Cardiovascular system

We have previously looked at the cardiomyogenic potential of AFSCs in vitro and in vivo. AFSCs cultured in cardiomyocyte induction media or in co-culture with cardiomyocytes demonstrated expression of proteins specific for cardiomyocytes (atrial natriuretic peptide and α-myosin heavy chain), endothelial (CD31, CD144) and smooth muscle cells (α-smooth muscle actin). In our first experience with xenogeneic transplantation, human AFSCs were transplanted in a rat model of myocardial infarction (MI). Cells of the immune system were recruited including T cells, B cells, natural killer (NK) cells and macrophages resulting in cell rejection. We speculated that this may be due to AFSC expression of the B7 co-stimulatory molecules CD80 and CD86, as well as macrophage marker CD68 . Next, we attempted allogeneic rat AFSC cardiac therapy by intracardiac transplantation in rats with ischaemia–reperfusion injury (IR). A portion of the cells acquired an endothelial or smooth muscle phenotype, and a smaller number had cardiomyocyte characteristics, 3 weeks following transplantation. The left ventricular ejection fraction was improved in animals that received the AFSC injection, as quantified using magnetic resonance imaging (MRI), suggesting a paracrine therapeutic effect . The aim was to further investigate this paracrine effect using a rat model of MI and xenogeneic transplantation of human AFSCs administered intravascularly immediately following reperfusion. This was dissimilar to our previous attempt with xenogeneic cellular cardiomyoplasty, which involved intramuscular injection of human AFSCs within 20 min of coronary artery occlusion without reperfusion. Intravascular injection of human AFSCs and their conditioned medium showed a cardioprotective effect, improved cell survival and decreased the infarct size from 54% to around 40% in both cases. We have also shown that AFSCs secrete the actin monomer-binding protein thymosin beta-4 (Tβ-4), a paracrine factor with cardioprotective properties . Tβ-4 had also previously been implicated in cardioprotection in MI models that involved bone marrow-derived mesenchymal stem cell (BMMSC) injection .

In addition to models of myocardial IR injury, we have investigated the salutary effects of AFSCs in a rat model of right heart failure secondary to pulmonary hypertension. Following intravascular injection, AFSCs engrafted in the lungs, heart and skeletal muscle, reducing the levels of brain natriuretic peptide (BNP), a surrogate marker for heart failure, and pro-inflammatory cytokines. In addition, AFSCs differentiated into endothelial and vascular cells forming micro-vessels, capillaries and small arteries. A 35% decrease in pulmonary arteriole thickness accompanied the injection .

Gastrointestinal system

We have recently looked at the effect of AFSCs in a rat model of necrotising enterocolitis (NEC) that involved hypoxia, hypothermia, gavage feeding with a hyperosmolar formula, hypoxia and administration of lipopolysaccharide. After 24 h of life, NEC rats were randomised to treatments of either AFSCs, BMMSCs, myoblasts (as a committed negative control) or phosphate-buffered saline (PBS) via intraperitoneal injection. NEC rats treated with AFSCs showed significantly higher survival at 7 days in comparison to all the other groups, and had an improved NEC clinical status at 96 h. The MRI scans displayed significantly decreased peritoneal fluid accumulation (a surrogate marker for NEC grade) in the AFSC-treated rats. The improved clinical picture of the pups injected with AFSCs was also evident by measurement of intestinal permeability, contraction and motility.

The observational data were confirmed by histological analysis, demonstrating a decreased amount of villus sloughing, core separation and venous congestion. The relationship between these therapeutic effects and the presence of AFSCs in the intestine was confirmed by tracking AFSCs expressing green fluorescent protein (GFP). The cells were adherent to the mesentery at 48 h, found in the serosa and muscularis at 72 h and located in the villi at 96 h. The low cell numbers in these locations alongside the great clinical differences between treated groups suggested a paracrine effect. Accordingly, when we performed microarray analysis, we observed differences in a number of genes involved in inflammation and tissue repair, cell-cycle regulation and enterocyte differentiation. These results were corroborated by immunofluorescence analysis examining cell proliferation and apoptosis.

We then sought to investigate the paracrine mechanism by which AFSCs mediate their therapeutic effect, establishing that the number of cryptal cells expressing cyclo-oxygenase-2 (COX-2 ⁺ ) inversely correlated with the degree of intestinal damage. COX-2 ⁺ cells were diminished in rats treated with PBS, whereas they were maintained in rats treated with AFSCs. Interestingly, even though the total number of COX-2 ⁺ cells in villi was similar in AFSC-treated and control animals, cryptal COX-2 ⁺ cells were significantly increased in the AFSC rats in comparison to both control and pups treated with PBS. This dependence on COX-2 was confirmed when the effect of AFSCs was abolished by both selective COX-2 and non-selective COX inhibition (COX-1 and COX-2), but remained unaffected by a selective COX-1 inhibitor.

Haematopoietic system

Ditadi and colleagues were the first to demonstrate the haematopoietic potential of murine and human CD117 positive/lineage negative AFSCs . In vitro, the AFSC population in both species displayed multi-lineage haematopoietic potential, as demonstrated by the generation of erythroid, myeloid and lymphoid cells (haematopoietic lineages). In vivo, cells belonging to all haematopoietic lineages were found after primary and secondary transplantation of murine AFSCs into immunocompromised hosts, thus demonstrating the long-term haematopoietic repopulating capacity of these cells. The latter results support the idea that the AF may be a source of stem cells with potential for therapy of haematological disorders.

One of the most exciting applications of AFSCs in this setting is in the field of in utero transplantation (IUT). IUT has been proposed as a novel approach for the treatment of inherited haematological disorders (including thalassaemia and sickle cell disease) before birth . Clinical translation to date has been limited by competitive and immunological barriers associated with IUT of adult bone-marrow-derived haematopoietic stem cells (BMHSCs) . The use of AFSCs for IUT could address many of these limitations. AFSCs are of foetal origin and, as a result, should be able to compete better against host cells in comparison to adult stem cells (potentially overcoming competitive barriers to engraftment). Due to the tolerogenic properties of the placenta, AFSCs are non-immunogenic to the foetus at any gestational age and are also unlikely to result in maternal immunisation. The IUT of AFSCs would involve harvesting the cells from the AF, in vitro gene therapy to correct the genetic defect and transplantation back to the donor foetus. Such a combined autologous stem cell–gene transfer approach would also address some of the risks associated with administering gene therapy directly to the foetus (in utero gene therapy; IUGT) . The possibility of performing in vitro gene transfer to harvested ASFCs would allow cells to be checked for insertional mutagenesis prior to transplantation, and it would obviate the risk of germ-line transmission of transgenes. In recent proof-of-principle studies, Shaw and colleagues showed that in utero transplantation of autologous (isolated using ultrasound-guided amniocentesis), expanded and transduced AFSCs resulted in widespread tissue engraftment (including the haematopoietic system) in the ovine foetus . At present, we are investigating the haematopoietic potential of freshly isolated and expanded AFSCs following intravenous transplantation in immunocompetent foetal mice, and we have obtained stable, multi-lineage engraftment at near-therapeutic levels using relatively small donor cell numbers (Loukogeorgakis et al.; unpublished data). However, despite encouraging preliminary data, whether in utero stem cell–gene therapy with AFSCs would be therapeutic in models of haematological and other congenital disorders remains to be determined.

Musculoskeletal system

We have investigated the osteogenic potential of AFSCs after they were cultured in a medium containing dexamethasone, β-glycerophosphate and ascorbic acid-2-phosphate. Following seeding in a collagen/alginate scaffold, they were implanted subcutaneously in immunodeficient mice. At 18 weeks, micro-computed tomography (CT) revealed highly mineralised tissues and blocks of bone-like material . In a study by Sun et al. osteogenic differentiation of human amniotic fluid stem cells (hAFSCs) was achieved using bone morphogenetic protein-7 (rhBMP-7) and seeding on nanofibrous scaffolds, as evident by alkaline phosphatase (ALP) activity, calcium content, von Kossa staining and the expression of osteogenic genes. Implantation into the subcutaneous space led to bone formation in 8 weeks with positive von Kossa staining and a radioopaque profile on X-ray .

A series of experiments by the Goldberg laboratory investigated osteogenesis of AFSCs following seeding on a poly-(ε-caprolactone) (PCL) biodegradable polymer. Cells that were differentiated in a three-dimensional (3-D) PCL scaffold deposited mineralised matrix and were viable after 15 weeks of culture. It was also shown that pre-differentiated cells in vitro produced seven times more mineralised matrix when implanted subcutaneously in vivo. The authors obtained some remarkable results when comparing AFSCs and BMMSCs for osteogenesis following seeding on scaffolds and long-term in vitro culture. Although BMMSCs differentiated more rapidly than AFSCs, their growth and production of mineralised matrix halted at 5 weeks. In sharp contrast, AFSCs continued producing matrix for up to 15 weeks, thus leading to an overall production that was five times larger in comparison to BMMSC-seeded scaffolds. In another comparison, PCL scaffolds coated with adeno-associated viral vector encoding bone morphogenetic protein 2 (scAAV2.5-BMP-2) were seeded with AFSCs and BMMSCs. Whereas BMP-2 was shown to have increased production in both populations after 2 weeks in vitro, it was only the BMMSCs that had significantly increased mineral formation. It was interesting how both acellular-coated and BMMSC-seeded scaffolds led to increased mineral formation and mechanical properties when used to cover large femoral defects.

More recently, we demonstrated for the first time the functional and stable long-term integration of AFS cells into the skeletal muscle of HSA-Cre SmnF7/F7 mutant mice, which closely replicate the clinical features of human muscular dystrophy . AFSCs were obtained from E11.5–13.5 GFP ⁺ mice through immediate CD117 selection after AF collection. Approximately, 25,000 freshly isolated AFSCs were directly injected into the tail vein of each animal without previous expansion in culture. Transplanted mice displayed enhanced muscle strength, improved survival rate by 75% and restored muscle phenotype in comparison to untreated animals. Not only was the dystrophin distribution in GFP ⁺ myofibres similar to that of wild-type animals but GFP+ cells were also found engrafted into the muscle stem cell niche as demonstrated by their sub-laminal position and Pax7 and alpha7integrin expression. Functional integration of AFSCs in the stem cell niche was confirmed by successful secondary transplants of GFP ⁺ satellite cells derived from AFSC-treated mice into untreated SmnF7/F7 mutant mice. In order to progress towards their application for therapy, the therapeutic potential of cultivated AFSCs was also investigated, and 25,000 AFSCs previously expanded in the culture were intravenously injected into SmnF7/F7 mice. Despite maintaining their therapeutic potential, cultivated AFSCs regenerated approximately 20% of the recipient muscle fibres versus 50% when using freshly isolated AFSCs, thus, highlighting the importance of optimising cell expansion conditions . In summary, the available data support the potential of AFSCs (1) to differentiate towards the myogenic lineage; (2) to participate in muscle regeneration concerning animal models with muscle injury; and (3) to engraft and participate in the muscle stem cell niche. Therefore, AFSCs constitute a promising therapeutic option for musculoskeletal and muscle degenerative diseases. Understanding the process of AFSCs’ myogenesis in vivo and improving cell expansion in vitro will be essential requisites before this could happen.

Nervous system

We have previously investigated human AFSC injection in the brain of twitcher mice (model of Krabbe globoid leucodystrophy, associated with progressive oligodendrocyte and neuronal loss). Human AFSCs engrafted into the lateral cerebral ventricles, differentiated to cells similar with the surrounding environment and survived for up to 2 months. It was also demonstrated that engraftment was variable, with 70% of AFSCs surviving in the brain of twitcher mice, in contrast to only 30% of AFSCs shown to survive in the brain of normal mice . A study by Prasongchean et al. indicated that treatment with small molecules that normally lead to neuronal differentiation and grafting of AFSCs into environments such as organotypic rat hippocampal cultures and the embryonic chick nervous system led to no expression of neural cell markers. However, AFSCs reduced haemorrhage and increased survival in a chick embryo model of extensive thoracic crush injury. Survival rates did not improve through mesenchymal cells, neural cells or desmopressin. The authors explained this effect in association with the secretion of paracrine factors as evidenced by a trans-well co-culture model .

Respiratory system

AFSCs have been shown to have plastic regenerative properties in the lungs, by differentiating into different lineages according to the type of lung injury taking place in animal models of disease. AFSCs injected intravascularly into nude mice subjected to hyperoxia-induced pulmonary injury, migrated to the lung and expressed the human pulmonary epithelial differentiation marker thyroid transcription factor 1 (TTF1) and type-II pneumocyte marker surfactant protein C (SPC). Following naphthalene injury to Clara cells, AFSCs expressed the Clara cell-specific 10-kDa protein .

Beneficial effects of cell therapy with AFSCs have also been reported in the setting of lung hypoplasia, with potential applications in patients with congenital diaphragmatic hernia or prematurity. A study by Pederiva and colleagues showed, in the rat nitrofen model of pulmonary hypoplasia, that early prenatal administration of expanded AFSCs improved lung growth, bronchial motility and innervation, both in vitro and in vivo . The salutary effects of AFSC transplantation were related to release and paracrine activity of growth factors including fibroblast growth factor 10 (FGF10) and vascular endothelial growth factor alpha (VEGF-α).

Urinary system

The first evidence in terms of a nephrogenic potential of AFSCs arose from a series of experiments involving the ex vivo growth of murine embryonic kidneys that were injected with labelled AFSCs. Whilst AFSCs were viable for up to 10 days of growth, they were also shown to contribute to a number of components of the developing kidneys, such as the renal vesicle, S- and C-shape bodies. In addition, the extracellular matrix (ECM) and surrounding cells induced renal differentiation, with the AFSCs expressing kidney markers (zona occludens-1, glial-derived neurotrophic factor and claudin) .

In a mouse model of acute tubular necrosis (ATN) that involved glycerol injection, luciferase-labelled AFSCs that were injected intra-renally homed to the kidney. This decreased creatinine and blood urea nitrogen (BUN) levels and reduced the number of damaged tubules, whilst increasing proliferation of tubular epithelial cells. Interestingly, AFSCs injected during the acute phase of ATN (between 48 and 72 h) had no effect on creatinine and BUN levels, whereas AFSCs injected into the kidneys on the same day of glycerol injection resulted in no observed peaks in creatinine or BUN. The authors speculated that this may be due to AFSCs accelerating the proliferation of partially damaged epithelial tubular cells, whilst, in addition, preventing apoptosis . Another laboratory confirmed the protective effect of AFSCs in the same mouse model of ATN, whilst comparing their effect with MSCs. In addition to results in terms of amelioration concerning the effect of glycerol, MSCs were reported to be more efficient in inducing proliferation, and AFSCs were more anti-apoptotic. To assess the effect of AFSCs in renal fibrosis, Sedrakyan et al. used Col4a5(-/-) mice as a murine model of Alport syndrome . Early intracardiac administration of AFSCs delayed interstitial fibrosis and progression of glomerular sclerosis, and prolonged animal survival. However, AFSCs did not exhibit differentiation into podocytes, suggesting a paracrine mechanism that mediated the positive effects on the basement membrane .

Cardiovascular system

We have previously looked at the cardiomyogenic potential of AFSCs in vitro and in vivo. AFSCs cultured in cardiomyocyte induction media or in co-culture with cardiomyocytes demonstrated expression of proteins specific for cardiomyocytes (atrial natriuretic peptide and α-myosin heavy chain), endothelial (CD31, CD144) and smooth muscle cells (α-smooth muscle actin). In our first experience with xenogeneic transplantation, human AFSCs were transplanted in a rat model of myocardial infarction (MI). Cells of the immune system were recruited including T cells, B cells, natural killer (NK) cells and macrophages resulting in cell rejection. We speculated that this may be due to AFSC expression of the B7 co-stimulatory molecules CD80 and CD86, as well as macrophage marker CD68 . Next, we attempted allogeneic rat AFSC cardiac therapy by intracardiac transplantation in rats with ischaemia–reperfusion injury (IR). A portion of the cells acquired an endothelial or smooth muscle phenotype, and a smaller number had cardiomyocyte characteristics, 3 weeks following transplantation. The left ventricular ejection fraction was improved in animals that received the AFSC injection, as quantified using magnetic resonance imaging (MRI), suggesting a paracrine therapeutic effect . The aim was to further investigate this paracrine effect using a rat model of MI and xenogeneic transplantation of human AFSCs administered intravascularly immediately following reperfusion. This was dissimilar to our previous attempt with xenogeneic cellular cardiomyoplasty, which involved intramuscular injection of human AFSCs within 20 min of coronary artery occlusion without reperfusion. Intravascular injection of human AFSCs and their conditioned medium showed a cardioprotective effect, improved cell survival and decreased the infarct size from 54% to around 40% in both cases. We have also shown that AFSCs secrete the actin monomer-binding protein thymosin beta-4 (Tβ-4), a paracrine factor with cardioprotective properties . Tβ-4 had also previously been implicated in cardioprotection in MI models that involved bone marrow-derived mesenchymal stem cell (BMMSC) injection .

In addition to models of myocardial IR injury, we have investigated the salutary effects of AFSCs in a rat model of right heart failure secondary to pulmonary hypertension. Following intravascular injection, AFSCs engrafted in the lungs, heart and skeletal muscle, reducing the levels of brain natriuretic peptide (BNP), a surrogate marker for heart failure, and pro-inflammatory cytokines. In addition, AFSCs differentiated into endothelial and vascular cells forming micro-vessels, capillaries and small arteries. A 35% decrease in pulmonary arteriole thickness accompanied the injection .

Gastrointestinal system

We have recently looked at the effect of AFSCs in a rat model of necrotising enterocolitis (NEC) that involved hypoxia, hypothermia, gavage feeding with a hyperosmolar formula, hypoxia and administration of lipopolysaccharide. After 24 h of life, NEC rats were randomised to treatments of either AFSCs, BMMSCs, myoblasts (as a committed negative control) or phosphate-buffered saline (PBS) via intraperitoneal injection. NEC rats treated with AFSCs showed significantly higher survival at 7 days in comparison to all the other groups, and had an improved NEC clinical status at 96 h. The MRI scans displayed significantly decreased peritoneal fluid accumulation (a surrogate marker for NEC grade) in the AFSC-treated rats. The improved clinical picture of the pups injected with AFSCs was also evident by measurement of intestinal permeability, contraction and motility.

The observational data were confirmed by histological analysis, demonstrating a decreased amount of villus sloughing, core separation and venous congestion. The relationship between these therapeutic effects and the presence of AFSCs in the intestine was confirmed by tracking AFSCs expressing green fluorescent protein (GFP). The cells were adherent to the mesentery at 48 h, found in the serosa and muscularis at 72 h and located in the villi at 96 h. The low cell numbers in these locations alongside the great clinical differences between treated groups suggested a paracrine effect. Accordingly, when we performed microarray analysis, we observed differences in a number of genes involved in inflammation and tissue repair, cell-cycle regulation and enterocyte differentiation. These results were corroborated by immunofluorescence analysis examining cell proliferation and apoptosis.

We then sought to investigate the paracrine mechanism by which AFSCs mediate their therapeutic effect, establishing that the number of cryptal cells expressing cyclo-oxygenase-2 (COX-2 ⁺ ) inversely correlated with the degree of intestinal damage. COX-2 ⁺ cells were diminished in rats treated with PBS, whereas they were maintained in rats treated with AFSCs. Interestingly, even though the total number of COX-2 ⁺ cells in villi was similar in AFSC-treated and control animals, cryptal COX-2 ⁺ cells were significantly increased in the AFSC rats in comparison to both control and pups treated with PBS. This dependence on COX-2 was confirmed when the effect of AFSCs was abolished by both selective COX-2 and non-selective COX inhibition (COX-1 and COX-2), but remained unaffected by a selective COX-1 inhibitor.

Haematopoietic system

Ditadi and colleagues were the first to demonstrate the haematopoietic potential of murine and human CD117 positive/lineage negative AFSCs . In vitro, the AFSC population in both species displayed multi-lineage haematopoietic potential, as demonstrated by the generation of erythroid, myeloid and lymphoid cells (haematopoietic lineages). In vivo, cells belonging to all haematopoietic lineages were found after primary and secondary transplantation of murine AFSCs into immunocompromised hosts, thus demonstrating the long-term haematopoietic repopulating capacity of these cells. The latter results support the idea that the AF may be a source of stem cells with potential for therapy of haematological disorders.

One of the most exciting applications of AFSCs in this setting is in the field of in utero transplantation (IUT). IUT has been proposed as a novel approach for the treatment of inherited haematological disorders (including thalassaemia and sickle cell disease) before birth . Clinical translation to date has been limited by competitive and immunological barriers associated with IUT of adult bone-marrow-derived haematopoietic stem cells (BMHSCs) . The use of AFSCs for IUT could address many of these limitations. AFSCs are of foetal origin and, as a result, should be able to compete better against host cells in comparison to adult stem cells (potentially overcoming competitive barriers to engraftment). Due to the tolerogenic properties of the placenta, AFSCs are non-immunogenic to the foetus at any gestational age and are also unlikely to result in maternal immunisation. The IUT of AFSCs would involve harvesting the cells from the AF, in vitro gene therapy to correct the genetic defect and transplantation back to the donor foetus. Such a combined autologous stem cell–gene transfer approach would also address some of the risks associated with administering gene therapy directly to the foetus (in utero gene therapy; IUGT) . The possibility of performing in vitro gene transfer to harvested ASFCs would allow cells to be checked for insertional mutagenesis prior to transplantation, and it would obviate the risk of germ-line transmission of transgenes. In recent proof-of-principle studies, Shaw and colleagues showed that in utero transplantation of autologous (isolated using ultrasound-guided amniocentesis), expanded and transduced AFSCs resulted in widespread tissue engraftment (including the haematopoietic system) in the ovine foetus . At present, we are investigating the haematopoietic potential of freshly isolated and expanded AFSCs following intravenous transplantation in immunocompetent foetal mice, and we have obtained stable, multi-lineage engraftment at near-therapeutic levels using relatively small donor cell numbers (Loukogeorgakis et al.; unpublished data). However, despite encouraging preliminary data, whether in utero stem cell–gene therapy with AFSCs would be therapeutic in models of haematological and other congenital disorders remains to be determined.

Musculoskeletal system

We have investigated the osteogenic potential of AFSCs after they were cultured in a medium containing dexamethasone, β-glycerophosphate and ascorbic acid-2-phosphate. Following seeding in a collagen/alginate scaffold, they were implanted subcutaneously in immunodeficient mice. At 18 weeks, micro-computed tomography (CT) revealed highly mineralised tissues and blocks of bone-like material . In a study by Sun et al. osteogenic differentiation of human amniotic fluid stem cells (hAFSCs) was achieved using bone morphogenetic protein-7 (rhBMP-7) and seeding on nanofibrous scaffolds, as evident by alkaline phosphatase (ALP) activity, calcium content, von Kossa staining and the expression of osteogenic genes. Implantation into the subcutaneous space led to bone formation in 8 weeks with positive von Kossa staining and a radioopaque profile on X-ray .

A series of experiments by the Goldberg laboratory investigated osteogenesis of AFSCs following seeding on a poly-(ε-caprolactone) (PCL) biodegradable polymer. Cells that were differentiated in a three-dimensional (3-D) PCL scaffold deposited mineralised matrix and were viable after 15 weeks of culture. It was also shown that pre-differentiated cells in vitro produced seven times more mineralised matrix when implanted subcutaneously in vivo. The authors obtained some remarkable results when comparing AFSCs and BMMSCs for osteogenesis following seeding on scaffolds and long-term in vitro culture. Although BMMSCs differentiated more rapidly than AFSCs, their growth and production of mineralised matrix halted at 5 weeks. In sharp contrast, AFSCs continued producing matrix for up to 15 weeks, thus leading to an overall production that was five times larger in comparison to BMMSC-seeded scaffolds. In another comparison, PCL scaffolds coated with adeno-associated viral vector encoding bone morphogenetic protein 2 (scAAV2.5-BMP-2) were seeded with AFSCs and BMMSCs. Whereas BMP-2 was shown to have increased production in both populations after 2 weeks in vitro, it was only the BMMSCs that had significantly increased mineral formation. It was interesting how both acellular-coated and BMMSC-seeded scaffolds led to increased mineral formation and mechanical properties when used to cover large femoral defects.

More recently, we demonstrated for the first time the functional and stable long-term integration of AFS cells into the skeletal muscle of HSA-Cre SmnF7/F7 mutant mice, which closely replicate the clinical features of human muscular dystrophy . AFSCs were obtained from E11.5–13.5 GFP ⁺ mice through immediate CD117 selection after AF collection. Approximately, 25,000 freshly isolated AFSCs were directly injected into the tail vein of each animal without previous expansion in culture. Transplanted mice displayed enhanced muscle strength, improved survival rate by 75% and restored muscle phenotype in comparison to untreated animals. Not only was the dystrophin distribution in GFP ⁺ myofibres similar to that of wild-type animals but GFP+ cells were also found engrafted into the muscle stem cell niche as demonstrated by their sub-laminal position and Pax7 and alpha7integrin expression. Functional integration of AFSCs in the stem cell niche was confirmed by successful secondary transplants of GFP ⁺ satellite cells derived from AFSC-treated mice into untreated SmnF7/F7 mutant mice. In order to progress towards their application for therapy, the therapeutic potential of cultivated AFSCs was also investigated, and 25,000 AFSCs previously expanded in the culture were intravenously injected into SmnF7/F7 mice. Despite maintaining their therapeutic potential, cultivated AFSCs regenerated approximately 20% of the recipient muscle fibres versus 50% when using freshly isolated AFSCs, thus, highlighting the importance of optimising cell expansion conditions . In summary, the available data support the potential of AFSCs (1) to differentiate towards the myogenic lineage; (2) to participate in muscle regeneration concerning animal models with muscle injury; and (3) to engraft and participate in the muscle stem cell niche. Therefore, AFSCs constitute a promising therapeutic option for musculoskeletal and muscle degenerative diseases. Understanding the process of AFSCs’ myogenesis in vivo and improving cell expansion in vitro will be essential requisites before this could happen.

Nervous system

We have previously investigated human AFSC injection in the brain of twitcher mice (model of Krabbe globoid leucodystrophy, associated with progressive oligodendrocyte and neuronal loss). Human AFSCs engrafted into the lateral cerebral ventricles, differentiated to cells similar with the surrounding environment and survived for up to 2 months. It was also demonstrated that engraftment was variable, with 70% of AFSCs surviving in the brain of twitcher mice, in contrast to only 30% of AFSCs shown to survive in the brain of normal mice . A study by Prasongchean et al. indicated that treatment with small molecules that normally lead to neuronal differentiation and grafting of AFSCs into environments such as organotypic rat hippocampal cultures and the embryonic chick nervous system led to no expression of neural cell markers. However, AFSCs reduced haemorrhage and increased survival in a chick embryo model of extensive thoracic crush injury. Survival rates did not improve through mesenchymal cells, neural cells or desmopressin. The authors explained this effect in association with the secretion of paracrine factors as evidenced by a trans-well co-culture model .

Respiratory system

AFSCs have been shown to have plastic regenerative properties in the lungs, by differentiating into different lineages according to the type of lung injury taking place in animal models of disease. AFSCs injected intravascularly into nude mice subjected to hyperoxia-induced pulmonary injury, migrated to the lung and expressed the human pulmonary epithelial differentiation marker thyroid transcription factor 1 (TTF1) and type-II pneumocyte marker surfactant protein C (SPC). Following naphthalene injury to Clara cells, AFSCs expressed the Clara cell-specific 10-kDa protein .

Beneficial effects of cell therapy with AFSCs have also been reported in the setting of lung hypoplasia, with potential applications in patients with congenital diaphragmatic hernia or prematurity. A study by Pederiva and colleagues showed, in the rat nitrofen model of pulmonary hypoplasia, that early prenatal administration of expanded AFSCs improved lung growth, bronchial motility and innervation, both in vitro and in vivo . The salutary effects of AFSC transplantation were related to release and paracrine activity of growth factors including fibroblast growth factor 10 (FGF10) and vascular endothelial growth factor alpha (VEGF-α).

Urinary system

The first evidence in terms of a nephrogenic potential of AFSCs arose from a series of experiments involving the ex vivo growth of murine embryonic kidneys that were injected with labelled AFSCs. Whilst AFSCs were viable for up to 10 days of growth, they were also shown to contribute to a number of components of the developing kidneys, such as the renal vesicle, S- and C-shape bodies. In addition, the extracellular matrix (ECM) and surrounding cells induced renal differentiation, with the AFSCs expressing kidney markers (zona occludens-1, glial-derived neurotrophic factor and claudin) .

In a mouse model of acute tubular necrosis (ATN) that involved glycerol injection, luciferase-labelled AFSCs that were injected intra-renally homed to the kidney. This decreased creatinine and blood urea nitrogen (BUN) levels and reduced the number of damaged tubules, whilst increasing proliferation of tubular epithelial cells. Interestingly, AFSCs injected during the acute phase of ATN (between 48 and 72 h) had no effect on creatinine and BUN levels, whereas AFSCs injected into the kidneys on the same day of glycerol injection resulted in no observed peaks in creatinine or BUN. The authors speculated that this may be due to AFSCs accelerating the proliferation of partially damaged epithelial tubular cells, whilst, in addition, preventing apoptosis . Another laboratory confirmed the protective effect of AFSCs in the same mouse model of ATN, whilst comparing their effect with MSCs. In addition to results in terms of amelioration concerning the effect of glycerol, MSCs were reported to be more efficient in inducing proliferation, and AFSCs were more anti-apoptotic. To assess the effect of AFSCs in renal fibrosis, Sedrakyan et al. used Col4a5(-/-) mice as a murine model of Alport syndrome . Early intracardiac administration of AFSCs delayed interstitial fibrosis and progression of glomerular sclerosis, and prolonged animal survival. However, AFSCs did not exhibit differentiation into podocytes, suggesting a paracrine mechanism that mediated the positive effects on the basement membrane .