,*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Preparation and growth of functional macrophages derived from hiPS and THP-1.
To establish a hiPS-Mφ model, hiPS was induced to EB and matured after 8-11 days of culture. When EB was completely adherent, it began to differentiate toward monocytes. After culturing for about 12-20 days, suspended cells gradually appeared and were collected every 4-5 days into a new gelatin-coated petri dish and the remaining adherent cells continued to be cultured in fresh monocyte medium. Macrophage medium containing two cytokines of IL-3 and M-CSF at the same concentration was added to the collected suspension cells to differentiate monocytes into macrophages. The matured macrophages were observed to have typical macrophage morphology under the microscope ( Figure.1A). Meanwhile, THP-1 cells were initially circular suspension cells, and were cultured in the original medium for several days after being treated with PMA for 24 hours. The cells were adherent and had various forms, such as ellipse, fusiform, irregular shape, etc., which conformed to the morphology of typical macrophages. The collected hiPS-Mφ and THP-1-Mφ with typical macrophage morphological characteristics were identi ed by Giemsa stain assay. After staining, the color of the cells changed to purple and the shape of the cells displayed round, oval, spindle-shaped and irregularly shaped. Some cells were observed as convex pseudopods with other typical Mφ morphological characteristics upon bright-eld microscopy (Fig. 1B). The phagocytic ink granule level was determined by performing an ink phagocytosis test on the obtained macrophage-like cells. As a result, the macrophage-like cells had ink particles in the cytoplasm, indicating that they had a phagocytic function (Fig. 1C). All results further con rmed the establishment of the functionally active hiPS-Mφ and THP-1-Mφ model.

BCG induced autophagy in hiPS-Mφ and THP-1-Mφ
To verify whether BCG could induce autophagy in hiPS-Mφ and THP-1-Mφ, hiPS-Mφ, THP-1-Mφ, hiPS-Mφ infected with BCG and THP-1-Mφ infected with BCG were observed by scanning electron microscope (SEM) and transmission electron microscope (TEM). SEM observation showed that after 24 hours of BCG infection of hiPS-Mφ, the cells began to shrink and severe lysis occurred. The distribution of rod-shaped bacteria was observed in the extracellular cells ( Fig. 2A). At the same time, similar changes occurred in THP-1-Mφ after 24 h of BCG infection (Fig. 2B). TEM observation showed that after 24 hours of BCG infection of hiPS-Mφ, the endoplasmic reticulum was swollen and the number of autophagosomes increased, but it was not obvious (Fig. 3A). When BCG infected THP-1-Mφ cells for 24 h, the number of autophagosomes increased signi cantly, but the endoplasmic reticulum did not show obvious swelling like hiPS-Mφ (Fig. 3B).
To ascertain whether BCG would induce autophagy in hiPS-Mφ and THP-1-Mφ, hiPS-Mφ, hiPS-Mφ infected with BCG, THP-1-Mφ and THP-1-Mφ infected with BCG were transfected with Ad-mCherry-GFP-LC3B to monitor autophagic ux. In the case of non-autophagy, Ad-mCherry-GFP-LC3B will cause the cells to exhibit diffuse yellow uorescence due to the superposition of mCherry and GFP. While in the case of autophagy, the GFP signal is unstable in the acidic conditions of the lysosome lumen and mCherry is acid-stable. Thus, colocalization of both GFP and mCherry uorescence (yellow puncta) indicates that an autophagosome has not fused with a lysosome, whereas a mCherry signal without GFP (red puncta) indicates an autolysosome formation 51 . Laser confocal microscopy showed that after infecting with Ad-mCherry-GFP-LC3B for 24 h and 48 h, the inhibition or promotion of autophagy was more obvious after 48 h of infection than that of 24 h (Fig. 4, 5). Besides, we also studied the effect of the 3-methyladenine (3-MA) which is a class III PI3K inhibitor and can be used to inhibit autophagy, laser confocal microscopy showed that the uorescence of the control cells and the 3-MA-treated hiPS-Mφ were weak and diffusely presented in the cytoplasm, indicating that the number of intracellular autophagosomes is small (Fig. 4, 5). The uorescence of hiPS-Mφ and THP-1-Mφ in the BCG treatment group was mostly aggregated spots, and the uorescence intensity was strong, indicating that the cells were autophagic, and the number of autophagosomes was high. At the same time, some cells had only red puncta, indicating the formation of autolysosomes and the cells were in the late stage of autophagy. Besides, the uorescence expression of hiPS-Mφ and THP-1-Mφ in the 3-MA + BCG treatment group was similar to that in the BCG treatment group, except that the uorescence intensity and the spots were slightly lower, indicating the degree of autophagy was slightly lower than that of the BCG treatment group (Fig. 4, 5).
The results are shown in Table 1. According to the analysis, the proportion of valid data in the six samples was 97.88%, 97.94%, 98.14%, 97.96%, 97.93%, and 97.47% respectively, and Q30 was 97.33%, 97.39%, 97.41%, 97.19%, 97.44% and 97.37% respectively. According to the above data, the sequencing quality is excellent and the data can be further analyzed (Table 1). At the same time, Control THP-1-Mφ samples (Control_1, Control_2, and Control_3) and BCG-infected THP-1-Mφ samples (BCG _1, BCG _2, and BCG _3) were also sent to Lianchuan Biotech for same tests. The results are shown in Table 2. According to the analysis, the proportion of valid data in the six samples was 9 98.72% 98.73% 98.71% 98.81% 98.80% and 98.37% respectively, and Q30 was 96.91% 97.00% 96.94% 97.36% 96.68% and 96.49% respectively. According to the above data, the data can be further analyzed ( Table 2). The abundance values of gene expression were measured using FPKM (Fragments Per Kilobase of exon model per Million mapped reads). FPKM represents the number of sequencing fragments contained in the bases per thousand transcripts sequenced per million sequencing bases. Brie y, the FPKM value can be understood as the amount of expression of a gene. Plotting the FPKM data of each sample into a box plot can intuitively understand the level of gene expression from the overall level, and for samples with biological replicates, it can also be used to initially judge the repeatability of the designed sample. In the gure, the abscissa is the name of the sample, and the ordinate is log10 (FPKM). The box plot of each region corresponds to ve statistics (top to bottom, maximum, upper quartile, median, lower four quantile and minimum). It can be seen from the gure that the hiPS-Mφ and THP-1-Mφ has sample repeatability (Fig. 6A, B). The value of log10 (FPKM) of the different expression levels of different samples was plotted as an expression density map, which can compare the changes in expression trends between different samples. In the gure, the abscissa is log10 (FPKM) and the ordinate is the density of the gene. It can be seen from Fig. 6A.b and Fig. 6B.b that the bio-repetition of hiPS-Mφ and THP-1-Mφ tends to be the same, but the expression trends of these two cells are different. Differential expression analysis was performed on all genes according to the signi cant difference threshold of |log2 (Fold Change, FC)| ≥1, p < 0.05 (where Fold change indicates the fold difference). The statistical results showed that compared with the control hiPS-Mφ, BCG-treated hiPS-Mφ had 832 signi cantly differentially expressed genes, of which 572 genes were up-regulated and 260 genes were down-regulated (Fig. 7A). Compared with the control THP-1-Mφ, BCG-treated THP-1-Mφ had 397 signi cant differentially expressed genes, of which 211 genes were up-regulated and 186 genes were down-regulated (Fig. 7B). As shown in the Fig. 7A.a and Fig. 7B.a, the volcano map can understand the overall distribution of differentially expressed genes. Therefore, the volcano map is further analyzed for the differential gene expression level. The volcano map uses log2 (Fold Change, FC) as the abscissa and -log10 (p-value) (represents the statistical signi cance of the difference in gene expression change) as the ordinate. All genes in the expression analysis were added for mapping. Among them, red represents a signi cantly differentially expressed gene that is upregulated, blue represents a signi cantly differentially expressed gene that is down-regulated, and gray represents a non-signi cantly differentially expressed gene.
Differential gene cluster analysis can be used to determine the clustering pattern of genes under different experimental conditions. According to the similarity of the gene expression pro les of the samples, the genes were clustered and analyzed to visually show the expression of the genes in different samples, and then the biological related information was obtained. To better re ect the cluster expression pattern, log10 (FPKM + 1) was used for gene expression display, and the differential gene FPKM was displayed by Z value. The results are shown in Fig. 8, in which the abscissa is the sample name, the ordinate is the name of the differentially expressed gene, the different colors indicate different gene expression levels, and the color is expressed in blue from white to red to indicate the expression level from low to high. Red indicates a highly expressed gene and dark blue indicates a low expressed gene. The results showed that in the regulation of autophagy-related genes, DNA damage-inducible transcript 4 (DDIT4) and Exocyst complex component 8 (EXOC8) were signi cantly up-regulated in hiPS-Mφ. That means, the expression of DDIT4 and EXOC8 genes in hiPS-Mφ was signi cantly increased after BCG infection compared with the expression in uninfected cells (P < 0.0001) (Fig. 8A). At the same time, cathepsin D (CTSD) and RAB7B were signi cantly down-regulated in THP-1-Mφ, that means, compared with uninfected THP-1-Mφ cells, the expression of CTSD and RAB7B genes in THP-1-Mφ was signi cantly inhibited after BCG infection (P < 0.0001) (Fig. 8B). The GO enrichment analysis results histogram is a distribution of the number of differential genes re ected in the GO term enriched in the cellular component, the biological process, and the molecular function. The result is shown in Fig. 9A. In the hiPS-Mφ group, GO enrichment analysis of differentially expressed genes indicated that it is mainly involved in processes such as transcriptional regulation, signal transduction, cell membrane composition and protein binding. While in the THP-1-Mφ group, GO enrichment analysis of differentially expressed genes indicated that it is mainly involved in signal transduction, in ammatory response, immune response, cell membrane composition and protein binding (Fig. 9B).
The Kyoto Encyclopedia of Genes and Genomes (KEGG) is a public database commonly used in genome deciphering. Genomic information is stored in a gene database, including intact and partially sequenced genomic sequences; more advanced functions are stored in the Pathway database, including graphical cell biochemical processes such as metabolism, membrane transport, signaling, and cell cycle. Therefore, the analysis of the pathway will help us better understand the biological functions of genes. KEGG is an important public database on Pathway that provides integrated pathway queries that provide not only all possible metabolic pathways, but also comprehensive and detailed annotation of the enzymes involved in the catalysis of each step. The KEGG enrichment analysis was performed using ggplot2, and the analysis results were plotted as scatter plots. The Rich Factor in the gure indicates the number of differential genes located in the KEGG / the total number of genes in the KEGG, the larger the enrichment factor value, the greater the degree of KEGG enrichment. Figure 10A shows a partial integrative metabolic pathway for KEGG enrichment analysis of hiPS-Mφ after BCG treatment. It can be seen that the pathway associated with in ammation, autophagy, apoptosis, etc. changes sigini cantly in hiPS-Mφ treated with BCG. At the same time, Fig. 10B shows a partial integrative metabolic pathway for KEGG enrichment analysis of THP-1-Mφ after BCG treatment. It can be seen from the gure that pathway associated with in ammation, autophagy, apoptosis, etc. also changes sigini cantly in THP-1-Mφ after BCG treatment.
Compared with THP-1-Mφ, hiPS-Mφ is enriched in tuberculosis, Toll-like receptors, TNF, PI3K-AKT, phagosomes, nodular receptors, NF-κB, mTOR, MAPK, IL-17, chemokine signaling pathway, autophagy pathway and apoptotic pathways, but the enrichment factors are different, indicating the number of differential genes is different. Based on this, it is speculated that there are some differences in the mechanism underlying anti-tuberculosis immunity in hiPS-Mφ and THP-1-Mφ after BCG infection.
The expression levels of autophagy-related genes LC3B, GABARAPL2, P62, DDIT4, EXOC8, CTSD and RAB7B were detected by qPCR after extracting total RNA of hiPS-Mφ with BCG infection. In general, when autophagy is promoted, LC3B rises, GABARAPL2 decreases, and P62 decreases. Conversely, when autophagy is suppressed, LC3B decreases, GABARAPL2 rises, and P62 rises. Compared with uninfected hiPS-Mφ, the expression of LC3B was signi cantly increased in hiPS-Mφ after BCG infection(P < 0.0001), the expression of GABARAPL2 was signi cantly decreased (P < 0.0001), and the expression of P62 was signi cantly decreased (P < 0.01) (Fig. 11A). Based on this, it is possible that the occurrence of autophagy was promoted by regulating the expression of the above genes in hiPS-Mφ with BCG infection. At the same time, the autophagy-related genes DDIT4 and EXOC8 genes are also signi cantly up-regulated in transcriptome sequencing, indicating that qPCR veri cation results are consistent with hiPS-Mφ sequencing results. Similarly, the expression of LC3B gene was signi cantly increased after BCG infection of THP-1-Mφ (P < 0.05), the expression of GABARAPL2 gene was signi cantly decreased (P < 0.001), and the expression of P62 gene was signi cantly decreased(P < 0.01). It can be speculated that the occurrence of autophagy was promoted by regulating the expression of CTSD and RAB7B in THP-1-Mφ with BCG infection. At the same time, CTSD and RAB7B are the signi cant down-regulated autophagy-related genes found in the THP-1-Mφ transcriptome sequencing. The qPCR veri cation results are consistent with the sequencing results (Fig. 11B).

PI3K-AKT-mTOR pathway is involved in hiPS-Mφ and THP-1-Mφ in response to tuberculosis infection
The In conclusion, this study showed that after BCG infection, the autophagy morphology and expression of autophagy-related genes in hiPS-Mφ and THP-1-Mφ cells changed signi cantly. It is preliminarily speculated that the expression of autophagy-related genes is regulated by PI3K/AKT/mTOR signaling pathway after BCG infection to promote the occurrence of autophagy. The difference is that for hiPS-Mφ cells, DDIT4 and EXOC8 genes are regulating the occurrence of autophagy, that is, the expression level increases when autophagy occurs, and for THP-1-Mφ, the RAB7B and CTSD genes negatively regulate the occurrence of autophagy, that is, the expression level decreases when autophagy occurs. The differences in the expression of genes and proteomes caused by BCG-induced autophagy in macrophages derived from different sources and their speci c regulatory mechanisms need further investigation.

Discussion
The macrophages currently used for immune responses are tumor cell lines-derived macrophages and primary cell lines-derived macrophages, but such cells have many disadvantages. The karyotype and function of macrophages derived from tumor cell lines (such as U937,THP-1) are easy to be abnormal.
Primary cells are di cult to be genetically manipulated because of the individual differences and low availablity, furthermore,they cannot be self-renewing. To solve all these problems, stem cells become an effective and practical source of macrophages. Macrophages derived from hES and hiPS are most commonly used in related researches but hES-Mφ have ethical problems, the selection of hiPS-Mφ as a research object can overcome the above-mentioned de ciencies. In addition, hiPS-derived monocytes and macrophages could potentially be a useful source of patient-speci c cells that are di cult or impossible to derive 52 . We have established the protocol to obtain macrophage derived from hiPS by using EB induced differentiation method. At the same time, THP-1-Mφ were obtained for comparison. We have performed a series of related identi cation tests on the hiPS-Mφ and THP-1-Mφ to determine whether we really get functional hiPS-Mφ and THP-1-Mφ. The main methods in our research include ink phagocytosis detection and Gimesa staining. As a result, it was con rmed that hiPS-Mφ and THP-1-Mφ can be dyed purple by the Giemsa Stain solution, and can also engulf ink particles. Based on the above experimental results, it can be determined that we have successfully obtained functionally active macrophages induced by hiPS and THP-1.
In addition, hiPS-Mφ and THP-1-Mφ was infected with BCG in this research and then subjected to transcriptome sequencing analysis, and the KEGG enrichment of BCG-infected hiPS-Mφ and THP-1-Mφ was further analyzed. The integrated metabolic pathway involved in ammation, apoptosis, self-induction and autophagy, etc., and is enriched in these aspects. Among them, in autophagy pathway, BCG infected hiPS-Mφ was signi cantly up-regulated by DDIT4 and EXOC8 genes (P < 0.0001). DDIT4 53,54 is present in the cytoplasm and is induced to be highly expressed under various adverse conditions such as hypoxia, insu cient nutrient supply, and oxidative stress. DDIT4 can promote autophagy of cells by inhibiting the target of mTOR kinase complex 1 (mTORC1) and regulating Atg4B activity. Exocyst Complex 8 (EXOC8) is one of the eight subunits of the exocyst complex 55,56 . The main role of the exocyst complex lies in the regulation of cell polarity, the targeted transport of vesicles, and the migration of cells. Studies have shown that the function of exocyst complex is regulated by small GTPase, mammalian EXOC8 is an effector of Ral GTPases, and Ral GTPases interact with EXOC8 and another subunit Sec5 of exocyst complex to regulate the composition of exocyst [57][58][59] . The results of our research suggest that DDIT4 and EXOC8 may be involved in the Bacillus Calmette-Guerin induced autophagy. At the same time, the CTSD and RAB7B genes were signi cantly down-regulated in BCG-infected THP-1-Mφ (P < 0.001). It is suggested that there is a difference in BCG infection between hiPS-Mφ and THP-1-Mφ in the autophagy signaling pathway. RAB7B belongs to the small GTPase, Studies have shown that RAB7B regulates LC3 processing by interacting with Atg4B 60,61 . And RAB7B can negatively regulate lipopolysaccharide-treated macrophages, induce the production of TNF-α, IL-6, NO, and IFN-β, and enhance lipopolysaccharideinduced activation of mitogen-activated protein kinase, NF-κB and IFN. Regulating the Factor 3 signaling pathway, in turn, promotes degradation of TLR4. The Toll-like signaling pathway was activated in our transcriptome results 34 , but there was no signi cant change in the expression of TLR4. It is speculated that THP-1-Mφ of BCG infection differs in innate immunity and autophagy. CTSD belongs to the endopeptidase, which is present in lysosomes and degrades peptides and proteins. It is involved in the regulation of physiological processes, such as the digestion of intracellular proteins, the activation of growth factors, peptides and hormones, the activation and processing of zymogens, the presentation of brain antigens and the programmed cell death 62 . Studies have found that when autophagy ux increases, lysosomes may be damaged and CTSD expression is inhibited 63 . Thus, in autophagy pathway, although the genes changed signi cantly after BCG infection of hiPS-Mφ and THP-1-Mφ were different, the signaling pathways involved are interrelated.
After BCG infection of macrophages, an increase in autophagosomes was also observed from the submicroscopic structure. Besides, the study also found that autophagy has a highly evolved nature, its occurrence and development are regulated by a variety of autophagy-related genes. At present, at least 30 autophagy-speci c genes and more than 50 related genes have been identi ed 64 . LC3B, GABARAP and P62 are typical autophagy markers 65,66 . This study found that the expression of LC3B gene in BCGinfected hiPS-Mφ and BCG-infected THP-1-Mφ was increased (P < 0.0001), the expression of GABARAPL2 and P62 genes was decreased (P < 0.01). Then the BCG-infected hiPS-Mφ and BCG-infected THP-1-Mφ was treated with 3-methyladenine (3-MA). 3-MA is a popular inhibitor of the autophagic pathway and has been reported to inhibit the activity of PI3-Kinase. It can block the formation of preautophagosome, autophagosome and autophagic vacuoles 67 . Researches have reported that in the human myeloid leukemia cell line, K562, crotoxin-induced apoptotic cell death was potentiated by 3-MA 68 . And 3-MA also has been reported as the contribution to the upregulation of oridonin-induced apoptosis in the human cervical carcinoma cell line, HeLa 69 . According to current researches, 3-MA or suppression of autophagy regulatory pathways may provoke apoptotic e ciency of chemotherapeutic agents in prostate 70 , breast 71 , colon 72 , lung 73 , HeLa cancer cells 69 and so on. From all these grounds, 3-MA can be de ned as a suitable inhibitor of the autophagic pathway and can be used in our research as a contrast to BCG.
According to the results, the expression level of the autophagy-related genes changed correspondingly, and the level of autophagy was also decreased after treated with 3-MA. At the same time, laser confocal detection of mCherry-GFP-LC3B uorescent protein expression also con rmed that when BCG infected hiPS-Mφ and THP-1-Mφ, the uorescence intensity and the number of autophagosomes increased, while the 3-MA + BCG group had weaker uorescence than that of the BCG group. It was shown that 3-MA can inhibit autophagy caused by BCG infection and the mechanism may be related to the inhibition of 3-MA on PI3K as 3-MA is a class III PI3K inhibitor. A large amount of data indicates that autophagy is closely related to the PI3K/AKT/mTOR signaling pathway and the expression of DDIT4 is related to the PI3K/AKT pathway 74,75 . In this study, the above autophagy-related proteomes were detected. The results showed that the LC3BII/I ratio was signi cantly increased after BCG infection of hiPS-Mφ and THP-1-Mφ (P < 0.0001), and the expression of p-AKT decreased sigini cantly (P < 0.0001). The expression of p-mTOR in hiPS-Mφ and THP-1-Mφ decreased after BCG infection, the difference is that it was signigicantly in THP-1-Mφ (P < 0.0001) but slightly in hiPS-Mφ (P < 0.05). The ratio of LC3BII/LC3BI was signi cantly decreased after 3-MA treatment of hiPS-Mφ and THP-1-Mφ (P < 0.01), and the expression of p-mTOR increased signi cantly (P < 0.01). The expression of p-AKT in hiPS-Mφ and THP-1-Mφ increased after 3-MA treatment, it was signigicantly in THP-1-Mφ (P < 0.0001) but only saw a small rise in hiPS-Mφ. All above results indicate that autophagy occurs in hiPS-Mφ and THP-1-Mφ, because post-translational modi cation of LC3 occurs during autophagy, lipiding LC3-I to LC3-II,LC3II is then modi ed at the Cterminus by phosphatidylethanolamine and tightly bound to the autophagosome membrane. In addition, in the transcriptome analysis, compared with the uninfected control cells, the expression of autophagyrelated genes DDIT4, EXOC8, RAB7B and CTSD in the hiPS-Mφ and THP-1-Mφ after BCG infection was signi cantly different (P < 0.0001). Therefore, this experiment performs validation analysis at the mRNA and proteome levels. It was found that the expression of DDIT4 and EXOC8 genes was signi cantly increased when hiPS-Mφ was infected with BCG (P < 0.0001). When the inhibition of autophagy occurred, the expression levels of DDIT4 and EXOC8 genes decreased (P < 0.0001). At the same time, after BCG infection of THP-1-Mφ, the expression of RAB7B and CTSD gene decreased when autophagy was promoted (P < 0.001). When treated with 3-MA, the expression of RAB7B and CTSD increased when autophagy was inhibited. (P < 0.001).
The macrophages currently used in related research on immune diseases are mainly derived from tumor cell lines and primary cell lines which have many defects and limiations. Under these circumstances, hiPS-Mφ with a large number of advantages are very eager to be studied. However, based on the above results, there are some differences and relevances in the mechanisms involved in the occurrence of autophagy induced by BCG in macrophages derived from different sources (hiPS-Mφ and THP-1-Mφ). In order to nd the pathogenesis and therapeutic targets of TB, establishing an in vitro TB model of hiPS-Mφ is critical. While regarding the autophagy induced by BCG infection in macrophages derived from different sources, there are still some issues needed to be further clari ed. For example, when BCG infection induces autophagy in hiPS-Mφ and THP-1-Mφ, the expression of some genes and proteomes is different, and the autophagy-related genes involved are also different, the reasons still remain unclear.
Furthermore, the speci c regulatory mechanisms involved and the differences of the mechanisms are still unclear and need further investigation. Before using hiPS-Mφ for TB pathogenesis and autophagy-related researches, a large variety of researches still urgently needed to be completed.

Conclusion
This study showed that after BCG infection, the autophagy morphology and expression of autophagyrelated genes and proteomes in hiPS-Mφ and THP-1-Mφ changed signi cantly, and it is preliminarily speculated that the expression of autophagy-related genes is regulated by PI3K/AKT/mTOR signaling pathway after BCG infection to promote the occurrence of autophagy. The difference is that for hiPS-Mφ, DDIT4 and EXOC8 genes are positively regulating the occurrence of autophagy, and for THP-1-Mφ, the RAB7B and CTSD genes negatively regulate the occurrence of autophagy (Fig. 13).       cells in the BCG treatment group was mostly aggregated spots, and the uorescence intensity was strong, the cells were autophagic, and the number of autophagosomes was high. The uorescence expression of hiPS-Mφ and THP-1-Mφ cells in the 3-MA+BCG treatment group was similar to that in the BCG treatment group, except that the uorescence intensity and the spots were slightly lower, the degree of autophagy was slightly lower than that of the BCG treatment group.       Uncorrected Fisher's least signi cant differences test: ns, non-signi cant; *p < 0.05, **p < 0.01,***p < 0.001, ****p < 0.0001. Fisher's least signi cant differences test: ns, non-signi cant; *p < 0.05, **p < 0.01,***p < 0.001, ****p < 0.0001.