ESCs exhibit pluripotency and differentiate into the three germ layers: ectoderm, mesoderm, and endoderm. These cells are derived from the ICM layer of the embryo's blastocyst. Mouse ESC was first derived in 1981 by Evans and Kaufman[33] in the United Kingdom and Martin[34] in the United States from the ICM of the blastocyst (d2.5). Human ESCs (hESCs) were derived by Thomson and colleagues isolated from preimplantation blastocysts[22]. hESCs have been an excellent source of pluripotent cells for therapeutic use[35]. Derivation of pluripotent ESCs from the blastocyst’s ICM layer is usually done by a standardized immunosurgery technique[36]. The ESCs are isolated and seeded on feeder layers in culture plates. The cells may be characterized for pluripotency markers by immunostaining using specific antibodies to octamer-binding transcription factor ¾ (Oct3/4), stage-specific embryonic antigen 3 (SSEA-3), SSEA-4, TRA-1-60, and TRA-1-81 or by assessing alkaline phosphatase activity. ESCs are karyotypically normal and possess a high telomerase activity[37,38]. However, there are severe concerns in using ESC in regenerative medicine despite being a promising candidate. Severe ethical concerns prevail in using human embryos for the isolation of hESCs[39]. Different legal guidelines are governing ESC research in various countries. In the United States, the destruction of human embryos for any form of research is banned. According to the guidelines, hESC-lines derived before August 9, 2001 can be used for research. The development of hESC therapies is restricted, and most research studies are focused on animals[40]. In the United Kingdom, research using hESCs derived from discarded embryos in in vitro fertilization clinics is allowed; however, hESC research is prohibited in Italy[41-43].

It is also essential to analyze the safety issues associated with ESCs in regenerative medicine. The ESCs can differentiate into any cell type of the body. However, when these undifferentiated cells are implanted in vivo, this plasticity poses the risk of developing teratomas and tumors[44-47]. The alternative method is to differentiate the undifferentiated ESCs in vitro into a specific cell type along a lineage and then transplant the differentiated cell in vivo. hESC-derived cardiomyocytes, when transplanted in mice, did not result in teratoma[48]. However, in certain instances, the transplanted progenitor cells continue to proliferate, as, for example, nestin+ dopaminergic neurons derived from hESCs continue to proliferate in the striatum[49]. The screening of the undifferentiated cells by specific markers and their subsequent purification before transplantation may solve the problem.

There are some reported methods to overcome the risk of tumorigenesis, such that ESCs can be used for regenerative therapies[50]. One of the recent findings involves cluster of differentiation 133 (CD133) (prominin 1), a transmembrane protein generally expressed on cancer stem cells is highly expressed on hESCs. CD133-deficient knock-out hESC line retained the capacity to differentiate into the three embryonic germ layers in vivo. Still, the proliferating potential is reduced and results in reduced teratoma formation[51]. Therefore, CD133 may be used to sort ESCs for transplantation[52].

Despite the safety issues, hESC-derived progenitor cells are still considered promising candidates in regenerative medicine under controlled conditions[53]. The first approval of the hESC trial for spinal cord injury was received in 2009, in which hESC-derived oligodendrocytes progenitor cells were used[54]. The hESC-based clinical trials have been performed for the treatment of macular degeneration with some positive results in follow-up studies[55-58], diabetes mellitus[59], and ischemic heart disease[60]. One clinical trial has been approved for Type 1 diabetes produced by the company ViaCyte (ClinicalTrials.gov Identifier: NCT03163511, NCT02239354)[61]. These pancreatic progenitor cells are produced from human pluripotent stem cells in vitro and differentiate into beta cells after transplantation in an immune isolation device in vivo[62]. Clinical trials for Parkinson’s disease with hESCs are being conducted in Australia (NCT02452723) and China (NCT03119636)[63].

iPSCs were first successfully generated by Takahashi and Yamanaka[64] in 2006 by inserting the reprogramming factors known as “Yamanaka factors,” Oct4/3, Sox2, Klf4, and c-Myc in mouse fibroblast cells. Initially, Yamanaka and his group (Okita et al[65]) started transducing mouse fibroblasts with a recombinant retrovirus carrying 24 genes responsible for maintaining the ESC characteristics. The mouse fibroblasts were selected by antibiotic-resistant gene cassette under the promoter, Fbx15, which is active only in the ESCs. The number of genes was narrowed down to ten and finally to four genes, Oct3/4, c-Myc, Sox2, and Klf4. The first generation induced pluripotent cells selected by Fbx15 possessed unlimited self-renewal and differentiation capacity, produced embryoid bodies and fetal chimeras but failed to produce adult chimeras. However, the DNA methylation pattern, post-translational modifications, and epigenetic changes revealed that the generated iPSCs were intermediate between fibroblasts and ESCs. Yamanaka and his group further generated second-generation iPSCs using the selection for Nanog instead of Fbx15 selection. The second-generation iPSCs showed greater ES cell-like characteristics, DNA methylation pattern, and germline competence[65]. However, 20% of the chimeric mice developed cancer as two of the genes, c-Myc, and Klf4, are oncogenic. Human-induced pluripotent cells were generated from somatic cells in 2007 by two independent groups simultaneously by introducing Oct3/4 and Sox2 with either Klf4 and c-Myc or Nanog and Lin28[66]. The latter group has shown that reprogramming of human somatic cells is possible even when the reprogramming factors are not integrated into the genome. The use of non-integrated episomal vectors makes these cells more suitable for clinical use[67]. iPSCs have also been derived from peripheral blood mononuclear cells[68]. Adenovirus has also been used vector for the delivery of reprogramming factors.

However, virus-mediated delivery systems sometimes threaten iPSCs' clinical use due to insertional mutagenesis, mainly caused by cMyc[69-71]. RNA, proteins, and small molecules enhance iPSCs' efficiency and safety. Reprogramming of somatic cells by mRNA or microRNA became very successful and effective[72-75]. It has been reported that activation of the innate immune system enhances the efficiency of induced pluripotent cells by mRNA transfection[76]. Insertion of recombinant proteins such as Oct 4, sex-determining region Y-box 2 (Sox2), Kruppel-like factor 4 (Klf4), and cMyc can be used to reprogram somatic cells to induce pluripotent cells[77-79]. The protein-based approach has been used to generate dopaminergic neurons from iPSCs to treat Parkinson’s disease in rats[80].

The iPSCs have potential similar to that of ESCs, and additionally overcome the ethical concerns associated with ESC research and clinical use. The iPSCs have gained much attention in recent years because of their advancement in regenerative medicine, organoid formation, and scope for personalized therapies. The first human clinical trial with iPSCs was done for macular degeneration in which retinal pigmented epithelial cells generated from autologous iPSCs were transplanted into a patient[81]. There have been several in vitro[82,83] and in vivo preclinical studies investigating the safety and efficacy of iPSCs[84-87]. A human clinical trial was recently conducted to treat Parkinson’s disease in Kyoto, Japan, by transplanting dopaminergic neurons generated from iPSCs. The clinical-grade iPSCs are produced from cells taken from healthy volunteers[88]. Autologous iPSC-derived dopaminergic neurons transplanted in a patient with Parkinson’s disease retained the function until 2 years without any adverse effect[89]. Autologous iPSCs are more advantageous to avoid immunological rejection, but the development of autologous iPSC is time-consuming and costly.

The iPSC technology has enormous potential in regenerative medicine. However, more interventional studies[90] must be conducted to address the challenges of routine clinical applications of these cells e.g., genomic instability[91], carcinogenicity[92], immunological rejection[93]. Humanized mouse models, e.g., a mouse with human immune cells, may be developed in the future to investigate the immunogenicity of human pluripotent stem cells[94].

iTS are produced by incomplete reprogramming of somatic cells by transient overexpression of reprogramming factors by plasmids and performing a tissue-specific selection. These cells have the potential to self-renew but also express tissue-specific markers. iTS have been produced from mouse pancreatic cells, which can self-renewal and express pancreatic tissue-specific transcription factor, Pdx1[13]. These generated iTS can differentiate into insulin-producing cells more efficiently than ESCs or iPSCs and, probably, can be utilized to treat diabetes. iTS with neural stem cell-like characteristics have also been reported[95-98]. Teratoma formation is not reported when iTS are transplanted in nude mice. In this respect, iTS are advantageous over the ESCs or iPSCs in terms of clinical applications due to the risk of tumorigenicity associated with the use of pluripotent stem cells. These cells are “incompletely reprogrammed” cells and have different methylation patterns than ESCs or iPSCs. The iTS retain the donor tissue’s epigenetic memory and can be explored as a potential candidate for cell replacement therapy.

Fetal stem cells are collected from aborted fetal tissues and extraembryonic structures like amniotic fluid, umbilical cord, Wharton’s jelly, and placenta[99]. The fetal stem cells are multipotent HSCs or MSCs. HSCs collected from fetal bone marrow or umbilical cord express CD34 and CD45 Like adult HSCs but show greater proliferating capacity, low immunogenicity, and lower risk of graft vs host disease (GvHD) compared to adult HSCs. Fetal MSCs can be isolated from fetal blood, bone marrow, liver, lung, or pancreas. These cells have more differentiation capacity than the adult MSCs. Fetal MSCs have active telomerase and express low levels of human leukocyte antigen (HLA) I and lack intracellular HLA II[100]. First trimester fetal MSCs express baseline levels of pluripotent stem cell markers such as Oct4, Nanog, Rex1 SSEA3, SSEA4, Tra-1-60, and Tra-1-81. Umbilical cord blood MSCs are easy to harvest and can be stored under controlled conditions for longer periods for future clinical use[101]. Although the ethical concerns associated with fetal stem cells are minimal, there are several reports on the adverse effects of fetal stem cells in vitro in animals and humans[102]. Thromboembolism has been reported in patients who received transplantation of umbilical cord blood MSCs[103]. Placental-derived MSCs express higher levels of tissue factor[104,105], which aggravates the thrombotic events in patients infused with these MSCs for treatment of Crohn’s disease[106]. The ASCs are less prevalent and undifferentiated cells present in various adult tissues with a primary role of repair and maintenance of residing tissues.

ASCs are derived from different tissues and have limited potency compared to the ESCs or iPSCs. ASCs are named depending on the tissue of origin, e.g., HSCs, pancreatic stem cells, corneal stem cells, etc. HSCs were the first multipotent ASCs isolated from the bone marrow[107]. HSC transplantation is used as therapy for several malignant and non-malignant disorders and autoimmune diseases. These cells are also used for the recovery of patients undergoing chemotherapy and radiotherapy[108]. The allogeneic transplantation requires matching HLA, Class I, and Class II between donor and recipient[109]; however, there are risks of GvHD[110]. The emergence of more modern and less toxic methods of treatments replaces HSC transplantation in hematologic malignancies. Recent reports suggest that risks of bloodstream infections caused by Gram-negative bacteria are associated with allogeneic hematopoietic transplantation[111,112]. Hemorrhagic cystitis is another complication that has been reported in patients post-HSC transplantation[113].

Cell-based therapy using MSCs is currently an essential domain of research. MSCs may be isolated from various sources, including bone marrow, adipose tissues, dental pulp, peripheral blood, synovium, and extraembryonic sources, as described earlier. These cells are plastic adherent and multipotential and can differentiate into bones, cartilage, fat tissues, muscles, and tendons[114]. The International Society for Cellular Therapy defined MSCs as plastic adherent cells expressing CD73, CD105, CD90 (≥ 90%), and not expressing hematopoietic markers CD34, CD45, CD14, CD19, and HLA-DR (≤ 2%) and have the potential of multilineage differentiation to osteogenic, adipogenic and chondrogenic lineage[115]. There is, however, no single marker to identify MSCs from various sources[116-118]. Figure 2 shows immune-modulatory characteristics and the ability of differentiation of MSCs.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Characteristics and therapeutic potential of mesenchymal stem cells. Sources of mesenchymal stem cells (MSCs) are shown as adipose tissue, umbilical cord, and bone marrow. MSCs are characterized by positive and negative markers and may be differentiated for clinical applications. The immunomodulatory properties of MSCs make them a promising candidate for cell-based therapies. NK cell: Natural killer cell.

The three primary characteristics of MSCs are: Differentiation into specific cell types and their incorporation into tissues that make it suitable for regenerative medicine; secretion of cytokines and exosomes that stimulate cell growth and proliferation and modulate inflammation; and direct contact with the host tissue and regulate effector function[119,120]. MSCs are therapeutically more successful due to the multipotentiality, immunomodulatory, anti-inflammatory, efficient homing capacity to injured sites, and minimum ethical issues[121-123]. However, there are disparities in MSC potency and pharmacological functionality, depending on tissue sources, cell handling, method of harvest, cultural expansion, dose, and route of delivery[124].

The most important sources of MSCs for clinical trials are bone marrow and adipose tissues[125]. Although MSCs isolated from the bone marrow have been extensively studied. However, there are several challenges in the clinical use of these cells. Cells collected from bone marrow are contaminated with HSCs, and a very small fraction (0.001%-0.01%) of MSCs are harvested from bone marrow. The cells isolated also show early signs of senescence during culture[126]. Although the collection of bone marrow aspirate is considered safe, certain complications and morbidity have been reported during collection from the sternum and posterior iliac crest[127]. Adipose tissue-derived stem cells are another potential source of MSCs by less invasive procedures[128]. Adipose tissue is abundant in the body and can be easily collected by liposuction, and the yield of stem cells is comparatively greater without any adverse effects. Adipose tissue-derived stem cells have more stability in the culture and have greater differentiation potential to osteocytes, chondrocytes, adipocytes, cardiomyocytes, and neurocytes[129]. Several studies have reported adipose tissue-derived MCSs' efficacy and safety for regenerative medicine[130,131]. The other sources include dental pulp, tendon, or from the perivascular fraction of any tissue[132].

MSCs have immunomodulatory properties and can reduce the inflammatory response. These cells can modulate the function of the innate and adaptive immune response. MSCs have very low levels of major histocompatibility complex and reduced expression of FasL (Fas Ligand) or costimulatory signals like B7, CD40, or CD40L[133]. MSCs secrete extracellular vesicles containing various growth factors and cytokines that suppress B lymphocyte and T lymphocyte function and maturation of dendritic cells while activating T regulatory cells. MSCs also secrete angiogenic, antiapoptotic, and antioxidative effects[134]. Table 1 summarizes the discovery time, sources, advantages, disadvantages, current clinical applications and prospects of various stem cells.

Regenerative medicines offer enormous potentials for better treatment of patients and quick recovery. However, there are certain drawbacks due to inefficient production, lengthy and complex processes, and human errors due to excessive human efforts. Understanding genetic components that influence the development of shape, size and orientation of an organ is extremely vital. Although the mechanism of most of the regenerative models is available with diagrams depicting the gene regulations, however, the stepwise dynamics to produce a particular shape of an organ are lacking. The AI-driven models and constructive algorithms could be a powerful solution for a deeper understanding of such mechanisms. These models could make the development of regenerative medicines automated and minimize human error factors.

The regenerative medicines' manufacturing has its own set of challenges, viz. efficient, cost-effective, large-scale production, lack of automation and quality control systems, and absence of closed and modular systems[181].

A vast number of datasets are published every year for experimental regenerative biology, but there is no international guideline for standard and high-quality datasets for regenerative medicines. Also, the tools for analyzing available datasets to get deeper insights and meaningful patterns are lacking. There have been some limited non-AI-based computational methods, platforms, and tools available to assist regenerative therapies. In an effort to derive such computational methods, Lobo and Levin developed a computational method to understand the physiological controls in planarian regeneration[182]. The exceptional ability of planaria to regenerate its body parts could be a potential model for leading regenerative medicine research. Another computational platform, KeyGene[183], can predict the tissue origin of various cell types. It could find out the equivalent stage of human PSC differentiation products along with the identification of stem cell derivatives. The KeyGene algorithm applies next-generation sequencing and microarray datasets and could be used to predict human adult tissue identity. This tool also helps monitor the cell-differentiation conditions and evaluate in-vitro cell-differentiation efficacy, thereby fetching improved protocol outcomes. CellNet[184] is among the recent computational tools that provide cell identity parameters, evaluate cell fate conversions, and rank suitable candidates for future interventions. However, it cannot differentiate between cell subtypes and cell heterogenicity, which remains a challenge to solve to date.

Nevertheless, AI-driven methods have emerged as an important component of stem cell research. Over the last decade, AI algorithms have advanced very rapidly, and along with enormous progress, methods to apply them have also enhanced subsequently. Many algorithms and tools can be expected in the recent future, which could efficiently assist stem cell-based regenerative medicines development, outcome prediction, and decision support to healthcare providers.