Suicide gene‐enabled cell therapy: A novel approach to scalable human pluripotent stem cell quality control

There are an increasing number of cell therapy approaches being studied and employed world‐wide. An emerging area in this field is the use of human pluripotent stem cell (hPSC) products for the treatment of injuries/diseases that cannot be effectively managed through current approaches. However, as with any cell therapy, vast numbers of functional and safe cells are required. Bioreactors provide an attractive avenue to generate clinically relevant cell numbers with decreased labour and decreased batch to batch variation. Yet, current methods of performing quality control are not readily scalable to the cell densities produced during bioreactor scale‐up. One potential solution is the application of inducible/controllable suicide genes that can trigger cell death in unwanted cell types. These types of approaches have been demonstrated to increase the quality and safety of the resultant cell products. In this review, we will provide background on these approaches and how they could be used together with bioreactor technology to create effective bioprocesses for the generation of high quality and safe hPSCs for use in regenerative medicine approaches.


INTRODUCTION
Regenerative medicine employs cells, tissues, biomaterials and/or gene therapy to replace, repair or regenerate tissues/organs in patients that require intervention beyond the scope of current medical practices. [1,2]The concept of regenerative medicine was initially conceived in the 1920s, when Alexis Carrel was awarded the Nobel Prize for sustaining thyroid gland viability in ex vivo culture demonstrating that tissues can be maintained outside of the body with the potential to be re-introduced at a later time. [3]A series of medical advances led these successes in humans.Particularly stem cells have received extensive attention (and not all positive) as the therapeutic approach for treating disease and disability. [9]em cells were first observed by Till and McCulloch after nodules began growing on the spleen injecting bone marrow cells into irradiated mice, and these nodules had the ability to self-replicate as well as differentiate into the three primary blood lineages. [10]Today, hematopoietic stem cells (HSCs) are the only current curative therapy for a wide range of haematological malignant and non-malignant conditions.
While HSCs were the first cells identified with potency in vitro and in vivo, [11] there are numerous classes of stem cells that has been employed in regenerative medicine research over the last 60+ years.Leveraging off a molecular understanding of early development, Embryonic stem cells (ESCs) a pluripotent cell population derived from the inner cell mass (in mice) were first cultured and expanded in vitro in the 1980s. [12]The first human ESC lines were isolated in the 1990s [13] and the first clinical trials with these cells were undertaken in the 2000s. [14]Yet, issues around quality control, tumourigenesis and the ethical/moral concerns around the derivation of these cells, translational research using ESCs has stagnated.Adult/tissue resident/mesenchymal stem cells (MSCs) then gained significant traction in the field and while a number of clinical trials have been undertaken with these cells, no treatment has been approved yet by the FDA.There are also issues around the clinical expansion of these cells, since they only retain limited self-renewal in vitro.In 2012, John Gurdon and Shinya Yamanaka were awarded the Nobel Prize in Physiology or Medicine for the discovery of induced pluripotent stem cells (iPSCs) -somatic cells that are reprogrammed to a pluripotent-like state. [15]These cells have comparable self-renewal and differentiation potential to ESCs but avoid the ethical concerns around embryo destruction to obtain cells. [16]However, they still carry the risk of generating tumours posttransplant and theoretically even more so than ESCs because of the genetic and/or epigenetic modifications required to revert these cells to take on an early developmental phenotype.
Therefore, the purpose of this study is to discuss key methods to increase the safety of iPSCs while still being able to generate clinically relevant cell numbers.

ADVANCES IN REGENERATIVE MEDICINE
Over the past two decades, there has been a push to translate stem cell research into the clinical theatre.Most of these studies have employed HSCs and MSCs as a cell source.
Both autologous and allogeneic HSCs typically harvested from bone marrow may be used as an immunotherapy against haematological malignancies.Autologous HSCs may be gene-corrected to then be delivered to compromised tissues and organs [17] while allogenic HSCs may be delivered to compromised tissues and organs to deliver a graft-versus-tumour (GvT) effect to reverse malignancies. [18]However, the use of HSCs in allogeneic cell therapies are limited by human leukocyte antigen (HLA) donor matching in order to prevent graft-versus-host disease (GvHD) and increase the chance of survival. [19]Cs harvested from cord blood reduce the need for HLA matching and decrease the risk of graft rejection [20] ; however, the small population of donor cells and limited proliferative capacity can prevent their use in therapeutic applications.Exposing cord blood cells to pyrimidoindole derivatives such as UM171 during cellular expansion has been demonstrated to increase their proliferative capacity in an effort to make a less invasive HSC therapy that does not require HLA matching. [21]4] Yet, despite all the recent advancements in MSC-based therapies, there has been significant failures in MSC trials that have highlighted barriers in the widespread translation of this cell source for regenerative medicine strategies. [25]pically thought of one of the main advantages of donor-type (allogeneic) MSCs treatment is the absence of HLA-II, the low amount of HLA-I, and the costimulatory proteins CD40, CD80, and CD86, which are all believed to contribute to the low immunogenicity of MSCs [8] ; there have been a number of independent reports demonstrating MSC immunogenicity due to inadequate culture and processing conditions. [8]This issue can be compounded in treatment schedules that require multiple MSC infusions. [8,26]MSCs are also recognised for their multipotency and specific ability to differentiate into connective tissue cells.Yet, serial passaging and prolonged cell culture of MSCs results in senescence-related protein expression and downregulation of core transcriptional pathways involved in self-renewal and potency.Therefore, to provide clinically relevant numbers of MSCs for both trials and application this hurdle will need to be overcome. [27,28] addition to issues surrounding expansion of MSCs, the heterogeneity of these cell populations can be significant and impacted by; donor, tissue source, age, health condition, culture condition and other factors. [9]Therefore, many pre-clinical and clinical studies have suggested that allogenic therapies may be more practical as an off the self-solution, which would also decrease the cost per therapy. [29]This change in paradigm has also had the effect of softening the stance on genetic manipulation of the cells to increase their regenerative potential, immunomodulation ability, reduction of immunogenicity and/or self-life. [30]One consequence of these new directions in the MSC field is the re-emergence of allogenic therapies using human embryonic stem cells (hESCs) and their genetically and/or epigenetically modified cousins -iPSCs (hiPSCs).

HUMAN PLURIPOTENT STEM CELLS
The generation of the first hESC line [31] acted as a catalyst for innovative yet rapid research in the field.Initially, there was a burst in the development of methodologies aimed at directed differentiation to drive hESCs into clinically relevant cell types.Many saw these pluripotent cells as a one-stop-shop for regenerative medicine as they demonstrated unlimited self-renewal and can become any cell type within the human body.One of the early pre-clinical and clinical targets for application of this cell source was the differentiation to neural cells as MSCs do not have the ability to undergo neurogenesis under physiological conditions.With a significant amount of interest and funding in this area the first clinical trial for spinal cord injury was conducted in less than 10 years from the first published paper on hESCs. [32]This clinical trial was run by Geron and used oligodendrocyte progenitor cells that had been differentiated in vitro and reported to have lost the ability to form teratomas in the host.This trial was met with skepticism in the field because of a lack of independent reproducibility at the in vitro level and scepticism regarding the ability of these differentiated cells to induce injury repair in vivo.In 2009, the trial was halted due to animals developing spinal cysts at the injection site and even when the trial was resumed, only four patients received the therapy with no beneficial outcome before the trial was aborted due to lack of funding. [14,32]ile not the outcome hoped for by the community for the first hESC clinical trial, this experience highlighted a number of issues that needed to be addressed before future attempts.At the cell production level it was clear that robust quality control measures were needed to purge undifferentiated or partially differentiated cells while at the same time developing bioprocesses to generate the massive quantities of cells required for clinical application.
Since Geron, there have been ∼90 clinical studies registered for embryonic and induced human pluripotent stem cells (hPSCs) globally.Unlike hESCs, hiPSC isolation does not have the same moral, ethical and legal issues. [16]Nevertheless, the presence of undifferentiated hPSCs in the hPSCs and derived therapies, raises the danger of tumour development following implantation of hPSCs into patients. [33]oto-oncogenes, tumour suppressor genes and/or pluripotency regulatory factors may undergo variety of mutations that affect the regulation of these genes and eventual transformation of hPSCs resulting in tumourigenesis. [34]Individual hPSCs may carry a number of chromosomal abnormalities and copy number variations (CNVs) as a consequence of de novo mutations that arise at random -some of which may give these cells a selective advantage of their wildtype neighbours. [34]hiPSCs can acquire these genomic and epigenetic alterations and accumulate them during both the reprogramming process and during proliferation.Prolonged passaging during long-term culture can generate distinct populations each with biases in their differentiation ability. [35]For instance, karyotypic aberrations on chromosomes 8, 12, 17, 20 and X have been identified in hPSCs [36,37] with some of these mutations conferring growth advantages and selection dominance.These issues are not isolated to DNA mutation as reprogramming and passaging can also induce epigenetic modifications thought hypermethylation and consequent gene expression silencing. [35]re specific to hiPSCs, loss of imprinting during reprogramming and prolonged cell culture has been shown to result in epigenetic heterogeneity.X inactivation, cell cycle heterogeneity, positional heterogeneity (position of cells within a colony) all can lead to heterogeneity.Even under the optimal culture conditions, distinct responses to differentiation stimuli are possible. [35,36]Of particular concern for the translation of hiPSCs is the immunogenicity of the cells and/or their differentiated products, making the assessment of immunological phenotype of cells a key criterion prior to their clinical application.
Specifically, it has been demonstrated that hESCs are incapable of directly stimulating T cells [38,39] and as patient-derived cells, autologous hiPSC and their derivatives would be expected to be well tolerated by the host immune system, and immune system-related complications should be avoided.However, when iPSC-derived differentiated cells are transplanted into the host body, they have the potential to become immunogenic. [39,40]Epigenetic aberrations and gene mutations may be the underlying mechanisms that compromise the immunological privilege status of hiPSCs and their offspring.In addition, the accumulation of mutations might generate immunogenic aberrant proteins or peptides. [39,40]To decrease the tumourigenic potential of hPSC implantation in patients, it is necessary to develop innovative screening technologies while also considering the necessity of production of clinically relevant numbers. [36]

BIOREACTORS OFFER A SCALABLE PLATFORM FOR HIPSC EXPANSION
Clinical trials often require large therapeutic doses of hiPSCs ranging from 10 9 to 10 12 cells per patient. [41]For hiPSCs to achieve these clinically relevant cell densities, a robust, controlled, well-defined bioprocess is required.Every step of the bioprocess must be critically evaluated to maximise the therapeutic potential of the cells.Culture conditions such as media composition, pH, osmolality, oxygenation, metabolite concentrations, substrate interactions and shear forces are all considered to be important when generating clinically relevant quantities of high-quality cells and their differentiated products.
Traditional culture of hPSCs involves using small-scale static vessels with extracellular matrix (ECM) coatings and growth factors that support pluripotency and self-renewal. [42]These vessels are commonly T-flasks or petri dishes coated with an attachment matrix to allow hPSCs to grow as monolayer colonies.Feeder layers (embryonic fibroblasts) are utilised for their contribution of growth factors and ECM proteins; however, this practice can result in large heterogeneities as well as posing a risk of transfer of animal pathogens. [42]trigel is also commonly used as a feeder-free replacement when culturing hPSCs. [43]However, Matrigel suffers from batch-to-batch variation and does not allow for xeno-free culture since it is mouse derived.Batch-to-batch variation is a serious concern when striving for current good manufacturing practices (cGMP) certification and by extension use in clinical therapies.A number of animal-componentfree alternatives have been developing including laminin, vitronectin or Synthemax, which all facilitate hPSCs adherence. [44,45]en when all the reagents can be made animal free and/or GMP compliant, heterogeneity still results from the large number of static vessels required to reach clinically relevant cell densities.Small differences in ECM coating, concentration gradients of growth factors and colony size/morphology within vessels results in variability and therefore can compromise product quality.Static culture vessels such as T-flasks or Petri dishes do not allow for process monitoring or F I G U R E 1 Bioreactors offer a scalable technology to expand and differentiate induced pluripotent stem cells (iPSCs) in a controlled, homogenous environment.When compared to traditional static culture, the absence of control mechanisms combined with flask-flask heterogeneity allows for the spontaneous differentiation of cells or the differentiation to harmful hybridoma causing cells.control of many critical bioprocess variables, and are labour-intensive to scale up.This results in difficulties designing reproducible and robust processes, as many variables are changing and influencing cell quantity and quantity.For these reasons, the expansion and differentiation of hPSCs under controlled conditions is essential when manufacturing clinical cell therapies.Complexities such as heterogeneous starting populations, the transient nature of subpopulations, cell location within colonies and the compound interactions between culture conditions such as growth factors, dissolved oxygen and cellcell interactions must be monitored for the generation of high-quality and reproducible hPSC products. [46,47]As an alternative to traditional static culture, bioreactors offer a scalable, well-mixed vessel which can be monitored throughout the expansion process to reproducibly produce large numbers of high-quality cells.Various bioreactor geometries, feeding regimes and oxygen sparging techniques exist; each with goals of optimising conditions such as media availability, hydrodynamics and oxygen tension to generate useful cell products.Bioreactors also offer the ability to have all cells experiencing the same conditions, eliminating variability between smaller flasks or between specific areas within static vessels.Employing a bioreactor platforms also allows for the optimisation of available space/surface area for cellular expansion, [48] specifically suspension bioreactors (stirred tank or vertical wheel [VW]), to offer a well-mixed environment coupled with culture regulation and optimisation [49] (Figure 1).
When utilising bioreactors for hPSC expansion, culture conditions change from a two-dimensional (2D, e.g., cells grown on micro-carries) environment to a three-dimensional (3D, e.g., cells grown as aggregates) environment.The 3D culture environment allows hPSC aggregates to form as well as introduces mechanical forces which change the interaction networks of cells [48] ; therefore, these factors must be considered when scaling up vessels to reach clinically relevant cell densities.[52] By increasing the bioreactor scale, the hydrodynamic profile changes which results in altered product quality. [52]hiPSCs are extremely sensitive to shear stress and for this reason, the VW bioreactor is commonly used for hiPSC expansion.VW bioreactors are available from 0.1 to 15 L, and it has been shown it is possible to scale up hiPSC expansion in VW bioreactors while maintaining cell quality using computational fluid dynamics (CFD) modelling. [50]Scale-up using CFD is done by investigating key hydrodynamic variables, such as the volume average velocity profile as well as the energy dissipation rate and optimising agitation regimes to determine an operating range for hiPSC aggregates across multiple scales. [50]However, the effects of hydrodynamic forces on protein production, cell-cell signalling and cell-matrix interactions must still be investigated to completely understand 3D culture of hiPSCs.
Bioreactors are not only useful when generating high cell densities, they are also essential for maintaining cGMP compliance when translating hiPSC therapies from bench to bedside. [49,53]To meet cGMP conditions, the entire manufacturing process of hiPSCs must meet current standards including; tissue collection, cell reprogramming, cell expansion, cell banking and quality control testing. [54]Stem cell media, cell culture matrix, culture conditions, oxygen levels and cell density must be defined.Monitoring and controlling the levels of these essential components with automated systems is the only feasible way for the wide scale manufacturing and application of cell therapies.Identifying the attributes that influence the quality and safety/efficacy of the product, identifying parameters that influence these attributes, and then defining how variability in these parameters impact quality attributes are important steps in designing a GMP compliant bioprocess. [55]Computer-controlled bioreactors offer the ability to define a set of operating conditions while also preventing deviation from these set conditions through probes and feedback loops.However, one of the current limitations of bioreactor process control is the inability to measure specific cellular responses in a timely and accurate straightforward manner. [47]Undetected deviations from setpoints during the exponential growth phase of cells can result in propagated and unpredictable heterogeneity within the final cell product [56] ; therefore, quality-by-design principles must be employed to ensure process robustness in the anticipation of any deviations. [55,56]This involves the continuous design, monitoring and validation of control strategies with respect to their effects on cell phenotype, purity, potency and yield.
Current large-scale processes using bioreactors involve the manufacturing of protein therapeutics using cells as factories, and current expansion of hPSCs as therapeutics involves static culture vessels.A significant technological gap exists between the current technology available and the GMP expansion of hPSCs in bioreactors.There is a need for scalable quality control methods to ensure the safety and efficacy of hPSC products developed in bioreactors to harness their true therapeutic potential.While cellular expansion capabilities, process control mechanisms and scalability of bioreactors make them an attractive candidate for clinical applications of hPSCs; the current focus must shift from achieving the highest cell density [48] and refocus on achieving high-quality and safe hPSC-derived products.One critical gap in the focus on quality versus quantity is the labour-intensive and time-consuming methods of determining quality that are currently not suitable to scale-up.

QUALITY CONTROL OF CELL PRODUCTS
Clinical hPSC therapy is not possible without defined criteria, standardised guidelines and quality control assessment for functional capacity and safety evaluation of cells (Figure 2).Despite the fact that health authorities and regulatory organisations have created strict guidelines for regulatory frameworks to attain optimal results, the regulations for therapeutic cells vary by jurisdiction. [57]tection methods should be able to identify genetic and epigenetic alterations that occur during in vitro expansion which may confer cells with a proliferation advantage and/or apoptosis resistance. [58]Karyotype analysis, such as G-banding, can be used to estimate genome stability, but this is a terminal cell assay and can only speak to relative percentage of mutant cells in the population.Copy number variant (CNV) and other common mutations can be detected using techniques such as microarray, PCR and sequencing analysis, [58] but these assayed are typically a candidate approach and you would only examine mutations you are looking for in the first place.For cell line authentication, short tandem repeat (STR) loci are evaluated in the event of cell line cross-contamination.There are fewer validated tests for the evaluation of the risk of tumourigenicity, whole-genome or exome sequencing has been proposed, [59,60] but it remains unclear on how these would be integrated into a cell therapy workflow.In the case of hiPSCs, residual vector clearance needs to be confirmed before the cells are transplanted.In terms of functional assays, directed differentiation, and teratoma assays can examine the differentiation capability of hPSC [59,60] but these methods are laborious and time-consuming.
A number of quality control methods exist for the removal of undifferentiated hPSCs from the final cell product.The two main ways to accomplish this involve the use of cell sorting and monoclonal antibodies.Although these methods have been proven to be effective using in vitro and in vivo models, the cost, safety, scalability and efficiency of these methods remain to be evaluated.Magnetic-activated cell sorting (MACS) and fluorescent-activated cell sorting (FACS) are techniques to isolate remaining pluripotent cells from a differentiated hPSC population.Using these methods, pluripotent cells can not only be identified (based on cell surface marker expression) but also isolated from the bulk population.Although these methods are quick and easy to use, they cannot be used in vivo.In addition, they are not effective at removing all pluripotent cells from a population.For instance, it has been reported that utilising MACS alone allows for hESC retention efficiencies of 81%-83%. [61]e binding affinity and cytotoxicity of specific monoclonal antibodies can be utilised to target and eliminate specific cells.The overexpression of CD20 antibody on human T Cells has been demonstrated to induce apoptosis when exposed to anti-CD20 antibody Rituximab. [62]is overexpression is done using retroviral vectors and was developed to prevent GvHD in allogenic T cell therapies.However, endogenous CD20 presenting cells are also killed when exposed to anti-CD antibodies, limiting this system in therapeutic applications.Direct binding of antibodies may also be used to directly eliminate targeted cells.
Cytotoxic monoclonal antibody mAb 84 has been demonstrated to eliminate undifferentiated ESCs prior to transplantation by binding to podocalyxin-like protein-1 (PODXL) on hESCs [63] ; however, mAb 84 is an IgM molecule making it difficult to penetrate clumps of PSCs.
To improve efficacy of this treatment, antibody fragments of mAb 84 have been demonstrated to bind efficiently to single cells and clumps F I G U R E 2 Methods to determine quality control and understand cellular heterogeneity.There are a number of methods that have been employed to understand and control cellular heterogeneity in human pluripotent stem cells (hPSCs).However, to date, there are still significant barriers to obtaining pure, safe cell populations for clinical application. of hESCs to prevent tumour formation. [61]Cytotoxic antibodies may also be engineered to target and eliminate cells by binding to a specific pluripotent marker.A monoclonal antibody designed to target antistage-specific embryonic antigen SSEA-5, which is a found on hPSCs, was able to successfully eliminate pluripotent cells to prevent tumour formation in vitro and in vivo. [64]mbining cell sorting and cytotoxic monoclonal antibody administration is a current method of quality control before administering pluripotent cells to patients.Due to the high cost of monoclonal antibodies, sorting the cells with FACS or MACS first allows for a lower concentration, making the treatment more affordable.As pluripotent cells are more likely to remain after cell sorting and teratomas can form in vivo with low numbers of cells, these methods must be combined to prevent tumour formation.However, this combination of quality control methods does not offer scalability as we generate high cell densities.

SUICIDE GENES
Gene therapy and transgenic approaches have evolved significantly from the days of delivering genetic information to cells in order to mod-ify their genomes.Gene-directed enzyme-prodrug treatment (GDEPT) is a suicide gene method that inserts a transgene into cells that modify a non-toxic delivered prodrug into a toxic metabolite with the intention to induce apoptosis. [65]This type of suicide gene strategy consists of three components: the prodrug, the enzyme, and the gene delivery method [66] (Figure 3).This type of system is also highly adaptable since suicide gene expression and by extension prodrug activation can be regulated by cell/tissue specific promoters and/or inducible promoters that can be activated spatially and/or temporally. [67]The efficacy of the suicide gene strategy can be optimised by through changes to these three pillars and in theory is only limited to our current understanding of the genome and the creativity of the design.Furthermore, suicide gene strategies are not limited to prodrug delivery, by can also include; enzymes, poisons and/or proapoptotic genes.
In GDEPT strategies, there are two main modes or action.The first category employs enzymes that are normally expressed in human cells, such as p450.While these enzymes are unlikely to be immunogenic, using this approach increases the risk of off-target toxicity. [67]The second group consists of non-human enzymes originated from viruses, bacteria or yeast, such as thymidine kinase (TK, viral) and cytosine deaminase (CD, bacterial and yeast).For example, the Herpes simplex virus thymidine kinase (HSV-TK) gene converts a non-toxic drug, F I G U R E 3 Schematic of suicide gene approach.There are distinct design and delivery approaches involving suicide genes that can be employed to generate more homogeneous cells populations for cell therapy applications.ganciclovir (GCV) into a toxic metabolite that inhibits DNA synthesis and induces apoptosis. [68]However, this class of enzymes has the potential to be highly immunogenic. [67]nes that produce toxins are also commonly used as an option for suicide gene therapy.By inactivating signalling pathways, bacterial toxins such as diphtheria (Corynebacterium diphtheriae) can impede critical cell processes resulting in cell death. [68]Another approach is to deregulate apoptotic genes.For example, Caspase9 is an apoptosis initiator that can trigger programmed cell death under normal conditions.However, it can also be genetically modified to only be functional in the presence of chemical inducers of dimerisation (CID) class drugs that results in homo-dimerisation of iCaspase9 and initiates the apoptosis pathway. [68]ere are two major vector categories that can transmit the suicide gene: viral and non-viral vectors.Viral vectors convey foreign DNA by replacing non-essential viral DNA compartments with a foreign suicide system, such as DNA related to virus pathogenesis. [69]ral vectors for gene therapy may be created from five major viral vector types, including lentivirus, herpes simplex virus, retrovirus, adenoviruses and adenoviruses associated adenovirus. [70]Virus-based approaches of gene therapy have been effective in preclinical and clinical settings [15,22] due to their increased gene delivery and gene expression efficiency, high efficiency of gene transduction in a broad variety of dividing and non-dividing cell types. [71,72]However, immunogenicity and cytotoxicity, as well as insertional mutagenesis, are the major drawbacks of employing these viral vectors. [73]Non-viral vectors are more attractive for clinical applications because of their biosafety, low cost, flexibility of manufacture and low immunogenicity.However, low transduction efficiency and transient gene expression are issues with this approach. [74] addition to exploiting apoptotic machinery to trigger cell death, enzymes can also be exploited to transform prodrugs into cytotoxins.
For this approach to work, the prodrug must be an enzyme substrate that is selective and can efficiently activate the enzyme, [75,76] but it also must be stable and bio-inert before to activation; while becoming a toxic metabolite following activation.Depending on the activation mechanisms, prodrugs may be divided into two groups: direct-linked prodrugs and self-immolative drugs. [67]A prodrug with a direct-linked, such as GCV, can be converted into a lethal metabolite by a single modification.A self-immolative drug, such as doxorubicin, can be transformed to an intermediate product via enzymatic cleavage and then further chemical modifications are required before the active molecule is generated. [67]

CURRENT CELL SUICIDE GENE SYSTEM IN CLINICAL TRIALS
To date, HSV-TK is the widely employed suicide mechanisms in human cells.HSV encodes a TK gene that is distinct from endogenous human TK genes.The HSV-TK transgene produces an enzyme that transforms prodrugs, such as GCV, into monophosphorylated nucleoside analogues.In the case of GCV triphosphate, the molecule is integrated into nascent DNA strands during replication, limiting DNA chain elongation thereby initiating apoptosis. [77]This approach has been employed in clinical studies.In a phase III clinical study, HSV-TK was integrated into haploidentical HSCs for high-risk acute leukaemia.In this study (TK008, NCT00914628), patients received a 4-month monthly infusion of TK modified cells to prevent or control the development of GvHD. [78]ile HSV-TK is a promising method for eliminating undesirable cells, it does have certain disadvantages.Specifically, the activation of HSV-TK requires ganciclovir which can trigger immune responses; cell mediated apoptosis through HSV-TK is relatively slow and requires days to complete; and this approach also prevents the use of ganciclovir to treat CMV infections in recipients. [79]To avoid these limitations, the iCaspase9 suicide gene approach has been developed and employed in clinical studies.The iCaspase9 transgene was generated by substituting the Caspase recruitment domain (CARD) of Caspase9 with a mutant FKBP12F36V (a FK506 drug receptor) which can now bind the bioinert and non-toxic drugs Ap20187 and Ap1903. [80]ese lipid-permeable synthetic ligand (chemical inducer of dimerisation/CID/Ap20187/Ap1903) force dimerisation of two iCaspase molecules to activate apoptosis. [81]benefit of this system is that it employs caspase9 which normally exists in all cells of the body.In addition, CID is a nonpeptide synthetic ligand and therefore has low immunogenicity.After exposure to CID drug, the dimerisation and induction of apoptosis occurs rapidly and kills transfected dividing and non-dividing cells equally. [82,83]However, while up to 99% of cells engineered with iCaspase9 system can be eradicated after CID administration; if 1% of cell remain, there is still a chance that these cells can repopulate over time. [84,85]Some have suggested this incomplete efficiency is due to epigenetic alterations, such as transgenic promoter hypermethylation which can suppress transgene expression.Overall, this could lead to a decrease in iCaspase9 expression in transduced cells. [84]A number of groups are investigating if this limitation can be overcome by exposing the cells to drugs that modify the epigenetic status of the cells.
Since the safety and therapeutic efficacy of patient-specific hiPSC transplantation in clinical trials remains unclear, [86] it can be advantageous to use a suicide gene prior to initiating reprogramming to make hiPSCs and their derivatives trackable.Remaining residual of reprogramming transgenes or reactivation of these transgenes after reprogramming, and integration of viral factors into the genome and activation of adjacent oncogene promoters have increased the need for trackable stem cells that are genetically distinct from the recipient.
In addition, during reprogramming, the entire genome's transcriptional program is reprogrammed and this can result in a tendency/bias towards uncontrolled behaviour.In addition, clonal hiPSCs can be generated with trackable suicide gene compared to polyclonal populations.
These clonal hiPSCs can have the integration site can be mapped, and the possibility of insertional activation of nearby oncogenes can be determined.
The silencing of inserted transgenes in the case of suicide gene systems in hPSCs may ultimately lead to the presence of tumours in vivo.To prevent this silencing and generate effective suicide gene systems, multiple fail-safe methods are being explored to ensure reliable expression.Linking the cell-division locus CDK1 and HSV-TK can ensure that cells which have the ability to divide will express the suicide gene was mathematically predicted to lower the probability of one non-safe batch in the number of therapeutic batches required for clinical therapy. [87]Biallelic insertion of the suicide gene iCaspase9 was similarly used to mathematically predict the possibility of transgene silencing causing unsafe cell batches.By inserting iCaspase9 into both alleles of the AASV1 safe harbour locus of chromosome 19 in hiP-SCs, it was predicted that the probability of simultaneous methylation of both alleles causing silencing had an extremely low frequency of 9 × 10 −16 . [88]Inserting suicide genes such as iCaspase9 using genetic editing tools such as clustered regularly interspaced short palindromic repeats (CRISPR) may also cause significant changes in downstream cell behaviour and must therefore undergo the required sequencing before being translated to the clinical setting. [89] is also important to recognise that these suicide gene strategies can be undertaken without genetic modification of cells.[92] These EVs can be internalised by tumour cells and induce apoptosis in tumour cells following treatment with prodrug such as ganciclovir or 5-fluorocytosine.These types of approaches will no doubt expand the fields options for developing and delivering suicide gene strategies.

APPLICATION OF SUICIDE GENE SYSTEMS FOR BIOREACTOR QUALITY CONTROL
Bioreactors offer a proven, scalable, well-mixed system to facilitate the expansion and differentiation of hPSCs.This technology is becoming more attractive in order to meet the high cell dosage required for clinical cell therapies as well to manufacture cells in a controlled environment to meet GMP standards.However, the lack of pre-defined regulatory pathways has limited the translation of hPSCs for clinical use. [93]A robust manufacturing process with appropriate quality controls will provide a path for the large-scale production of hPSCs used in clinical therapies.Due to time and labour and overall cost constraints required to accurately purify populations, current product characterisation methods such as flow cytometry are not scalable to the clinical densities needed for therapy.Currently, in-process controls cannot regulate the purity of a cell population.Therefore, scalable quality control mechanisms such as suicide gene systems allow for control over population heterogeneity and are likely to improve the overall safety of the final product by removing cells with the potential to form teratomas and/or unwanted tissues in vivo.Therefore, integrating suicide gene systems into bioprocess quality controls bridges a significant gap in the manufacturing of hPSCs for cell therapies.