Generation of Hematopoietic-Like Stem Cells from Adult Human Peripheral Blood Following Treatment with Platelet-Derived Mitochondria

Adult stem cells represent a potential source for cellular therapy to treat serious human diseases. We characterized the insulin-producing cells from adult peripheral blood (designated PB-IPC), which displayed a unique phenotype. Mitochondria are normally located in the cellular cytoplasm, where they generate ATP to power the cell’s functions. Ex vivo and in vivo functional studies established that treatment with platelet-derived mitochondria can reprogram the transformation of adult PB-IPC into functional CD34+ hematopoietic stem cells (HSC)-like cells, leading to the production of blood cells such as T cells, B cells, monocytes/macrophages, granulocytes, red blood cells, and megakaryocytes (MKs)/platelets. These findings revealed a novel function of mitochondria in directly contributing to cellular reprogramming, thus overcoming the limitations and safety concerns of using conventional technologies to reprogram embryonic stem (ES) and induced pluripotent stem (iPS) cells in regenerative medicine.


Introduction
Stem-cell research has the potential to revolutionize treatments for certain life-changing injuries and devastating human diseases such as diabetes and Alzheimer's disease. To date, researchers have characterized multiple types of human stem cells with varying levels of potential for regeneration. Animal studies and human trials have demonstrated stem cells' translational capability to treat human diseases. For over three decades now, the most common stem-cell therapy approved by the FDA has been hematopoietic cell transplantation (HCT) (also termed hematopoietic stem cell transplantation, HSCT) for the treatment of bone marrow failure, malignant blood disorders, post-chemotherapy and/or -radiation cell regeneration, genetically based blood disorders, and autoimmune diseases [1][2][3]. However, several major limitations have restricted the broad clinical application of allogeneic HCT, including the difficulty in identifying a fully human leukocyte antigen (HLA)-matched or haploidentical donor, the scarcity of CD34 + hematopoietic stem cells (HSCs) amongst all sources of harvested cells (≤1%) [4], and, in particular, the incidence of graft-versus-host disease (GVHD), opportunistic infections, relapse of primary disease, and toxicities associated with immunosuppressive drugs and radiation. An autologous source of HSC would address the problems of matching and GVHD, but engraftment could still be hampered by the limited number of CD34 + HSCs. Since the success rate of engraftment for clinical HCT is correlated with the number of functional CD34 + hematopoietic progenitor cells (HPCs) and HSCs in the transplant [5,6], researchers have evaluated whether embryonic stem (ES) cells and induced pluripotent stem (iPS) cells can be manipulated to produce HSCs through reprogramming by small molecules or by viral transduction of transcription factors [7][8][9][10][11][12]. Thus far, these approaches have been limited by an inability to generate true functional HSCs in sufficient numbers for therapeutic use, as well as safety and ethical concerns and potential immune rejection issues to ES or iPS derivatives [5,13]. Alternative approaches are needed to circumvent these limitations.
We previously characterized a new type of multipotent stem cell from human cord blood (designated cord-blood-derived multipotent stem cells, CB-SCs) [14] that is distinguished from other known stem-cell types, including HSCs and mesenchymal stem cells (MSCs). Based on the comprehensive immune modulation characteristics of CB-SCs [15], we developed the stem cell educator (SCE) therapy, which consists of isolating the patient's blood cells through a blood cell separator, co-culturing the patient's immune cells with adherent CB-SCs in vitro, and returning the "educated" immune cells to the patient's circulation [15][16][17]. Our clinical trials indicated that SCE therapy reverses multiple immune dysfunctions and corrects autoimmune memory [17], promotes regeneration of islet β cells, improves metabolic control for the treatment of type 1 diabetes (T1D) [15,17,18] and type 2 diabetes (T2D) [16,18], and can be used to treat other autoimmune diseases such as for hair regrowth in alopecia areata (AA) [19]. Using a similar approach previously utilized for an isolation of CB-SC [15][16][17][18][19], we characterized adult peripheral blood insulin-producing cells (PB-IPC) [20] from adult peripheral blood by virtue of their ability to adhere to hydrophobic surfaces with positive charge. While investigating the mechanisms underlying the effects of SCE therapy in ex vivo studies [18], we observed the migration of platelet-released mitochondria to pancreatic islets, where they can enhance islet-β-cell function and C-peptide (an insulin byproduct) production, supporting clinical observations of the long-lasting improved health status in subjects with T1D and T2D after receiving SCE therapy [18]. Subsequently, we sought to determine whether platelet-derived mitochondria could directly influence the functions or reprogram the differentiation of adult PB-IPCs. Notably, our recent ex vivo studies established that mitochondrion-induced PB-IPCs (miPB-IPCs) can give rise to retinal pigment epithelium (RPE) cells and neuronal cells in the presence of different inducers [21]. Here, we found that platelet-derived mitochondria may be used to transform PB-IPCs into CD34 + -hematopoietic-stem-cell-like cells, giving rise to multiple blood cell lineages.
Next, we induced miCD34 + HSCs to granulocyte colony-stimulating factor (G-CSF). After 3 days, 81.14% ± 3.7% of treated cells displayed the granulocyte-specific marker CD66b, with a reduced nuclear-cytoplasmic ratio and multi-lobed nuclei shown by Wright-Giemsa staining ( Figure 3B). However, the untreated miCD34 + HSCs failed to express CD66b and displayed large nuclei, with a large nuclear-cytoplasmic ratio ( Figure 3B, n = 4). Moreover, after being treated with erythropoietin (EPO) for 5 days, miCD34 + HSCs turned into nucleated cells strongly positive for the erythroid (Er) lineage marker CD235a and facilitated RBC maturation via expulsion of their nuclei with additional EPO treatment, exhibiting a distinctive biconcave shape and enucleated RBCs ( Figure 3C, right). In total, 41.4% ± 11.46% of cells were terminally differentiated into enucleated CD235a + CD45 − hemoglobin + RBCs ( Figure 3D, n = 4). However, untreated cells failed to show these differentiations, or only expressed background levels of these markers. Further flow cytometry analysis demonstrated that the level of hemoglobin expression was increased in the matured RBCs, with the mean fluorescence intensity of hemoglobin + CD45 − mature RBCs at 13.61 ± 4.29, while hemoglobin + CD45 + immature RBCs was 8.29 ± 1.61 (p = 0.044, n = 4).
Additionally, we examined the commitment of miCD34 + HSCs to MKs and platelets, which are critical for blood clotting. After treatment with FLT-3 ligand and thrombopoietin (TPO) for 7 days, production of CD42 + MKs was achieved with typical polyploidization (mostly from 2N to 7N) ( Figure

Ex Vivo Differentiation of miCD34 + HSCs into Other Hematopoietic Lineages
To further examine their potential for differentiation, we treated miCD34 + HSCs with macrophage colony-stimulating factor (M-CSF) for 3 days, whereupon they became adherent, well-

Discussion
The use of autologous stem cells for regenerative medicine is more ethically acceptable and likely to be more successful than the use of other stem cells, and such therapies would avoid many of the immune rejection and safety concerns associated with other stem cells (e.g., ES-or iPS-based therapies). PB-IPCs can be easily isolated from peripheral blood and expanded in serum-free culture

Discussion
The use of autologous stem cells for regenerative medicine is more ethically acceptable and likely to be more successful than the use of other stem cells, and such therapies would avoid many of the immune rejection and safety concerns associated with other stem cells (e.g., ES-or iPS-based therapies). PB-IPCs can be easily isolated from peripheral blood and expanded in serum-free culture medium to avoid the painful and invasive procedures required to withdraw bone marrow. Using autologous PB-IPCs from patients as a starting material, mitochondrial treatment can generate functional autologous miCD34 + HSCs on a large scale, giving rise to different blood cell lineages such as T cells, MΦ, granulocytes, erythrocytes, and MKs/platelets. Current data indicated that miCD34 + HSCs exhibited a rapid multiple potential for differentiations post-engraftment into irradiated NSG mice. Thus, these cells offer great promise as a solution for the current bottlenecks associated with conventional stem-cell transplants and have tremendous potential for patient benefit in the clinic. Generation of functional autologous miCD34 + HSCs from PB-IPCs may address an unmet medical need.
Both ex vivo and in vivo studies demonstrated the T-cell differentiation of miCD34 + HSCs at high efficiency. To exclude the possibility of the expansion of contaminated cells, we carefully designed experiments and analyzed the differentiated cells through morphology and flow cytometry with multiple different T-cell specific markers such as CD3, CD4, CD8, TCRα/β, TCRγ/δ, and Th1 and Th2 cytokines. There were a few CD8 + T cells adhered to PB-IPCs after overnight (12 h) culture in non-tissue-culture-treated Petri dishes, but no CD4 + T cells. After the expansion of PB-IPCs for 7 days, PB-IPCs were attached to the bottom of the Petri dish with high purity (>97%) and remained negative for CD4, with a few CD8 + T cells (~2.63%). Before the treatment with platelet-derived mitochondria, all floating cells and cellular debris were washed away twice with PBS. The purified miCD34 + HSCs were utilized for transplantation. Thus, the possibility of transplanting mature PBMCs and/or lymphocytes was very limited in our protocol. Normally, NSG mice fail to support human T-cell growth and differentiation. A few CD8 + T cells might be temporarily expanded after transplant in mice; however, flow cytometry demonstrated that most of the T cells were CD3 + CD4 + T cells in the miCD34 + -HSC-transplanted NSG mice. Therefore, the engrafted CD3 + CD4 + T cells were differentiated from miCD34 + HSC. Additionally, our ex vivo studies demonstrated the high efficiency of CD3 + CD4 + CD8 − CD38 + T-cell differentiation from miCD34 + HSCs, with an expression of Th1and Th2-associated cytokines (IL-4, IL-5, and IL-12). To confirm that the CD4 expression was on T cells, not on monocytes/macrophages, the T-cell-specific marker CD3 was utilized for cellular gating during the flow cytometry. Therefore, the CD4 + T cells were derived from the differentiation of miCD34 + HSCs.
Notch receptors act as key regulators in both CD34 + HSC and T-cell development and maturation at different stages [27,28]. After binding to its specific ligands, the Notch intracellular domain (NICD) is cleaved off by the enzyme γ-secretase and translocated to the nucleus, where it binds to the transcription factor recombination signal sequence-binding protein Jkappa (RBP-J), leading to gene regulation. To understand the mechanism underlying miCD34 + HSC and T-cell differentiations, the current study revealed the upregulation of Notch receptors 1-4 on miPB-IPC, while platelet-derived mitochondria expressed Notch ligands Jagged 1 (JAG1), JAG3 and Delta-like 3 (DLL3), but not DLL1 and DLL4, highlighting that the Notch signaling pathway may contribute to the miCD34 + HSC and T-cell differentiations after treatment with platelet-derived mitochondria. Notably, the percentage of miCD34 + HSCs was significantly upregulated after mitochondrial treatment with addition of the γ-secretase inhibitor DAPT. The data confirmed that the canonical Notch pathway contributed to the differentiation of miCD34 + HSCs through the interaction of Notch receptors on PB-IPCs and their ligands on platelet-derived mitochondria. Due to the action of the Notch pathway in the maintenance of quiescent feature of hematopoietic stem and progenitor cells [30], blocking with DAPT promoted the miCD34 + HSC differentiation of PB-IPCs, which was consistent with a previous report [30]. Additional detailed molecular mechanisms need to be explored to better understand the individual and synergistic effects during the interaction of mitochondrial Notch ligands (JAG1, JAG3, and DLL3) with Notch receptors 1-4 on PB-IPCs.
The roadmap of cell differentiation and maturation is tightly modulated through the systematic integration of distinct activating or repressing signaling pathways located both in the cytoplasm and nucleus, specific to the hematopoietic system, underlying the hierarchical differentiation model [22,26,31]. Our previous work [18] demonstrated that highly purified adult peripheral blood (PB)-derived platelets (>99% purity) strongly displayed ES cell-associated pluripotent gene markers such as transcription factors OCT3/4 and SOX2, with little or no expression of NANOG. Real-time PCR array revealed the expressions of human-stem-cell-related transcription factors and human-stem-cell-associated markers in the mitochondria of human PB-platelets [18], highlighting that the stemness markers are localized in platelets' mitochondria, which may contribute to the induction of multipotency of PB-IPCs after treatment with PB-derived mitochondria. Our recent mechanistic studies confirmed that mitochondria enter cells and directly penetrate the nuclei of PB-IPCs after treatment with platelet-derived mitochondria, where they can produce profound epigenetic changes [21]. Due to the essential role of mitochondria in the reprogramming of somatic cells to iPS cells [32], additional molecular mechanisms underlying the mitochondrial reprogramming of PB-IPCs need to be determined. In conclusion, innovative reprogramming of adult PB-IPCs by treatment with platelet-derived mitochondria may overcome the limitations and safety concerns associated with using conventional transgenic technologies to reprogram ES and iPS cells in the clinical setting.

PB-IPC Cell Culture
Human buffy coat blood units (N = 51; mean age of 48.97 ± 14.11; age range from 18 to 72 years old; 24 males and 27 females) were purchased from the New York Blood Center (New York, NY, USA, http://nybloodcenter.org/). Human buffy coats were initially added to 40 mL chemical-defined serum-free culture X-VIVO 15 TM medium (Lonza, Walkersville, MD, USA) and mixed thoroughly with a 10 mL pipette, and then used for isolation of peripheral-blood-derived mononuclear cells (PBMCs). PBMCs were harvested as previously described [33]. Briefly, mononuclear cells were isolated from buffy coat blood using Ficoll-Paque TM PLUS (γ = 1.007, GE Healthcare, Chicago, IL, USA), followed by removing the red blood cells using red blood cell lysis buffer (eBioscience, San Diego, CA, USA). After three washes with saline, the whole PBMCs were seeded in 150 × 15 mm Petri dishes (BD Falcon, NC, USA) at 1 × 10 6 cells/mL, 25 mL/dish in chemical-defined serum-free culture X-VIVO 15 TM medium (Lonza, Walkersville, MD, USA) without any other added growth factors, and incubated at 37 • C in 8% CO 2 [34]. Seven days later, PB-IPCs were growing and expanded by adhering to the hydrophobic bottom of Petri dishes. Subsequently, PB-IPCs were washed three times with saline and all floating cells were removed. Next, the serum-free NutriStem ® hPSC XF culture medium (Corning) was added to continue cell culture and expansion at 37 • C in 8% CO 2 . The expanded PB-IPCs were usually applied for experiments within 7-14 days.

Isolation of Mitochondria from Platelets
The mitochondria were isolated from PB-platelets using the Mitochondria Isolation kit (Thermo Scientific, Rockford, IL, USA, Prod: 89874) according to the manufacturer's recommended protocol [18]. Adult human platelet units (N = 16; mean age of 30.81 ± 8.64; age range from 16 to 40 years old; 9 males and 7 females) were purchased from the New York Blood Center (New York, NY, USA, http://nybloodcenter.org/). The concentration of mitochondria was determined by the measurement of protein concentration using a NanoDrop 2000 Spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA). The isolated mitochondria were aliquoted and kept in a −80 • C freezer for experiments.
For mitochondrial staining with fluorescent dyes, mitochondria were labeled with MitoTracker Deep Red FM (100 nM) (Thermo Fisher Scientific, Waltham, MA, USA) at 37 • C for 15 min according to the manufacturer's recommended protocol, followed by two washes with PBS at 3000 rpm × 15 min [18].

In Vitro Differentiation of PB-IPCs into miCD34 + HSCs
PB-IPCs were treated with 100 µg/mL platelet-derived mitochondria for 7-14 days in the non-treated 24 well plates or Petri dishes with the serum-free NutriStem ® hPSC XF culture medium (Corning, New York, NY, USA), at 37 • C and 8% CO 2 . According to our current protocol, the miCD34 + HSCs were purified from mitochondria-treated PB-IPCs by immunomagnetic sorting with Miltenyi Biotech CD34 MicroBead Kit (Miltenyi Biotech, Gladbach, Germany, catalog #130-097-047) according to the manufacturer's instructions. The differentiation of miCD34 + HSCs was characterized by flow cytometry.

Flow Cytometry
Flow cytometric analyses of surface and intracellular markers were performed as previously described [18].
To test T-cell differentiation, the purified miCD34 + HSCs (1 × 10 5 cells/mL) were planted in 24 well non-treated plates in the presence of HSC-Brew GMP Basal Medium (Miltenyi Biotec, Gladbach, Germany) with addition of cytokines 25 ng/mL hFlt3L and 25 ng/mL rhIL -7 (R&D Systems, Minneapolis, MN, USA), at 37 • C in 5% CO 2 . After the treatment for 3-7 days, cells were photographed and analyzed by confocal microscopy and flow cytometry using different T-cell markers such as CD3, CD4, CD8, TCR α/β, CD38, Th1 cytokines (IL-4 and IL-5) and Th2 cytokines (IFN-γ and IL-12). Untreated miCD34 + HSCs served as negative controls. T cells from healthy donors served as positive controls. For combined immunocytochemistry, the differentiated cells were fixed in 24 well plates with 4% paraformaldehyde for 20 min and permeabilized with 0.5% triton X-100 (Sigma, Saint Louis, MO, USA) for 5 min, blocking non-specific binding with 2.5% horse serum, and followed by immunostaining with FITC-conjugated mouse anti-human CD4 and CD8 (Beckman Coulter, Brea, CA, USA). After covering with mounting medium with DAPI (Vector Laboratories, Burlingame, CA, USA), cells were photographed with a Nikon A1R confocal microscope on a Nikon Eclipse Ti2 inverted base, using NIS Elements Version 4.60 software.
For the differentiation of miCD34 + HSCs to macrophages, the purified miCD34 + HSCs (1 × 10 5 cells/mL) were treated with 50 ng/mL M-CSF (Sigma, St. Louis, MO, USA) in 24 well non-treated plates in the presence of HSC-Brew GMP Basal Medium, at 37 • C in 5% CO 2 . After treatment for 2-3 days, cells were analyzed with phagocytosis and by flow cytometry with macrophage marker CD11b (Beckman Coulter, Brea, CA, USA) and CD209 (BD Biosciences, San Jose, CA, USA). Untreated miCD34 + HSCs served as negative controls. To detect the function of differentiated macrophages, fluorescence latex beads (Sigma, Saint Louis, MO, USA) were added to M-CSF-treated and untreated miCD34 + HSC cultures. After 4 h of incubation with latex beads, cells were washed three times with PBS. The phagocytosis was viewed and evaluated under microscopy. The positive cells had a minimum of five beads per cell.
To differentiate the miCD34 + HSCs into granulocytes, the purified miCD34 + HSCs (1 × 10 5 cells/mL) were treated with 25 ng/mL hFlt3L + 100 ng/mL G-CSF (R&D Systems) in the presence of HSC-Brew GMP Basal Medium, in the 24 well non-treated plates, at 37 • C in 5% CO 2 . After the treatment for 3-5 days, cells were photographed and analyzed by flow cytometry with granulocyte marker CD66b and staining with Wright-Giemsa (Sigma, Saint Louis, MO, USA) according to the manufacturer's instructions. Untreated miCD34 + HSCs served as negative controls. PBMCs from healthy donors served as positive controls.
To differentiate miCD34 + HSCs into RBCs, purified miCD34 + HSCs (1 × 10 5 cells/mL) were initially treated with 25 ng/mL hFlt3L + 3 units/mL EPO (R&D Systems, Minneapolis, MN, USA) in the presence of HSC-Brew GMP Basal Medium, in the 24 well non-treated plates, at 37 • C in 5% CO 2 . After this treatment for 5 days, cells were re-treated with 3 units/mL EPO for an additional 3-7 days. Subsequently, cells were photographed and analyzed by flow cytometry with erythrocyte markers CD235a and hemoglobin. Untreated miCD34 + HSCs served as negative controls. For intracellular flow cytometry, all floating cells were collected and centrifuged at 2700× g 15 min. First, after blocking non-specific binding with Fc Blocker (BD Biosciences, San Jose, CA, USA), cells were fixed and permeabilized using the PerFix-nc kit (Beckman Coulter, Brea, CA, USA) according to the manufacturer's recommended protocol. Second, cells were incubated with rabbit anti-human hemoglobinβ/γ/δ polyclonal antibody (Santa Cruze, Dallas, TX, USA) at 1:100 dilution, room temperature for 30 min, and then washed with PBS at 2700× g 15 min. Next, cells were labeled with Cy5-conjugated AffiniPure donkey anti-rabbit 2nd Ab (Jackson ImmunoResearch Laboratories, West Grove, PA, USA), in combination with staining with mouse anti-human CD235a-FITC (Beckman Coulter, Brea, CA, USA) and CD45-PE-CY7 mAbs for 30 min, and followed by flow cytometry analysis.
To differentiate miCD34 + HSCs into megakaryocytes and platelets, the purified miCD34 + HSCs (1 × 10 5 cells/mL) were initially treated with 25 ng/mL hFlt3L + 100 ng/mL TPO (R&D Systems, Minneapolis, MN, USA) in the presence of HSC-Brew GMP Basal Medium, in the 24 well non-treated plates, at 37 • C in 5% CO 2 . After this treatment for 3-7 days, cells were photographed and collected for flow cytometry with MK/platelet marker CD42a (Beckman Coulter, Brea, CA, USA). Untreated miCD34 + HSCs served as negative controls. For the analysis of polyploidization, viable TPO-treated miCD34 + cells were first stained with CD42a mAb and Hoechst 33342 (Sigma, Saint Louis, MO, USA) and photographed under a confocal microscope. Secondly, using healthy donor-derived matured T cells (1N) and platelets (0N) as controls, the polyploidy of differentiated CD42a + MKs was analyzed by flow cytometry after staining with propidium iodide (PI) (Abcam, Cambridge, MA, USA) according to the manufacturer's recommended protocol.
To morphologically determine the differentiation of miCD34 + HSCs to granulocytes, RBCs, and megakaryocytes/platelets, Wright-Giemsa staining was performed on the treated and untreated cells, which were then observed and photographed under an inverted Nikon ECLIPSE Ti2 microscope.
For the DAPT-blocking experiment, PB-IPCs were treated with 100 µg/mL mitochondria + 10 µM DAPT (Sigma, Saint Louis, MO, USA, Catalog# D5942) for 7-14 days in the non-tissue-culture-treated 24 well plates or Petri dishes with the serum-free NutriStem ® hPSC XF culture medium (Corning, New York, NY, USA), at 37 • C and 8% CO 2 . Subsequently, both treated and untreated PB-IPC were collected to examine the expression of CD34 by flow cytometry.

Statistics
Statistical analyses were performed with GraphPad Prism 8 (version 8.0.1) software. The normality test of samples was evaluated using the Shapiro-Wilk test. Statistical analyses of data were performed using the two-tailed paired Student's t-test to determine statistical significance between untreated and treated groups. The Mann-Whitney U test was utilized for non-parametric data. Values are given as mean ± SD (standard deviation). Statistical significance was defined as p < 0.05, with two sided.