Human primary CD34⁺ cells transplantation for critical limb ischemia

Abstract

Background

The goal of this study was to characterize the properties of human CD34⁺ cells in culture and investigate the feasibility and efficacy of CD34⁺ transplantation in a mouse model of limb ischemia and in patients with no‐option critical limb ischemia.

Methods

Human CD34⁺ cells isolated from peripheral blood and grown in culture for up to four passages stained positively for the surface markers CD34 and CD133 and showed high viability after cryopreservation and recovery. Seven days after surgery to induce limb ischemia, ischemic muscles of nude mice were injected with CD34⁺ cells. Two weeks later, mice were scored for extent of ischemic injury, and muscle tissue was collected for immunohistochemical analysis of vascular endothelial cells and RT‐PCR analysis of cytokine expression.

Results

Injury scores of CD34⁺‐treated, but not control, mice were significantly different before and after transplantation. Vascular density and expression of VEGF and bFGF mRNAs were also significantly increased in the treated mice. Patients with severe lower extremity arterial ischemia were injected with their own CD34⁺ cells in the affected calf, foot, or toe. Significant improvements were observed in peak pain‐free walking time, ankle‐brachial index, and transcutaneous partial oxygen pressure. These findings demonstrate that growth of human CD34⁺ cells in vitro and cryopreservations are feasible.

Conclusion

Such cells may provide a renewable source of stem cells for transplantation, which appears to be a feasible, safe, and effective treatment for patients with critical limb ischemia.

1. INTRODUCTION

Efficient selection of hematopoietic stem and progenitor cells is important for both experimental investigation and the clinical strategies of stem cell transplantation and gene therapy. Over the past decade, the use of CD34 as a progenitor surface marker has resulted in rapid development of more sophisticated purification techniques and clinical trials in the field of hematopoiesis.1, 2

The CD34 antigen, first described by Civin et al in 1984,3 is a cluster of differentiation (CD) molecule present on the surface of certain cells, including hematopoietic stem cells, endothelial progenitor cells (EPCs), and vascular endothelial cells (ECs)4 in humans. CD34 is a single‐pass membrane glycoprotein and functions as a cell‐cell adhesion factor. CD34⁺ cells represent approximately 1.15% (1%‐3%) of bone marrow mononuclear cells (BM‐MNCs), 0.12%‐0.15% of cells in umbilical cord blood, and 0.1% of peripheral blood cells.5 Recently, several companies have developed equipment for CD34⁺ cell purification using antibody‐coated magnetic beads, avidin‐coated beads, or antibody‐bound beads. Reinfusion of purified CD34⁺ cells after high‐dose chemotherapy has shown normal and durable engraftment, similar to total peripheral blood transfusion or bone marrow transplant.6, 7, 8, 9

Endothelial progenitor cells differentiate into endothelial cells in a process referred to as vasculogenesis.10 Vasculogenesis describes the development of new blood vessels from in situ differentiating endothelial cells,11 whereas angiogenesis is the process of remodeling and expanding this network. Until recently, vasculogenesis was considered to be restricted to embryogenesis; however, there now exists striking evidence that EPCs also circulate in adult peripheral blood and participate in ongoing neovascularization.12 Different cytokines and growth factors have a stimulatory effect on these bone‐marrow‐derived EPCs. Granulocyte macrophage colony‐stimulating factor (GM‐CSF) and vascular endothelial growth factor (VEGF) mobilize EPCs from the bone marrow into the peripheral circulation.13

Although their contribution to postnatal neovascularization needs to be documented, the expansion and mobilization of EPCs might represent an effective means of augmenting the resident population of ECs.14 This kind of cell therapy for tissue regeneration in ischemic cardiovascular diseases opens a novel and challenging clinical option in lieu of or in addition to the use of growth factors in gene therapy.15

Previously, most clinical studies used bone BM‐MNCs,16, 17, 18, 19 peripheral blood mononuclear cells (PBMCs),20 genes, and cytokines.21, 22, 23, 24 The present pilot study was initiated to evaluate the feasibility, safety, and efficacy of transplantation of fresh CD34⁺ cells, which were considered the predominant functional cells for vasculogenesis 25, 26, 27 that had been immunomagnetically isolated from PBMCs.28

2. MATERIALS AND METHODS

2.1. Experimental animals

In all, 36 male BALB/C athymic nude mice, 8‐week old and weighing 22.5‐27.5 g, were provided and housed by the Experimental Animal Center of Zhongshan Hospital, Fudan University. This study was approved by the Animal Care and Use Committee of Fudan University.

2.2. Patient demographics

Patients suffering from severe lower extremity arterial ischemia were enrolled in the study. All patients provided written informed consent. From May 2009 to July 2011, 25 patients were enrolled, 24 males and 1 female, with a mean age of 44 ± 12 years (range, 23‐75 years), and technique was successful in all of them. Demographic details are presented in Table 1. Of these, 18 showed thromboangiitis obliterans (TAO), 3 showed arteriosclerosis obliterans (ASO), and 4 had arteritis caused by systemic lupus erythematosus (SLE), Crohn's disease, or erythema nodosum. In all, 25 lower and three upper extremities were treated.

Table 1.

Injury scores before and after transplantation

	CD34 group	X‐VIVO group	Sham group
Before	1.18 ± 1.08	1.08 ± 1.17	0.09 ± 0.30
2 wk after	0.86 ± 1.46	1.00 ± 1.31	0.00 ± 0.00

2.3. Mobilization of CD34⁺ cells into the peripheral circulation

To mobilize movement of CD34⁺ cells from the bone marrow into the peripheral circulation, patients were injected subcutaneously with 300 μg granulocyte colony‐stimulating factor(G‐CSF) filgrastim) in the morning and 150 μg in the afternoon, for a total daily dose of 5‐10 μg/kg, for 5 days. White blood cell test was carried out once daily. Low molecular weight heparin (Clexane, 4000 IU) was injected subcutaneously once daily. Cilostazol (50 mg) was taken orally twice daily. Aspirin (100 mg) was taken once daily.

2.4. Collection of peripheral blood

Peripheral blood mononuclear cells were collected on day 5 of mobilization. A routine blood test was conducted, and 3 hours before collection, patients were injected with 300 μg filgrastim and 5 mg dexamethasone. Blood was collected using a COM.TEC blood cell separator (Fresenius Kabi, Bad Homburg, Germany, http://www.fresenius-kabi.com) at a flow rate of 45‐50 mL/min. Circulating blood volume was 10‐12 L (2‐2.5 cycles). Anticoagulant (sodium citrate) was administered at 1/12 the volume of circulating blood. Calcium gluconate was given to prevent hypocalcemia. Collected blood was transferred to an octopus bag in a large‐capacity centrifuge. Sample volumes were adjusted with PBS and 5 g/L human serum albumin, and samples were centrifuged at 1000 rpm for 15 minutes to remove platelet‐rich plasma.

2.5. CD34⁺ cell sorting

CD34⁺ cells were isolated from PBMCs by immunomagnetic separation using a CD34 MicroBead Kit and a CliniMACS sorter, sorting pipes, and phosphate buffer (all from Miltenyi Biotec, Cologne, Germany, https://www.miltenyibiotec.com). CD34⁺ sorting reagents (3.75 mL) were added, and samples were incubated at room temperature for 30 minutes with rotation. Samples were washed twice with PBS/EDTA to remove unbound reagents. The final product obtained was approximately 45 mL; 40 mL was transplanted into the patient, and the remainders of the cells were used for cell counting and classification, determination of viability and purity of CD34⁺ cell population, cell culture, and animal experiments.

2.6. Culture of primary human CD34⁺ cells

CD34⁺ cells were (5 × 10⁴‐1 × 10⁵) plated on fibronectin‐coated culture dishes inX‐VIVO 10 medium (Lonza, Allendale, NJ, USA, http://www.lonza.com) and incubated at 37°C in a humidified atmosphere of 5% CO₂. On day 9, nonadherent cells were collected and cryopreserved, and adherent cells were maintained in culture in Dulbecco's modified Eagle medium containing high glucose (DMEM‐H). Cell growth and morphologic features were observed by microscopy and phase contrast microscopy. CD34 and CD133 surface markers were detected by immunofluorescence.

2.7. Continuous live‐cell imaging and analysis

CD34⁺ cells (5 × 10⁴‐1 × 10⁵) were seeded in wells of six‐well plates and placed in the Cell‐IQ system (CM Technologies, Tampere, Finland, http://www.c-mtechnologies.com). Five fields in each well were selected for continuous imaging and were photographed every 45 minutes for 2 weeks. Cell proliferation and migration were documented by Cell‐IQ.

2.8. Growth curves

CD34⁺ cells (1 × 10⁴) were seeded in wells of six 24‐well plates and cultured in X‐XIVO medium, DMEM‐H, or X‐XIVO until day 9 and DMEM‐H thereafter. Medium was changed every 3 days, and cells were passaged once a week. Cells in three wells of each plate were counted each day, and the mean number of cells was determined. Growth curves were generated for each group of cells.

2.9. Cell cryopreservation and recovery

CryoStor preservation medium (Sigma‐Aldrich, St. Louis, MO, USA, http://www.sigmaaldrich.com) or dimethyl sulfoxide (DMSO) was added to CD34⁺ cells suspended in culture medium to 10% cell density, and cells were stored immediately at −80°C. Cells were thawed and recovered 1 year later, and cell viability was determined by trypan blue dye exclusion assay. Percent cell viability was calculated as the number of living cells/total number of cells) × 100%.

2.10. Immunohistochemical staining of CD34 and CD133

Slides seeded with cells were fixed in 4% paraformaldehyde for 20 minutes, rinsed with PBS containing 0.1% bovine serum albumin (BSA), and blocked in PBS containing 1% BSA, 10% donkey serum, and 0.1% Triton X‐100 for 45 minutes. Cells were then incubated with PE‐conjugated anti‐CD34 (or FITC‐conjugated anti‐CD133) primary antibody diluted 1:100 in PBS containing 20% fetal bovine serum (FBS) at room temperature in the dark for 1‐2 hours. Cells were incubated with secondary antibody in a humidified chamber at room temperature in the dark for 1 hours, and incubated with 1 mg/mL 4′,6‐diamidino‐2‐phenylindole (DAPI) for 5 minutes to counterstain nuclei.

2.11. Induction of limb ischemia and treatment with CD34⁺ cells

In all, 36 BALB/C nude mice were randomly divided into three groups of 12 mice each. Mice were anesthetized. A 2‐cm incision was made from the midpoint of the left groin to the knee, and the femoral nerve and the femoral artery and its branches were exposed. In the CD34⁺ treatment group and the X‐VIVO control group, the superficial femoral artery was ligated and stripped followed by suturing of the skin. In the sham group, skin was sutured without ligation and transection of the artery.

2.12. CD34⁺ cell transplantation

Seven days after surgery, mice in the CD34⁺ group were injected with 1 × 10⁵ CD34⁺ cells in 0.1 mL X‐VIVO medium in the ischemic muscle tissue of the left limb using a 26 gauge needle. Mice in the X‐VIVO control group were injected with 0.1 mL X‐VIVO medium alone, and mice in the sham group were injected with 0.1 mL saline.

2.13. Assessment of ischemic injury

One week after surgery and again 2 weeks after CD34 cell transplantation, mice were scored for extent of ischemic injury according to the following scale29: 3 points, if the hind limb dragged or showed gangrene; 2 points, if limb dragging was not obvious, but foot lacked plantar flexion, or limb was moderately discolored; 1 point, if tail was pulled, ipsilateral foot showed plantar flexion, or limb was mildly discolored; 0 points if animal appeared no different from a control mouse.

2.14. Immunohistochemistry

Samples of adductor and semimembranosus muscles were collected 2 weeks after CD34⁺ cell transplantation. Tissues were fixed in 10% formalin and embedded in paraffin. Sections were treated with anti‐CD31 and anti‐CD34 antibodies to stain vascular endothelial cells (ECs).

2.15. Real‐time PCR analysis of cytokine mRNA levels

Total RNA was extracted from ischemic and control muscle tissue. RNA purity and concentration were determined, and cDNA was prepared and used in real‐time PCR to assess expression of VEGF and bFGF. Primer sequences were VEGF,

5′‐GCACATAGAGAGAATGAGCTTCC‐3′ (forward) and

3′‐CTCCGCTCTGAACAAGGCT‐5′ (reverse); bFGF,

5′‐GCGACCCACACGTCAAACTA‐3′ (forward) and

3′‐CCGTCCATCTTCCTTCATAGC‐5′ (reverse); and 18S,

5′‐CGCCGCTAGAGGTGAAATTCT‐3′ (forward) and

3′‐CATTCTTGGCAAATGCTTTCG‐5′ (reverse).

2.16. CD34⁺ cell transplantation into patients

Spinal anesthesia was usually used when the lower limb was treated and general anesthesia was used when the limb was treated. Purified CD34⁺ cells (0.5 mL) were injected into calf muscle at 60 sites and into feet and toes at 20 sites; injection sites were separated by 3 cm. The total number of cells transplanted was 1 × 10⁵‐1 × 10⁶/kg. If the patient developed infected ulcers or gangrene, debridement or amputation was carried out, whereas dry gangrene unaccompanied by infection was treated next time. Cilostazol and aspirin would be taken for life after transplantation.

2.17. Patient follow‐up protocol and endpoints

All patients were followed up once monthly for the first 3 months after transplantation, once every 2 months at 4‐12 months after transplantation, and every 6 months thereafter. Additionally, a follow‐up phone call was made every week during the first month to identify when pain the rest began to subside. Safety endpoints included any perioperative adverse event, pathogenic angiogenesis in the retina, and abnormal blood cell count during the first 3 months. Efficacy endpoints included rate of survival without major amputation at 6 months, peak pain‐free walking time (PPFWT) on a treadmill at 2.5 km/h and 10% incline, Wong‐Baker FACES pain rating scale score (WFPRSS), ankle‐brachial index (ABI), transcutaneous partial oxygen pressure (TcPO2), and ulcer healing.

2.18. Statistical analysis

The Kaplan‐Meier method was used to estimate the rate of survival without major amputation at 6 months, and the Greenwood method was used to calculate the corresponding 95% confidence interval. One‐way ANOVA, Wilcoxon signed‐rank test, and Mann‐Whitney U test were used to compare groups. Quantitative indicators, such as cell viability, ABI, TcPO2, PPFWT, and WFPRSS, were expressed as mean ± SD, and analyzed via t test. P value < .05 was considered statistically significant. Statistical analysis was carried out using SPSS l6 software.

3. RESULTS

3.1. Morphology of primary CD34⁺ cells

Primary CD34⁺ cells were initially uniformly round and mononuclear (Figure 1A). After 2 days in culture cells began to increase in number and volume (Figure 1B), and a small number of irregularly shaped adherent cells appeared (Figure 1C). Colonies had formed by day 3. On day 7, the number of suspended cells increased significantly, and the spindle‐shaped adherent cells grew over one another in a multilayered arrangement (Figure 1D). At approximately day 9, a small amount of cellular debris appeared (Figure 1E) and most of the suspended cells died at day 14 (Figure 1F). Round cells located at the center of colonies of adherent cells were surrounded by spindle‐shaped cells (Figure 1G); finally, adherent cells increased rapidly and assumed a radial or spiral arrangement (Figure 1H).

Figure 1.

A shows the 1d of primary culture (40 times) when the cells are uniformly circular mononuclear cells. B shows the 3d of primary culture day (40 times) when suspension cells increased and enlarged. C shows the 5d of primary culture (100 times) when a small number of adherent cells were mainly round and cells with irregular shape can also be seen. D shows the 5d of primary culture (200 times) when the number of suspension cells increased significantly and the fuzzy cell clusters above were suspension cells of different levels, whereas the cells below were adherent shuttle fibroblast‐like cells. E shows the 9d of primary culture (100 times) when a small amount of cell debris can be seen among the suspension cells. F shows the 14d of primary culture (200 times) when most of suspension cells died and cell debris increased significantly. G shows the 9d of primary culture (400 times) when adherent cells began to form scattered colonies centralized by round cells and surrounded by spindle cells. H shows the 14d of primary culture (200 times) when adherent cell colonies were densely populated and the surrounding cells were arranged radially or spirally

3.2. Visualization and analysis of cell proliferation and migration

Continuous live‐cell imaging using Cell‐IQ showed that the suspended cells rapidly migrated to the center of the culture dish (Figure 2A‐F). Cell division was observed, and the average time between divisions was calculated to be 90 minutes (Figure 3).

Figure 2.

Cell migration dynamically observed by cell‐IQ CD34⁺ cells were dynamically observed near the middle of the culture dish, and the suspension cells continued to centralize within 4 d of continuous observation

Figure 3.

Cell proliferation dynamically observed by cell‐IQ. The average time of cell division was 90 min

3.3. CD34⁺ cell growth under different culture conditions

Growth curves for cultures of suspended CD34⁺ cells grown in X‐VIVO or DMEM‐H and for adherent cells grown in X‐VIVO until day 9 and thereafter in DMEM‐H are shown in Figure 4. In all groups, we found that days 1‐3 were the incubation period, days 5‐7 were the logarithmic growth phase and, after the logarithmic phase, cell growth gradually plateaued at 8‐9 days. The number of cells grown exclusively in X‐VIVO or DMEM‐H was not maintained at a constant level on days 9‐14, and the cells died gradually. However, after DMEM‐H replaced X‐VIVO on day 9 in the third group of cells, these cells entered another phase of logarithmic growth (Figure 4).

Figure 4.

CD34⁺ cell growth curves in different culture modes

3.4. CD34 and CD133 expression

Adherent primary (P1) and fourth passage (P4) CD34⁺ cells were positive for expression of the cell surface markers CD34 and CD133 (Figure 5).

Figure 5.

A, PE‐CD133 antibody staining of P1 cells (red); B, FITC‐CD34 antibody staining of P1 cells (green); C, image fusion of P1 cell suggests that surface markers CD133 and CD34 were double positive. D, PE‐CD133 antibody staining of P4 cells; E, FITC‐CD34 antibody staining of P4 cells; F, image fusion of P4 cells. Cell nuclei were stained by Hoechst and gave off blue fluorescence after the staining

3.5. Cryopreservation and recovery of CD34⁺ cells

CD34⁺ cells were stored at −80°C for 1 year in medium containing 10% Cryostor or 10% DMSO. The viability of recovered cells was greater for cells stored in Cryostor (89.0% ± 2.58) than in DMSO (84.9% ± 3.28) (P = .006) (Figure 6).

Figure 6.

Box plot of cell viability

3.6. Ischemic injury score and prognosis

The ischemic injury scores and prognosis for the three groups of mice before and after transplantation are shown in Table 1 and Figure 7. Before transplantation, both the CD34⁺ and X‐VIVO groups were significantly different from the sham group (P = .000), indicating that the model of hind limb ischemia was successful. The injury scores of the CD34⁺ group before and after transplantation were significantly different (P = .020), illustrating that limb ischemia was improved by CD34⁺ cell transplantation. The X‐VIVO group improved after transplantation compared to before transplantation, but the difference was not significant (P = 0.083). There was also no significant difference between the CD34⁺ and X‐VIVO groups after transplantation (P = 0.343).

Figure 7.

Hindlimb ischemia of CD34 nude mice. The typical hindlimb ischemia and healing in the CD34 group were observed at 3d, 7d, 10d, 14d, and 21d after modeling. It can be found that CD34⁺ cell transplantation has favorable therapeutic effects on hindlimb ulcers in nude mice, but poor effects on acute gangrene

3.7. Real‐time PCR analysis of cytokine mRNAs

mRNA expression levels of the angiogenic cytokines VEGF and bFGF were significantly higher in the CD34⁺ group than in the X‐VIVO and sham groups (P < .05) (Figure 8). These data suggest that CD34⁺ cells may upregulate VEGF and bFGF expression, thus promoting angiogenesis through paracrine effects.

Figure 8.

Logarithm of relative concentration of VEGF and BFGF in the three groups after amplification

3.7.1. Immunohistochemical staining of CD34 and CD31 in ischemic muscle

Vascular endothelial cells in sections of muscle taken from the three groups of mice showed positive staining with anti‐CD31 and anti‐CD34 antibodies (Figure 9). Staining of new capillaries was most extensive in the CD34⁺ group. New capillaries were rarely observed, but some larger vessels were seen in the sham group, whereas both new capillaries and larger vessels were rare in the X‐VIVO group.

Figure 9.

Immunohistochemical staining of the three groups. Few new capillaries of anti‐CD31 (18A × 400) and anti‐CD34 (18B × 200) were stained in sham group, whereas larger vessels were mostly stained where blood cells can be observed. More new capillaries of anti‐CD31 (18C × 200) and anti‐CD34 (18D × 200) were stained in CD34 group, and fewer new capillaries of anti‐CD31 (18E × 200) and anti‐CD34 (18F × 100) were stained in X‐VIVO group

3.8. Mobilization, collection, and sorting of PBMCs

Patient's WBC increased from (8.07 ± 2.31) × 10⁹/L before mobilization with G‐CSF to (42.86 ± 12.49) × 10⁹/L 5 days after mobilization. CD34⁺ cells represented 0.48 ± 0.13% (range, 0.31%‐0.67%) of WBC.

Collection time was 205 ± 28 minutes (range, 169‐241 minutes). The total circulating blood volume was 2.3 ± 0.4 (range, 1.8‐2.8) and blood volume was 10 185 ± 993 mL (range, 8728‐11578). The final product volume was 265 ± 24 mL (range, 235‐315). The concentration of mononuclear cells was (158.0 ± 61.2) × 10⁹/L (range, [86.8‐269.9] × 10⁹/L), and the purity of the CD34⁺ cell population was 0.48 ± 0.13% (range, 0.31%‐0.67%).

The volume of sorting was 43.7 ± 2.5 mL (range, 40‐48 mL). Cell viability by trypan blue dye exclusion was 99 ± 0.3% (range, 98%‐99%). The number of PBMCs was 1.6 ± 0.6 × 10⁹/L (range, 1.1‐2.4) × 10⁹/L), and CD34⁺ cell purity was 80.4% ± 12.1 (range, 70%‐98%). The number of cells transplanted was (7.5 ± 2.5) × 10⁵/kg (range 4.5 × 10⁵‐1 × 10⁶)/kg].

3.9. Patient follow‐up

Pain at rest was apparently relieved within 1 month of CD34⁺ cell transplantation in 19 of the 25 patients. The mean time to pain relief was 3 ± 1 weeks (range, 2‐8 weeks). WFPRSS decreased from 7 ± 2 to 3 ± 3 (P < .001) and 1 ± 2 (P < .001) at 1 and 2 months, respectively, and there was also a significant difference between 1 and 2 postoperative months (P = .02). In all, 12 cases were eligible for a test of PPFWT, which increased markedly from 5 ± 3 minutes to 12 ± 6 minutes (P < .001) and 19 ± 5 minutes (P < .001) at 3 and 6 months, respectively. ABI was considerably elevated from 0.44 ± 0.20 to 0.62 ± 0.17 (P = .004) and 0.66 ± 0.14 (P < .001) at 3 and 6 months, respectively, whereas the difference between 3 and 6 postoperative months was not statistically significant (P = .19). TcPO2 was significantly improved from 26 ± 11 mm Hg to 42 ± 11 mm Hg (P < .001) and 56 ± 12 mm Hg (P < .001) at 3 and 6 months, respectively. Also, the improvement from 3 to 6 months was statistically significant (P = .0001) (Table 2). Of the 16 cases with an ulcer, the ulcer healed within 5 ± 4 months (range, 2‐14 months) in 11 patients (69%).

Table 2.

The follow‐up items before and after transplantation

Items	Before	3 months after	P value	6 months after	P value
PPFWT (min)	5 ± 3	12 ± 6	P < .001	19 ± 5	P < .001
ABI	0.44 ± 0.20	0.62 ± 0.17	P = .004	0.66 ± 0.14	P < .001
TcPO2 (mm Hg)	26 ± 11	42 ± 11	P < .001	56 ± 12	P < .001

ABI, Ankle brachial index; PPFWT, peak pain‐free walking time; TcPO₂, transcutaneous partial oxygen pressure.

In the four cases with an unhealed ulcer, two ASO cases received above‐the‐knee amputation within 3 post‐transplantation months. The other two cases were still receiving wound care; one case had an ulcer on the pressure‐bearing bottom of the foot, and the other had extensive ulcers with calcaneus exposure. In the five cases with gangrene, one was lost to follow‐up and four underwent toe amputation. Of these, the wound healed in three cases within 3 ± 1 months (range, 3‐4 months) and remained unhealed in one case, who underwent a below‐the‐knee amputation 1 month after transplantation. Another ASO case underwent an above‐the‐knee amputation at 1 month because of progressive pain at rest. The Kaplan‐Meier estimate of the rate of survival without major amputation at 16 months was 84% (95% confidence interval, 0.63‐0.94).

3.10. Safety evaluation

Mobilization‐induced events, such as nausea, lower back pain, and low‐grade fever, occurred in 12 cases and were spontaneously relieved within 2‐3 days. Numbness of the lips observed in 13 patients during apheresis disappeared shortly after oral administration of calcium gluconate. No pathogenic retinal angiogenesis was observed throughout the follow‐up. No other morbidities or mortality developed either perioperatively or during the follow‐up.

4. DISCUSSION

Many studies agree that cells initially CD34⁺/VEGF receptor‐2⁺/CD133⁺may be the endothelial progenitor cells (EPCs),30, 31, 32 whereas CD34⁺ cells were mixed with hematopoietic stem cells (HSCs), EPCs, and ECs.4, 33, 34 In this study, CD34⁺ cells mobilized by G‐CSF treatment and sorted using immunomagnetic beads were a mixture of HSCs and EPCs.

Human primary CD34⁺ cells were initially round, isomorphous, suspended mononuclear cells and, following proliferation in culture, most died gradually by day 14, regardless of whether they were grown in X‐VIVO or DMEM‐H culture medium. This is in agreement with the findings of other studies.10, 35 By chance, we found that adherent CD34⁺ cells survived over a long period of time (>3 months) after 9 days of culture in X‐VIVO medium, whereas most cells growing in suspension died at 2 weeks. Based on this observation, adherent cells were cultured continually and immunohistochemistry was carried out to assess expression of the cell surface markers CD34 and CD133. The cells were positive for both markers; therefore, we hypothesized that the adherent cells were EPCs. We cryopreserved the suspended cells and recovered them after 1 year of storage, with 89% viability. This provided theoretic support for transplantation of cells after in vitro culture.

During live‐cell image analysis using Cell‐IQ, we observed the interesting phenomenon that the majority of CD34⁺ cells migrated to the center of the culture dish, where they were visible to the naked eye because of high density. This may be an indication that EPCs can migrate to the site of an injury in response to a specific signal. We reviewed the literature and found no relevant reports. Is there some factor that drives cells to the center? Further research is needed to identify the signaling pathways and underlying mechanisms involved in this phenomenon.

It has been proposed that there are two mechanisms of vasculogenesis.36, 37, 38 One is that EPCs differentiate and develop directly through migration and homing to damaged blood vessels. The other is that EPCs secrete cytokines, such as VEGF, bFGF, and angiopoietin‐1 (ang‐1), thus promoting vasculogenesis in ischemic tissues by a paracrine route.39

Generally, bFGF promotes the growth of new vessels early in vasculogenesis, and VEGF then performs the primary role of sustaining vasculogenesis.40 Ang‐1 and VEGF perform complementary functions in vasculogenesis and angiogenesis. VEGF promotes endothelial cell proliferation and migration to establish a uniform vascular network, and Ang‐1 remodels vessels to complete the mature vascular network with normal function in later stages.41 Some studies also found that ischemic limb perfusion can be significantly increased and damage repair improved by perfusing immunodeficient mice with EPCs, but only a small number of EPCs integrated into the new capillaries.42, 43, 44, 45 This phenomenon illustrates that EPCs may release angiogenic factors through paracrine signaling.

In this study, we found that levels of VEGF and bFGF mRNAs were significantly higher in mice treated with CD34⁺ cells than in those treated with X‐VIVO alone and the sham group. This finding indicates that EPCs, stimulated by unknown cell signaling pathways, secrete cytokines that promote the proliferation and migration of vascular ECs and smooth muscle cells around the ischemic limb and increase collateral vessel formation and remodeling.

Limb ulcers in mice healed, and injury score improved significantly 2 weeks after CD34⁺ cell transplantation, further supporting the idea that CD34⁺ cell transplantation can significantly increase angiogenesis and promote the formation of collateral vessels to increase ischemic tissue perfusion. But CD34⁺ cell transplantation showed only a weak effect on acute limb ischemia and necrosis, and may require more time for neovascularization. An improvement, but no significant difference (P = .343), was observed between the CD34⁺ and X‐VIVO groups 2 weeks after transplantation. It is possible that (1) acute limb necrosis in nude mice affected the overall treatment, or that (2) VEGF and other cytokines in X‐VIVO medium promoted vascularization.

As for cell therapy, investigators initially preferred to use BM‐MNCs because the number of CD34⁺ cells was approximately 500‐fold higher than in PB‐MNCs.16, 17 However, some disadvantages of BM‐MNC transplantation, such as the need for aspiration of approximately 500 mL bone marrow and general anesthesia, prompted doctors to seek new approaches to obtain target cells. With the application of G‐CSF for mobilization, the number of CD34⁺ cells acquired from PB‐MNCs dramatically increased to close to the number collected from bone marrow, thus promoting the use of G‐CSF‐mobilized PB‐MNCs for transplantation.20, 46 CD34 was deemed a common surface marker of EPCs, hematopoietic stem cells, and mature ECs, whereas the hematopoietic stem cells could differentiate into EPCs in an ischemic setting.47, 48

In our study, 25 patients were not eligible for bypass surgery or endovascular therapy and who had undergone ineffective drug treatment for more than 3 months, but their ABI, PPFWT, TcPO2, and WFPRSC improved significantly after transplantation of purified CD34⁺ cells. It was noted that results with all three ASO cases were unsatisfactory, suggesting that CD34⁺ cell transplantation might be less effective in patients with ASO than in those with TAO or arteritis. A similar finding was reported after transplantation of BM‐MNCs or PB‐MNCs, with overall efficacy rates of 50% and 80% in ASO and TAO patients, respectively.17, 18, 49 The proposed causes were (1) factors causing endothelial dysfunction or depletion, such as aging, diabetes, and hypercholesterolemia, were much more common in ASO patients; and (2) plasma levels of angiogenic cytokines, such as VEGF, a critical factor in vasculogenesis and angiogenesis, were attenuated in older patients.50, 51, 52, 53 In summary, transplantation of purified CD34⁺ cells appears to be feasible, safe, and effective in treatment of no‐option CLI, particularly in younger patients with TAO or collagen diseases.

In short, there are many properties of CD34 cells that still must be explored, and elucidation of the mechanism of vasculogenesis requires further research. Further clinical research, employing a greater number of patients and longer follow‐up periods, are necessary.

CONFLICT AND INTEREST

The authors declare that they do not have any conflict and interest.

Lian W, Hu X, Pan L, et al. Human primary CD34⁺ cells transplantation for critical limb ischemia. J Clin Lab Anal. 2018;32:e22569 10.1002/jcla.22569