Our results demonstrate that ADSCs isolated from human earlobe fat express MSC-specific markers and can be differentiated into neural cells via bFGF- and forskolin-dependent pathways. In addition, NI-hADSCs have neural markers and functional neuron-like characteristics.
In the present study, the phenotypic expression of hADSCs is consistent over culture passages and the morphological features are the same as those previously reported [12, 34, 48, 49]. Adipose tissue-derived stem cells are understood to express surface markers of CD9, CD10, CD13, CD29, CD44, CD49d, CD49e, CD54, CD55, CD59, CD73, CD90, CD105, CD146, CD166, and STRO-1, and lack hematopoietic lineage markers CD11b, CD14, CD19, CD34, and CD45 [20, 34, 50–53]. The hADSCs expressed MSC-specific cell type markers including CD13, CD44, CD90, and CD166, however, did not express CD14, CD34, and CD45 indicating that hADSCs in this study were of mesenchymal origin. Adipose tissue-derived stem cells also seem to possess the capacity to differentiate into multiple mesodermal lineages such as bone, fat, and cartilage [31, 34]. The current study supports this hypothesis, characterizing the expression of multiple lineage-specific genes and proteins including adipocytes, osteoblasts, and chondrocytes. This observation has led us to speculate that adipose tissue may be a valuable source of mesodermal stem cells.
Reports indicate that the neural differentiation of ADSCs is achieved with different experimental protocols, Protocols include using chemical agents like β-mercaptoethanol[34, 35], a mix of valproic acid, butylated hydroxyanisole, insulin, hydrocortisone [38, 39, 54], azacytidine [55, 56], or a cocktail of isobutylmethylxanthine, indomethacin, and insulin [48, 57, 58]. A mixture of glial growth factors , a mixture including bFGF and platelet-derived growth factor , or brain-derived neurotrophic factor (BDNF) with retinoic acid  also evoked the neural induction of ADSCs in vitro. In addition, rodent ADSCs are differentiated into Schwann cell-like cells by a procedure that involves making floating neurospheres . In bone marrow-derived MSCs, neural differentiation was induced via transfection of proneural genes , treatment with bFGF, forskolin and ciliary neurotrophic factor (CNTF)  or co-culture with neural cells [2, 46]. In the present study, we used bFGF and forskolin to induce the neural differentiation of hADSCs because bFGF is known to generate neural precursor cells with a greater capacity for neuronal differentiation [9, 26, 60]. Contrary to bFGF, epidermal growth factor and cilliary neurotrophic factor are reported to restrict astrocyte lineages. Forskolin is a commonly used agent to increase the intracellular levels of cyclic adenosine monophosphate (cAMP) by activating the enzyme adenylyl cyclase. Furthermore, forskolin is reported to induce the neuron-like morphology and expression of NSE, NFH, and Tuj1 in human MSCs cultured in serum-free conditions [61, 62]. Our experiments demonstrate that NI-hADSCs express increased immunoreactivities for neuronal markers Tuj1, MAP2, NFL, NFM, NFH, NSE, NeuN, GAP43, and SNAP25, as well as the increased mRNA expression of Tuj1, MAP2, NFL, NFM, NSE, GAP43, and SNAP25 compared to primary hADSCs (Figure 3), indicating that hADSCs differentiate into neural cells via bFGF and forskolin-mediated differentiation. Much like neuronal cells derived from other MSCs or embryonic stem cells [63, 64], in vitro-transdifferentiated hADSCs also exhibited neuronal cell properties.
According to previous reports about neurogenic differentiation, ADSCs exhibit neuron-like morphology and express several proteins and genes consistent with the neuronal phenotype [20, 35, 54, 57–59, 65]. However, reliance on neural marker expression as an indicator of neurally differentiated MSCs has become unreliable because undifferentiated MSCs express several neural markers at both the mRNA and protein levels [42, 66]. Adult mesenchymal stem cells constitutively express native immature neural proteins (nestin and Tuj1), whereas more mature neuronal and glial proteins (tyrosine hydroxylase, MAP2, and GFAP) are expressed in increasing passage numbers [67, 68]. Undifferentiated ADSCs also express markers characteristic of neural cells such as NSE, vimentin, and NeuN . In addition, inter-donor variability of expression of neural marker genes in MSC samples needs to be considered . We therefore performed electrophysiological studies to investigate whether NI-hADSCs demonstrate the functional properties of mature neurons. Through patch-clamp recordings, NI-hADSCs were identified to generate prominent TTX-sensitive voltage-dependent sodium currents and outward potassium currents, both of which are hallmarks of mature neurons and crucial for signal transmission in the nervous system. In contrast to bone marrow-derived MSCs, in which combination of bFGF, forskolin, and CNTF was not sufficient to induce voltage-dependent sodium current , hADSCs expressed the sodium current by treatment of bFGF and forskolin. Furthermore, NI-hADSCs exhibited about -58 mV of resting membrane potential, indicating that they also have functional characteristics of neurons.
Recently, Anghileri et al. reported electrophysiological evidence of neuronal differentiation . After differentiation with BDNF and retinoic acid, hADSC exhibited immunocytochemical evidence of neuronal differentiation in only 57% of cells and mean peak amplitude of approximately -189 pA for voltage-dependent Na+ currents in differentiated hADSCs. Ashjian et al. also demonstrated that supplementation with isobutylmethylxnthine (IBMX), indomethacin and insulin induced transdifferentiation of human processed lipoaspirate cells into neuron-like cells . However, they could not observe inward sodium currents. According to our immunocytochemical studies, our results indicate that more than 80% of hADSCs appear to differentiate into neuron-like cells under specific in vitro culture conditions with bFGF and forskolin, and express several proteins specific to the neuronal phenotype that exhibit neuronal morphology (Figure 3a, 3b). Furthermore, approximately 75% of hADSCs demonstrate neural differentiation properties under electrophysiological study, with about -605 pA of voltage-dependent sodium currents being recorded in NI-hADSCs. These results suggest that bFGF and forskolin may be more effective than BDNF, retinoic acid, IBMX, indomethacin, and insulin in inducing the neural differentiation of ADSCs.
Although ADSCs have been used for years in the investigation of cell replacement therapy and differentiation, information about ion channel expression remains undocumented. Undifferentiated bone marrow-derived hMSCs are known to express the TTX-sensitive sodium channel gene (NE-Na), potassium channel genes (MaxiK, Kv1.4, Kv4.2, Kv4.3, and Eag1) and the calcium channel gene (CACNA1C) [56, 57]. However, this study indicates that primary hADSCs also express ion channel mRNAs, including potassium channel genes (MaxiK, Kv1.4, Kv4.2, and Eag2), calcium channel genes (CACNA1C and CACN1G), and the TTX-insensitive sodium channel gene (SCN5A), but do not exhibit the TTX-sensitive sodium channel gene (NE-Na) and other voltage-dependent potassium channel genes (Kv4.3 and Eag1) (Figure 4d). Since the primary hADSCs did not display voltage-dependent sodium currents prior to neural differentiation, gene expression results are consistent with electrophysiological data. However, neural induction with bFGF and forskolin increased the expression of these ion channel genes, particularly those expressed in the primary hMSCs, and induced three novel functional ion channel genes (NE-Na, Kv4.3, and Eag1), indicating the differentiation of hADSCs towards neuronal cells. This result is an initial finding on the expression of ionic channel genes in both primary- and NI-hADSCs.