Abstract

   Transplanted adipose stem cells (ASC) have a low survival rate in the
   body, and there are not many ASC that can be effectively used, which
   weakens their tissue repair function. Based on this status quo, a new
   type of copper‐based metal–organic network (Cu‐MON) was used to
   pretreat cells to regulate cell activity in order to improve the
   efficacy of cell therapy or reduce the number of cells used, thus
   reducing the cost of clinical treatment. Gene expression changes before
   and after Cu‐MON treatment of normal donor adipose stem cells (ND‐ASC)
   and type 2 diabetes mellitus adipose stem cells (T2DM‐ASC) were
   evaluated through RNA sequencing, KEGG and GO enrichment analysis. The
   results showed that Cu‐MON improved ASC cell quality by regulating
   immune response and promoting paracrine secretion. IL‐17 signaling
   pathway and IL‐6, CXCL8, and MMP‐9 were key pathways and necessary
   genes that affected the ability of stem cells. In addition, Cu‐MON also
   improved stem cell antiviral ability through Type I interferon
   signaling pathway. Our research showed that Cu‐MON had improved the
   cell quality of ASC by regulating immune response, promoting paracrine
   secretion, and improving antiviral capabilities. This approach to
   biomaterial pretreatment is fast, convenient, and relatively safe, and
   provides new strategies for improving the efficiency of cell therapies.

   Keywords: gene expression, RNA sequence analysis, stem cells,
   transcriptome, type 2 diabetes mellitus
     __________________________________________________________________

   RNA sequencing, KEGG and GO enrichment analysis showed that Cu‐MON
   improved ASC cell quality by regulating immune response and promoting
   paracrine secretion through IL‐17 signaling pathway. Cu‐MON also
   improved stem cell antiviral ability through Type I interferon
   signaling pathway. This approach to biomaterial pretreatment is fast,
   convenient, and relatively safe, and provides new strategies for
   improving the efficiency of cell therapies.

   graphic file with name FBA2-7-e1485-g005.jpg

1. INTRODUCTION

   Stem cells have shown promising prospects in the treatment of various
   diseases,[30] ^1 , [31]^2 including chronic kidney disease,
   osteoarthritis, autoimmune diseases, limb ischemia, cancer, myocardial
   infarction, and so on. Among all the stem cells, adipose stem cells
   (ASC) have been favored by more and more researchers because of their
   easy access and abundance. They are used to promote tissue repair and
   regeneration[32] ^3 , [33]^4 and inhibit aging.[34] ^5 , [35]^6
   Mechanisms supporting the therapeutic potential of ASC include local
   differentiation after entering the human body[36] ^7 ; affecting the
   differentiation and activation of immune cells through direct action or
   paracrine action of immune cells, and rebuilding the immune balance of
   the body.[37] ^8 , [38]^9 However, there are still some questions to be
   answered about their application in clinical therapy, including: Choose
   allograft or autologous transplantation? How to control the quality of
   adipose stem cells from different sources? Limited by the disease
   microenvironment, stem cells are inactivated a few hours after entering
   the body.[39] ^10 How can their vitality be improved? The number of
   cells required for stem cell therapy is huge.[40] ^11 Can we optimize
   the culture conditions to improve its preparation efficiency or
   therapeutic activity?

   Regarding the first question, it has been reported that autologous ASC
   is superior to allogeneic ASC in treating acute pyelonephritis in
   rabbits, improving cerebral infarction, and promoting burn wound
   healing.[41] ^12 , [42]^13 Moreover, autologous stem cells are derived
   from the patient's own body, and the risk of immune rejection is very
   low. Therefore, the choice of autologous cell transplantation may get
   more satisfactory treatment results. However, for patients with
   disease, their cell function is impaired. For example, due to long‐term
   “immersion” in high blood sugar, the blood vessels of diabetic patients
   have increased reactive oxygen and free radicals in their blood
   vessels,[43] ^14 , [44]^15 causing vascular cells to be oxidized and
   damaged. Specifically, compared with ASC from normal donors, ASC from
   diabetic patients have reduced stemness,[45] ^16 reduced cell paracrine
   function,[46] ^17 and reduced metabolic capacity. Moreover, stem cells
   isolated from patients with autoimmune diseases also have impaired cell
   cycle and immune regulation functions.[47] ^18 , [48]^19 In addition,
   cell quality of ASC is closely related to the culture system.[49] ^20 ,
   [50]^21 Studies have shown that cells that have undergone a
   cryopreservation and resuscitation process are more likely to be
   cleared by immune cells in the body and are more likely to age.[51] ^22
   , [52]^23 And aged stem cells significantly lose their progenitor cell
   characteristics and reduce antioxidant capacity during in vitro
   expansion. After cell transplantation, limited by the influence of the
   microenvironment of the disease site, less than 5% of the transplanted
   cells survive within 10 days after injection.[53] ^10 Therefore, it is
   essential to improve the efficacy of stem cell therapy by optimizing
   culture conditions, biomaterial treatment, or gene modification.

   In recent years, scientists have used various ways to improve cell
   efficacy, including genetic modification,[54] ^24 , [55]^25
   biomaterials to change the external environment,[56] ^4 , [57]^26 ,
   [58]^27 and using the secretome of stem cells.[59] ^28 , [60]^29 In our
   previous studies, Cu‐MON combined baicalein and copper ions to achieve
   antioxidant and inflammatory inhibition,[61] ^30 promoting partial
   recovery of type 1 diabetes mellitus adipose stem cells (T1DM‐ASC)
   transcriptome to normal donor adipose stem cells (ND‐ASC) by enhancing
   the stemness and paracrine effects, thus promoting angiogenesis
   hindlimb motor function recovery in mice with critical limb
   ischemia.[62] ^31 Among them, baicalein played the role of antioxidant
   and immune regulation,[63] ^32 , [64]^33 and copper ion played a role
   of promoting angiogenesis and redox homeostasis.[65] ^34 Based on the
   above findings, this biomaterial was used for the pretreatment of
   ND‐ASC. The results showed that Cu‐MON improved ND‐ASC quality by
   regulating immune response and promoting paracrine, which was expected
   to be used in tissue repair and blood vessel regeneration. In addition,
   the effect of Cu‐MON on type 2 diabetes mellitus adipose stem cells
   (T2DM‐ASC) was further explored to determine the quality‐promoting
   effect of Cu‐MON, a biomaterial independently developed by the research
   team, on mesenchymal stem cells. This biomaterial pretreatment method
   is fast, convenient, and relatively safe, providing new strategies for
   improving the efficiency of cell therapy or reducing the number of
   cells used, thus reducing the clinical treatment cost.

2. MATERIALS

2.1. Cell lines

   Human peripancreatic adipose tissues were obtained from T2DM (n = 2)
   and nondiabetic organ donors (n = 2). ND‐ASC and T2DM‐ASC were
   extracted from the indicated adipose tissues according to the guideline
   of Declaration of Helsinki. All the experimental procedures were
   approved by the Ethics Committee of Tianjin First Central Hospital of
   Nankai University and informed consents were obtained (2017N080KY). The
   information of the ND and T2DM donors is available in Table [66]1.

TABLE 1.

   Donor information.
        N (Male/Female) Age (y) HbA1c (%) BMI (kg/m^2)
   ND   Female          31      —         20.83
   ND   Female          32      —         19.10
   T2DM Male            49      7.7       26.30
   T2DM Male            49      5.8       27.78
   [67]Open in a new tab

2.2. ND‐ASC and T2DM‐ASC culture

   ND‐ASC and T2DM‐ASC were cultured by serum‐free medium (NC0103, Yocon
   Biology Technology Company, Beijing, China) supplemented with 1%
   mesenchymal stem cell serum‐free medium addition3 (NC0104.S, Yocon
   Biology Technology Company, Beijing, China), and incubated at 37°C and
   5% CO[2]. The pancreatic enzymes used were stem cell mild digestive
   enzymes (NC1004.2, Yocon Biology Technology Company, Beijing, China).

2.3. Synthesis of Cu‐MON

   The synthesis of Cu‐MON was as reported in the literature.[68] ^30 In
   simple terms, 2.7 mg BAI was dispersed in 1 mL water, and 1.7 mg
   CuCl[2]·2H[2]O was dissolved in 1 mL water. The two solutions were
   mixed. 2 mL PBS (50 mM, neutral) was added to regulate the reaction,
   and the reaction was stirred at 600 rpm for 1 h, then centrifugally
   washed.

2.4. RNA sequencing

   Each patient listed in Table [69]1 corresponded to a biological
   duplicate sample. Cells from both patients were mixed and used as a
   third biological replication sample for follow‐up experiments. The
   purpose of using small samples is to accumulate evidence and make
   “sound choices” for large samples among validated paradigms that are
   sensitive to interindividual and intraindividual correlated variation.
   Therefore, ND‐ASC and T2DM‐ASC were seeded in a 6‐well plate and
   incubated for 24 h (n = 3). Then cells were incubated with Cu‐MON for
   48 h and harvested, and then sent to Wuhan BGI Technology Co., Ltd. for
   transcriptomic sequencing. Then genes were extracted by
   Trizol‐chloroform‐isopropanol, OD[260/280] was detected by NanoDrop,
   and RNA integrity was determined by agarose gel electrophoresis. The
   tested qualified Total RNA samples were subjected to DNase I digestion
   and enriched with mRNA using Oligo (dT) magnetic beads. Interrupting
   the mRNA, random primers were added for the first‐strand synthesis of
   the cDNA. The thesis of cDNA two strands was performed with dUTP
   instead of dTTP. The amplified cDNA was end‐repaired and connected with
   “A”, and the ligation products were amplified by PCR. After denaturing
   the PCR product into single strands, circularization was performed to
   obtain a single‐stranded circular DNA library. The final library was
   obtained after digestion of the uncyclized linear DNA molecules.
   Single‐stranded circular DNA molecules were replicated through the
   rolling ring to form DNA nanospheres (DNB) and were sequenced.

   Data obtained by sequencing were called raw reads and subsequently
   filtered using SOAPnuke software (v1.5.2, ‐l 15 ‐q 0.2 ‐n 0.05) for
   quality control (QC). The resulting clean reads were aligned to the
   reference sequence (Homo_sapiens, NCBI, GCF_000001405.39_GRCh38.p13) by
   HISAT (Hierarchical Indexing for Spliced Alignment of Transcripts,
   v2.0.4, ‐‐sensitive ‐‐no‐discordant ‐‐no‐mixed ‐I 1 ‐X 1000 ‐p 8
   ‐‐rna‐strandness RF). Then a deeper mining and analysis were performed
   on the BGI website ([70]https://biosys.bgi.com/#/report/login) through
   gene quantitative analysis and gene expression level analysis
   (principal components, correlation, differential gene screening,
   pathway significant enrichment analysis, clustering and protein
   interaction network, etc.).

   Where gene quantitative analysis was obtained by alignment of clean
   reads to the reference gene sequence using Bowtie2 (v2.2.5, ‐q
   ‐‐sensitive ‐‐dpad 0 ‐‐gbar 99999999 ‐‐mp 1,1 ‐‐np 1 ‐‐score‐min
   L,0,‐0.1 ‐p 16 ‐k 200) and then calculating using RSEM (v1.2.8, ‐p 8
   ‐‐forward‐prob 0 ‐‐paired‐end). The DEseq2 differential gene (Q value
   ≤0.05 and |log2(fold change)| ≥ 1) was tested on the basis of the
   negative binomial distribution principle. Hierarchical clustering
   analysis was performed using the R package pheatmap (Parameter:
   Default).

   Pathway enrichment analysis included GO enrichment analysis and KEGG
   enrichment analysis. Among them, GO enrichment analysis mapped all
   candidate genes to each entry in the Gene Ontology database
   ([71]http://www.geneontology.org/), calculated the number of genes for
   each entry, and then applied hypergeometric test to find GO entries
   significantly enriched in candidate genes compared to all background
   genes in this species. The p value was calculated using the underlying
   function phyper of R
   ([72]https://stat.ethz.ch/R‐manual/R‐devel/library/stats/html/Hypergeom
   etric.html). When the p value was more positive, the correction package
   was q value
   ([73]https://bioconductor.org/packages/release/bioc/html/qvalue.html).
   Finally, using Q value (corrected p value) < = 0.05 as the threshold,
   the GO term satisfying this condition was defined as the GO term
   significantly enriched in the candidate genes.

   KEGG enrichment analysis applied hypergeometric test to identify
   pathway significantly enriched in candidate genes compared to the whole
   genome background. A final pathway of Q value ≤0.05 was defined as the
   pathway significantly enriched in the differentially expressed genes.
   The most important biochemical metabolic pathways and signal
   transduction pathways involved by the candidate genes were identified
   by pathway significant enrichment.

   External database expansion was completed based on the online widget of
   Dr. Tom system ([74]https://biosys.bgi.com/#/report/login), and its
   specific mechanism was as follows: Expression was converted into
   qualitative data. The expression of genes in a sample classification
   was transformed from numerical to qualitative expression level. All the
   genes under each classification label were ordered from most to least
   expressed. The expression levels were as follows: high, middle high,
   middle, middle low, and low (Table [75]2).

TABLE 2.

   Different proportion of genes were given corresponding expression
   levels.
   Expression level The proportion of total gene concentration
   High             0%–1%
   Middle high      1%–10%
   Middle           10%–30%
   Middle low       30%–50%
   Low              50%–100%
   [76]Open in a new tab

2.5. Secretion activity of ND‐ASC

   Human Angiogenesis Panel 1 (10‐plex) with V‐Bottom Plate (Biolegend
   740698, California, USA) was used to analyze the proteins in the
   culture medium. Briefly, a selected panel of capture beads were mixed
   and incubated with 25 μL of culture medium for 2 h. After washing, a
   biotinylated detection antibody cocktail was added and incubated for
   1 h. Then streptavidin‐phycoerythrin (SA‐PE) was then added and
   incubated for 30 min. Specific populations were segregated and PE
   fluorescence signal was quantified on a flow cytometer. The
   concentration of a particular analyte was determined using a standard
   curve .

2.6. Statistical analysis

   Data were presented as mean ± standard deviation (Mean ± SD). One‐way
   ANOVA was used to analyze data with only one variable. Two‐way ANOVA
   was used to analyze the data with two variables. All analyses were
   performed using GraphPad Prism 8 software, and the significance was
   expressed as follows: *p < 0.05, **p < 0.01, ***p < 0.001,
   ****p < 0.0001.

3. RESULTS

3.1. Quality assessment of the sequencing data

   After processing the sequencing data, Total Clean Bases was above 6.71
   Gb, with Q20 ranging from 96.47 to 96.72 and Q30 at 91.31% to 91.91%.
   The Clean Reads Ratio of the six samples was greater than 88%
   (Table [77]3). After obtaining the clean reads, HISAT and Bowtie2 were
   used to align the clean reads to the reference genome sequence and the
   reference gene sequence. The alignment results were counted as follows.

TABLE 3.

   Comparative results of the clean reads and the reference genome.
   Sample ND‐ASC1 ND‐ASC2 ND‐ASC3 Cu‐MON‐treated1 Cu‐MON‐treated2
   Cu‐MON‐treated3
   Total raw reads (M) 49.08 49.08 49.08 49.08 50.83 49.08
   Total clean reads (M) 44.75 45.29 45.14 44.94 44.99 44.71
   Total clean bases (Gb) 6.71 6.79 6.77 6.74 6.75 6.71
   Clean reads Q20 (%) 96.52 96.72 96.47 96.68 96.69 96.7
   Clean reads Q30 (%) 91.47 91.91 91.31 91.82 91.89 91.89
   Clean reads ratio (%) 91.17 92.27 91.97 91.56 88.51 91.11
   Total mapping (%, HISAT) 92.42 92.65 92.59 92.63 92.71 92.44
   Uniquely mapping (%, HISAT) 87.13 87.24 87.28 87.19 87.21 87.17
   Total mapping (%, Bowtie2) 80.75 80.55 79.7 82.45 83.4 81.52
   Uniquely mapping (%, Bowtie2) 77.31 76.99 76.29 78.71 79.75 77.97
   [78]Open in a new tab

3.2. Screening of the differentially expressed genes

   According to the principal component analysis map, the three samples of
   ND‐ASC and the three samples of Cu‐MON‐treated were far apart,
   indicating a difference between the two sample groups (Figure [79]1A).
   Comparing ND‐ASC and Cu‐MON‐treated by differential expression
   analysis, 153 differentially expressed genes were obtained. Among them,
   37 genes were upregulated and 116 genes were downregulated in
   Cu‐MON‐treated compared to ND‐ASC (Figure [80]1B, Table [81]4).

FIGURE 1.

   FIGURE 1
   [82]Open in a new tab

   Differences in transcriptome between ND‐ASC and Cu‐MON‐treated. (A) PCA
   plot illustrating the variances of ND‐ASC and Cu‐MON‐treated. (B)
   Volcano plot showing that 153 genes are differentially regulated
   between ND‐ASC and Cu‐MON‐treated. The log2(Fold‐Change) was estimated
   by DESeq2. Q value ≤0.05 and |log2(fold change)| ≥ 1. (C) Kyoto
   encyclopedia of Genes and Genomes (KEGG) analysis of differentially
   modulated genes classified by their biological functions and arranged
   according to their statistical significance. Protein–protein
   interaction networks based upon the IL‐17 signaling pathway (D) and
   viral protein interaction with cytokine and receptor (E). Heatmap
   representing expression (F) and KEGG networks (G) of the eight
   differentially expressed genes from IL‐17 signaling pathway.

TABLE 4.

   mRNAs regulated by Cu‐MON in Cu‐MON‐treated ND‐ASC, compared to ND‐ASC.
   Gene ID Gene symbol log2 Q value Gene ID Gene symbol log2 Q value Gene
   ID Gene symbol log2 Q value
   100133220 “GOLGA6L3” 2.34 2.01E‐02 3164 “NR4A1” −1.29 2.62E‐06 60676
   “PAPPA2” −1.04 2.55E‐07
   100528030 “POC1B‐GALNT4” 1.29 2.61E‐02 3321 “IGSF3” −1.20 4.27E‐02 6236
   “RRAD” −1.55 5.60E‐06
   100533496 “TVP23C‐CDRT4” −1.34 5.24E‐03 3371 “TNC” 1.06 7.95E‐08 6288
   “SAA1” −1.68 2.90E‐19
   100534012 “TNFAIP8L2‐SCNM1” −5.70 7.31E‐04 342035 “GLDN” −2.05 1.13E‐04
   6364 “CCL20” −1.25 3.60E‐07
   10437 “IFI30” −1.30 4.23E‐07 3429 “IFI27” −1.37 7.70E‐11 6387 “CXCL12”
   −1.32 2.22E‐02
   10561 “IFI44” −1.14 5.63E‐05 3434 “IFIT1” −1.45 2.67E‐12 63926 “ANKEF1”
   −1.20 1.92E‐02
   10630 “PDPN” −1.05 6.33E‐09 3437 “IFIT3” −1.22 7.53E‐23 6445 “SGCG”
   −2.17 4.70E‐04
   10964 “IFI44L” −2.32 2.75E‐10 3488 “IGFBP5” 1.15 1.54E‐12 6450 “SH3BGR”
   −2.51 4.52E‐05
   11095 “ADAMTS8” 2.49 6.56E‐03 3569 “IL6” −2.24 4.24E‐16 650 “BMP2”
   −1.29 3.19E‐06
   11197 “WIF1” 1.12 3.31E‐02 3576 “CXCL8” −1.00 6.20E‐29 6615 “SNAI1”
   1.09 1.41E‐09
   11341 “SCRG1” −1.55 3.71E‐02 375616 “KCP” −2.31 1.66E‐02 6648 “SOD2”
   −1.13 2.62E‐35
   113730 “KLHDC7B” −2.15 4.45E‐02 3866 “KRT15” 1.22 1.24E‐06 6696 “SPP1”
   −1.82 2.01E‐11
   114108587 “ATF7‐NPFF” 1.20 2.48E‐03 386618 “KCTD4” −2.13 6.05E‐03 684
   “BST2” −1.74 3.84E‐03
   114902 “C1QTNF5” −1.23 3.18E‐04 3880 “KRT19” 1.41 6.68E‐16 7079 “TIMP4”
   −1.53 1.12E‐09
   114907 “FBXO32” −1.29 2.75E‐10 3936 “LCP1” −1.16 1.85E‐06 718 “C3”
   −1.38 1.58E‐39
   119391 “GSTO2” −1.09 4.18E‐02 3965 “LGALS9” −1.63 1.00E‐02 72 “ACTG2”
   1.76 3.86E‐02
   125111 “GJD3” −1.84 2.51E‐03 402415 “XKRX” 1.13 1.09E‐02 723790
   “H2AC19” −22.61 4.75E‐07
   1286 “COL4A4” −1.04 1.76E‐03 4312 “MMP1” −1.67 2.76E‐140 728 “C5AR1”
   −1.18 1.83E‐03
   129607 “CMPK2” −1.44 1.47E‐03 4314 “MMP3” −2.65 4.71E‐74 79173
   “C19orf57” −4.13 4.13E‐02
   131578 “LRRC15” −1.72 2.02E‐37 4317 “MMP8” −2.02 4.26E‐02 79789 “CLMN”
   1.05 1.12E‐02
   1440 “CSF3” −1.24 2.31E‐43 4318 “MMP9” 1.00 4.96E‐07 79812 “MMRN2” 1.10
   1.18E‐03
   146850 “PIK3R6” 1.96 7.50E‐03 4324 “MMP15” −1.18 4.27E‐09 79961
   “DENND2D” −1.43 1.51E‐02
   1475 “CSTA” −1.14 1.00E‐03 440804 “RIMBP3B” −1.48 2.38E‐02 8362 “H4C12”
   4.96 1.39E‐02
   151258 “SLC38A11” 1.74 2.74E‐02 4494 “MT1F” −4.20 2.33E‐05 84419
   “C15orf48” −1.95 3.17E‐02
   1520 “CTSS” −1.25 1.72E‐12 4495 “MT1G” −4.11 9.44E‐43 84624 “FNDC1”
   −1.45 7.70E‐03
   154664 “ABCA13” −1.31 1.86E‐02 4496 “MT1H” −4.85 1.69E‐02 84707 “BEX2”
   −2.39 7.38E‐04
   165904 “XIRP1” −1.26 1.06E‐02 4599 “MX1” −1.80 2.85E‐07 8477 “GPR65”
   −1.69 3.67E‐02
   169044 “COL22A1” 2.60 3.95E‐02 4600 “MX2” −1.17 1.57E‐09 84898 “PLXDC2”
   −1.05 9.51E‐07
   183 “AGT” −4.18 2.87E‐05 4693 “NDP” 1.24 3.69E‐03 84935 “MEDAG” −1.05
   2.73E‐09
   2028 “ENPEP” 1.47 4.45E‐08 4884 “NPTX1” 1.49 1.19E‐04 84952 “CGNL1”
   −2.06 4.13E‐02
   2034 “EPAS1” −1.46 8.70E‐21 4923 “NTSR1” 1.02 2.71E‐37 8510 “MMP23B”
   −1.06 2.09E‐05
   2047 “EPHB1” −1.15 2.13E‐06 4938 “OAS1” −2.14 1.04E‐06 8519 “IFITM1”
   −1.09 7.33E‐03
   2150 “F2RL1” 1.30 4.27E‐09 493861 “EID3” −1.66 4.82E‐02 8532 “CPZ”
   −1.99 4.18E‐02
   23213 “SULF1” −1.14 4.26E‐03 4939 “OAS2” −1.08 2.04E‐14 85407 “NKD1”
   −1.03 2.47E‐06
   2357 “FPR1” −1.57 1.29E‐02 50509 “COL5A3” 1.54 8.53E‐89 8542 “APOL1”
   −1.39 4.88E‐08
   241 “ALOX5AP” −2.90 1.98E‐07 5069 “PAPPA” −1.41 6.45E‐43 8637
   “EIF4EBP3” −1.18 2.78E‐02
   2495 “FTH1” −1.25 7.38E‐95 51171 “HSD17B14” −1.34 6.07E‐06 8638 “OASL”
   −1.45 1.11E‐05
   2517 “FUCA1” −1.57 3.30E‐02 51191 “HERC5” −1.50 1.35E‐02 8714 “ABCC3”
   −1.57 1.13E‐25
   2537 “IFI6” −1.35 2.94E‐05 53831 “GPR84” −2.23 4.96E‐02 8794
   “TNFRSF10C” −1.50 1.56E‐02
   25849 “PARM1” −2.23 1.46E‐04 55008 “HERC6” −1.06 7.99E‐05 8854
   “ALDH1A2” 1.24 8.31E‐30
   259307 “IL4I1” −1.03 4.02E‐04 55024 “BANK1” −2.13 6.51E‐03 8878
   “SQSTM1” −1.12 4.97E‐49
   27074 “LAMP3” −3.36 1.46E‐10 55301 “OLAH” −2.34 3.14E‐08 8900 “CCNA1”
   1.34 3.32E‐14
   28231 “SLCO4A1” 1.05 4.49E‐03 55601 “DDX60” −1.10 5.85E‐08 8942 “KYNU”
   −1.67 4.48E‐03
   2825 “GPR1” −2.03 1.66E‐11 5646 “PRSS3” 1.24 5.68E‐13 9077 “DIRAS3”
   2.48 1.88E‐25
   283537 “SLC46A3” −1.03 7.02E‐05 57221 “ARFGEF3” −1.86 4.46E‐02 91543
   “RSAD2” −2.46 9.17E‐04
   2878 “GPX3” −1.08 5.55E‐08 5730 “PTGDS” −1.93 4.51E‐05 93145 “OLFM2”
   1.22 2.25E‐02
   29114 “TAGLN3” 1.32 8.91E‐07 5874 “RAB27B” −1.16 8.58E‐03 94240
   “EPSTI1” −1.12 2.36E‐06
   2920 “CXCL2” −1.54 4.30E‐06 59277 “NTN4” −1.09 1.11E‐03 9472 “AKAP6”
   1.34 3.76E‐03
   2977 “GUCY1A2” −1.23 5.40E‐04 59353 “TMEM35A” 1.04 3.88E‐18 9518
   “GDF15” −1.31 1.72E‐37
   29775 “CARD10” 1.10 7.79E‐16 597 “BCL2A1” 1.32 1.53E‐07 9619 “ABCG1”
   −3.72 4.30E‐03
   3082 “HGF” 1.11 2.83E‐18 5973 “RENBP” −2.48 4.30E‐06 9734 “HDAC9” −1.05
   1.38E‐14
   [83]Open in a new tab

3.3. KEGG pathway enrichment analysis and protein interaction network
analysis

   By KEGG analysis of the 153 differentially expressed genes, the
   pathways were significantly enriched at IL‐17 signaling pathway
   (enrichment number 8), rheumatoid arthritis (enrichment number 7), and
   viral protein interaction with cytokine and cytokine receptor
   (enrichment number 6) (Figure [84]1C). Two protein action networks were
   based on genes under the IL‐17 signaling pathway and viral protein
   interaction with cytokine and cytokine receptor pathways. The larger
   the nodes, the more interactive the relationship is with the other
   nodes. As shown in Figure [85]1D,E, IL‐6, CXCL8, and MMP‐9 were the
   most prominent nodes. Combined with the heatmap analysis of IL‐17
   signaling pathway differential genes drawn in Figure [86]1F, decreased
   IL‐6 expression, decreased CXCL8 expression, and increased MMP‐9
   expression suggested suppressing inflammatory factor expression,[87]
   ^35 , [88]^36 remodeling the extracellular matrix,[89] ^37 and
   promoting angiogenesis[90] ^38 after Cu‐MON treatment. A KEGG network
   map was plotted against the eight differentially expressed genes in
   Figure [91]1F, including transcriptional misregulation in cancer, TNF
   signaling pathway, amoebiasis, pathways in cancer, NOD‐like receptor
   signaling pathway, legionellosis, IL‐17 signaling pathway, viral
   protein interaction with cytokine and receptor, rheumatoid arthritis,
   and cytokine‐cytokine receptor interaction (Figure [92]1G).

3.4. GO pathway enrichment analysis and protein interaction network analysis

   By performing a GO analysis of the biological processes for the
   differentially expressed 153 genes, the top 20 significantly enriched
   biological processes included: Type I interferon signaling pathway,
   immune system process, immune response, defense response to virus,
   response to virus, negative regulation of viral genome replication,
   collagen catabolic process, innate immune response, cellular protein
   metabolic process, response to interferon‐alpha, regulation of
   neuroinflammatory response, extracellular matrix disassembly,
   regulation of ribonuclease activity, extracellular matrix organization,
   negative regulation of growth, cytokine‐mediated signaling pathway,
   response to bacterium, positive regulation of vascular endothelial
   growth factor production, activation of MAPK activity, and cell
   chemotaxis (Figure [93]2A).

FIGURE 2.

   FIGURE 2
   [94]Open in a new tab

   (A) GO analysis of differentially modulated genes classified by their
   Biological Process and arranged according to their statistical
   significance. Protein–protein interaction networks based upon Type I
   interferon signaling pathway (B) and immune system process (C).

   Two protein action networks, based on genes under the Type I interferon
   signaling pathway and immune system process pathways were mapped. The
   larger the nodes, the more interactive the relationship is with the
   other nodes. There were no prominent nodes in Figure [95]2B, and in
   Figure [96]2C, LAMP3, IFI27, and IFITM1 were all large, suggesting that
   Cu‐MON treatment enhanced the antiviral capacity.[97] ^39 , [98]^40 ,
   [99]^41 , [100]^42 , [101]^43 , [102]^44 Specifically, LAMP3 mediated
   the endosomal‐lysosomal pathway for the uptake of herpes simplex virus
   2 (HSV‐2) by human vaginal epithelial cells,[103] ^39 and the
   ATF4‐LAMP3 pathway upregulated by ORF45 promoted Caposi's
   sarcoma‐associated herpesvirus replication.[104] ^40 IFI27 was
   associated with the progression of HIV infection[105] ^41 and preterm
   respiratory syncytial virus infection.[106] ^43 IFITM1‐enhanced
   cholesterol transport to the Golgi, thereby accumulating cholesterol at
   Golgi‐derived replication sites, enabling efficient replication of the
   genome of non‐enveloped RNA viruses.[107] ^44 After Cu‐MON treatment,
   low expression levels of LAMP3, IFI27, and IFITM1 reduced viral
   replication and thereby attenuated the virulent capacity of T2DM‐ASC.

   Moreover, IFI6, LAMP3, IFI27, IFIT3, LGALS9, OAS1, HERC5, DDX60, and
   IFITM1 were all large, suggesting that Cu‐MON treatment enhanced the
   antitumor ability.[108] ^45 , [109]^46 , [110]^47 , [111]^48 , [112]^49
   , [113]^50 , [114]^51 , [115]^52 , [116]^53 Specifically,
   downregulation of IFI6 reversed oxaliplatin resistance in colorectal
   cancer cells by activating the p38/MAPK signaling induced pathway by
   ROS.[117] ^45 LAMP3 interacted with RPL21 to promote the invasion and
   metastasis of colorectal cancer by regulating focal adhesion
   formation.[118] ^46 IFIT3 accelerated the progression of head and neck
   squamous cell carcinoma by targeting PD‐L1 to activate the PI3K/AKT
   signaling pathway.[119] ^48 LGALS9 was a member of galectin family that
   regulated immune homeostasis and tumor cell survival through its
   interaction with its receptor Tim‐3.[120] ^49 Overexpression of OAS1
   led to CTL dysfunction and M2 macrophages polarization. Knockdown of
   OAS1 decreased the invasive capacity of pancreatic cancer cells and
   increased the apoptosis rate of Pancreatic cancer cells.[121] ^50
   HERC5/IFI16/p53 signaling mediated breast cancer cell proliferation and
   migration.[122] ^51 DDX60 promoted the migration and invasion of head
   and neck squamous cell cancer cells through the NF‐ κB/IFI27 signaling
   pathway.[123] ^52 After Cu‐MON treatment, low expressed IFI6, LAMP3,
   IFI27, IFIT3, LGALS9, OAS1, HERC5, DDX60, and IFITM1 inhibited cancer
   cell survival, thus attenuated the tumorigenic capacity of T2DM‐ASC.

   By performing a GO analysis of the cellular components for the
   differentially expressed 153 genes, significantly enriched cell
   fractions included: extracellular space (enrichment number 43),
   extracellular region (enrichment number 52), collagen‐containing
   extracellular matrix (enrichment number16), extracellular matrix
   (enrichment number 12), collagen trimer (enrichment number 7), tertiary
   granule lumen (enrichment number 5), endoplasmic reticulum lumen
   (enrichment number 10), and postsynaptic endocytic zone membrane
   (enrichment number 2) (Figure [124]3A). Based on genes under the
   extracellular space pathway, protein action networks were mapped. The
   larger the nodes, the more interactive the relationship is with the
   other nodes. IL‐6, CXCL8, CXCL12, SPP1, HGF, and MMP‐9 were all large
   nodes (Figure [125]3B), suggesting that Cu‐MON treatment regulated
   immune response,[126] ^35 , [127]^36 shaped extracellular matrix,[128]
   ^37 and participated in angiogenesis,[129] ^38 which was consistent
   with the results of KEGG enrichment analysis (Figure [130]1D–F). By GO
   analysis of the molecular functions for the differentially expressed
   153 genes, significantly enriched molecular functions included:
   metallopeptodase activity, metalloendopeptidase activity,
   2′‐5′‐oligoadenylatesynthetase acivity, cytokine activity, CXCR
   chemokine receptor binding, peptidase activity, growth factor activity,
   chemokine activity, fibronectin binding, endopeptidase activity, and
   zinc ion binding (Figure [131]3C). Two protein action networks, based
   on genes under the metallopeptodase activity and cytokine activity
   pathways, were mapped. The larger the nodes, the more interactive the
   relationship is with the other nodes. MMP‐9 nodes were large in
   Figure [132]3D, and IL‐6, CXCL8, CXCL12, and CCL20 were all large in
   Figure [133]3E, suggesting that Cu‐MON treatment regulated immune
   response,[134] ^35 , [135]^36 , [136]^54 shaped the extracellular
   matrix,[137] ^37 and participated in angiogenesis,[138] ^38 which was
   consistent with the results of cellular components GO enrichment
   (Figure [139]3A,B).

FIGURE 3.

   FIGURE 3
   [140]Open in a new tab

   (A) GO analysis of differentially modulated genes classified by their
   cellular components and arranged according to their statistical
   significance. (B) Protein–protein interaction networks based upon
   extracellular space. (C) GO analysis of differentially modulated genes
   classified by their molecular functions and arranged according to their
   statistical significance. Protein–protein interaction networks based
   upon metallopeptidase activity (D) and cytokine activity (E).

3.5. External database extension

   Two databases of The Cancer Genome Atlas (TCGA) and all RNA‐seq and
   ChIP‐seq sample and signature search (ARCHS4) were used to evaluate the
   gene expression of Cu‐MON‐treated T2DM‐ASC in different systems and
   cells, which helped to reveal the potential relationship between
   mesenchymal stem cells and other systems and cells in human body.
   Hematopoietic tissue was composed of network cells and network fibers,
   and the mesh was filled with various blood cells at different
   developmental stages. The reticuloendothelial system included reticular
   cells in the spleen and lymph nodes, covering endothelial cells in the
   liver, bone marrow, adrenal cortex, sinusoid space of the anterior
   pituitary, and free tissue cells in other organs.[141] ^55 , [142]^56 ,
   [143]^57 Among the 153 differentially expressed genes, 33 genes were
   middle or high expressed in hematopoietic and reticuloendothelial
   system, indicating that Cu‐MON treatment contributed to enhancing the
   ability of MSCs to repair the hematopoietic system. Among the 153
   differentially expressed genes, 71, 90, 80 genes were middle or high
   expressed in Muscular system, Immune system, Cardiovascular system
   Medium, suggesting that mesenchymal stem cells regulated muscle
   regeneration, immunity and repaired blood vessels (Table [144]5).
   Moreover, 65, 64, 67, 87, 55 genes were middle or high expressed in
   HUVEC, myoblast, smooth muscle, macrophage, and myofibroblast
   (Table [145]6), suggesting that these systems and cells share common
   provascular repair and muscle regeneration roles, consistent with the
   reported functions of mesenchymal stem cells.[146] ^3 , [147]^4

TABLE 5.

   High, middle high and middle expression genes in Hematopoietic and
   reticiloendothelial system, Muscular system, Immune system,
   Cardiovascular system.
   High Middle high Middle
   Hematopoietic and reticiloendothelial system LCP1 (1) CSTA, CTSS,
   ALOX5AP, FTH1, IFI6, HGF, CXCL8, LGALS9, BST2, C5AR1, IFITM1, CCNA1
   (12) IFI44, PIK3R6, FPR1, FUCA1, SLC46A3, CXCL2, NR4A1, IFIT3, MMP8,
   OAS1, OAS2, PTGDS, RAB27B, BCL2A1, RENBP, CXCL12, DENND2D, PLXDC2,
   APOL1, SQSTM1 (20)
   Muscular system EPAS1, SULF1, FTH1, IGFBP5, CXCL12, SOD2, SQSTM1 (7)
   FBXO32, IFI6, TNC, IFIT1, IFIT3, CXCL8, MMP1, MX1, PAPPA, PTGDS, NTN4,
   C3, ACTG2, CCNA1 (14) IFI30, IL6, IFI44, MMP3, PDPN, MX2, IFI44L, OAS1,
   ADAMTS8, OAS2, C1QTNF5, COL5A3, COL4A4, HERC5, LRRC15, HERC6, CSF3,
   DDX60, XIRP1, RAB27B, AGT, TMEM35A, CPZ, FUCA1, RRAD, APOL1, PARM1,
   SAA1, OASL, GPR1, BMP2, ABCC3, SLC46A3, BST2, KYNU, GPX3, H2AC19,
   RSAD2, CARD10, FNDC1, EPSTI1, HGF, PLXDC2, AKAP6, NR4A1, MEDAG, GDF15,
   IFI27, IFITM1, HDAC9 (50)
   Immune system IFI30, CTSS, FTH1, FUCA1, IFI6, IFIT3, CXCL8, LCP1, MMP9,
   MX1, SOD2, SPP1, C5AR1, SQSTM1, RSAD2 (15) IFI44, MT1H, PDPN, MX2,
   IFI44L, OAS1, CMPK2, OAS2, EPAS1, HERC5, FPR1, GPR84, ALOX5AP, DDX60,
   IL4I1, BCL2A1, LAMP3. RENBP, SLCO4A1, CCL20, CXCL2, BST2, TNC, C3,
   IFI27, PLXDC2, IFIT1, IFITM1, IGFBP5, APOL1, IL6, OASL, LGALS9, ABCC3,
   MMP1, TNFRSF10C, MT1F, KYNU, MT1G, EPSTI1 (40) FBXO32, CSF3, PIK3R6,
   CSTA, XIRP1, EPHB1, HERC6, SULF1, PTGDS, PARM1, RRAD, SLC46A3, SNAI1,
   GPX3, H2AC19, HGF, CLMN, NR4A1, DENND2D, IGSF3, GPR65, KCP, EIF4EBP3,
   MMP8, ALDH1A2, MMP15, CCNA1, NDP, AKAP6, COL5A3, GDF15, PAPPA, ABCG1,
   HSD17B14, HDAC9 (35)
   Cardiovascular system XIRP1, EPAS1, FTH1, GPX3, IGFBP5, PTGDS, SOD2,
   C3, SQSTM1 (9) FBXO32, AGT, SULF1, IFI6,PARM1, NR4A1, TNC, IFI27,
   MMP15, COL5A3, PAPPA, NTN4, RRAD, CXCL12, SH3BGR, SPP1, MMRN2, CGNL1,
   PLXDC2, IFITM1, AKAP6 (21) IFI30, IL6, IFI44, CXCL8, PDPN, LCP1,
   IFI44L, LGALS9, C1QTNF5, MMP1, COL4A4, MMP9, CTSS, MX1, ENPEP, MX2,
   ALOX5AP, OAS1, FUCA1, OAS2, SLCO4A1, HERC6, FNDC1, GPR1, DDX60, MEDAG,
   SLC46A3, RENBP, MMP23B, CXCL2, SGCG, APOL1, GUCY1A2, BST2, ABCC3,
   CARD10, TIMP4, ALDH1A2, HGF, ACTG2, EPSTI1, IGSF3, H2AC19, GDF15,
   IFIT1, C5AR1, ABCG1, IFIT3, CLMN, HDAC9 (50)
   [148]Open in a new tab

TABLE 6.

   High, middle high, and middle expression genes in HUVEC, myoblast,
   smooth muscle, macrophage, and myofibroblast.
   High Middle high Middle
   Myoblast FTH1, SOD2 (2) IFI44L, FBXO32, EPAS1, SULF1, IFI6, TNC, IFIT1,
   IFIT3, IGFBP5, MX1, PAPPA, PTGDS, H2AC19, SQSTM1, CCNA1 (15) IFI30,
   OAS1, IFI44, OAS2, PDPN, COL5A3, C1QTNF5, HERC5, COL4A4, HERC6, CSF3,
   DDX60, CTSS, RAB27B, XIRP1, NTN4, FPR1, RRAD, FUCA1, SAA1, PARM1,
   CXCL12, GPR1, BST2, SLC46A3, C3, GPX3, PLXDC2, KYNU, CARD10, MEDAG,
   DIRAS3, IFI27, IFITM1, RSAD2, CXCL8, CPZ, EPSTI1, MMP1, APOL1, AKAP6,
   MMP3, OASL, GDF15, MX2, ABCC3, HDAC9 (47)
   Myofibroblast EPAS1, FTH1, IGFBP5, ACTG2, SQSTM1 (5) PDPN, C1QTNF5,
   LRRC15, SULF1, HGF, TNC, CXCL8, NTN4, BMP2, SOD2, SPP1, IFITM1, CPZ,
   ABCC3, GDF15 (15) IFI30, F2RL1, FUCA1, IFI6, PARM1, SLC46A3, PAPPA,
   GPX3, HERC6, CARD10, DDX60, NR4A1, RAB27B, IFI27, CXCL12, IFIT1, SNAI1,
   IFIT3, C3, IL6, H2AC19, KRT19, FNDC1, MMP1, PLXDC2, MMP3, APOL1, MMP15,
   OLFM2, MX1, EPSTI1, MX2, AKAP6, COL5A3, HDAC9 (35)
   HUVEC EPAS1, FTH1, MMP1, SQSTM1 (4) C1QTNF5, SULF1, IFI6, GPX3, IFI27,
   CXCL8, MX1, NTN4, SOD2, MMRN2, CGNL1, GDF15 (12) TVP23C‐CDRT4, LGALS9,
   IFI44, MMP15, IFI44L, MX2, FBXO32, OAS1, CMPK2, OAS2, CTSS, PAPPA,
   F2RL1, HSD17B14, FUCA1, HERC6, IL4I1, DDX60, LAMP3, PRSS3, SLCO4A1,
   ARFGEF3, CXCL2, TMEM35A, OASL, CARD10, BMP2, TNFRSF10C, NR4A1, SNAI1,
   ALDH1A2, IFIT1, BST2, CCNA1, IFIT3, H2AC19, KYNU, IGFBP5, CLMN, RSAD2,
   IL6, MEDAG, EPSTI1, KRT19, IFITM1 ABCG1, LCP1, APOL1, HDAC9 (49)
   Smooth muscle EPAS1, SULF1, FTH1, IGFBP5, PAPPA, CXCL12, SOD2, SQSTM1
   (8) ADAMTS8, FBXO32, IFI6, GPX3, HGF, TNC, CXCL8, MMP1, PTGDS, NTN4,
   BMP2, C3, ACTG2, MEDAG, IFITM1, CPZ, APOL1, GDF15 (18) TVP23C‐CDRT4,
   MMP3, IFI30, MMP15, IFI44, MX1, PDPN, OAS2, C1QTNF5, COL5A3, COL4A4,
   HSD17B14, LRRC15, HERC6, CSF3, DDX60, AGT, TMEM35A, F2RL1, SNAI1,
   FUCA1, SPP1, GPR1, BST2, SLC46A3, H2AC19, NR4A1, FNDC1, GLDN, PLXDC2,
   IFI27, MMP23B, IFIT1, ABCC3, IFIT3, EPSTI1, IL6, AKAP6, KRT19, ABCG1,
   HDAC9 (41)
   Macrophage IFI30, CTSS, FTH1, FUCA1, IFI6, CXCL2, IFIT3, CXCL8, LCP1,
   MMP9, MX1, SOD2, SPP1, C5AR1, SQSTM1, RSAD2 (16) FBXO32, CSF3, PIK3R6,
   CSTA, XIRP1, FPR1, PTGDS, PARM1, RRAD, SLC46A3, SNAI1, GPX3, H2AC19,
   HGF, CLMN, NR4A1, DENND2D, IGSF3, GPR65, IGFBP5, IFITM1, KCP, EIF4EBP3,
   MMP8, ALDH1A2, MMP15, CCNA1, NDP, AKAP6, COL5A3, GDF15, HSD17B14,
   ABCG1, ARFGEF3, HDAC9 (35) IFI44, PDPN, IFI44L, CMPK2, EPAS1, OAS2,
   ALOX5AP, HERC5, IL4I1, GPR84, LAMP3, HERC6, SLCO4A1, DDX60, TNC,
   BCL2A1, IFI27, RENBP, IFIT1, CCL20, IL6, BST2, LGALS9, C3, MMP1,
   PLXDC2, MT1F, APOL1, MT1G, OASL, MT1H, ABCC3, MX2, KYNU, OAS1, EPSTI1
   (36)
   [149]Open in a new tab

3.6. Cytokine secretion

   Protein expression associated with angiogenesis was examined. As shown
   in Figure [150]4C, Cu‐MON treatment enhanced the expression of bFGF,
   EGF and decreased the expression of AngII and IL‐8 when compared with
   ND‐ASC. This result indicated that Cu‐MON attenuated the inflammatory
   response and activated the expression of angiogenic factors, consistent
   with the sequencing results of transcriptomics.

FIGURE 4.

   FIGURE 4
   [151]Open in a new tab

   Gene number and level of the 153 differentially expressed genes
   expressed in Hematopoietic and reticiloendothelial system, Muscular
   system, Immune system, Cardiovascular system (A) and HUVEC, myoblast,
   smooth muscle, macrophage, myofibroblast (B). (C) bFGF, EGF, AngII,
   IL‐8 expression of ND‐ASC and Cu‐MON‐treated. Analyses were performed
   using GraphPad Prism 8 software, and the significance was expressed as
   follows: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

3.7. Effect of Cu‐MON on T2DM‐ASC

   To clarify the effect of Cu‐MON on T2DM‐ASC, transcriptome sequencing
   analysis of adipose‐derived stem cells from type 2 diabetic donors
   (T2DM‐ASC) and T2DM‐ASC treated with Cu‐MON were performed. According
   to the principal component analysis map, the three samples of T2DM‐ASC
   and the three samples of Cu‐MON‐treated were far apart, indicating a
   difference between the two sample groups (Figure [152]5A). Comparing
   T2DM‐ASC and Cu‐MON‐treated by differential expression analysis, 15
   differentially expressed genes were obtained, whose expression results
   were shown in Figure [153]5B. Among them, MT1G showed the most
   significant change, with log2 (Cu‐MON‐treated/T2DM‐ASC) =5.90,
   indicating that Cu‐MON had functions of regulating heavy metal binding
   to proteins,[154] ^58 regulating cell growth,[155] ^25 free radical
   scavenging, and detoxification,[156] ^58 which was consistent with our
   previous results.[157] ^30 Furthermore, other significantly highly
   expressed genes DUSP 2 was proved to regulate cell signaling, cell
   proliferation, and differentiation,[158] ^59 ARC was proved to promote
   memory,[159] ^60 and H4C11 was proved to regulate transcription and
   expression of genes.[160] ^61 Based on the GO pathway enrichment
   analysis, the pathways regulated by Cu‐MON were mainly focused on:
   cellular response to zinc ion, detoxification of copper ion, negative
   regulation of growth, cellular response to copper ion, cellular zinc
   ion homeostasis, and cellular response to cadmium ion (Figure [161]5C).
   In addition, it also involved regulation of serine
   C‐palmitoyltransferase activity, vesicle‐mediated intercellular
   transport, negative regulation of triglyceride biosynthetic process,
   and so on.

FIGURE 5.

   FIGURE 5
   [162]Open in a new tab

   Differences in transcriptome between T2DM‐ASC and Cu‐MON‐treated. (A)
   PCA plot illustrating the variances of T2DM‐ASC and Cu‐MON‐treated. (B)
   Heatmap expression of the 15 differentially expressed genes. (C) GO
   analysis of differentially modulated genes classified by their
   Biological Process and arranged according to their statistical
   significance.

4. DISCUSSION

   Human adipose tissue contains about 2% ASC, while bone marrow‐derived
   stem cells in the bone marrow are only 0.001%–0.01%.[163] ^62 ASC have
   received widespread attention due to the advantages of rich sources,
   low immunogenicity, and minimal damage to the human body.[164] ^63
   Their therapeutic potential is attributed to their strong
   differentiation,[165] ^64 homing effect, immunomodulation, and
   paracrine functions.[166] ^9 , [167]^65 Stem cells used for therapy can
   be autologous, using the patient's own cells, or allogeneic, namely
   using cells from a healthy donor. Currently, internationally approved
   adipose stem cell products are mainly used to treat connective tissue
   diseases and Crohn's disease fistulas, as well as the treatment of
   complex anal fistulas in patients with Crohn's disease.[168] ^66
   However, for patients with underlying diseases, the cell quality of
   autologous ASC is not sufficient. For example, ASC from diabetic
   patients have decreased stem cell quality, reduced secretion function,
   and reduced metabolic capabilities.[169] ^16 , [170]^17 The cell cycle
   and immune regulation function from patients with immune deficiencies
   are also impaired.[171] ^18 , [172]^19 Therefore, using ASC from normal
   donors, performing adipose stem cell allogeneic transplantation will
   provide more possibilities for curing patients with underlying
   diseases.

   In fact, the survival time of ASC after transplantation depends on
   their microenvironment adaptation in the body. Adequate blood supply
   and oxygenation help the survival and function of stem cells.[173] ^67
   , [174]^68 The use of immunosuppressive agents can reduce rejection
   after heterogeneous stem cell transplantation and extend their survival
   time.[175] ^69 , [176]^70 However, in diseased areas, ASC may face
   ischemia, hypoxia, inflammation, and high ROS microenvironment, which
   damages their viability and reduces therapeutic effects.[177] ^71 ,
   [178]^72 Therefore, in recent years, to improve therapeutic
   effectiveness, scientists have begun to explore physical regulation or
   biomaterials to overcome the current challenges. For instance, Gwangjun
   Go et al.[179] ^73 used magnetic drivers to improve the targeted
   efficiency of ASC in tissue regeneration. Zhijie Guo et al.[180] ^64
   used radiofrequency‐based methods as a toroidal stent to guide the
   nerve differentiation of ASC as conductive stents. Xiaoyan Qu
   et al.[181] ^7 used light to activate MXENE (Ti3C2TX) nanomaterials to
   activate heat shock protein 70 (HSP70) through the ERK signaling
   pathway to regulate the osteogenesis of ASC. Yu Bin Lee et al.[182] ^74
   used continuous optical cross‐linking of oxidized methacrylic (OMA) and
   methacrylic gel (GelMA) to prepare a double‐layer hydrogel with
   controllable geometric changes to control ASC differentiation into bone
   and cartilage. These methods have expanded the prospects of asc in
   regenerative medicine applications, but they also introduced stents or
   other biomaterials, which may easily cause an immune response. When the
   immune system is excessively activated, the inflammatory reaction
   becomes strong, which may lead to tissue damage and inflammatory
   diseases.[183] ^75 In severe cases, it will damage lung tissue and
   cause acute respiratory distress syndrome (ARDS).[184] ^76 , [185]^77
   Therefore, pretreatment of ASC with physical stimulation or
   biomaterials is safer when considering transplantation.

   In our previous research, a new type of copper‐based metal–organic
   framework (Cu‐MON) was developed.[186] ^30 Unlike previous biomaterials
   used to enhance stem cell function, Cu‐MON was composed of baicalein
   and copper ions, where the valence state of the copper ion was
   adjustable, giving it REDOX properties rather than just antioxidant
   properties. In addition, Cu‐MON, microflower, was incubated with
   adipose stem cells as a component of the medium, then the medium was
   discarded and fully washed, and the modified adipose stem cells were
   obtained without biological material, which helped to reduce the immune
   response caused by cell transplantation. In previous study, after
   Cu‐MON pretreatment, stemness, and paracrine effects of T1DM‐ASC were
   enhanced, thereby effectively promoting ischemic tissue repair.[187]
   ^31 , [188]^34 However, different types of ASCs' response to Cu‐MON
   still need to be determined. To understand whether Cu‐MON effectively
   improved the quality of ASC, this study evaluated gene expression
   changes before and after Cu‐MON treatment through RNA sequencing and
   transcriptional comparative analysis. The results showed that
   Cu‐MON‐treated ND‐ASC had 153 differentially expressed genes compared
   to ND‐ASC. Based on KEGG enrichment analysis and protein interaction
   network analysis, IL‐17 signaling pathway and three hub genes
   associated with inflammation (IL‐6, CXCL8, and MMP‐9) were important
   pathway and key genes affecting stem cell capacity. In this study, the
   expression levels of IL‐6 and IL‐8 (CXCL8) were downregulated,
   suggesting that Cu‐MON inhibited the inflammatory response by
   inhibiting IL‐17 signaling pathway, which was consistent with the
   results reported in the literature[189] ^77 that IL‐17 was able to
   induce IL‐6 and IL‐8 secretion by synoviocytes and mesenchymal cells.

   Based on different gene biological processes, the GO enrichment
   analysis and protein interaction network analysis showed that the Type
   I interferon signaling pathway, immune system process, and 14 genes
   related to antiviral or immune systems (IFI6, IFIT3, IFI27, IFITM1,
   LAMP3, etc.) were important pathways and key genes affecting stem cell
   ability. Among them, IFITM1, IFI27, and LAMP3 are not only involved in
   viral replication,[190] ^39 , [191]^40 , [192]^41 , [193]^42 , [194]^43
   , [195]^44 but also in tumor invasion and metastasis.[196] ^46 ,
   [197]^47 , [198]^53 In this study, after Cu‐MON treatment, IFITM1
   decreased by 2.13 times, IFI27 decreased by 2.58 times, and LAMP3
   decreased by 10.27 times, suggesting that Cu‐MON treatment improved the
   antiviral and antitumor ability of ND‐ASC. GO enrichment analysis also
   includes cellular component and molecular functions. Analysis of
   cellular component and molecular function showed that important
   pathways affecting the ability of stem cells also included
   extracellular space, metallopeptidase activity, and cytokine activity.
   The protein interaction network results showed that the key genes
   affecting these three processes were IL‐6, CXCL 8, and MMP‐9, which was
   consistent with the results of KEGG enrichment analysis. In addition,
   the differences in gene expression were analyzed with external
   databases. Among the 153 differentially expressed genes, there were 33,
   71, 90, and 80 genes highly expressed in the hematopoietic and
   reticiloendothelial system, muscular system, immune system,
   cardiovascular system, respectively (Table [199]5), and 65, 64, 67, 87,
   and 55 genes highly expressed in HUVEC, myoblast, smooth muscle,
   macrophage, and myofibroblast, respectively (Table [200]6), indicating
   common immune regulation and repair of these systems and cells.
   Examination of cytokines secreted from ND‐ASC and Cu‐MON‐treated
   ND‐ASC, the expression of angiogenic factors (bFGF, EGF) were increased
   and the expression of inflammatory cytokines (AngII and IL‐8) were
   decreased, which was consistent with the results of the gene enrichment
   analysis described above. Cu‐MON exerted a pro‐repair effect by
   regulating immune response .

   To further clarify the immune regulation effect of Cu‐MON,
   transcriptomic sequencing analysis was performed on T2DM‐ASC and Cu‐MON
   treated T2DM‐ASC. Differentially expressed genes showed that the most
   critical regulatory gene was MT1G. MT1G had been reported to regulate
   the combination of heavy metals and protein,[201] ^58 regulation of
   cell growth,[202] ^25 and clearance of free radicals and
   detoxification,[203] ^58 consistent with our earlier findings on
   Cu‐MON's antioxidant properties,[204] ^30 promoting the proliferation
   and migration of endothelial cells. In addition, genes with
   significantly high expression of Cu‐MON treated T2DM‐ASC included
   DUSP2, ARC, and H4C11, which were reported to regulate cell signaling,
   cell proliferation, and differentiation.[205] ^59 , [206]^60 , [207]^61
   In conclusion, Cu‐MON promoted paracrine, regulated immune responses,
   as well as regulated cell signal transduction and antiviral
   capabilities, showing the potential of its own or combined ASC in
   tissue repair, antiviral, and other fields (Table [208]7).
   Interestingly, in our previous study, Cu‐MON promoted partial recovery
   of the transcriptome of T1DM‐ASC cells, making their gene expression
   levels more similar to ND‐ASC.[209] ^31 However, this phenomenon was
   not observed at the gene expression level of T2DM‐ASC, and few genes
   were altered in Cu‐MON‐treated T2DM‐ASC. This may be due to the higher
   age of type 2 diabetes compared to type 1 diabetes, and its relative
   cellular senescence is more severe. Treatment with an equivalent dose
   of Cu‐MON was not sufficient to convert T2DM‐ASC. In addition, the
   existing sample size was small, and the sample size should be expanded
   in future studies to improve the accuracy of the study. Animal models
   such as limb ischemia, tumor, or bacterial infection can be established
   for validation in vivo. Moreover, transmembrane peptides can be
   modified on the surface of Cu‐MON to make it more easily swallowed by
   aging T2DM‐ASC. The dose and time of Cu‐MON administration can also be
   adjusted to restore the transcriptome of T2DM‐ASC to ND‐ASC.

TABLE 7.

   Summary of key gene and pathway changes, application scenarios,
   shortcomings, and improvements of Cu‐MON for ND‐ASC and T2DM‐ASC.
   Cu‐MON‐treated Key pathway Key gene Application Shortcoming Outlook
   ND‐ASC Cytokine activity bFGF↑, EGF↑, CXCL12↓, AngII↓ Angiogenesis
   Small sample size Increase the sample size, establish animal models
   such as limb ischemia, tumor or bacterial infection for validation
   in vivo
   IL‐17 signaling pathway IL‐6↓, CXCL8↓, MMP‐9↑, CCL20↑ Immunomodulation
   Type I interferon signaling pathway IFI6↑, IFITM1↑, IF127↓, LAMP3↓
   Antiviral, antitumor
   T2DM‐ASC Cellular response to zinc/copper ion MT1G↑, MT1H↑
   Detoxification Adjust Cu‐MON administration dose and time, modify
   transmembrane peptides on Cu‐MON surface
   Growth regulation DUSP2↑, ARC↑, H4C11↑ Signal transduction
   [210]Open in a new tab

5. CONCLUSION

   In this study, transcriptome sequencing and gene enrichment analysis
   were performed on ND‐ASC, Cu‐MON treated ND‐ASC, T2DM‐ASC, and Cu‐MON
   treated T2DM‐ASC. KEGG and GO enrichment indicated that IL‐6, CXCL8,
   and MMP‐9 were the key genes affecting stem cell capacity. Moreover,
   Cu‐MON improved the antiviral ability and immune regulation ability of
   stem cells through IL‐17 signaling pathway and Type I interferon
   signaling pathway. This was the molecular mechanism of Cu‐MON to
   improve the function of adipose stem cells, providing valuable
   information and a theoretical basis for biomaterials to enhance cell
   quality. For clinical diabetic patients, the pretreatment of damaged
   adipose stem cells with this biomaterial can reduce the response of the
   immune system to the cell transplantation while improving the cell
   quality, helping to reduce the cost of clinical treatment and improve
   the safety of cell transplantation. In the future, the study sample
   size should be further expanded, and the corresponding animal
   experiment should be verified, so as to provide more information
   support for clinical application.

AUTHOR CONTRIBUTIONS

   Jing Yang, Ruixin Pang, and Kaijing Liu conceived the project and
   designed the experiments. Kaijing Liu synthesized the material. Ruixin
   Pang and Kaijing Liu performed most experiments and prepared the
   figures. Ruixin Pang and Kaijing Liu analyzed and discussed the
   results. Ruixin Pang and Kaijing Liu wrote the manuscript. Jing Yang
   and Biou Liu reviewed and edited the manuscript. All authors have given
   approval to the final version of the manuscript.

CONFLICT OF INTEREST STATEMENT

   The authors declare no conflicts of interest.

ACKNOWLEDGMENTS