Posts Tagged ‘skeletal muscle’

Transplantation of modified human adipose derived stromal cells expressing VEGF165

Author: Larry H. Bernstein, MD, FCAP and Curator: Aviva Lev-Ari, PhD, RN

https://pharmaceuticalintelligence.com/2013-11-03/larryhbern/Transplantation of modified human adipose derived stromal cells expressing VEGF165 

This contribution to the series on stem cells and regenerative medicine deals with transplantation of modified human adipose tissue to repair  ischemic damaged skeletal muscle by apparently increase neovascularization, essentrial for laying down the circulation that feeds the tissue.

Transplantation of modified human adipose derived stromal cells expressing VEGF165 results in more efficient angiogenic response in ischemic skeletal muscle

Evgeny K Shevchenko1*Pavel I Makarevich12Zoya I Tsokolaeva1,Maria A Boldyreva1Veronika Yu Sysoeva2Vsevolod A Tkachuk23 andYelena V Parfyonova12
1Laboratory of angiogenesis, Russian Cardiology Research and Production Complex;  2Lomonosov Moscow State University; 3Laboratory of molecular endocrinology, Russian Cardiology Research and Production Complex, Moscow,  Russia.
Journal of Translational Medicine 2013, 11:138.   http://www.translational-medicine.com/content/11/1/138   http://dx.doi.org/10.1186/1479-5876-11-138
This is an Open Access article distributed under the terms of the Creative Commons Attribution License   http://creativecommons.org/licenses/by/2.0



Modified cell-based angiogenic therapy has become a promising novel strategy for ischemic heart and limb diseases. Most studies focused on myoblast, endothelial cell progenitors or bone marrow mesenchymal stromal cells transplantation. Yet adipose-derived stromal cells (in contrast to bone marrow) are abundantly available and can be easily harvested during surgery or liposuction. Due to high paracrine activity and availability ADSCs appear to be a preferable cell type for cardiovascular therapy. Still neither genetic modification of human ADSC nor in vivo therapeutic potential of modified ADSC have been thoroughly studied. Presented work is sought to evaluate angiogenic efficacy of modified ADSCs transplantation to ischemic tissue.

Materials and methods

Human ADSCs were transduced using recombinant adeno-associated virus (rAAV) serotype 2 encoding human VEGF165. The influence of genetic modification on functional properties of ADSCs and their angiogenic potential in animal models were studied.


We obtained AAV-modified ADSC with substantially increased secretion of VEGF (VEGF-ADSCs). Transduced ADSCs retained their adipogenic and osteogenic differentiation capacities and adhesion properties.

  • The level of angiopoetin-1 mRNA was significantly increased in VEGF-ADSC compared to unmodified cells yet
    • expression of FGF-2, HGF and urokinase did not change.

Using matrigel implant model in mice it was shown that

  • VEGF-ADSC substantially stimulated implant vascularization with paralleling increase of capillaries and arterioles.

In murine hind limb ischemia test we found

  • significant reperfusion and revascularization after intramuscular transplantation of VEGF-ADSC compared to controls with no evidence of angioma formation.


Transplantation of AAV-VEGF- gene modified hADSC resulted in stronger therapeutic effects in the ischemic skeletal muscle and may be a promising clinical treatment for therapeutic angiogenesis.


Therapeutic angiogenesis; Cell therapy; Gene modified cells; Adipose stromal cells; Vascular endothelial growth factor; Adeno-associated virus; Ischemia


Despite advances in revascularization techniques, the treatment of ischemic heart and limb diseases remains a worldwide problem. Therapeutic angiogenesis represents alternative new strategy for ischemia resolution that utilizes regenerative capacity of human body and

  • stimulates natural process of
  1. vessel growth,
  2. remodeling and
  3. tissue revascularization [1].

Commonly adopted approaches for therapeutic angiogenesis include

  • direct introduction of recombinant growth factors and gene therapy.

Yet clinical trials have shown several drawbacks of these modalities. Thus low efficacy of recombinant protein administration is explained by

  • dissemination after injection and
  • rapid degradation of therapeutic agent, which
  • requires multiple and long-term infusions thus
    • leading to tremendous expenses [2,3].

Delivery of cDNA coding angiogenic factors via different expression mammalian vectors (plasmids, recombinant viruses) was found more feasible and allowed to achieve great improvement in some cases yet

  • efficacy was still not high enough especially in double blind placebo controlled trials [4].

Many authors discussed possible reasons of gene therapy low efficacy and most of them are univocal to emphasize transfection efficacy and transient transgene expression after plasmid delivery. This can be circumvented by administration of viral vectors but their use is limited due to possible danger of insertional mutagenesis and immune reactions [5,6].

Recently, autologous transplantation of bone marrow stromal cells or endothelial progenitor cells has been shown to enhance angiogenesis and peripheral blood flow [79]. However,

  • the regenerative capacity of these cells decreases with age and
  • in patients with co-morbidities such as diabetes mellitus which reduces efficacy of autologous cell administration, and
  • limited cell viability after transplantation into ischemic tissues also restricts their angiogenic potential [1012].

It was shown in several experimental studies that this problem could be circumvented by gene modified cell therapy strategy utilizing stem or progenitor cells overexpressing angiogenic proteins [13,14]. To develop a feasible and potent gene modified cell therapy for ischemic diseases

  1. the cells should be both effective and accessible in large numbers as well as
  2. the chosen viral vector should be both safe and effective in terms of gene delivery.

The majority of experimental studies have evaluated gene modified bone marrow stromal cells or endothelial progenitor cells for ischemia treatment [1517]. However, cells extracted from bone marrow or peripheral blood after mobilization are available in limited numbers and as for bone marrow cells painful aspiration procedure is required.

In contrast to bone marrow or myoblasts, stromal fraction of adipose tissue contains an abundant population of multipotent stem cells that can be easily harvested in high numbers by minimally invasive surgical techniques [1821]. These adipose –derived stromal cells (ADSCs)

  • share common properties with bone marrow stromal cells and represent a very convenient object for therapeutic use.

However the best development of ADSC for angiogenic therapy still needs to be determined.

As for genetic modification of cells the choice of safe and effective gene transfer vector as well as the appropriate transgene determines the quality and safety of the cell product affecting the efficacy of modified cell based therapy. Recombinant adeno-associated viruses (rAAV) are one of the most promising and versatile tools in this field due to

  1. low immunogenicity and
  2. high transduction potency in vitro

in many types of both – dividing and non-dividing mammalian cells. Besides that until now no human disease caused by AAV has been identified [22].

In this study we genetically modified human ADSCs with a key regulator of angiogenesis – VEGF165 [23] via rAAV-transduction and then evaluated effects of rAAV-transduction and VEGF165 overexpression on human ADSC

  1. growth,
  2. differentiation capacity,
  3. adhesion and
  4. angiogenic factor expression as well as
  5. revascularization and
  6. functional improvement

after intramuscular injection in a mice hind limb model.


Cell culture

(refer to doi:10.1186/1479-5876-11-138)

DNA constructs production of rAAV particles and cell transduction

(refer to doi:10.1186/1479-5876-11-138)

Western blotting and ELISA

(refer to doi:10.1186/1479-5876-11-138)

ADSC proliferation activity assay

To assess population doubling time (PDT) of gene modified (transduced with rAAV at passage 1) or untreated ADSC (passage 2) seeded on 6-well plates (2 × 104 cells/well). After a 9 day incubation average cell numbers for three wells were obtained using a hemocytometer chamber. PDT was calculated as follows:


 where t is period of incubation (hours), Nt – endpoint amount of cells, N0 – initial number of cells.

ADSC cell cycle stage analysis by flow cytometry

(refer to doi:10.1186/1479-5876-11-138)

ModFit LT 3.2 software (Verity Software House, USA) was used for analysis of cell distribution over cell cycle stages according to intensity of propidium iodide fluorescence in a wavelength range of 600–625 nm (excitation wavelength – 488 nm). Results are presented as a percentage of cells in S + G2/M stages.

Apoptosis assay

Analysis of spontaneous apoptosis frequency in ADSC culture was performed using Annexin-V FITC Apoptosis Kit (Invitrogen, USA) according to manufacturer’s protocol.

Adipogenic, osteogenic and endothelial differentiation of ADSC

To confirm adipogenesis intracellular lipid droplets were detected using Oil red O staining reagent (Millipore, USA) 2 weeks after induction. To confirm osteogenesis Alizarine Red C staining was used to detect extracellular matrix mineralization 2 and 3 weeks post induction. Endothelial cells were stained for CD31 and VEGFR2 surface antigens and cell counts were obtained using flow cytometry.

Cell attachment assay

(refer to doi:10.1186/1479-5876-11-138)

Flow cytometry

Antigen expression analysis was performed on cell sorter MoFlo (DakoCytomation, Denmark) or flow cytometry scanner BD FACS CantoTM II (BD Pharmingen, USA). 10 000 events were acquired and analyzed for antigen expression.

Quantitative polymerase chain reaction

Quantitative polymerase chain reaction (qPCR) was performed using primers specific for human VEGF165, ANGPT1, HGF, FGF2 and PLAU mRNAs.


8–10 week-old male BALB/c NUDE mice

Matrigel plug assay

(Refer to doi:10.1186/1479-5876-11-138)

Hind limb ischemia model

Ten week-old male BALB/с NUDE mice were anaesthetized by intraperitoneal injection of 0.3 ml of 2.5% avertin. Femoral artery was separated in its distal part and ligated proximal to its popliteal bifurcation (keeping v. femoralis and n. ischiadicus intact). ADSC, GFP-ADSC or VEGF-ADSC (5×105 cells per animal) were resuspended in 150 μl of PBS, and injected in 3 equally divided doses tom. tibialis anterior, m. gastrocnemius and m. biceps femoris to generate three experimental animal groups: “GFP-ADSC”, “VEGF-ADSC”, “ADSC” (14 animals per group). PBS (150 μl) was injected in negative control “PBS” group. Blood flow was subsequently measured by laser Doppler imaging.

Laser doppler imaging

(Refer to doi:10.1186/1479-5876-11-138)

Muscle explants

M. tibialis anterior explant culture was prepared on matrigel according to Jang et al. [26] protocol and cultured in M199 medium (Gibco, USA), containing 2% FBS. At day 3 and 7 medium was collected for determination of human VEGF165 concentration by ELISA.

Specimen preparation and histological analysis

At designated period (day 20 for muscles, day 14 for matrigel plugs) animals were sacrificed by lethal isoflurane dose followed by cervical dislocation. Afterwards m. tibialis anterior or matrigel implants respectively were harvested,

For muscle necrosis analysis we used routine hematoxylin-eosin staining of formalin-fixed muscle sections. Necrotic tissue was defined by loss of fiber morphology, cytoplasm disruption, inflammatory cells infiltration and fibrosis.

Statistical analysis

Results were analyzed in Statsoft Statistica 6.0 (Statsoft, USA).


Effective transduction of human ADSC by adeno-associated virus serotype 2

Low-passage human ADSC obtained from different donors were transduced using rAAV encoding GFP to assess gene delivery efficacy. Transduced to total cells ratio was counted by flow cytometry. GFP-positive ADSC (GFP-ADSC) were detected as early as day 2 after viral infection. Maximum number of positive cells (65.6±3%) and highest GFP-fluorescence intensity was reached by day 4–5 (Figure 1). GFP signal was detectable for at least 30 days. At day 15 and 30 flow cytometry showed that 45±2% and 25±1.5% of ADSC were GFP-positive respectively.

Figure 1. Human ADSC transduction by recombinant adeno-associated virus

Figure 1. Human ADSC transduction by recombinant adeno-associated virus

A. GFP-positive cell count by FACS in GFP-ADSC culture at day 4 after transduction by rAAV. B.Representative image of GFP-positive human ADSC (green) transduced by rAAV, 100 × magnification.

Increase of VEGF expression and secretion after rAAV transduction of human ADSC

To obtain gene modified ADSC we constructed rAAV vector encoding human VEGF165. In ADSC transduced by rAAV-VEGF (VEGF-ADSC) VEGF165 mRNA level increased 80±15-fold compared to basal expression in unmodified ADSC or GFP-ADSC (Figure 2A). Protein production was analyzed by Western blotting and ELISA. Data presented at Figure 2B, C shows that in VEGF-ADSC secretion of VEGF increased 45-50-fold (4.5±1.8 ng/ml/105 cells) compared to unmodified cells (0.1±0.02 ng/ml/105 cells) or GFP-ADSC (0.09±0.02 ng/ml/105cells). VEGF concentration in conditioned medium decreased over time during VEGF-ADSC cultivation but remained 30-fold higher (2.9±1.1 ng/ml/105 cells) than in controls (0.09 ± 0.02 ng/ml/105 cells) at day 30 post transduction. Material from a total of 10 donors was used to obtain mean values of VEGF expression increase.

 Figure 2. Validation of VEGF165 expression in AAV-modified VEGF-ADSC.
Figure 2. Validation of VEGF165 expression in AAV-modified VEGF-ADSC.

A. VEGFA expression level in human ADSC 10 days after AAV transduction determined by quantitative PCR. B, C. Analysis of VEGF secretion by GFP-ADSC, VEGF-ADSC and unmodified cells using ELISA (B) and immunoblotting (C). In immunosorbent assay protein content was determined in conditioned media samples obtained at days 7 and 30 post genetic modification of ADSC.

rAAV-mediated modification of human ADSC suppresses their proliferation activity yet does not influence apoptosis

We found that proliferation rate of VEGF-ADSC and GFP-ADSC was reduced compared to unmodified cells (Figure 3A). ADSC population doubling time was 61.3±7 h, while for GFP-ADSC and VEGF-ADSC it was 116.9±11 and 145.4±12 h respectively (n=5, p<0.01 vs unmodified cells). At the same time spontaneous apoptosis rate in all three cell cultures was comparable and comprised about 2±0.5% of total cell population.

Figure 3. . Proliferation of gene modified ADSC.

Figure 3. Proliferation of gene modified ADSC.
Population doubling time in GFP-ADSC, VEGF-ADSC and ADSC cultures. Data of five serial runs. B. Cells distribution in S-G2 cell cycle stages according to cytometry analysis of GFP-ADSC, VEGF-ADSC and ADSC. Data of three serial runs.

Analysis of cell cycle stages distribution in ADSC, GFP-ADSC and VEGF-ADSC cultures (Figure 3B) showed that number of cells in S-G2 stages was more than 1.5-fold lower in modified cells: GFP-ADSC (16±4% cells) and VEGF-ADSC (13±6% cells) compared to unmodified ADSC (25±3% cells; n=3; p<0.05 vs unmodified cells).

ADSC adhesion does not change after genetic modification

Interactions with extracellular matrix proteins play important role in incorporation and integration to recipient’s tissue, cell viability and their functional properties upon transplantation. ADSC did adhere on main extracellular protein collagen type 1 as well as vitronectin and fibronectin while almost none of cells attached to laminin-coated plastic. We did not observe statistically significant differences in adhesion properties between ADSC, GFP-ADSC and VEGF-ADSC cultures (Figure 4).

Figure 4. Data from comparative study of ADSC, GFP-ADSC and VEGF-ADSC adhesion on culture plates

 Figure 4. Data from comparative study of ADSC, GFP-ADSC and VEGF-ADSC adhesion on culture plates coated by collagen 1, vitronectin, fibronectin or laminin (n=4).

Modified ADSC retain their adipogenic, osteogenic and endothelial differentiation potential in vitro

To analyze potential influence of viral transduction and transgene overexpression on differentiation capacity of gene modified cells we performed experiments on adipogenic and osteogenic differentiation of ADSC.

Microscopic analysis of gene modified and untreated ADSC stained with Oil Red O reagent after 14 days of incubation in adipogenic media showed >30% of differentiated (visualized by intracellular lipid droplets accumulation) cells (Figure 5). Oil Red O+ cell count did not reveal statistically significant differences in both GFP-ADSC (33.7±8.1%) and VEGF-ADSC (34.1±11.5%) as well as unmodified ADSC (34.3±11.7%). Similar results were obtained in osteogenic differentiation assay of ADSC. It was confirmed by Alizarin Red C staining that detects extracellular matrix mineralization. At 14 days of incubation in osteogenic media we detected dye-positive cells in ADSC, GFP-ADSC, VEGF-ADSC culture. At day 21 it was followed by dramatic increase of extracellular matrix calcification in both – modified and untreated cells without significant differences (Figure 5).

Figure 5. Adipogenic and osteogenic differentiation of gene modified ADSC

Figure 5. Adipogenic and osteogenic differentiation of gene modified ADSC

Representative images of ADSC and VEGF-ADSC cultures stained by Oil Red O (lipid droplets detection, kjadipogenic differentiation, 100 × magnification) and Alizarine Red C (matrix mineralization, osteogenic differentiation, 100 × magnification for “day 14” and 50 × magnification for “day 21”) reagents after incubation in specific differentiation medium, n=3.

Taking into account mitogenic activity of VEGF we analyzed possible effect of genetic modification and VEGF overexpression on endothelial cell fraction in VEGF-ADSC. Using flow cytometry we determined amount of cells that carry CD31 and VEGFR2 endothelial markers in ADSC, GFP-ADSC and VEGF-ADSC (rAAV-modified at passage 1) cultures at passage 2. Less than 1.5% of CD31, VEGFR2-positive cells were detected in all three populations. Subsequently modified and untreated ADSC at passage 2 that reached >90% confluency were subject to incubation in EGM-2 medium to stimulate endothelial differentiation. After 14 days of cultivation in EGM-2 repeated analysis of CD31 and VEGFR2 expression showed that percentage of endothelial marker-positive cells did not change and remained about 1% in all assayed cultures.

Level of angiopoietin-1 mRNA increases in VEGF-ADSC

Using qPCR we studied potential impact of genetic modification and augmented VEGF secretion on expression activity of hepatocyte growth factor (HGF), fibroblast growth factor-2 (FGF2), angiopoietin-1 (ANGPT-1) and urokinase (PLAU) genes in VEGF-ADSC. As shown in Figure 6 we did not find any changes in FGF2 and HGF expression in GFP-ADSC and VEGF-ADSC compared to ADSC. We found a 3-fold increase in urokinase expression in VEGF-ADSC yet it was not statistically significant. At the same time increase of ANGPT-1 expression in VEGF-ADSC was significant and 5.3±0.6-fold higher than in unmodified cells or GFP-ADSC (n=6, p<0.05).

Figure 6. Comparison of ANGPT1, FGF2, PLAU and HGF genes expression

Figure 6. Comparison of ANGPT1, FGF2, PLAU and HGF genes expression by quantitative PCR in GFP-ADSC, VEGF-ADSC and unmodified ADSC. Charts represent relative expression for assayed genes from a total of 6 runs.

Analysis of VEGF and PDGF receptors expression on ADSC surface

Analysis of VEGF receptors expression on human ADSC was carried out to assess possible autocrine action of VEGF on VEGF-ADSC functional properties. Flow cytometry of ADSC and VEGF-ADSC (at passage 1–2) from different donors stained for VEGF receptor 1 and 2 showed <1% of positive cells (Figure 7). Taking into account observation of Ball et al. which indicated platelet-derived growth factor receptors (PDGFRα and PDGFRβ) as facultative receptors for VEGF165 [27] we analyzed the presence of cells which expressed PDGFRβ in human ADSC culture. Using specific monoclonal antibodies and subsequent flow cytometry we found that >90% of human ADSC were positive for PDGFRβ (Figure 7).

Figure 7. Analysis of VEGF and PDGF receptors expression on ADSC surface.

Figure 7. Analysis of VEGF and PDGF receptors expression on ADSC surface. VEGFR1, VEGFR2 or PDGFRβ-positive cell count by flow cytometry in ADSC culture.

Increased vascularisation of matrigel implants after VEGF-ADSC transplantation

We used matrigel plug assay to determine angiogenic properties of gene modified ADSC in vivo. At day 14 matrigel implants were harvested and subject to histological analysis (Figure 8). In negative control group we found only small sporadic capillaries (<1 capillary per FOV) were detected while in “ADSC”, “GFP-ADSC” and “VEGF-ADSC” groups formation of vessel network was more evident. Vessel counts revealed a 2.7-fold increase of CD31-positive vessels in group “VEGF-ADSC” (88.1±10.4 vessels per FOV) compared to “GFP-ADSC” (31.3±6.2 vessels per FOV) and “ADSC” (34.5±11.6 per FOV). Number of smooth muscle actin (SMA)-positive vessels was also 2.5-fold higher in “VEGF-ADSC” (1.7±0.24 vessels per FOV) than in “GFP-ADSC” (0.7±0.3 vessels per FOV) and “ADSC” (0.7±0.2 vessels per FOV). Thus capillaries/SMA+vessels ratio did not vary among experimental groups.

Figure 8. Effect of VEGF-ADSC or ADSC on vascularization of matrigel implants in nude mice.

Figure 8. Effect of VEGF-ADSC or ADSC on vascularization of matrigel implants in nude mice.
A.Representative images of matrigel sections from “VEGF-ADSC” and “ADSC” groups stained by antibodies against murine CD31 and SMA, 100× magnification. B. Capillaries and arterioles count in matrigel implants.

Blood flow recovery after VEGF-ADSC transplantation into ischemic murine limb

Perfusion assessment in hind limb ischemia model showed maximum blood flow recovery in “VEGF-ADSC” group (Figure 9). By day 20 spontaneous reperfusion of ischemic limb in «PBS» group was feeble and did not exceed 30%. In contrast we observed evident augmentation of blood supply in three experimental groups that received cell injections. At the end of experiment perfusion in “ADSC” and “GFP-ADSC” groups reached 50% and 55% respectively. Blood flow recovery after VEGF-ADSC transplantation was much more effective. At day 12 perfusion in group “VEGF-ADSC” significantly exceeded values in “ADSC” and “GFP-ADSC” by 15-20% and towards the end of experiment (day 20) it reached 80-90%. Thus transplantation of ADSC overexpressing VEGF was more effective than of untreated or GFP-ADSC.

Figure 9. Reperfusion of murine ischemic limb after ADSC administration.

Figure 9. Reperfusion of murine ischemic limb after ADSC administration.
A. Representative laser-doppler scans of subcutaneous blood flow in mice from “ADSC” and “VEGF-ADSC” groups obtained at days 4 and 20 after ischemia induction and cell transplantation. B. Dynamics of blood flow recovery in ischemic limbs within 20 days after intramuscular injection of ADSC, GFP-ADSC, VEGF-ADSC or PBS.

Transplantation of VEGF-ADSC reduces necrosis and stimulates stable vessel formation in ischemic muscle

Histological analysis of hematoxylin-eosin stained m. tibialis anterior specimens obtained at day 20 after and cell transplantation showed significant decrease in necrotic tissue span in «VEGF-ADSC» group (31.3±7%) compared to «ADSC» and «GFP-ADSC» groups (54.3±8.4% and 55.63±6.8%). Animals that received PBS injection as a negative control were characterized by the highest muscle necrosis span that reached 84±6.7% (Figure 10).

Figure 10. Morphometric analysis of tissue necrosis in ischemic muscle from study group animals.

Figure 10. Morphometric analysis of tissue necrosis in ischemic muscle from study group animals.
A. Images of hematoxylin-eosin stained m. tibialis anterior sections. Necrotic tissue is marked by black line. (N* – necrotic tissue, B* – border zone, H* – healthy or regenerating tissue). B. Representative images of muscle tissue from different zones of section. Labels: star – vasa in normal muscle tissue with; black dot – inflammatory demarcation zone between anucleic disrupted tissue and regenerating muscle fibers; triangle – regenerating round-shaped muscle fibers with multiple centrally located nuclei. C. Statistical data of necrotic tissue area in “PBS”, “ADSC”, “GFP-ADSC” and “VEGF-ADSC” groups. Measurements made in 4–5 animals per group.

To assess vascular density muscle tissue sections were stained by specific antibodies against mouse CD31 and SMA (Figure 11). Vessel count showed that in “ADSC” and “GFP-ADSC” groups capillary and arteriolar densities were similar reaching 129±11 and 125±14 capillaries/FOV, 1.35±0.12 and 1.37±0.09 arterioles/FOV respectively. In specimens from animals that received VEGF-ADSC capillary density was 189±19 per FOV (p<0.05) with arteriolar density of 3.1±0.2 per FOV (p<0.01). Furthermore, we found that arterioles/CD31+ vessels ratio was similar in all experimental groups and slightly higher in group “VEGF-ADSC” (1% vs 1.6%). In addition morphometric analysis of muscle tissue from group “VEGF-ADSC” did not reveal angioma or abnormal vessel formation.

Figure 11. Vascularization of murine ischemic muscles after ADSC administration.

Figure 11. Vascularization of murine ischemic muscles after ADSC administration.
A. Representative images ofm. tibialis anterior sections from “VEGF-ADSC” and “ADSC” groups stained by antibodies against murine CD31 and SMA, 100× magnification. B. Capillaries and arterioles count in m. tibialis anterior sections. Counts made in 5–6 animals per group.

ADSC retain viability and transgene expression after transplantation into ischemic muscle

To evaluate viability of transplanted ADSC after injection into ischemic tissue m. tibialis anterior specimens from “GFP-ADSC” group were harvested at day 7 after induction of ischemia and cell transplantation. Frozen muscle sections were analyzed using fluorescence microscopy that allowed to detect GFP-positive cells distributed throughout muscle (Figure 12A).

1479-5876-11-138-12  Figure 12. Human ADSC viability and VEGF expression

Figure 12. Human ADSC viability and VEGF expression after transplantation to ischemic murine muscle. 

A. Representative image of m. tibialis anterior section from “GFP-ADSC” group obtained at day 7 after ischemia induction and GFP-ADSC injection, 50× magnification. GFP-positive cells are distributed in tissue around injection site.B. Analysis of VEGF165 content by ELISA in explants culture medium from “ADSC”, “GFP-ADSC”, “VEGF-ADSC” groups obtained at days 3 and 20 after cell trasplantation.

Data from experimental studies indicates that prolonged expression of therapeutic transgene is essential for effective stimulation of angiogenesis and ischemic tissue recovery. Muscle explant model was carried out to confirm the presence of viable and functionally active human ADSC overexpressing VEGF in ischemic muscle at hind limb ischemia experiment endpoint. M. tibialis anteriorwere harvested from “ADSC”, “GFP-ADSC” and “VEGF-ADSC” group animals at day 3 and 20 after cell transplantation and cultured as explant in matrigel. In culture medium samples collected after 3 days of “VEGF-ADSC” explant incubation (obtained at day 3 after cell transplantation) human VEGF165 concentration determined by ELISA reached 2.86±0.21 ng/ml (Figure 12B). Protein concentration was expectedly lower (0.145±0.015 ng/ml) in conditioned medium from muscle explants harvested at day 20. In addition comparison of VEGF concentration in culture medium samples collected at day 3 and 7 post incubation of explant culture revealed accumulation of VEGF. It indirectly confirms presence of functionally active human VEGF-ADSC in ischemic muscle up to 20 days post transplantation. In contrast to “VEGF-ASDC” human VEGF165 concentration in explant cultures from “GFP-ADSC” and “ADSC” groups was below limit of detection.


Gene modified cell-based therapy for ischemic disorders: myocardium infarction and limb ischemia is a rapidly evolving trend in experimental and regenerative medicine. Promoting angiogenesis in ischemic tissues via paracrine action of transplanted modified cells is an emerging alternative modality for patients who are unsuitable for surgical and interventional revascularization. Still choice of

  1. appropriate cell type,
  2. angiogenic factor and
  3. gene delivery tool

are crucial issues for efficacy and safety of the method.

Regarding type of cells there are certain issues concerning their derivation and preparation prior to grafting. Thus, embryonic stem cells application is doubtful due to

  1. ethical reasons,
  2. potential risks of teratogenesis and
  3. immune response to their differentiated progenies [28].

Use of endothelial progenitor cells from peripheral blood and bone marrow are limited by

  • expensive procedures of isolation and difficulties in obtaining sufficient amount of cells.

Regarding the latter point it is known that prolonged incubation of cells in vitro prior to transplantation is associated with

  1. potential risks of malignancy,
  2. proliferation decrease and
  3. commitment to terminal differentiation.

Use of skeletal myoblasts or bone marrow derived mesenchymal stromal cells (BMMSC) is associated with painful isolation procedure of muscle biopsy and suprailiac puncture respectively.

ADSC used in our study share a lot of similar properties and characteristics with BMMSC, while they are easier to obtain in sufficient quantity using minimally invasive liposuction procedure. Various data suggests that up to 1.5 × 10adipose stromal cells can be isolated from 1 ml of adipose tissue [29,30]. This allows to reduce the time of cell propagation in vitro prior to transplantation. As for therapeutic angiogenesis,

human ADSC produce a wide spectrum of biologically active molecules – angiogenic growth factors, cytokines, proteases etc. [31,32].

Multiple experimental studies accumulate data on relatively high therapeutic potential of ADSC for tissue regeneration and stimulation of angiogenesis [21,33,34]. However well-known reduction of cell regenerative potential with age and among patients with severe co-morbidities is also relevant for ADSC. Donor age-associated decrease of proliferation activity and differentiation capabilities was shown for human ADSC [35,36]. Angiogenic potential of ADSC also decreases with ageing and is characterized by reduced secretion of

  • VEGF,
  • HGF,
  • angiopoietin-1 and other angiogenic factors [37].

Thus, attempts to improve regenerative potential of ADSC are reasonable.

We have shown high efficacy of rAAV-mediated genetic modification of human ADSC. Using rAAV encoding VEGF165 we obtained human ADSC with increased level of VEGF165 secretion which retained for at least 30 days. VEGF-A and particularly its most abundant 165-amino acid isoform triggers multiple reactions promoting new vessel formation and growth [23] that supported our choice of therapeutic gene in presented study. Observed gradual decrease of transgene expression can be attributed to proliferation activity of ADSC together with known episomal subsistence of rAAV [38]. Moreover cellular mechanism of addressed

  • methylation can be activated after transduction leading to
  • suppression of cytomegalovirus promoter which triggers
  • transgene expression in our vector [39].

Potential influence of genetic modification and transgene expression on cell behavior and functional activity is frequently kept out of consideration while this issue is of great importance, especially for potential clinical application. We examined possible effects of rAAV-transduction and VEGF overexpression on functional properties of ADSC which included

  • proliferation,
  • spontaneous apoptosis,
  • adhesion and
  • differentiation capability.

We observed a decline in ADSC proliferation after modification by rAAV that was evident by

  • increase of population doubling time as well as
  • decrease in number of cells in S–G2 stages of cell cycle.

At the same time spontaneous apoptosis rate did not exceed 2% in modified and unmodified cells. These results contribute to previously published data that showed transient cell cycle arrest after AAV transduction of embryonic fibroblasts and BMMSC [40]. This effect was observed whenever

  • wild-type,
  • recombinant or
  • genome-empty AAV particles were used.

It was suggested that changes in expression profile and decreased proliferation were related to initial stage of virus entry and caused by capsid proteins interaction with cellular signaling pathways [40]. Growth inhibitory effect was transient and

  • proliferation restored to normal level over time of cell passaging [41].

It appears that proliferation decline of rAAV-modified ADSC occurs by a common mechanism.

ADSC are known to be able to differentiate into

  • adipocytes, chondrocytes, osteoblasts, myocytes, neural cells, cardiomyocytes, endothelial and liver cells
  • when cultured in special induction medium [42,43].

Analyzing data from our differentiation experiments we concluded that rAAV-mediated genetic modification of human ADSC and VEGF overexpression did not alter their adipogenic and osteogenic differentiation properties.

There are several observations indicating ability of ADSC for endothelial differentiation [44,45] as well as evidence for presence of small amount of endothelial cells in ADSC population at early passages [18,19]. In our experiments we did not find an increase in amount of cells positive for endothelial markers CD31 and VEGFR2 in VEGF-ADSC compared to unmodified ADSC population. This suggests that VEGF overexpression

  • neither induces endothelial differentiation of modified ADSC
  • or stimulates proliferation of preexisting endothelial cells in ADSC culture.

Adhesion tests conducted in our study were based on a fact that

  • interaction with extracellular matrix proteins is a key factor
  • that contributes to cell viability and integration into host tissue after transplantation [46].

We found that both modified and untreated ADSC showed very common adhesion on collagen type 1, vitronectin and fibronectin. Thus we can suggest that

  • rAAV-mediated genetic modification did not alter expression of adhesion molecules on cell surface of ADSC.

Our results showing low ADSC adhesion on laminin are not surprising taking into account published observations which indicate diminished or lack of α6, α7 and ß1 integrins expression in ADSC-components of α6/ß1 and α7/ß1 receptors for laminin [47,48].

Since VEGF can regulate multiple signaling pathways [23] we next determined whether expression of HGF, FGF2, urokinase and angiopoietin-1 might be altered in VEGF-ADSC. HGF and FGF2 are mitogens and chemoattractants for both endothelial and mural cells and directly participate in angio- and arteriogenesis [4]. Angiopoietin-1 is characterized as a stabilizing factor that provides formation of functionally mature vessel network [49]. Urokinase plasminogen activator is a key regulator of extracellular proteolysis which is

  • responsible for cleavage activation of growth factors and migration of endothelial cells during vessel growth [50,51].

We found almost 3-fold yet not statistically significant increase of urokinase expression while expression of HGF and FGF2 did not change. Another interesting finding is a 5-fold increase of angiopoietin-1 expression in VEGF–ADSC compared to GFP–ADSC or unmodified cells. We assumed that up-regulation of angiopoietin-1 expression occurs due to autocrine action of VEGF165 produced by VEGF-ADSC. However according to our data supported by other studies [30,52] cultured human ADSC population contains <1% of cells that express receptors to VEGF165 – VEGFR1 and VEGFR2. At the same time we found that

  • >90% of ADSC carry receptor to platelet-derived growth factor – PDGFRβ.
There is a published observation that
  • PDGFRα and PDGFRβ can act as a facultative receptor for VEGF [27].
  • it is also known that PDGFR activation leads to increase of angiopoietin-1 expression [53].

Considering that more than 90% of human ADSC are PDGFRβ-positive

  • we can speculate that increased expression of angiopoetin-1 in VEGF-ADSC could be attributed to PDGFRβ-mediated autocrine action of VEGF.
In our study we evaluated therapeutic potential of gene modified human ADSC in terms of their ability to induce angiogenesis in ischemic muscle tissue. It was found that matrigel implants after transplantation of VEGF-ADSC had higher vascular density than after delivery of untreated cells or ADSC transduced by a reporter gene. Along with

  • capillary formation we also found
  • proportional increase in amount of mature blood vessels characterized by smooth-muscle wall.

This can occur due to the fact that cells transplanted in matrigel produce other angiogenic factors besides VEGF that can promote vessel maturation and stabilization.

Key angiogenic property of cell therapies in experimental study is ability to induce reperfusion of ischemic tissue in appropriate animal models. We used hind limb ischemia model to show that

  • VEGF-ADSC transplantation led to significantly higher perfusion restoration than
  • after untreated of GFP-transduced cell administration.

It was also found that intramuscular injection of VEGF-ADSC had a tissue-protective effect and led to vivid decrement of necrosis span. VEGF is known to be significant antiapoptotic factor that can enhance cell survival. We suggest that

  • increased VEGF content during the first days after onset of acute ischemia and cells administration leads to promotion of cell survival and thus to reduction of necrotic disruption in muscle tissue.

We should also point that during the experiment we did not observe any blood flow decrease after cell administration or rapid “plateau” formation like it was previously described for plasmid-mediated gene delivery due to short-term transgene expression [4]. This can be explained by

  • presence of viable and functionally active ADSC that produced VEGF throughout the experiment.

In our muscle explant experiments we showed that VEGF-ADSC retain functional activity even at long terms after injection (up to 27 days) and produce VEGF in detectable quantities. Thus we can confidently attribute

  • tissue protection and restoration of blood flow in mice that received VEGF-ADSC to increased long-term VEGF production by modified cells.

As for decrease of human VEGF content in murine tissue by day 20 we suggest that cells undergo apoptosis over time. Besides that methylation of CMV promoter which drives VEGF expression in our vector could take place. Taking into account that Nude mice were used we find it hard to assume possible rejection of transplanted cells as far as this animal strain lacks T-cells immunity which plays a crucial role in graft rejection. Still, it seems that produced amount of VEGF is sufficient to trigger angiogenesis and relief tissue ischemia via restoration of blood flow.

Histological analysis of ischemic muscle injected with modified VEGF-ADSC revealed that

  • capillary density was significantly higher than in specimens from animals that received untreated cells or GFP-ADSC.

We noticed that this increase was not only due to higher capillary count, but also to SMA-positive blood vessels of arteriolar type. Furthermore arteriole/capillary ratio was constant throughout experimental groups that indicated formation of a stable mature vascular network. Thus, despite high level of VEGF produced by modified ADSC we did not observe any evidence for abnormal tumour-like vascular structures in muscle as it was previously shown e.g. in studies of adenovirus-mediated delivery of VEGF gene [54]. In contrast to matrigel implants experiment in case of skeletal muscle we do not state that increase of vascular density in experimental groups was only due to de novo formed vessels. Besides promoting endothelial cell proliferation VEGF also prevents endothelial apoptosis leading to survival of preexisting vessels. There was surely a vast amount of persisted capillaries in the muscles due to VEGF anti-apoptotic effect of VEGF.

It is often speculated that low efficacy reported in clinical trials using gene delivery of VEGF alone can be explained by its high mitogenic activity which is not supported by vessel stabilizing stimuli and consequently ends up with dissociation of formed capillaries [55]. This led to a concept of combined gene delivery

  • indicating that combinations of angiogenic and vascular stabilizing factors should be used to treat ischemic tissues [5558].

Cell therapy for ischemic disorders has a valuable advantage since transplanted cells produce a whole “cocktail” of biologically active molecules which render combined effect in impaired tissue. We suggest that stable vessel formation observed in our study is

  • mediated by aforementioned ADSC ability to produce a wide spectrum of angiogenic factors
  • including ones responsible for vessel stabilization and maturation: angiopoietin-1, TGF-β, PDGF,
  • which can act synergistically with increased production of VEGF165 by modified cells.

Besides that, genetic modification can alter cell’s expression profile. Observed increase in expression of angiopoietin-1 in VEGF-ADSC can further contribute to

  • formation of mature vascular network that also
  • supports therapeutic effect of transplanted cells.
Increased concentration of VEGF in ischemic tissue plays a substantial role in vessel stabilization and therapeutic effect if maintained over a significant period of time, which was achieved in our study and exceeded a substantial term of 3 weeks.


Thus we can conclude that human ADSC with their accessibility and angiogenic paracrine activity is an appropriate and preferable type of cells for therapeutic angiogenesis. Obtained results indicate that relatively safe rAAV holds great potential for gene transfer into human ADSC. Taken together, we suggest that

  • the use of AAV-modified ADSC overexpressing VEGF165 is a feasible and effective approach for stimulation of stable vascular network formation in ischemic muscle and

can be implied for therapeutic angiogenesis or tissue-engineered transplants. Further study and improvements in vector design, regulated transgene expression, cell preparation and propagation conditions are still to be completed to allow clinical application of modified cell-based therapeuticals.


ADSC: Adipose derived stromal cells; BMMSC: Bone marrow derived mesenchymal stem cells; CGS: Cell growth supplement; DMEM: Dulbecco’s modified Eagle’s medium; ELISA: Enzyme-linked immunosorbent assay; FBS: Fetal bovine serum; FGF2: Fibroblast growth factor 2; FOV: Field of view; GFP: Green fluorescent protein; HEK293T: Human embryonic kidney 293 T; HGF: Hepatocyte growth factor; PDGF: Platelet derived growth factor; PDT: Population doubling time; PBS: Phosphate buffer saline; rAAV: Recombinant adeno-associated virus; SMA: Smooth muscle actin; VEGF: Vascular endothelial growth factor

Read Full Post »

Source of Stem Cells to Ameliorate Damaged Myocardium (Part 2)

Author and Curator: Larry H. Bernstein, MD, FCAP


Curator: Aviva Lev-Ari, PhD, RN


This is Part 2 of a 3 part series of perspectives on stem cell applications to regenerating damaged myocardium.

Progenitor Cell Transplant for MI and Cardiogenesis  (Part 1)
Larry H. Bernstein, MD, FCAP and Aviva Lev-Ari, PhD, RN

Source of Stem Cells to Ameliorate Damage Myocardium (Part 2)
Larry H. Bernstein, MD, FCAP and Aviva Lev-Ari, PhD, RN
https://pharmaceuticalintelligence.com/2013-10-29/larryhbern/Source_of_Stem_Cells_to_Ameliorate_ Damaged_Myocardium/

An Acellular 3-Dimensional Collagen Scaffold  Induces Neo-angiogenesis (Part 3)
Larry H. Bernstein, MD, FCAP and Aviva Lev-Ari, PhD, RN
https://pharmaceuticalintelligence.com/2013-10-29/larryhbern/An_Acellular_3-Dimensional_Collagen_Scaffold _Induces_Neo-angiogenesis/

This series of articles discusses the difficulties we have encountered in adopting stem cell research to clinical therapeutics in regeneration of cardiac tissue damaged post myocardial infarct.  Enormous problems have been encountered in the selection of progenitor cells, the growth into compatible and functional myocardial tissue, and the survival of the myocardium.  Part I went into some detail on a method of obtaining suitable cells, growing them in sheets, and transferring the sheets to the surface for regeneration and repair, which is now going into clinical trials.  Part I will be confined to the importance of source of progenitor cells, whether adult stem cells or umbilical cord blood.

Do Adult Stem Cells Ameliorate the Damaged Myocardium? Human Cord Blood as a Potential Source of Stem Cells

Elise M.K. Furfaro and Mohamed A. Gaballa
Dept Internal Med, Sarver Heart Center, University of Arizona, College of Medicine, Tucson, AZ
Current Vascular Pharmacology, 2007, 5, 27-44  © 2007 Bentham Science Publishers Ltd.

Abstract: The heart does not mend itself after infarction. Cell-based strategies have promising therapeutic potential. Recent clinical and pre-clinical studies demonstrate varying degrees of improvement in cardiac function using different adult stem cell types such as bone marrow (BM)-derived progenitor cells and skeletal myoblasts. However, the efficacy of cell therapy after myocardial infarction (MI) is inconclusive and the cellular source with the highest potential for regeneration is unclear. Clinically, BM and skeletal muscle are the most commonly used sources of autologous stem cells. One major pitfall of using autologous stem cells is that the number of functional cells is generally depleted in the elderly and chronically ill. Therefore, there is an urgent need for a new source of adult stem cells. Human umbilical cord blood (CB) is a candidate and appears to have several key advantages. CB is a viable and practical source of progenitor cells. The cells are naïve and what’s more, CB contains a higher number of immature stem/progenitor cells than BM.

We review recent clinical experience with adult stem cells and explore the potential of CB as a source of cells for cardiac repair following MI. We conclude that there is a conspicuous absence of clinical studies utilizing CB-derived cells and there is a pressing need for large randomized double-blinded clinical trials to assess the overall efficacy of cell-based therapy.

Keywords: Umbilical cord blood, adult stem cell, myocardial infarction, congestive heart failure, human bone marrow, skeletal muscle, angiogenesis


There is an urgent need for new and effective therapy for congestive heart failure (CHF). Heart cells may have a limited capacity to regenerate after myocardial infarction (MI), therefore the use of stem cells for cardiac repair is a logical option. In the past three years, clinical and pre-clinical stud-ies examined the potential of a variety of adult stem cells from different sources as therapy for cardiac disease [1-40]. Adult stem cells are typically chosen in clinical studies be-cause their use avoids the ethical problems associated with embryonic cells. Furthermore, adult stem cells were reported to be pluripotent, capable of differentiating to different cell types [41-45]. Bone marrow-derived hematopoietic stem cells, for example, appear to differentiate into brain cells, skeletal muscle cells, liver cells and cardiomyocytes [42-45]. However, the conclusions of the studies have been recently challenged [10-21, 45].

Regardless of the source, stem cells are difficult to iden-tify because they are hard to distinguish from other cells. No techniques are available to reliably identify stem cells other than surface markers. However, cell surface markers are fickle in that none of them appear to be unique to stem cells. For example, stem and progenitor cells of a varying degree of maturity all express the CD34+ surface marker.. Stem cells are typically recovered by isolation of mononuclear cells (MNCs) and subsequent enrichment for a subset of cells that express certain surface markers such as CD34+ or CD133+, etc. These precursors are commonly sorted using the fluores-cence activated sorting system [1-45].

Direct intramyocardial injection of stem cells into the myocardium is the common route of delivery during surgical intervention. This technique of local delivery of stem/ pro-genitor cells to the myocardium has been shown to be feasi-ble and safe in patients with heart disease [1-4, 10-12, 13, 20, 22, 28]. Other than open-heart surgery, the intra-coronary route appears to be the preferred approach in clinical studies because the stem cells are delivered directly to the affected area without traumatizing the myocardium or submitting the body to the systemic side effects of stem cell mobilization [5-9, 14-19, 21]. A complementary approach to increase the efficiency of progenitor cell transplantation is to enhance cell recruitment and retention in the infarcted heart. For example, stromal cell-derived factor (SDF-1α) has recently been shown to play a critical role in stem cell recruitment to the heart after MI [46].

Although there are other sources of adult stem cells such as adipose tissue [47, 48] and cardiac tissue [49, 50], this review briefly discusses clinical trials using BM stem cells and skeletal muscle myoblasts and pre-clinical studies that used cord blood (CB) cells for heart repair carried out during the past three years. This time period was chosen due to the plethora of excellent published reviews that serve as a foun-dation for this work [51-54]. In addition, the reader may re-fer to several recently published reviews [55-63]. Current clinical experience purports the safety and feasibility of BM stem cells and skeletal muscle myoblasts as autologous cell-therapy for cardiac disease [1-20, 22-30]. However, these cell sources have limitations. For example, recovering sufficient numbers of functional BM progenitor cells is a problem in the elderly and ill [64]. Cardiovascular diseases such as diabetes are associated with BM cell dysfunction [64]. Cardiac calcifications were reported in patients following BM stem cell transplantation [64]. Bone marrow-derived mesenchymal cells (MSCs) have been suggested to play a role in myocardial scarring [64]. Skeletal myoblasts have been associated with arrhythmias and have failed to establish gap junctions with native myocardial cells [64]. Furthermore, the efficacy of these cells in repairing damaged myocardium in clinical settings is still not clear partially due to the lack of protocol standardization as well as the use of adjunct treatment. Different diseases, cell types, cell numbers, routes of cell delivery, end point measurements, and the small number of patients included in these studies make it difficult to draw conclusions about the efficacy of stem cell therapy. Larger clinical trials are now underway to assess the risks and benefits of cell-transplantation using stem cells from BM and skeletal muscle [65].

Another emerging source of stem cells is human umbilical CB. CB has the advantage of being readily available. Numerous CB banks already exist and their number is on the rise [64, 66]. CB is obtained by a non-invasive procedure, and contains a larger portion of immature and non-committed cells than BM. Stem cells derived from CB are expandable ex vivo, appear to be more resistant to apoptosis and the risk of transmission of infection is low [64, 67]. In addition, transplantation of CB cells is associated with a lower incidence and risk of graft-versus-host disease [68, 69]. Similar to previous studies that reported beneficial effects of stem cells isolated from BM and skeletal muscle, CB stem cells also show promise for cardiac repair [1, 3-9, 10¬12, 14, 15, 17-23, 25, 27-29]. Over four thousand CB transplants worldwide have been performed for the treatment of other diseases such as leukemia and immune deficiencies [70]. In contrast, to date, no clinical trials using CB-derived stem cells for transplant after MI have been reported.

The following is an update on recent clinical trials that used BM and skeletal muscle stem cells and preclinical studies that used CB cells to repair the injured myocardium. The emphasis is to evaluate CB as a potential and practical source of stem cells for heart repair after MI.


Being the first cell type used clinically, it seems logical to start by discussing the use of skeletal myoblasts, or skeletal muscle satellite cells, as cell therapy after MI. The advantages of these cells are that they are readily available from muscle biopsies, they are contractile cells, and they can be expanded ex vivo before delivery into the myocardium. Moreover, they appear to have an increased resistance to ischemia [55, 71]. Cell transplantation was usually performed concomitant to revascularization or in patients with previous revascularization [1, 2, 4-6]. Most of the studies used direct injection as the delivery route [1-4]. The number of patients in each study ranged from five to 30 and patients were followed up from 68 days to four years. Except for one study, transplantation of satellite cells was shown to improve left ventricular ejection fraction (LVEF) in all recent clinical studies [1, 3-6].
Several of these studies showed improvement in New York Heart Association (NYHA) class. Interestingly, Pagani et al. showed enhanced angiogenesis after cell transplantation, but they did not measure cardiac function or ventricular remodeling. Unfortunately, it appears that the incidence of arrhythmia and ventricular tachycardia, necessitating the implementation of prophylactic amiodarone or implanted cardioverter defibrillator as an adjunct treatment, is commonplace among these trials [2-6]. Further undermining the clinical use of skeletal myoblasts is the reported lack of cardiomyogenesis and electrical coupling with native cardiac cells that would be necessary to maintain a healthy and functioning heart [55, 72]. Detailed descriptions of these most recent clinical studies using skeletal muscle satellite cells are included in Table 1 (not shown).

[It is not surprising to this reader that the inadequacy of skeletal muscle donor cells is found to be inadequate for maintaining normal cardiac contractility.  Even though contraction of skeletal muscle, smooth muscle, and heart muscle share a basic motif involving CaMKII, the generation of a calcium spark triggering contraction involves a specific relationship between CaMKIIδ and the RyR2 receptor.   CaMKIIδ is specific to the cardiomyocyte.  The other consideration is that the heart is a syncytium, and it has a relationship to neurohumoral control, distinctly different than that in skeletal muscle  This is perhaps the most telling observation in the observed lack of cardiomyogenesis and electrical coupling with native cardiac cells that would be necessary to maintain a healthy and functioning heart [55, 72]].


To date, only small-scale clinical trials, including five to 69 patients, have been performed using bone marrow-derived stem cells (BM-SCs) for transplantation. Three different types of BM-SCs are typically used in recent clinical trials, namely un-fractionated MNCs, CD34+ cells and MSCs. These cells were proposed to treat acute or old MI as well as heart failure [7-21]. Intracoronary injection is the delivery route of choice for these cells [7-9, 14-21]. Revascularization with percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) is commonly used concomitant to cell treatment [13, 15, 16, 18-21]. Several recent trials purported improvement in cardiac function and/or ventricular remodeling three to 12 months after cell treatment [7-11, 15, 17, 18-21]. Some of these studies reported additional enhancement in clinical parameters such as

  • end diastolic (EDV),
  • end systolic volume (ESV)
  • and/or myocardial perfusion [7-9, 10, 17-20].

A small number of studies reported no benefits from BM transplantation [12-14, 16]. In one study, bone marrow transplantation was complicated by coronary artery re-occlusion [21]. The primary endpoint of most of these trials was to assess the safety and feasibility of BM-SC transplantation as a treatment for ischemic heart disease, however these studies are underpowered. In addition, the efficacy of bone marrow cell therapy is difficult to ascertain from clinical studies, at least in part, due to common utilization of adjunct therapy such as revascularization. More detailed descriptions of bone marrow clinical studies are found in Tables 2-5 (not shown).


Since transplantation of autologous BM-SCs leads to improvement in cardiac function, mobilization of BM-SCs using cytokines to increase the number of circulating cells was utilized in succeeding studies. Granulocyte colony stimulating factor (G-CSF) is the most common cytokine used to mobilize BM-SCs in clinical studies [22-31]. The feasibility and safety of G-CSF has been reported by several investigators. The number of patients in the G-CSF studies ranged from five to 114 and they were followed for up to 52 weeks. Clinical studies in the last three years have shown that cardiac function improved in about half of the trials using G-CSF to mobilize BM-SCs [22, 23, 27-29]. The remaining half of G-CSF studies reported no effects on cardiac function [24-26, 30, 31]. In one study, an unexpected reduction in LVEF was reported [31]. Adverse effects of G-CSF treatment were reported in almost all the recent clinical studies [22, 24-27, 29, 31]. Detailed descriptions of G-CSF stud¬ies are shown in Tables 6-7.


Amidst the flurry of clinical studies utilizing BM and skeletal muscle SCs, it is a wonder why no trials are reported using CB cell transplantation in humans. However, several pre-clinical studies using various animal models demonstrated the potential use of CB stem cells for cardiac repair after MI [32-40]. Conserved commonalities of cardiac function improvement exist in these studies despite dissimilarity of protocols used [32-40]. The following is a description of the pre-clinical studies which used different subsets of CB-derived stem cells to treat MI. In this review, the pre-clinical studies are categorized according to the type of stem cell administered.

We first start with studies that used CB-derived MNCs. Ma et al. reported that intravenous injection of six million CB-MNCs into non-obese diabetic severe combined immunodeficiency (NOD/SCID) mice 24 h post-MI resulted in an increase in capillary density and decrease in both infarct size and collagen deposition three weeks after treatment [38]. No myogenesis was observed. Human DNA was identified in 10 out of the 19 mice that underwent induction of MI. Direct myocardial injection of one-sixth of the amount of cells used in the above study in rats also reduced infarct size and increased both ventricular wall thickness and LVdP/dt and ejection fraction up to six months after treatment [34].

Similar to CB-MNCs, transplantation of two hundred thousand CD34+ cells, a subset of MNCs, within 20 min after MI

  • increased vascular density,
  • reduced LV dilation, and
  • improved cardiac function four weeks after treatment [35].

However, only about one percent of the injected cells were incorporated into the vessels of the rat myocardium, which suggests that angiogenic factors released by these cells may contribute to the observed angiogenesis [35]. A subset of CD34+, CD34+ KDR+ cell fraction, was proposed to be the subset of cells responsible for angiogenesis induction and improvement in cardiac function after treatment with either MNCs or CD34+ cells [32]. Two hundred thousand of either CD34+ or MNCs, or two thousand of either CD34+ KDR+ or CD34+ KDR- cells were injected in a NOD/SCID mouse model of MI. Compared to transplantation of MNCs or PBS, CD34+ cells

  • increased LVdP/dt,
  • decreased LV end diastolic pressure and
  • infarct size up to five months after MI.

Treatment with two thousand CD34+KDR+ cells, which is two log less than the number of CD34+ cells, resulted in more
angiogenesis compared to either MNC or CD34+ [32].

An immature subset of CB-MNCs, CD133+ cells, were also reported to improve cardiac function after transplantation into MI mice [37]. One to two million CD133+ cells were intravenously injected into athymic nude rats seven days after MI. Four weeks after transplantation,

  • reduction in both scar thinning and
  • LV systolic dilation, and
  • increase in LV fractional shortening were observed.

In contrast to other studies, vessel density did not differ between the cell-treated and control rats [37]. Similarly, transplantation of a subset of these immature CD133+ cells, CD34+ CD133+ cells, into a mouse model of hindlimb ischemia resulted in angiogenesis induction [40]. Transplantation of one hundred thousand CD34+ CD133+ cells into ischemic limbs of immunosup-pressed mice increased both vessel and muscle fiber densities fourteen days after injection. In contrast, administration of CD34+ cells resulted in increased vessel density only. Neither of these findings was observed after administration of CD34- cells [40].

An alternative subset of progenitor cells, called endothelial progenitor cells (EPCs) from either CB or adult peripheral blood (PB), was also shown to induce angiogenesis in ischemic hindlimb [39]. EPCs were derived from MNC CD34+ cells and identified in culture as attaching cells that exhibit spindle-shape. These cells

  1. incorporated acetylated-low density lipoprotein,
  2. released nitric oxide, and
  3. expressed KDR, VE-cadherin, CD31, and vW factor and CD45-.

Not only were the CB-derived EPCs more abundant (10 fold increase) than those derived from PB, they also further in-creased capillary density when injected into ischemic tissue [39].

Finally, another CB-derived cell subset, denoted as human unrestricted somatic stem cells (USSCs), was shown to engraft in the infarcted heart and improve cardiac perfusion [36]. USSCs were defined as negative for the following surface markers:

  • CD14, CD31, CD33, CD34, CD45, CD56, CD133 and human leukocyte antigen class II and
  • positive for CD13, CD29, CD44, and CD49e.

In a porcine model of MI, one hundred million USSCs were directly injected into the infarcted heart four weeks after MI.

  1. Regional perfusion,
  2. LVEF,
  3. scar thickness, and
  4. wall motion increased four weeks after transplantation [36].

In addition to cell transplantation alone, the combination of gene and cell therapy was shown to be a potential treatment for MI [33]. For example,

CD34+ cells transduced with the adeno associated viral vector that encoded either human angiopieotin-1 or vascular endothelial growth factor (VEGF) were intramyocardially injected in a mouse model of MI. Improved cardiac function and increased capillary density were observed with CD34+ cells alone.

However, exaggerated improvements were obtained with the combined therapy of CD34+ cells transfected with Angiopieotin-1 and or VEGF. Compared with CD34+ treatment alone,

  • the combined therapy further increased capillary density and decreased infarct size [33].

Taken together, based on the pre-clinical studies, a common feature of transplantation of human CB-derived cells is

  • induction of angiogenesis and cardiac function improvement in animal models of ischemia.

Myogenesis does not seem to be a mechanism of the beneficial effects of CB transplantation.

Compared with adult stem cells, CB cell treatment has limitations. The practical and crucial difference between stem cells obtained from adult human donors and from CB is quantitatively, not qualitatively based. It is uncommon that more than several million stem cells can be isolated from CB. That amount may be too small for transplantation to an adult. Children appear to be ideal recipients when utilizing this source of stem cells since they are smaller patients and require fewer cells per kilogram of body weight [71]. However, ex vivo expansion of these cells may overcome this limitation [73, 74]. There is another concern that the use of CB for transplantation presents a higher risk of transmitting opportunistic infections [75]. The human herpes viruses are common pathogens found in transplant recipients. Currently, it is routine to test for the presence of anti-cytomegalovirus immunoglobin M. However, screening prospective CB donors for these pathogens reduces the risk of transmission of infection [75].

(Tables from published document are to be viewed in that document.)


Although early clinical studies suggest that bone marrow and skeletal myoblast transplantation into the infarcted heart improves cardiac perfusion and function, there is an urgent need for large randomized double-blinded clinical trials that assess the overall efficacy of cell-based therapy. In addition, little is known about the mechanisms by which stem cells render their positive effects. Cardiac regeneration by bone marrow cells is an obvious mechanism. However, a small number of experimental studies have purported the occurrence of myocardial regeneration by bone marrow cells. Furthermore, substantial evidence demonstrates that cell types other than cardiomyocytes improve cardiac function, suggesting that the beneficial effects of cell therapy may be independent of cardiac regeneration [76-89]. Enhanced vascularization, on the other hand, is a common finding after bone marrow cell transplantation. Cell engrafment to the vascular wall as well as angiogenic factors released by transplanted cells may be responsible for the enhanced vascularization. Obviously, there remain a considerable number of unanswered questions that must be addressed in basic science laboratories before stem cell therapy becomes standard practice. For example, what are the mechanisms of improvement in cardiac function? Which cell type is best-suited for transplantation? What is the optimal cell concentration that should be used for transplant and what is the most effective route of delivery?

The target patient population which would draw clinical benefit from cell-based therapy must also be defined and the optimal time of injection after the onset of infarction has to be determined. Currently, it is difficult to assess the efficacy of stem cell treatment of MI. This is in part due to lack of standardization among clinical as well as pre-clinical studies. Therefore, in order to accomplish these objectives, there is great need for communication among the various research groups concerned with stem cells and clinical studies.

Here we add yet another source of stem cells, namely the umbilical CB. This source of stem cells had many advantages mentioned in the preceding sections. In addition, pre-clinical studies indicate the efficacy of CB cells in myocardial repair. However, the fate and benefits of these cells need to be tested in clinical settings.


[1]  Herreros J, Prosper F, Perez A, Gavira JJ, Garcia-Velloso MJ, Barba J, et al. Autologous intramyocardial injection of cultured skeletal muscle-derived stem cells in patients with non-acute myocardial infarction. Eur Heart J 2003; 24: 2012-20.

[2]  Pagani FD, DerSimonian H, Zawadzka A, Wetzel K, Edge AS, Jacoby DB, et al. Autologous skeletal myoblasts transplanted to ischemia-damaged myocardium in humans. Histological analysis of cell survival and differentiation. J Am Coll Cardiol 2003; 41: 879-88.

[3]  Smits PC, van Geuns RJ, Poldermans D, Bountioukos M, Onder-water EE, Lee CH, et al. Catheter-based intramyocardial injection of autologous skeletal myoblasts as a primary treatment of ischemic heart failure: clinical experience with six-month follow-up. J Am Coll Cardiol 2003; 42: 2063-9.

[4]  Dib N, Michler RE, Pagani FD, Wright S, Kereiakes DJ, Lengerich R, et al. Safety and feasibility of autologous myoblast transplantation in patients with ischemic cardiomyopathy: four-year follow-up. Circulation 2005; 112: 1748-55.

[5]  Siminiak T, Kalawski R, Fiszer D, Jerzykowska O, Rzezniczak J, Rozwadowska N, et al. Autologous skeletal myoblast transplantation for the treatment of postinfarction myocardial injury: phase I clinical study with 12 months of follow-up. Am Heart J 2004; 148: 531-7.

[6]  Siminiak T, Fiszer D, Jerzykowska O, Grygielska B, Rozwadowska N, Kalmucki P, et al. Percutaneous trans-coronary-venous trans-plantation of autologous skeletal myoblasts in the treatment of post-infarction myocardial contractility impairment: the POZNAN trial. Eur Heart J 2005; 26: 1188-95.

[7]  Assmus B, Schachinger V, Teupe C, Britten M, Lehmann R, Dobert N, et al. Transplantation of Progenitor Cells and Regeneration Enhancement in Acute Myocardial Infarction (TOPCARE-AMI). Circulation 2002; 106: 3009-17.

[8]  Schachinger V, Assmus B, Britten MB, Honold J, Lehmann R, Teupe C, et al. Transplantation of progenitor cells and regeneration enhancement in acute myocardial infarction: final one-year results of the TOPCARE-AMI Trial. J Am Coll Cardiol 2004; 44: 1690-9.

[9]  Britten MB, Abolmaali ND, Assmus B, Lehmann R, Honold J, Schmitt J, et al. Infarct remodeling after intracoronary progenitor cell treatment in patients with acute myocardial infarction (TOPCARE-AMI): mechanistic insights from serial contrast-enhanced magnetic resonance imaging. Circulation 2003; 108: 2212-8.

[10]  Perin EC, Dohmann HFR, Borojevic R, Silva SA, Sousa ALS, Mesquita CT, et al. Transendocardial, autologous bone marrow cell transplantation for severe, chronic ischemic heart failure. Circulation 2003; 107: 2294–302.

Human umbilical cord blood stem cells, myocardial infarction (and stroke)

Nathan Copeland, David Harris and Mohamed A Gaballa
Nathan Copeland, Research Associate and Medical Student, University of Arizona Medical School, Tucson, Arizona; David Harris, Professor of Microbiology and Immunology, University of Arizona, Tucson, Arizona; Mohamed A Gaballa, Director, Center for Cardiovascular Research, Sun Health Research Institute, Sun City, Arizona; Section Chief of Basic Science, Cardiology Section, Banner GoodSam Medical Center, Phoenix, Arizona
Clinical Medicine 2009, Vol 9, No 4: 342–5

ABSTRACT – Myocardial infarction (MI) and stroke are the first and third leading causes of death in the USA accounting for more than 1 in 3 deaths per annum. Despite interventional and pharmaceutical advances, the number of people diagnosed with heart disease is on the rise. Therefore, new clinical strategies are needed. Cell-based therapy holds great promise for treatment of these diseases and is currently under extensive preclinical as well as clinical trials. The source and types of stem cells for these clinical applications are questions of great interest. Human umbilical cord blood (hUCB) appears to be a logical candidate as a source of cells. hUCB is readily available, and presents little ethical challenges. Stem cells derived from hUCB are multipotent and immunologically naive. Here is a critical literature review of the beneficial effects of hUCB cell therapy in preclinical trials.
KEY WORDS: animal models, cerebral infarction, myocardial infarction, stem cells, umbilical cord blood


The study of stem cell therapies to address some of the most daunting medical challenges, including heart disease and stroke, has advanced steadily over the last three years. The majority of preclinical studies of stem cells as a potential therapy for either myocardial or cerebral ischaemia were positive on average. Small clinical trials, however, show either no or modest improvement in cardiac function after myocardial infarction (MI). Currently, there are two major types of autologous cells that are clinically used for MI and stroke. The first is skeletal myoblasts, harvested from skeletal muscle. These cells can be expanded in culture. Positive outcomes were recently reported in a phase 1 clinical trial using catheter-based injection of myoblasts to the endocardium (CAUSMIC, American Heart Association (AHA) Scientific Sessions 2007). The second is bone marrow cells (BMCs). Intracoronary injection of BMCs improve global left ventricular function (IC-BMC, AHA Scientific Sessions 2007). However, direct injection of BMC administration into scarred myocardium does not alter cardiac contractility of the injured area (IC/IM-BMC, AHA Scientific Sessions 2007). The effects of stem cell therapy can only be addressed using clinical trials that:

•             are randomised, blinded, placebo controlled and adequately sized

•             use standardisation of autologous stem cell processing protocols

•             use robust endpoints of efficacy and safety

•             ensure that follow-up is complete and of adequate duration.

It is becoming clear that realisation of the full potential of the therapeutic benefit of stem cells will require understanding the biology of these undifferentiated cells. A successful therapy will require a source with plentiful supply of multipotent stem cells with minimal or no immune rejection. Several sources of stem cells were explored such as

  • adipose tissue,1–3
  • cardiac tissue,4
  • skeletal muscle biopsies,5,6 and
  • hUCB.

Whether these subpopulations of cells are best suited to treat a disease is still unanswered.

Currently, the only confirmed source for totipotential cells is embryonic. However, there are ethical and scientific obstacles to unbridled use of such cells. For clinical application, autologous adult stem cells are the obvious choice. To date, only adult stem cells derived from a patient’s own bone marrow are being used in clinical trials.

Autologous BMC therapy is not without problems. The majority of instances of MI and cerebral ischaemia (CI) occur in the elderly. Since the quantity and function of BMCs decrease with age, an allogeneic younger donor may be used to source BMCs. This may hinder the efficiency of such a treatment and suffer rejection, therefore another source of stem cells is needed.

Cryopreserved stem cells derived from human leukocyte antigen (HLA)-matched and unmatched unrelated donor hUCB were realised as a sufficient source of transplantable hematopoietic stem cells with high donor-derived engraftment and low risk of refractory acute graft-versus-host disease. However, the use of hUCB cells as treatment for either MI or CI has only been recently investigated in preclinical models.

There are several outstanding review articles on stem cells derived from cord blood in MI7–11 and stroke.12–17 This article adds depth to the debate by providing an updated review as well as presenting an integrated overview of studies involving MI and CI cell-based therapy. In the preparation of this review, every effort was made to include all relevant publications since 2005. Due to space limitations, the number of articles cited has been limited.

Cardiovascular disease

Since 2005, several studies have explored the use of various sub-populations of hUCB stem cells for regenerative therapy. Five types of UCB-derived stem cells were investigated: umbilical cord derived stem (UCDS), unrestricted somatic stem cells (USSC), mononuclear progenitor cells (MNCs), CD133+ and CD34+ subpopulations. The experimental parameters of the studies varied. The majority of studies, however, were performed using the rat animal model and utilising the left antero-lateral descending (LAD) coronary artery ligation model of MI with intramyocardial injection of the stem cells. The laboratory used a similar model to determine the efficacy of stem cell derived from hUCB to improve cardiac function after ischemia and reperfusion. The data indicated that intracoronary administration of mononuclear or CD34+ cells derived from hUCB improved cardiac function after MI by inducing neovascularisation and retarding left ventricular (LV) remodelling.37

The majority of reported studies using hUCB cells showed improvement in the outcomes.18–25 Cardiac functional improvements were almost universally reported as evaluated by:

  • increased ejection fraction;
  • improved wall motion;
  • lowered LV end-diastolic pressure; and
  • increased cardiac contraction as determined by the maximum slope of LV pressure.18–21,23–25

There were conflicting reports on the effects of stem cells on LV fractional shorting. One study reported improved shortening while another reported that BM but not UCB cells produced improved shortening.22,23 Improvements in

  • myocardial perfusion, evaluated by increased capillary density, were repeatedly demonstrated as were
  • reductions in infarct size and the number of apoptotic cells.18–25

Retardation or reduction in LV remodelling were also reported.18,21,22 Although the vast majority of studies showed positive outcomes, HLA matching and further study are still needed before UCB stem cell therapies can become safe and effective treatments in humans. A prime example of the need for further elucidation of these emerging therapies can be illustrated by the findings in a study by Moelker.26 This study used intracoronary administration of unrestricted somatic stem cells (USSCs) in a balloon left circumflex artery (LCX) occlusion ischaemia-reperfusion porcine model of MI. They found that treatment did not improve outcome and actually increased infarct size. Their histological analysis revealed that the injected cells worsened the infarct by obstructing vessels downstream.

Furthermore, the mechanisms of the observed benefits of UCB stem cell therapy in MI are under investigation:

  • improved myocardial perfusion,
  • attenuation of cardiac remodelling,
  • reduction of inflammatory responses by
    • limiting expression of TNF-a, MCP-1, MIP and INF–y, and cardiac regeneration.18–5

Tissue regeneration may be mediated by incorporation of delivered cells in the target tissue.18–21,23 An in vitro study confirmed that mononuclear cells were migrated toward homogenised infarcted myocardium and that the greatest migration occurred at two and 24 hours post-MI.20 Paracrine effect, ie the delivered cells release factors that promote neovascularisation, was also reported. Indeed, the study laboratory has shown that hUCB cells release angiogenic factors in vitro under hypoxic conditions. The data are consistent with a previous report that showed

  • increased expression of VEGF 164 and 188 accompanied by
  • angiogenesis and improved remodelling after administration of hUCB mononuclear cells into the myocardium.21

Identifying subpopulations of progenitor cells with the highest potential for tissue repair is another unanswered ques¬tion prior to widespread application of this therapy in clinical settings. Previous studies showed that UCB-derived endothe¬lial progenitor cells (EPC) to be a promising subset of stem cells for treatment of MI; however their number may be insufficient to treat adult patients. This problem can be addressed by expanding these cells in culture prior to transplant. Techniques are being developed to culture clinically significant quantities (60 population doublings) of EPCs from UBC CD.25 Transplantation of these expanded cells improved ejection fraction (EF) and vascular density in vivo, demonstrating that such a culture method may be a viable option to produce EPCs for future use in humans. Another study evaluated the use of gene therapies in conjunction with UCB stem cell therapy.24 CD34+ cells were transfected with AAV-Ang1 and/or AAV-VEGF 165. The gene-modified stem cells resulted in greater increases in capillary density and cardiac performance along with larger reduction in infarct size compared to CD34+ cell therapy alone.


1 Valina C, Pinkernell K, Song YH et al. Intracoronary administration of autologous adipose tissue-derived stem cells improves left ventricular function, perfusion, and remodelling after acute myocardial infarction. Eur Heart J 2007;28:2667–77.

2 Zhang DZ, Gai LY, Liu HW et al. Transplantation of autologous adipose-derived stem cells ameliorates cardiac function in rabbits with myocardial infarction. Chin Med J (Engl) 2007;120:300–7.

4 Hoogduijn MJ, Crop MJ, Peeters AM et al. Human heart, spleen, and perirenal fat-derived mesenchymal stem cells have immunomodulatory capacities. Stem Cells Dev 2007;16:597–604.

5 Payne TR, Oshima H, Okada M et al. A relationship between vascular endothelial growth factor, angiogenesis, and cardiac repair after muscle stem cell transplantation into ischemic hearts. J Am Coll Cardiol 2007;50:1677–84.

6 Herreros J, Prósper F, Perez A et al. Autologous intramyocardial injection of cultured skeletal muscle-derived stem cells in patients with non-acute myocardial infarction. Eur Heart J 2003;24:2012–20.

7 Goldberg JL, Laughlin MJ, Pompili VJ. Umbilical cord blood stem cells: implications for cardiovascular regenerative medicine. J Mol Cell Cardiol 2007;42:912–20.

8  Wu KH, Yang SG, Zhou B et al. Human umbilical cord derived stem cells for the injured heart. Med Hypotheses 2007;68:94–7.

9 Zhang L, Yang R, Han ZC. Transplantation of umbilical cord blood-derived endothelial progenitor cells: a promising method of therapeutic revascularisation. Eur J Haematol 2006;76:1–8.

18 Wu KH, Zhou B, Yu CT et al. Therapeutic potential of human umbil¬ical cord derived stem cells in a rat myocardial infarction model. Ann Thorac Surg 2007;83:1491–8.

19 Kim BO, Tian H, Prasongsukarn K et al. Cell transplantation improves ventricular function after a myocardial infarction: a preclinical study of human unrestricted somatic stem cells in a porcine model. Circulation 2005;112:I96–104.

20 Henning RJ, Burgos JD, Ondrovic L et al. Human umbilical cord blood progenitor cells are attracted to infarcted myocardium and sig-nificantly reduce myocardial infarction size. Cell Transplant 2006;15:647–58.

21 Hu CH, Wu GF, Wang XQ et al. Transplanted human umbilical cord blood mononuclear cells improve left ventricular function through angiogenesis in myocardial infarction. Chin Med J (Engl)  2006;119:1499–506.

22 Ma N, Ladilov Y, Moebius JM et al. Intramyocardial delivery of human CD133+ cells in a SCID mouse cryoinjury model: Bone marrow vs. cord blood-derived cells. Cardiovasc Res 2006;71:158–69.

23 Leor J, Guetta E, Feinberg MS et al. Human umbilical cord blood-derived CD133+ cells enhance function and repair of the infarcted myocardium. Stem Cells 2006;24:772–80.

24 Chen HK, Hung HF, Shyu KG et al. Combined cord blood stem cells and gene therapy enhances angiogenesis and improves cardiac perfor-mance in mouse after acute myocardial infarction. Eur J Clin Invest 2005;35:677–86.


Read Full Post »