Ex vivo conditioning of peripheral blood mononuclear cells of diabetic patients promotes vasculogenic wound healing

Abstract The quality and quantity of endothelial progenitor cells (EPCs) are impaired in patients with diabetes mellitus patients, leading to reduced tissue repair during autologous EPC therapy. This study aimed to address the limitations of the previously described serum‐free Quantity and Quality Control Culture System (QQc) using CD34+ cells by investigating the therapeutic potential of a novel mononuclear cell (MNC)‐QQ. MNCs were isolated from 50 mL of peripheral blood of patients with diabetes mellitus and healthy volunteers (n = 13 each) and subjected to QQc for 7 days in serum‐free expansion media with VEGF, Flt‐3 ligand, TPO, IL‐6, and SCF. The vascular regeneration capability of MNC‐QQ cells pre‐ or post‐QQc was evaluated with an EPC colony‐forming assay, FACS, EPC culture, tube formation assay, and quantitative real time PCR. For in vivo assessment, 1 × 104 pre‐ and post‐MNC‐QQc cells from diabetic donors were injected into a murine wound‐healing model using Balb/c nude mice. The percentage of wound closure and angio‐vasculogenesis was then assessed. This study revealed vasculogenic, anti‐inflammatory, and wound‐healing effects of MNC‐QQ therapy in both in vitro and in vivo models. This system addresses the low efficiency and efficacy of the current naïve MNC therapy for wound‐healing in diabetic patients. As this technique requires a simple blood draw, isolation, and peripheral blood MNC suspension culture for only a week, it can be used as a simple and effective outpatient‐based vascular and regenerative therapy for patients with diabetes mellitus.


| INTRODUCTION
Since the successful isolation of endothelial progenitor cells (EPCs) from peripheral blood in 1997, 1

autologous mononuclear cells (MNCs)
have been derived from bone marrow (BM) 2,3 and used for clinical vascular regenerative therapy. Autologous BM and peripheral blood MNC (PbMNC) therapy with functional EPCs are clinically effective in ischemia and have been used successfully in nondiabetic individuals. 2,4 Furthermore, autologous granulocyte-colony stimulating factor (G-CSF)mobilized peripheral blood CD34+ cell therapy is safe, feasible, and effective in diabetic patients with chronic non-healing ulcers. 5 However, clinical trials of autologous cell therapy with freshly isolated CD34+ cells have limitations in the isolation and expansion of cells.
The number and functional capacity of EPCs (CD34+ cells) are severely impaired in diabetic patients, limiting their clinical value for autologous EPC therapy. [5][6][7][8] Since vasculogenic cells, such as CD34+ or CD133+ cells, constitute a very small percentage of PbMNCs (0.01%) and BM-MNCs (0.1%), a large amount of BM aspiration or peripheral blood apheresis and injection of G-CSF is needed for cell therapy applications. 5 Additionally, PbMNCs are heterogenous and may include cells such as B cells, Killer T cells, NK cells, and M1 macrophages, which can cause inflammatory side effects of cell therapy. 9,10 Ameliorating functional deficits and increasing cell counts to a sufficient level are, thus, central to successful autologous EPC therapy in diabetic patients. 11 Therefore, isolation techniques with a simple methodology and minimal effort are needed. We identified a serum-free Quantity and Control culture method (QQc) that could restore the functionality of BMderived murine diabetic c-kit+Sca-1+lin− (KSL) cells and peripheral blood CD34+ cells of patients with diabetes. 12 In subsequent study, we observed post QQc treatment, human peripheral blood CD34+ cells demonstrated significantly high vasculogenic and wound healing potency in mouse model. 13 However, isolating clinically sufficient number of CD34 + cells from small amount of human peripheral blood is difficult as apheresis can be a large burden for the patients. Therefore, we aimed to develop a method to accumulate highly regenerative cells from a simple cell isolation technique.
A novel QQc for PbMNCs obtained from serum-free expansion media with five cytokines enhanced the vasculogenic potential of EPCs derived from healthy subjects and facilitated the preparation of PbMNCs for regenerative phenotype activation. [14][15][16] Without any requirement of isolation of CD34+ cells, this new technique is more practical than CD34-QQ cell therapy. In this study, we hypothesized that QQc would restore the functions of human PbMNCs derived from patients with diabetes and therapy with the MNC-QQ cells (monocytes, CD34+, and lymphocytes) would demonstrate vasculogenic, anti-inflammatory, and wound-healing potential in diabetic patients with ischemic disease and non-healing wounds. Leveraging the prominent clinical potential of PbMNC-QQ, this is one of the first pre-clinical studies for vascular regenerative therapy using blood from patients with diabetes.

| Participants
Patients with type 2 diabetes were selected from the outpatient department of Juntendo University Hospital, Tokyo, Japan. Nonsmoking males, between 20 and 80 years of age with HbA 1C < 8.0 g/dL were included.
Patients with severe heart failure, hemodialysis, peritoneal dialysis, infectious disease, hematologic disease, inflammatory disease, or malignant tumors were excluded. Healthy volunteers were recruited to serve as controls (Table 1).

| Preparation of PbMNCs
Peripheral blood from healthy volunteers and patients with diabetes mellitus (DM) was obtained with written informed consent under the institutional approval of the Clinical Investigation Committee of Juntendo University School of Medicine and Tokai University School of Medicine. The preparation of PbMNCs was performed as described previously. 14 Peripheral blood samples (50 mL/subject) were drawn into tubes containing EDTA-2Na (Venoject II, Terumo, Tokyo, Japan).
The whole blood was placed into 50-mL tubes and centrifuged (400g, 25 C, 10 minutes). PbMNCs were prepared from the collected buffy coat suspended with EDTA-PBS by density gradient centrifugation (400g, 25 C, 30 minutes) using Histopaque-1077 (Sigma, St. Louis, Missouri). After washing with EDTA-PBS, the cells were treated with NH 4 Cl solution (pH 7.3), washed twice with ETDA-PBS, and, finally, suspended in an appropriate medium or ETDA-PBS.

| Ex vivo expansion culture
PbMNCs prepared from healthy donors and patients with DM were processed in an ex vivo serum-free expansion culture system named QQc, as described previosuly. 17 For this, fresh PbMNCs (pre-QQc) were seeded at a density of 2 × 10 6 cells/well in six-well Primaria plates (BD Falcon, Franklin Lakes, New Jersey) with 2 mL/well of Stemline II medium (Sigma, St. Lois, Missouri) supplemented with recombinant human vascular endothelial growth factor (rhVEGF; 50 ng/mL), rh interleukin-6 (rhIL-6; 20 ng/mL), rh Fms-related tyrosine kinase-3 ligand (rhFlt-3L; 100 ng/mL), rh thrombopoietin (rhTPO 20 ng/mL), rh stem cell factor (rhSCF; 100 ng/mL) (all from PeproTech, Rocky Hill, New Jersey), and an antibiotic cocktail (Invitrogen), and cultured for 7 days at 37 C in a 5% CO 2 atmosphere. After 7 days, without subculture or re-feeding, post-QQc cells were harvested by gently pipetting and washing the wells with EDTA-PBS, and were then suspended in an appropriate medium.

| RNA Seq
MNCs were lysed in RLT solution (Qiagen, Germany) and total RNA was prepared following the manufacturer's protocol. All RNA-seq was performed using a TruSeq Stranded RNA-seq kit (Illumina) following the manufacturer's protocol. The ribosomal RNA depletion method was followed using a Ribo-Zero kit (Illumina) or NEBNext rRNA depletion kit for samples smaller than 10 ng. All libraries were applied on a HiSeq2500 sequencer for 50 paired-end reads in the high throughput mode with HiSeq Flow Cell v3. The obtained reads were aligned with the reference sequence of the human genome, GRCh37, using STAR. 18 The alignments with a mapping quality of more than 20 were counted per gene of Gencode V19 19 using feature counts. 20 Gene expression levels were quantified as a logarithm of CPM (counts per million) base 2 through RLE normalization using edgeR. 21 Hierarchical clustering of gene expression was estimated by hclust in R, with the complete linkage method based on the distance of 1-Pearson's correlation coefficient.

| Tube formation assay
The tube formation assay was performed as described previously. 17 Briefly, PbMNCs were labeled with low-density lipoprotein from incubated at 37 C in a 5% CO 2 atmosphere for 4 hours. Wells containing only HUVECs were used as controls. The wells were photographed using a phase-contrast microscope. Total DiI-Ac-LDL-labeled cells incorporated into the tubes were analyzed using fluorescence microscopy (BZ-9000; Keyence, Osaka, Japan).

| Quantitative real time PCR
Pre-and post-QQc PbMNCs were lysed in RLT solution (Qiagen, Germany) and total RNA was prepared following the manufacturer's protocol.
Complimentary DNA (cDNA) was synthesized from the total RNA using Superscript II (Invitrogen) and was used as a template for PCR. PCR was carried out using cDNA mixed with TaqMan fast universal PCR master mix and TaqMan probes on an ABI 7500 fast real-time PCR system. TaqMan Olympus, Tokyo, Japan).  the DM group, respectively. There were no significant differences in the numbers of PbMNCs between healthy and DM subjects before or after QQc ( Figure 1A). PbMNCs also showed an increase in total colony number after QQc  Figure 1E).  Information Table 1).

| Significant increases in the expression of genes related to angiogenesis, wound-healing, and anti-inflammation after QQc
RNA-seq was performed to investigate genome-wide changes in the expression of genes related to angiogenesis, wound-healing, and anti-inflammation between pre-and post-QQ cells in patients with DM. Their hierarchical clustering indicated a clear shift in the expression of genes associated with angiogenesis after QQc (Supporting Information Figure S2). RT-PCR was carried out to confirm the mRNA levels of angiogenesis-related genes (Supporting Information   Table S2). Prior to QQc, the expression of angiogenesis-related genes, such as VEGF-A, VEGF-B, Ang-1, Ang-2, PGDF, and HGF ( Figure 3A), as well as wound-healing genes, such as FGF2 and IGF-1 ( Figure 3B), were lower in DM PbMNCs compared to those in healthy controls.
TGF-β and MMP-9 mRNA expression levels were significantly upregulated in both euglycemic and diabetic wounds in post-QQ healthy and diabetic PbMNCs compared to pre-QQc PbMNCs ( Figure 5C).

| MNC-QQc cells increase M2 wound macrophages
We previously reported that diabetic wounds are glucose-induced inflammatory environments with a high polarization of M1 macrophages but low in M2 macrophages. 23  Although the function of healthy PbMNCs was more pronounced in F I G U R E 3 Levels of mRNA and proteins related to angiogenesis, wound-healing, and anti-inflammation after Quantity and Quality Control Culture System (QQc) treatment. Levels of mRNA related to, A, angiogenesis, B, anti-inflammation, and C, wound-healing measured by quantitative PCR. Levels of proteins related to, D, angiogenesis, E, macrophages, and F, wound-healing measured by cytokine multiplex analysis. *P < .05, **P < .005. DM, diabetic; MNC-QQc, mononuclear cell treated with quality-quantity culture; Sl2 + 5G, Stemline II Hematopoietic Stem Cell +5 cytokines euglycemic wounds compared to diabetic wounds, post-QQc PbMNCs of both diabetic and healthy groups showed significantly higher in vivo angiogenic potential by the increase of vascular density in both wounds ( Figure 6A).

| DISCUSSION
Emerging evidence suggests that EPC dysfunction in diabetes leads to poor vasculogenesis and impaired wound regeneration. [24][25][26] In this study, we hypothesized that a serum-free QQc culture system We developed a serum-free QQc culture system using PbMNCs that can significantly increase the number of total and differentiated colony-forming EPCs conditioned for anti-inflammatory and regenerative phenotypes. 14 The efficacy of QQc observed in the PbMNCs of diabetic patients was unprecedented. RNA-sequencing of DM QQ-PbMNCs was performed to identify whether the gene expression changes were similar to healthy QQ-PbMNCs prior to investigation with DM QQ-PbMNCs. Our results confirmed that QQc drastically changed the PbMNC phenotype to pro-angiogenic and anti-inflammatory. 14 This study is the first, to our knowledge, to show that the ex vivo culture of PbMNCs from patients with diabetes develops proangiogenic, anti-inflammatory, and regenerative functions.
We previously demonstrated that human diabetic CD34+ cells could be expanded ex vivo to restore their impaired functions post-QQc, as denoted by a markedly higher potential for wound angiogenesis and vasculogenesis compared to pre-QQc healthy or pre-QQ DM CD34+ cells. 13 However, CD34+ cells constitute only 0.01% of peripheral MNCs and are very difficult to isolate. Therefore, for the clinical application of CD34+ cell QQ therapy, the patient must undergo apheresis for PbMNC isolation and magnetic bead-mediated CD34+ cell isolation, which is physically strenuous. An ideal regenerative therapy would be one where only a small amount of blood needs to be drawn and the cells can be processed ex vivo to generate a population of regenerative cells. There were no significant differences between QQc cells of healthy donors and diabetic patients, demonstrating that the PbMNCs of diabetic patients possess similar potential to healthy MNC-QQc, but with comparatively higher angiogenic properties. Furthermore, the expression levels of pro-angiogenic and wound-healing genes in both healthy and diabetic PbMNCs were It is currently undergoing clinical trials (unpublished). The present study presents the first pre-clinical trial using blood from diabetic patients. The benefits of the therapy include the ease of blood collection by a simple venipuncture sampling procedure and the simplicity of the isolation process. Limitations of the study, which should be addressed in future work, include small sample sizes, the use of animal models, which have limited clinical relevance to human disease settings, and a limited generalizability of results owing to the narrow patient selection criteria.

| CONCLUSION
This study demonstrates the therapeutic potential of a novel serumfree QQc culture system using MNCs derived from patients with diabetes. Vasculogenic, anti-inflammatory, and wound-healing effects were demonstrated in both in vitro and in vivo models. This system addresses the insufficient efficiency and efficacy of the current naïve MNC therapy for wound-healing in patients with diabetes. Through this technique, we expect to establish a simple, safe, and effective outpatient-based vascular and regenerative therapy for diabetic patients.