
==== Front
Mol Ther
Mol Ther
Molecular Therapy
1525-0016
1525-0024
American Society of Gene & Cell Therapy

S1525-0016(24)00396-4
10.1016/j.ymthe.2024.06.014
Original Article
Thymidylate synthase disruption to limit cell proliferation in cell therapies
Sartori-Maldonado Rocio 1
Montaser Hossam 1
Soppa Inkeri 1
Eurola Solja 1
Juutila Juhana 23
Balaz Melanie 1
Puttonen Henri 6
Otonkoski Timo 14
Saarimäki-Vire Jonna 1
Wartiovaara Kirmo kirmo.wartiovaara@helsinki.fi
15∗
1 Stem Cells and Metabolism Research Program, Faculty of Medicine, University of Helsinki, 00290 Helsinki, Finland
2 Faculty of Biological and Environmental Sciences University of Helsinki, 00790 Helsinki, Finland
3 Institute of Biotechnology, Helsinki Institute of Life Science, University of Helsinki, 00790 Helsinki, Finland
4 Children’s Hospital, University of Helsinki and Helsinki University Hospital, 00290 Helsinki, Finland
5 Clinical Genetics, Helsinki University Hospital, 00290 Helsinki, Finland
6 Department of Pathology, Helsinki University Hospital, 00290 Helsinki, Finland
∗ Corresponding author: Kirmo Wartiovaara, Stem Cells and Metabolism Research Program, Faculty of Medicine, University of Helsinki, 00290 Helsinki, Finland. kirmo.wartiovaara@helsinki.fi
12 6 2024
07 8 2024
12 6 2024
32 8 25352548
15 4 2024
10 6 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by/4.0/ This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Stem and progenitor cells hold great promise for regenerative medicine and gene therapy approaches. However, transplantation of living cells entails a fundamental risk of unwanted growth, potentially exacerbated by CRISPR-Cas9 or other genetic manipulations. Here, we describe a safety system to control cell proliferation while allowing robust and efficient cell manufacture, without any added genetic elements. Inactivating TYMS, a key nucleotide metabolism enzyme, in several cell lines resulted in cells that proliferate only when supplemented with exogenous thymidine. Under supplementation, TYMS−/−-pluripotent stem cells proliferate, produce teratomas, and successfully differentiate into potentially therapeutic cell types such as pancreatic β cells. Our results suggest that supplementation with exogenous thymidine affects stem cell proliferation, but not the function of stem cell-derived cells. After differentiation, postmitotic cells do not require thymidine in vitro or in vivo, as shown by the production of functional human insulin in mice up to 5 months after implantation of stem cell-derived pancreatic tissue.

Graphical abstract

A genetic modification in stem cells makes their proliferation possible only under thymidine supplementation. This method is proposed to reduce the risk of uncontrolled cell proliferation for stem cell-based therapies.

Keywords

TYMS
DNA synthesis
proliferation
safety switch
cell therapy
human pluripotent stem cells
diabetes
beta-cell differentiation
==== Body
pmcIntroduction

Cell therapies offer new possibilities for previously untreatable or challenging medical conditions. Among the cell types used in such therapies, human induced pluripotent stem cells (hiPSCs) have garnered significant attention due to their availability from patient-specific somatic cells1,2,3,4 and their capability to differentiate into potentially therapeutic cell lineages, such as glucose-responsive insulin-secreting pancreatic islets.5,6,7,8 hiPSCs circumvent the ethical concerns associated with the use of embryonal stem cells,9,10 and, unlike donor-derived mesenchymal stem cells, their autologous source diminishes the risk of immune rejection and graft-versus-host disease.11,12 Furthermore, the available technologies for iPSC gene transfer and genome editing enable the development of genetically modified cell-based therapies with enhanced therapeutic potential.13,14

Stem cells’ flexibility and tolerance for modifications have broadened their applications to include, for example, roles as medical delivery vehicles.15 They further permit silencing or modification of genes involved in, for example, immune recognition.16 These features may enhance their compatibility for allogeneic cell therapies, reducing the need for immunosuppression compared to other cell sources. However, although the universal off-the-shelf cell products decrease the risk for immune rejection, the diminished immunogenicity raises concerns related to immune evasion and potential oncogenic events.17

On top of creating biologically useful cells for a desired clinical purpose, a successful biotechnological therapeutic product needs to meet the criteria for safety and scalable manufacturing. These two goals, however, often compromise each other, since manufacturing benefits from robust cell growth, but intensive or unlimited proliferation poses a threat when the cells are transferred to patients.18,19

Attempts to create safer cells for therapy have yielded different strategies.20,21 These safety systems aim to prevent uncontrolled proliferation or selectively eliminate the transplanted cells if necessary.22,23 Some of the existing safety strategies include the use of inducible apoptotic systems, suicide genes, antibody-mediated cell depletion, and gene editing-based switches (Table 1). Often, these switches work by reversibly activating or inactivating genes upon exposure to a small molecule (on- and off-switches, respectively). However, they all present remarkable challenges: susceptibility to genetic silencing or alteration, transgenic or viral-derived origin, slow activation speed, rapid remission, and/or potential neurotoxicity.21,24,25Table 1 Summary table of selected “safety” systems, and potential advantages and disadvantages

Type	Function	Examples	Advantages	Disadvantages	
Enzyme/prodrug25,46,47,48,49	Suicide transgenes into cells that transform prodrug into toxic metabolite	HSV-TK + ganciclovir50,51	Some have the ability to eliminate the whole graft
Could be combined with other strategies
Safe
Generally efficient	Transgenes derive from viruses, bacteria or yeast
Immuno-rejection of the therapeutic cells
Toxicity of the prodrug
Bystander effect
Slow activation time
Silencing of transgene	
CD + 5-FC52	
NTR + CB195453	
PNP + MEP54	
mTMPK + AZT55	
Monoclonal antibody mediated49,56	Engineer a gene targeted by an antibody into the transplanted cells	CD20t + rituximab	Avoid graft-versus-host disease57
Efficient
Different payloads for ADC	CD20 present in endogenous B cells	
hEGFRt + cetuximab
58	
c-myc tag + anti-c-myc
tag59	
X + ADC anti-X60	
Inducible dimerization49	Insertion of a modified caspase-9 The recruitment sequence is replaced by a dimerization domain of FK506 binding protein	iCas961 + CID	Low immunogenicity
Inert small molecule
Fast action
Tested to eliminate iPSCs62	Silencing of transgene
CID resistance63	
Metabolic	Endogenous disruption of pyrimidine de novo synthesis	UMPS knockout20	Transgene free	May affect RNA synthesis	
ADC, antibody-drug conjugate; AZT, azidothymidine; CD, cytosine deaminase; CID, chemical inducer of dimerization; EGFR, epidermal growth factor receptor; HSV-TK, herpes simplex virus thymidine kinase; mTMPK, mutated thymidylate monophosphate kinase; NTR, nitroreductase; PNP, purine nucleoside phosphorylase.

Here, we envisioned an endogenous safety strategy by disrupting a key metabolic gene, whose function can be compensated by administering an easily available compound. Using CRISPR-Cas9, we genetically inactivated thymidylate synthase (TYMS), the only known enzyme in charge of the de novo thymidylate (dTMP) synthesis.26 This knockout generates a rescuable auxotrophy to thymidine and restricts the synthesis of DNA, without altering normal RNA or protein production (Figure 1A). This means that dividing cells depend on the external supplementation of thymidine to proliferate and achieve robust manufacturing of the cells of interest. However, once the cells have exited the cell cycle and terminally differentiated, they do not require supplementation to synthesize nucleic acids. While preventing dividing cells from proliferating uncontrollably, this approach does not require the use of small molecules or the insertion of large transgenic elements. Hence, it cannot be silenced by mutagenesis, and it does not result in leakiness or immunogenicity.Figure 1 Disruption of TYMS makes proliferative cells dependent on thymidine supplementation

(A) Diagram showing the nucleotide pathways for DNA (orange) or RNA (blue) synthesis. (B) Schematic representation of the gene of interest (TYMS), showing the 2 gRNAs, and the amino acids that form the active site of the protein over the corresponding exon that code for them. The lower line represents the change of frame upon DNA repair. (C) TYMS expression in wild-type versus knockout hiPSC (HEL24.3), A431, A549, and HSC3 (n = 3/cell line). Results shown as ratio of TYMS mRNA (primer targeting exon 5) over glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA contents. (D) Western blot showing the absence of TYMS upon inactivation of its gene. (E) Confluency-based growth curve of wild-type and knockout cancer cells supplemented with 0 or 25 μM dTMP. Results shown as average confluency ± SE. (F) Confluency-based growth curve of wild-type and knockout iPSCs supplemented with 0 or 25 μM dTMP. Results shown as average confluency ±SE. (G) Confluency at day 5 of wild-type and knockout hiPSCs under different concentrations of dTMP (0–5000 μM). Results shown as average confluency per image ± SE. (H) Cell count of wild-type and knockout iPSCs supplemented with 0 or 20 μM dTMP at 24 and 144 h postseeding. (I) Number of wild-type and knockout monoclonal colonies (N/48) supplemented with 0 or 20 μM dTMP. (J) Number of wild-type and knockout monoclonal colonies supplemented with 0 or 20 μM dTMP that reached over 90% confluency. (K) Summary of flow cytometry results for cell-cycle analysis. Statistical significance in (C), (H), and (J) based on Wilcoxon test; ns (not shown) p > 0.05; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. See also Figure S1.

Our results provide evidence that the culture and proliferation of TYMS−/− cells can be regulated with exogenous dTMP in vitro and in vivo. Furthermore, the characterization and differentiation of TYMS−/−-hiPSCs show that they provide a functional source for cell therapy that does not sustain unwanted proliferation without external dTMP supplementation.

Results

Disruption of TYMS makes proliferative cells dependent on thymidine supplementation

The human thymidylate synthase (PDB: 5X5D) is evolutionarily highly conserved. The functional enzymatic ligand pocket is encoded by exons 4 (amino acid R175), 5 (P193, C195, Q214, S216, D218, R225), and 6 (H256, Y258).27 To disrupt TYMS activity, we deleted the exon 5 with a dual single-guide RNA (sgRNA) Cas9 knockout approach targeting both intron 4 (I4) and I5 (Figure 1B). This removed the active site of the enzyme, generated an early stop codon in exon 6, and depleted the enzyme expression at both mRNA and protein levels (Figures 1C, 1D, S1A, and S1B) in all tested cell lines.

We hypothesized that TYMS−/− cells could be cultured and manufactured normally with external dTMP supplementation, but they would cease to proliferate in its absence. Therefore, we followed the confluency of TYMS−/− cancer cell lines and hiPSCs (HEL24.3)28 without or with dTMP supplementation (25 μM). The untreated (TYMS+/+) cell lines were used as controls. The results suggest that TYMS−/− cells proliferate at a rate comparable to that of TYMS+/+ when supplemented with thymidine, while the proliferation capacity of non-supplemented TYMS−/− cells was drastically impaired (Figures 1E and 1F). To determine the optimal concentration of dTMP for culturing TYMS−/−-hiPSCs, we ran a dose-dependent assay using a range of 0.3 μM–5 mM exogenous dTMP to explore their tolerance limits (Figures 1G and S1C). Contrarily to TYMS+/+ lines, TYMS−/−-hiPSCs could not sustain growth at physiological concentrations of thymidine (reference: 0.5–1.4 μM),29 and by day 5 had largely died, with only a few differentiated cells remaining (Figure S1D). On the other end of this spectrum, we observe a shift in the cytotoxic threshold at higher concentrations; while the growth rate of TYMS+/+ cells decreases at concentrations higher than 200 μM, knockout cells suffer comparable cytotoxic effects at concentrations higher than 1 mM (1,000 μM) (Figure 1F). For the hiPSCs we employed throughout this article (HEL24.3), we confirmed these results by quantifying the cell number of TYMS−/− and TYMS+/+ at 24 (1 day) and 144 (6 days) after plating with or without 20 μM dTMP (Figure 1H). In yet another proof-of-concept experiment, we plated TYMS−/− and TYMS+/+-hiPSCs in 48-well plates as single cells and followed their expansion with and without thymidine (Figures 1I and 1J). While all TYMS+/+ cells and supplemented TYMS−/− reached 90% confluency between days 15 and 27 postplating, the few non-supplemented TYMS−/− colonies differentiated early and eventually collapsed (data not shown). Cell-cycle analysis by flow cytometry of TYMS−/−-hiPSCs reveals that in the first 24 h of dTMP withdrawal, the cells accumulate in G0–G1 and S phase, and then, as they should continue to G2, accumulate in the subG0–G1 phase (Figures 1G and S1E). qRT-PCR of CDKN1A (P21) suggests the cells largely die during the first 24 h after supplementation withdrawal (Figure S1F).

In summary, our results show that TYMS−/− cells grow normally and can be expanded extensively under dTMP supplementation. Thymidine withdrawal, however, stalls cells in S phase and inhibits progression to G2/M, likely due to replicative stress from nucleotide imbalance.

TYMS knockout leads to few transcriptomic and metabolic changes

We explored the molecular effects of the genetic knockout through bulk RNA sequencing (RNA-seq) and untargeted metabolomic analysis of the supplemented TYMS−/− and non-supplemented TYMS+/+-hiPSCs. The results of the RNA-seq further confirmed the downregulation of TYMS expression and identified some undescribed transcripts that might be playing a role in the gene regulation (e.g., LINC02864, ENSG00000286456) (Figures S2A and S2B). Enrichment analysis suggested changes in ion binding, mostly due to changes in metallothioneins of class MT1 (Figure S2C). These changes might relate to response to stimuli,30 likely linked to metabolic changes from thymidine supplementation. Interestingly, the results also showed a downregulation of CDKN1A (p21) in TYMS−/− compared to TYMS+/+ (Figures S2B and S2C). The metabolomic analysis exposes a downregulation of thymidine subproducts (e.g., thymine), and an upregulation of metabolites that directly or indirectly participate in thymidine synthesis (Figures S2D–S2F and S3A). Taken together, these analyses show little difference between TYMS−/− and TYMS+/+. Moreover, the differences seem to arise as a response to increased dTMP, supported by the upregulation of genes like MAT2A, FBXO9, and PPP1R17. These genes act on cell-cycle protein dynamics and methionine metabolism,31,32,33 which in turn may lead to the upregulation of AMP, anserine, hypoxanthine, N-methylglutamate, and N,N-dimethylarginine.

TYMS knockout hiPSCs maintain pluripotency and genomic integrity

The TYMS−/−-hiPSC lines performed equally well on standard pluripotency analysis as their non-edited counterparts. They showed positive staining for pluripotency markers NANOG, OCT4, SOX2, and SSEA4 and comparable mRNA expression of SOX2, NANOG, and OCT4 to TYMS+/+-hiPSC (Figures S1G and S1H). Furthermore, their chromosomal integrity remained unaltered by the editing process in prolonged culture, tested in several time points for up to 25 passages (Figure S1I). The differentiation into the three germ layers with or without dTMP supplementation revealed that the cells successfully differentiate into ectoderm, mesoderm, and endoderm when supplemented with dTMP at the early stages of differentiation (Figures 2A–2C). Lack of supplementation from the TYMS−/− cells during days 0–14 produced significantly less embryoid bodies at day 14 (Figure 2B). By day 28, however, all attached cells from supplemented and non-supplemented cells expressed characteristic markers of the three germ layers (Figures 2C and S3B).Figure 2 Thymidine deficiency drastically reduces proliferation and teratoma formation in mice transplanted with TYMS−/−-hiPSCs

(A) Graphical representation of trilineage differentiation protocol. (B) Number of resulting 300-μm aggregates from wild-type and knockout hiPSC with and without supplementation during the first stage of differentiation (days 0–14). (C) Immunocytochemistry analysis of markers for ectoderm (β-3-tubulin), mesoderm (smooth muscle actin, SMA), and endoderm (SOX17) in cells derived from wild-type and knockout hiPSCs under dTMP supplementation during the first stage of differentiation. (D) Graphical representation of teratoma formation experiment. (E) H&E staining showing dense and cystic structures within the tumors. (F) Immunohistochemistry against TYMS. (G) Immunohistochemistry of characteristic markers of endoderm (SOX17), ectoderm (β-3-tubulin), and mesoderm (SMA). (H) Immunohistochemistry against Ki67 (proliferation marker, green) and total nuclei (Hoechst, blue). (I) Quantification of Ki67 index from immunohistochemistry, shown as positive nuclei over total nuclei. (J) TUNEL assay (apoptotic cells, green) and total nuclei (Hoechst, blue). (K) Quantification of TUNEL+ nuclei over total nuclei. Statistical significance in (B), (I), and (K) based on Wilcoxon test; ns p > 0.05; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. See also Figure S3.

Thymidine deficiency drastically reduces proliferation and teratoma formation in mice transplanted with TYMS−/−-hiPSCs

Next, we aimed to translate the in vitro trilineage differentiation to an in vivo setting. We implanted undifferentiated TYMS+/+- and TYMS−/−-hiPSC aggregates subcutaneously in mice, without or with dTMP supplementation (100 and 250 mg/kg/day) in the drinking water (Figure 2D). After 6 weeks, we retrieved the teratomas from the euthanized mice and analyzed their composition and proliferation (Figures 2E–2K).

TYMS+/+-hiPSC produced teratoma-like growth in all mice regardless of the thymidine supplementation, while TYMS−/−-hiPSCs produced visible teratomas only under dTMP supplementation. Only one mouse presented a small TYMS−/−-derived teratoma without supplementation (Figure 2E), composed mainly of primitive glial, epithelial, and mesenchymal tissue. As expected, immunohistochemical analysis of TYMS−/−-hiPSC-derived tumors was negative for TYMS, while TYMS+/+-hiPSCs present a rather homogeneous expression of the protein (Figure 2F). Additionally, analogous to the in vitro experiment, we could identify regions expressing markers for the three germ layers (Figure 2G) in all analyzed tumors. Importantly, although we did not find differences in the Ki67 index (Ki67+ nuclei over total nuclei) (Figures 2H and 2I), likely due to the reduced cell density in non-supplemented TYMS−/− cells (Figure S3C), we observed a significantly higher ratio of apoptotic cells from the TUNEL assay (as TUNEL+ nuclei over total nuclei) (Figures 2J and 2K).

In summary, we found that TYMS−/− cells produce proliferative teratomas in vivo, similar to those obtained with TYMS+/+ cells only under dTMP supplementation. Without it, however, the teratomas present lower cellular density and an increased percentage of apoptotic cells.

TYMS−/−-hiPSC cultures efficiently terminally differentiate in vitro

Once we ensured that the genetic disruption did not affect the pluripotency of our stem cells, we explored their directed differentiation into a potentially therapeutic cell type. In this context, diabetes represents a promising candidate for cell replacement therapy, in which stem cell-derived insulin-producing β cells act as the therapeutic source of insulin. We evaluated the safety, quality, and function of TYMS−/−-hiPSC-derived pancreatic islets using a well-defined, multi-step protocol for β cell differentiation (Figure 3A).5,6Figure 3 TYMS−/−-hiPSC cultures efficiently terminally differentiate in vitro

(A) Graphical representation of in vitro β cell differentiation from hiPSCs. Most tests were carried out S7W4, after 4 weeks from dTMP withdrawal. (B) mRNA expression analysis of TYMS (primers target exon 5) at different stages of differentiation. Results shown as ratio of TYMS mRNA over GAPDH mRNA. (C) mRNA expression analysis of maturation markers at different stages. Results shown as ratio of marker mRNA over GAPDH mRNA. (D) Dynamic insulin secretion in response to different stimuli at S7W4. (E) In vitro insulin secretion in response to low (2.8) and high (16.8) glucose concentration and glucose plus KCl (2.8 KCl) at S7W4 (E). (F) Immunohistochemistry and quantification of INS (magenta) and GCG (yellow) in hiPSC-derived islets at S7W4. (G) Immunohistochemistry and quantification of SYP (magenta) and SLC18A1 (yellow) in hiPSC-derived islets at S7W4. (H) Immunohistochemistry of Ki67 (magenta) in hiPSC-derived islets at S7W4 and quantification of proliferation index, as number of Ki67+ nuclei over the total nuclei. Statistical significance based on Wilcoxon test; ns p > 0.05; ∗p < 0.05; ∗∗p < 0.01; ∗∗∗∗p < 0.0001. Scale bar, 100 μM. See also Figure S4.

We differentiated TYMS−/− -hiPSCs (HEL24.3) in vitro, supplemented with thymidine (40 μM) through the whole differentiation experiment. Meanwhile, as expected, non-supplemented TYMS−/− cells did not survive thymidine withdrawal at early stages of differentiation, making further characterization of this condition impossible. Non-supplemented TYMS+/+-hiPSCs were used as controls. We did not observe differences between TYMS+/+-hiPSCs differentiated with or without dTMP supplementation, suggesting that the addition of dTMP at our working concentrations (up to 40 μM) does not affect TYMS+/+ cells (data not shown). Hereafter, we did not supplement TYMS+/+-hiPSCs during the differentiation.

We evaluated the progression of the differentiation by flow cytometry, qRT-PCR, and immunocytochemical analysis of cell identity markers at several developmental stages. With qRT-PCR analysis, we found that TYMS−/−-derived islets presented lower (close to zero) TYMS mRNA levels than their TYMS+/+ counterparts across all stages (Figure 3B). Nevertheless, both the TYMS−/−- and TYMS+/+-derived islets expressed similar levels of markers at stages definitive endoderm (DE) and pancreatic endocrine progenitor (S4), and markers of mature β cells at week 4 of stage 7 (S7W4) (Figure 3C). Accordingly, TYMS+/+- and supplemented TYMS−/−-hiPSCs showed comparable in vitro expression levels of markers at the DE stage (CXCR4), S4 (PDX1, NKX6.1), and endocrine maturation stage (S7) by flow cytometry and immunocytochemistry (Figure S4A–S4D).

dTMP withdrawal from differentiated TYMS−/−-β cell does not affect their function

At S7, we withdrew dTMP supplementation from the islets cultured for in vivo implantation. By the last stage before implantation (S7W4), TYMS−/− and TYMS+/+ hiPSC-derived islets showed similar in vitro levels of glucose-, exendin 4-, and KCl-induced functional insulin secretion at S7W4 (Figure 3D and 4E), and even after 3 months in culture (S7W16; Figure S4E). The TYMS−/−-hiPSC islets showed normal morphology and standard percentages of insulin, glucagon, synaptophysin (SYP), and SLC18A1-expressing cells (Figure 3F and 4G), while displaying a significantly decreased proliferative index by Ki67 analysis (Figure 3H). Thus, these results overall indicate that TYMS−/−-hiPSC lines efficiently differentiate to functional stem cell-derived (SC)-islets, comparable to their control counterparts, while displaying reduced proliferation upon dTMP withdrawal.

TYMS−/−-hiPSC-derived β cells without dTMP supplementation efficiently regulate blood glucose in vivo

We next evaluated the function of mature TYMS−/− β cells (S7) in vivo by engrafting them under the kidney capsule of immunosuppressed mice without dTMP supplementation (Figure 4A). We implanted mice with TYMS+/+- or TYMS−/−-hiPSC-derived islets, and monitored their body weight, glucose levels, and the grafts’ production of human C-peptide for 5 months.Figure 4 TYMS−/−-hiPSC-derived β cells without dTMP supplementation efficiently regulate blood glucose in vivo

(A) Diagram of S7W4 islet implantation setup. (B and C) Follow-up of human C-peptide secretion (in serum) and blood glucose in mice up to 5 months from implantation (month 0). (D) Blood glucose of mice under ITT at every 15 min for 1 h, at month 3. (E) Human C-peptide secretion (in serum) during ITT before and after 30 min and 1 h from insulin administration. (F) Quantification of area under the curves of human C-peptide secretion. (G) Immunohistochemistry of engrafted kidney 1 and 3 months postimplantation. Insulin (INS) (magenta), glucagon (GCG) (yellow). (H) Quantification of INS and GCG area from the islets in (G). (I) Immunohistochemistry against Ki67 in grafts after 1, 3, and 5 months. (J) Proliferation index, based on the quantification of Ki67+ nuclei over the total number of nuclei from (I). Statistical significance based on Wilcoxon test; ns p > 0.05; ∗p < 0.05; ∗∗p < 0.01. K, kidney. Scale bar, 50 μM. See also Figure S4.

Both TYMS−/− and TYMS+/+ control islets produced comparable levels of human C-peptide and lowered the mouse blood glucose (Figures 4B and 4C). All mice maintained their normal body weights (Figure S4F) and did not show any other signs of health concerns.

Three months postimplantation, we performed an insulin tolerance test (ITT) by treating the mice with insulin to test the ability of the grafts to shut down their bona fide insulin secretion in response to low glucose levels. Upon the decline of circulating glucose levels in response to the exogenously injected insulin (Figure 4D), our TYMS−/− grafts stopped their insulin secretion at rates comparable to those of their TYMS+/+ counterparts, suggesting an appropriate regulation of this process (Figures 4E and 4F).

Immunohistochemistry on the implanted grafts showed a comparable percentage of insulin- and glucagon-producing cells in TYMS+/+ and TYMS−/− grafts at 1, 3, and 5 months postimplantation (Figures 4G and 4H).

Additionally, we ran an immunohistochemical analysis against Ki67 to quantify the proliferation of the cells in the grafts in vivo through time (1, 3, and 5 months). Interestingly, despite finding no significant differences in functionality, we did observe a reduction in active proliferation of TYMS−/− cells (by Ki67 index) during the first 3 months after implantation (Figures 4I, 4J, and S4G).

To summarize, we demonstrate here that TYMS−/−-hiPSCs differentiate into pancreatic β cells and function normally in vitro and in vivo, as seen by prolonged regulated human insulin secretion.

Discussion

We have developed a method to generate controllable cells without inserting any external genetic elements. By disrupting the rate-limiting DNA-specific nucleotide synthesis reaction, we have obtained cells auxotrophic to thymidine. This novel safety mechanism allows TYMS−/−-hiPSCs to expand for undefined periods showing no signs of exhaustion, but only when externally supplemented with thymidine. In its absence, contrarily, the cells fail to proliferate. Moreover, this method does not affect the function of the therapeutic differentiated cells in vitro or in vivo.

The TYMS−/−-hiPSCs present only a few transcriptomic and metabolomic differences with their wild-type counterpart, largely related to the metabolism of the supplemented dTMP. They retain their pluripotency and differentiate successfully into different cell types. Non-supplemented TYMS−/−-hiPSCs generate small or no teratomas in vivo with a significantly increased percentage of apoptotic cells (TUNEL), pointing to the gradual loss of proliferative cells under physiological concentrations of thymidine. Previous reports by Diehl et al.34 suggest that thymidine withdrawal from TYMS−/− hiPSCs leads to replicative stress and cell-cycle arrest from nucleotide starvation. Indeed, dTMP withdrawal quickly affects the cell cycle of hiPSCs, with a concurrent increase in p21 expression, suggesting the activation of apoptosis processes or cell-cycle arrest.35,36 Supported by the results of our cell-cycle analysis, we hypothesize that cells that have not exited the cell cycle likely die soon after this replicative stress is sensed, at the S-G2 checkpoint.34,37 However, we observed an extended survival time of non-supplemented TYMS−/− cells when they form three-dimensional aggregates both in in vitro and in vivo differentiation experiments (embryoid bodies and teratoma formation, respectively). As TYMS activity is required for the synthesis of DNA but not that of RNA, this may hint at the recycling of thymidine from apoptotic cells until cell-cycle exit.38 In combination with the Ki67 analysis, the TUNEL assay on our teratomas may indeed suggest an active degradation of cells, which in turn could serve as a thymidine source for the Ki67+ cells in the absence of external supplementation.

Once out of the cell cycle, the cells may survive as differentiated progeny for extended periods. Accordingly, we see no functional alterations in the terminally differentiated cells without dTMP supplementation in vitro or in vivo. The hiPSC-derived pancreatic islets survived implantation and proved to successfully secrete human insulin and lower blood glucose up to the measured 5 months after implantation, without affecting the overall well-being of the mice.

The advances in basic science and technological developments simultaneously aid and challenge the development of clinical applications. On the one hand, improved gene and cell manipulation protocols have increased the efficiency and widened the applications of somatic cell and gene therapies, tissue-engineered medicines, and CRISPR-based products. On the other hand, the safety concerns of therapeutic cells with altered DNA challenge further clinical applications. Hence, strategies to make cellular products safer become a need. Ideally, a safety mechanism for cell therapies selectively targets cells that retain or have acquired the ability for dangerous proliferation, without affecting the functional therapeutic cells. Although attempts to control the activity and proliferation of the therapeutic cells have proven relatively efficient, they present some inherent disadvantages. As most safety switches alter DNA by including exogenous genetic material, they often present as complex, immunogenic, or susceptible to mutations or silencing, ultimately affecting the partial/remissive function of the system.25

In this regard, we believe our system brings added value to the efficiency and versatility of safety mechanisms. By a simple genetic manipulation of endogenous DNA, we generated cells auxotrophic toward thymidine. This approach not only evades the insertion of external genetic sequences but it also allows the control of cellular proliferation by the addition of a simple compound (thymidine). Thus, it facilitates the mass expansion and generation of therapeutic cells, which lose their proliferative capacity, but not their function, under in vivo concentrations of thymidine. Moreover, in vivo thymidine administration may control the function and efficiency of some therapeutic cells, such as immunotherapies, expanding the applications of this strategy.

The genetic disruption applied in this research brings about an additional advantage: the use of the TYMS locus as a genomic safe harbor. This opens the doors to future additions and combinatory developments with other selection and safety systems. We see this versatile feature as a solution to current limitations and an aid to expand the applications of TYMS disruption to other aims and other therapeutic cell types. We expect the replication and utilization of these results to contribute to the optimization of hiPSC-derived and other cellular products and to the future development of safer efficient cell-based therapies.

Materials and methods

Culture of A431, A549, HSC3

We cultured epidermoid carcinoma cell line A431 (RRID: CVCL_0037), adenocarcinoma human alveolar basal epithelial cells A549 (RRID: CVCL_0023), and human oral squamous carcinoma cell line HSC3 (RRID: CVCL_1288) cells in DMEM containing 10% fetal bovine serum (FBS), 100 μg/mL penicillin-streptomycin, and 2 mM GlutaMAX (all Thermo Fisher Scientific). After TYMS knockout, cell media were additionally supplemented with 5–20 μM dTMP (Thermo Fisher Scientific). We kept the cell lines at 37°C and 5% CO2, with media change every other day until splitting with TrypLE select (Gibco). All cell lines tested negative for mycoplasma.

IPSCs culture and characterization

The hiPSC line HEL24.3 (RRID: CVCL_9T96) was cultured on Matrigel (Corning)-coated plates in E8 medium (Thermo Fisher Scientific) and split using 0.5 mM EDTA. We kept the cell lines at 37°C and 5% CO2, with media change every other day. All cell lines tested negative for mycoplasma. After TYMS knockout, cell media were additionally supplemented with 5–20 μM dTMP. The karyotyping was carried out by Ambar Anàlisis Mèdiques (Barcelona, Spain) by G-banding. We treated edited (passages 20 and 45) and non-edited (passage 20) hiPSCs with Colcemid for 4 h, 37°C, 5% CO2 and prepared the cells as recommended by the service provider.

Genome editing

To knock out exon 5, we designed gRNA targeting I4 and I5 using online tools (https://benchling.com, CRISPOR39; Table S1). We generated the sgRNA by incubating our gRNA (customized Alt-R CRISPR-Cas9 gRNA, Integrated DNA Technologies [iDT]) for 5 min at 95°C with Alt-R CRISPR-Cas9 tracrRNA, ATTO 550 (iDT).

For each stem cell electroporation experiment, we dissociated 2 × 106 cells into single cells with StemPro Accutase (Thermo Fisher Scientific). We complexed the sgAlt-R S.p. HiFi Cas9 Nuclease V3 and both sgRNAs to form the functional ribonucleoprotein (RNP) and delivered this RNP into the cells, along the Alt-R Cas9 electroporation enhancer (all from iDT), by electroporation with Neon transfection systems (1,100 V, 20 ms, 2 pulses).

Cells were plated onto Matrigel-coated plates containing E8 with 5 μM ROCK inhibitor (Y-27632, Selleckchem) and 20–40 μM dTMP and incubated at 37°C, 5% CO2.

For the electroporation experiment with cell lines HSC-3, A549, and A431, cells were dissociated into single cells with TrypLE. sgRNAs were prepared as previously described and delivered to the cells using Neon transfection systems. For HSC3 and A549, the electroporation settings were 1250 V, 10 ms, 1 pulse, while for A431 we used 1450 V, 20 ms, 2 pulses. After transfection, the cells were plated onto plates containing culture media (10% FBS in DMEM) supplemented with 20–40 μM dTMP and incubated at 37°C, 5% CO2.

At 24–48 h after electroporation, we single-cell sorted ATTO550+ cells for monoclonal expansion in 96-well plates containing their corresponding dTMP-supplemented culture media (hiPSCs additionally supplemented with 10% Clone R [STEMCELL Technologies]). The media were refreshed every 72 h until splitting. We individually screened monoclonal colonies by PCR. All PCR products of the selected edited clones were validated by Sanger sequencing, along with the integrity of the top seven off-target sequences (Table S2).

Embryoid body (EB) differentiation

To test the pluripotency for an embryonal lineage differentiation, we grew the cells to 90%–100% confluency and performed an EB) differentiation assay as previously described.2 The final plated EBs were fixed with 4% paraformaldehyde (PFA)-PBS and analyzed by immunocytochemistry for characteristic markers of the three germ layers (see list below).

Teratoma assay

Knockout and wild-type HEL24.3 cells were cultured in ultra-low attachment plates, with E8 supplemented with 5 μM ROCK inhibitor for 24–48 h to form aggregates. The medium for knockout cells was additionally supplemented with 40 μM dTMP until 2 h before implantation. Groups of approximately 500 spheres were collected in syringe cannulas for subcutaneous implantation in the legs of the mice. The drinking water was supplemented with 0, 100, or 250 mg/kg/day dTMP (Table S3). At 6–8 weeks postimplantation, the teratomas were extracted and fixed for immunocytochemistry and H&E staining as described below.

dTMP auxotrophy

Wild-type and knock-out lines from each tested cell line were split onto Matrigel-coated 6-well plates at a reason of a 25,000 cells per well. Each well contained the corresponding medium for the cell line (see Culture of A431, A549, HSC3 and IPSCs culture and characterization) supplemented with 0–5 mM dTMP. The growth of the cells was followed using IncuCyte imaging systems (Sartorius), taking 49 pictures per well per time point. Cells were washed and media were refreshed daily to clear away dead cells debris and keep the thymidine concentration as constant as possible. For statistical significance, 3 biological replicates were plated per condition and 20 random images per well per time point were considered for quantification. The results are shown as average confluency per image ±SE.

β cell differentiation

The in vitro β cell differentiation was carried out as previously described.5,6 Briefly, TYMS knockout and control HEL24.3 hiPSCs were seeded onto Matrigel-coated 10-cm plates, as previously stated. The differentiation protocol was started 24 h postseeding, when the medium was changed to D0 medium. Both cell lines were supplemented with 0 or 40 μM dTMP until maturation. At S7, dTMP was withdrawn or kept until analysis (weeks 4–16 from the first day of S7). The media were refreshed every 2–3 days until analysis. Three independent rounds of differentiation were conducted, including 3–5 biological replicates per genotype each.

Flow cytometry

S4 cells and S7 SC-islets were dissociated with TrypLE for 10 min at 37°C and resuspended in 5% FBS-containing PBS. Fixation and permeabilization were done using Cytofix/Cytoperm (BD Biosciences, catalog no. 554714) for 20 min at room temperature (RT). Then, samples were incubated overnight with primary antibodies at 4°C, followed by secondary antibodies for 30 min at RT in Perm/Wash buffer (BD Biosciences, catalog no. 554714) supplemented with 5% FBS. The cells were run on the FACSCalibur cytometer (BD Biosciences); data were collected with CellQuest Pro version 4.0.2 (BD Biosciences) and analyzed with FlowJo version 10.8 software (BD Life Sciences). The antibodies are listed in Table S4.

For cell-cycle analysis, wild-type and knockout hiPSCs were plated overnight on 6-well plates containing E8 supplemented with ROCK inhibitor and 20 μM dTMP. Then, the media were refreshed, and for some plates, the dTMP supplementation was stopped. Cells from individual wells were collected at 0, 2, 4, 10, 16, and 24 h after dTMP withdrawal and fixed using 4% PFA in PBS. The same time points were collected for the supplemented plates. After fixation, the samples were washed and incubated 30 min with 10 μg/mL DAPI diluted in 1% Triton X-100 (Sigma-Aldrich) in PBS. Then, they were immediately analyzed using the NovoCyte Quanteon 4025.

Insulin secretion analysis in vitro

For the static analysis of insulin secretion, a total of 30–50 SC-islets were picked and incubated for 90 min in a 12-well plate containing 2.8-mM glucose (G2.8) in Krebs-Ringer buffer (KRB) for equilibration. This was followed by sequential 30-min incubations of G2.8, 16.8 mM glucose (G16.8) and G2.8 + 30 mM KCl in KRB. After each incubation, 200-μL samples were taken for further insulin secretion analysis. Once all samples were taken, the islets were collected for DNA quantification. Dynamic insulin secretion test was carried out using a perifusion apparatus (Brandel Suprafusion SF-06) at a 0.25-mL min−1 flow rate, sampling every 4  min. A total of 50 SC-islets were handpicked and perfused with KRB; the sample collection started after 90 min of equilibration in G2.8. The insulin content of secretion fractions and SC-islet lysates was analyzed with ELISA (Mercodia).

Islet in vivo characterization

Animal care and experiments were approved by the National Animal Experiment Board in Finland (ESAVI/9734/2021). NOD-SCID-Gamma (NSG, Jackson Laboratories, catalog no. 0055577) mice were housed in the Biomedicum Helsinki conventional facility in 12-h light/dark cycle and fed standard chow. SC-islet implantations were done following a previously established described protocol.4 A total of 32 mice between 6 and 18 months old were used in 3 individual implantation experiments, each including islets from 3 differentiation experiment replicates. Non-fasted blood samples were collected monthly from the saphenous vein for glucose measurement and C-peptide secretion analysis. For further characterization of the SC-islet graft, the engrafted kidney was removed after 1 or 3 months and processed as described below.

For the ITT at 3 months postimplantation, 3 mice implanted with TYMS−/− and 3 with TYMS+/+ islets were weighted and injected with insulin (0.75 IU/kg). Blood glucose was measured every 15 min for 1 h and serum samples for human C-peptide secretion analysis were measured at 30 and 60 min.

H&E staining

Teratoma-like growths were fixed using 4% PFA-PBS overnight and then embedded in paraffin and sectioned. Dried sections were deparaffinized and stained using a standard protocol. Briefly, they were deparaffinized using xylene (3 × 10 min) and gradually hydrated using decreasing concentrations of ethanol (99% EtOH 3 × 4 min, 96% EtOH 1 × 4 min, 70% EtOH 1 × 2 min). Then, they were incubated in hematoxylin for 2.5–3 min, washed in indirect flowing tap water for 5–10 min, followed by a 1-min incubation in ultrapure water. The samples were incubated in eosin for 2 min and dehydrated using increasing concentrations of ethanol (96% EtOH 3 × 10 s, 99% EtOH 2 × 2 min) and xylene (2 × 4 min). Slides were mounted with coverslips using mounting media and imaged the next day.

Immunocytochemistry, immunohistochemistry, and image analysis

Samples from hiPSCs were fixed using 4% PFA-PBS for 20 min and permeabilized using 1% Triton X-100 in PBS. Samples from S4 and S7 SC-islets were fixed with 4% PFA in PBS for 2 h, and explanted SC-islet grafts and teratoma-like growths were fixed overnight. After fixation, these samples were embedded in paraffin and sectioned.

For immunohistochemistry and immunocytochemistry, 5-μm sections were deparaffinized and subjected to heat-induced antigen retrieval in 0.1 mmol l–1 citrate buffer. The cells and tissue slides were blocked with UV-block (Thermo Scientific, catalog no. TA-125-PBQ) and incubated with primary antibodies in 0.1% Tween 20 overnight at 4°C with the given dilutions (Table S4). After washing, secondary antibodies diluted in a similar manner were incubated at RT for 1 h. For immunocytochemistry, secondary antibody incubation was done in the presence of Hoechst 33342 for nuclear staining. For immunohistochemistry, after the secondary antibody, the slides were incubated with a 3,3′-diaminobenzidine solution for 2–10 min until a brown color developed. Then, they were counterstained with hematoxylin for 2.5–3 min and dehydrated following a standard protocol (see “H&E staining” above).

The cells and immunocytochemistry slides were imaged using Apotome II from a Zeiss AxioImager at the Biomedicum Imaging Unit, with the same exposure and export setting used on all slides of each targeted marker. Immunohistochemistry slides were imaged using an Olympus BX51 microscope.

Images were processed in Zen2 Blue Edition version 2 (Zeiss) and analyzed using Fiji40 with pipelines adapted from a previous report.41 The same pipeline settings were used on all images of each targeted marker.

Protein expression analysis

mRNA expression levels were analyzed by qRT-PCR on three biological replicates (Tables S5–S7). TYMS protein expression was assessed by standard western blot. Briefly, protein was extracted by using radioimmunoprecipitation assay lysis buffer (20–188, Millipore) supplemented with cOmplete Mini protease and phosphatase inhibitors tablets (Roche). Samples were sonicated (30 s, 50% duty cycle pulse) and centrifuged for soluble protein extraction. Protein quantification was done using a Pierce BCA protein assay kit (Thermo Fisher), and 20- to 30-μg total protein were run in Mini-PROTEAN TGX Precast gels (4561033, Bio-Rad). Transfer to a nitrocellulose membrane was done using 25 V for 6 min in the iBlot2 Gel Transfer device (IB21001, ThermoFisher). Then, the membrane was probed with a rabbit polyclonal primary antibody against TYMS (1:2,500, Proteintech; Table S4). For detection, we used an anti-rabbit horseradish peroxidase (HRP)-coupled secondary antibody (1:5,000, Cell Signaling Technology) and Clarity Western ECL Substrate (1705061, Bio-Rad). As a loading control, we probed the membrane with a β-actin-HRP-coupled antibody (sc-47778, Santa Cruz).

Bulk RNA-seq

The 3′primed RNA-seq service was provided by the Biomedicum Functional Genomics Unit (FuGu) at the Helsinki Institute of Life Science and Biocenter Finland at the University of Helsinki. Wild-type and supplemented TYMS knockout cells were seeded in Matrigel-coated 6-well plates and expanded to 70%–80% confluency. Cells of 6 replicates each were washed twice with PBS, lysed, and collected using lipopolysaccharide binding protein (LBP, 740906.125, Macherey-Nagel), and purified using a NucleoSpin RNA isolation kit (Macherey-Nagel). Quality control, Illumina sequencing, and processing of the raw data were carried out by FuGu. Normalization and differential expression was done in R using the DESeq2 package42; enrichment was carried out using g:Profiler.43

Metabolomics analysis

Wild-type and supplemented TYMS knockout cells were seeded in Matrigel-coated 6-well plates and expanded to 70%–80% confluency. Cells from 5 replicates per genotype were collected and analyzed on Thermo Q Exactive Focus Quadrupole Orbitrap mass spectrometer coupled with a Thermo Dionex UltiMate 3000 high-performance liquid chromatography (HPLC) system (Thermo Fisher Scientific). The HPLC was equipped with a hydrophilic ZIC-pHILIC column (150 × 2.1 mm, 5 μm) with a ZIC-pHILIC guard column (20 × 2.1 mm, 5 μm, Merck Sequant). A total of 5 μL of the samples were injected into the LC-mass spectrometry after quality controls in randomized order, having every tenth sample as blank. A linear solvent gradient was applied for separation, in decreasing organic solvent (80%–35%, 16 min) at 0.15 mL/min flow rate and 45°C column oven temperature. The mobile phases included aqueous 200 mmol/L ammonium bicarbonate solution (pH 9.3, adjusted with 25% ammonium hydroxide), 100% acetonitrile, and 100% water. The ammonium bicarbonate solution was kept at 10% throughout the run, resulting in a steady concentration of 20 mmol/L. Metabolites were analyzed using a mass spectrometer equipped with a heated electrospray ionization source using polarity switching and the following settings—resolution of 70,000 at m/z of 200; spray voltages: 3,400 V for positive mode and 3,000 V for negative mode; sheath gas: 28 a.u., auxiliary gas: 8 a.u.; temperature of the vaporizer: 280°C; and temperature of the ion transfer tube: 300°C. The instrument control was conducted with Xcalibur 4.1.31.9 software (Thermo Scientific). The peaks for metabolites were confirmed with commercial standards (Sigma-Aldrich). The data quality was monitored using a parallel in-house quality control cell line extracted similar to other samples. The final peak integration was undertaken with the TraceFinder 4.1 SP2 software (Thermo Scientific), and the peak area data were exported as a Microsoft Excel file for further analysis. The peak area of each metabolite was normalized to the sum of the absolute peak areas of all metabolites in the same sample. These data were then analyzed in R using the limma package,44 and enrichment analysis was run in MetaboAnalyst 6.0.45

Statistical analyses

Statistical analyses for all figures were carried out in R. A non-parametric Wilcoxon test was performed to determine the p values and statistical significance for the comparisons between genotypes. Blinding of all samples (for immunocytochemistry, immunohistochemistry, serum samples, blood glucose measurements, etc.) was done by coding the sample names based on numbers (e.g., of paraffin block or animal). No results were excluded from analysis.

Data and code availability

All main data are available in the main text, figures, or the supplemental information. Requests for reagents and cell lines used in this study should be directed to the corresponding authors.

Supplemental information

Document S1. Figures S1–S4 and Tables S1–S7

Document S2. Article plus supplemental information

Acknowledgments

We thank Vaino Lithovius, M.D., and Nidhi Madhusudan, M.Sc. for assistance in image analysis; James Clubb, M.Sc., and Dr. Akseli Hemminki for the donation of the HSC3 and A549 cell lines; Sami Jalil, M.Sc., Diego Balboa, Ph.D., and Juan Cruz Landoni, Ph.D. for scientific discussion and support; and Hazem Ibrahim, M.Sc., and Eliisa Vahankangas, M.Sc. for assistance with animal experiments. We thank the Flow Cytometry Unit and the animal facility at Biomedicum Helsinki for their contribution to this study. We thank the Helsinki University Hospital research funding, the 10.13039/501100005744 Foundation for Pediatric Research , the 10.13039/501100007417 Paulo Foundation , the 10.13039/501100004325 Signe and Ane Gyllenberg Foundation , the 10.13039/501100004155 Magnus Ehrnrooth Foundation , the 10.13039/501100009067 K. Albin Johansson Foundation , the 10.13039/100007634 Ida Montin Foundation , the Finnish Red Cross Blood Service Research Fund, and the Diabetes Foundation for financial support.

Author contributions

R.S.-M., T.O., and K.W. conceptualized the study. R.S.-M., H.M., J.S.-V., and K.W. selected the methods used in the study;. R.S.-M., H.M., S.E., I.S., M.B., J.J., and J.S.-V. conducted all experiments and data collection included in the article, under the supervision of T.O. and K.W. R.S.-M. conducted the analysis and visualization of the data. R.S.-M., T.O., and K.W. acquired the main funding for the study. R.S.-M. and K.W. directed the project and wrote the original and revised drafts. H.M., J.S.-V., H.P., and T.O. further contributed in the reviewing and editing of the original draft.

Declaration of interests

R.S.-M. and K.W. are marked as inventors in a patent application by the University of Helsinki.

Supplemental information can be found online at https://doi.org/10.1016/j.ymthe.2024.06.014.
==== Refs
References

1 Takahashi K. Tanabe K. Ohnuki M. Narita M. Ichisaka T. Tomoda K. Yamanaka S. Induction of Pluripotent Stem Cells from Adult Human Fibroblasts by Defined Factors Cell 131 2007 861 872 10.1016/j.cell.2007.11.019 18035408
2 Weltner J. Balboa D. Katayama S. Bespalov M. Krjutškov K. Jouhilahti E.M. Trokovic R. Kere J. Otonkoski T. Human pluripotent reprogramming with CRISPR activators Nat. Commun. 9 2018 2643 10.1038/s41467-018-05067-x
3 Hanna J. Wernig M. Markoulaki S. Sun C.-W. Meissner A. Cassady J.P. Beard C. Brambrink T. Wu L.-C. Townes T.M. Jaenisch R. Treatment of Sickle Cell Anemia Mouse Model with iPS Cells Generated from Autologous Skin Science 318 2007 1920 1923 10.1126/science.1152092 18063756
4 Michiko M. Akira W. Yasuo K. Yasuhiko H. Chikako M. Takashi D. Masashi F. Hiroshi A. Noriko S. Yumiko S. Autologous Induced Stem-Cell–Derived Retinal Cells for Macular Degeneration N. Engl. J. Med. 376 2017 1038 1046 10.1056/nejmoa1608368 28296613
5 Balboa D. Barsby T. Lithovius V. Saarimäki-Vire J. Omar-Hmeadi M. Dyachok O. Montaser H. Lund P.E. Yang M. Ibrahim H. Functional, metabolic and transcriptional maturation of human pancreatic islets derived from stem cells Nat. Biotechnol. 40 2022 1042 1055 10.1038/s41587-022-01219-z 35241836
6 Barsby T. and T. Ibrahim H. Lithovius V. Montaser H. Balboa D. Vähäkangas E. Chandra V. Saarimäki-Vire J. Otonkoski T. Differentiating functional human islet-like aggregates from pluripotent stem cells STAR Protoc. 3 2022 101711 10.1016/j.xpro.2022.101711
7 D’Amour K.A. Agulnick A.D. Eliazer S. Kelly O.G. Kroon E. Baetge E.E. Efficient differentiation of human embryonic stem cells to definitive endoderm Nat. Biotechnol. 23 2005 1534 1541 10.1038/nbt1163 16258519
8 D’Amour K.A. Bang A.G. Eliazer S. Kelly O.G. Agulnick A.D. Smart N.G. Moorman M.A. Kroon E. Carpenter M.K. Baetge E.E. Production of pancreatic hormone–expressing endocrine cells from human embryonic stem cells Nat. Biotechnol. 24 2006 1392 1401 10.1038/nbt1259 17053790
9 Picken A. Harriman J. Iftimia-Mander A. Johnson L. Prosser A. Quirk R. Thomas R. A Monte Carlo framework for managing biological variability in manufacture of autologous cell therapy from mesenchymal stromal cells therapies Cytotherapy 22 2020 227 238 10.1016/j.jcyt.2020.01.006 32113873
10 Aboul-Soud M.A.M. Alzahrani A.J. Mahmoud A. Induced Pluripotent Stem Cells (iPSCs)-Roles in Regenerative Therapies, Disease Modelling and Drug Screening Cells 10 2021 2319 10.3390/cells10092319
11 Themeli M. Kloss C.C. Ciriello G. Fedorov V.D. Perna F. Gonen M. Sadelain M. Generation of tumor-targeted human T lymphocytes from induced pluripotent stem cells for cancer therapy Nat. Biotechnol. 31 2013 928 933 10.1038/nbt.2678 23934177
12 Zhao T. Zhang Z.-N. Rong Z. Xu Y. Immunogenicity of induced pluripotent stem cells Nature 474 2011 212 215 10.1038/nature10135 21572395
13 Maldonado R. Jalil S. Wartiovaara K. Curative gene therapies for rare diseases J. Community Genet. 12 2021 267 276 10.1007/s12687-020-00480-6 32803721
14 Maeder M.L. Gersbach C.A. Genome-editing Technologies for Gene and Cell Therapy Mol. Ther. 24 2016 430 446 10.1038/mt.2016.10 26755333
15 Yin P.T. Shah S. Pasquale N.J. Garbuzenko O.B. Minko T. Lee K.-B. Stem cell-based gene therapy activated using magnetic hyperthermia to enhance the treatment of cancer Biomaterials 81 2016 46 57 10.1016/j.biomaterials.2015.11.023 26720500
16 Xu H. Wang B. Ono M. Kagita A. Fujii K. Sasakawa N. Ueda T. Gee P. Nishikawa M. Nomura M. Targeted Disruption of HLA Genes via CRISPR-Cas9 Generates iPSCs with Enhanced Immune Compatibility Cell Stem Cell 24 2019 566 578.e7 10.1016/j.stem.2019.02.005 30853558
17 Ichise H. Nagano S. Maeda T. Miyazaki M. Miyazaki Y. Kojima H. Yawata N. Yawata M. Tanaka H. Saji H. NK Cell Alloreactivity against KIR-Ligand-Mismatched HLA-Haploidentical Tissue Derived from HLA Haplotype-Homozygous iPSCs Stem Cell Rep 9 2017 853 867 10.1016/j.stemcr.2017.07.020
18 Lee A.S. Tang C. Cao F. Xie X. van der Bogt K. Hwang A. Connolly A.J. Robbins R.C. Wu J.C. Effects of cell number on teratoma formation by human embryonic stem cells Cell Cycle 8 2009 2608 2612 10.4161/cc.8.16.9353 19597339
19 Hentze H. Soong P.L. Wang S.T. Phillips B.W. Putti T.C. Dunn N.R. Teratoma formation by human embryonic stem cells: Evaluation of essential parameters for future safety studies Stem Cell Res. 2 2009 198 210 10.1016/j.scr.2009.02.002 19393593
20 Wiebking V. Patterson J.O. Martin R. Chanda M.K. Lee C.M. Srifa W. Bao G. Porteus M.H. Metabolic engineering generates a transgene-free safety switch for cell therapy Nat. Biotechnol. 38 2020 1441 1450 10.1038/s41587-020-0580-6 32661439
21 Bonifant C.L. Jackson H.J. Brentjens R.J. Curran K.J. Toxicity and management in CAR T-cell therapy Mol. Ther. Oncolytics 3 2016 16011 10.1038/mto.2016.11
22 Lim T.-T. Geisen C. Hesse M. Fleischmann B.K. Zimmermann K. Pfeifer A. Lentiviral vector mediated thymidine kinase expression in pluripotent stem cells enables removal of tumorigenic cells PLoS One 8 2013 70543 10.1371/journal.pone.0070543
23 Chen F. Cai B. Gao Y. Yuan X. Cheng F. Wang T. Jiang M. Zhou Y. Lahn B.T. Li W. Xiang A.P. Suicide gene-mediated ablation of tumor-initiating mouse pluripotent stem cells Biomaterials 34 2013 1701 1711 10.1016/j.biomaterials.2012.11.018 23218839
24 Volarevic V. Markovic B.S. Gazdic M. Volarevic A. Jovicic N. Arsenijevic N. Armstrong L. Djonov V. Lako M. Stojkovic M. Ethical and Safety Issues of Stem Cell-Based Therapy Int. J. Med. Sci. 15 2018 36 45 10.7150/ijms.21666 29333086
25 Sahillioglu A.C. Schumacher T.N. Safety switches for adoptive cell therapy Curr. Opin. Immunol. 74 2022 190 198 10.1016/j.coi.2021.07.002 34389174
26 Anderson D.D. Quintero C.M. Stover P.J. Identification of a de novo thymidylate biosynthesis pathway in mammalian mitochondria Proc. Natl. Acad. Sci. USA 108 2011 15163 15168 10.1073/pnas.1103623108 21876188
27 Chen D. Jansson A. Sim D. Larsson A. Nordlund P. Structural analyses of human thymidylate synthase reveal a site that may control conformational switching between active and inactive states J. Biol. Chem. 292 2017 13449 13458 10.1074/jbc.m117.787267 28634233
28 Trokovic R. Weltner J. Otonkoski T. Generation of iPSC line HEL24.3 from human neonatal foreskin fibroblasts Stem Cell Res. 15 2015 266 268 10.1016/j.scr.2015.05.012 26093941
29 Traut T.W. Physiological concentrations of purines and pyrimidines Mol. Cell. Biochem. 140 1994 1 22 7877593
30 Ruttkay-Nedecky B. Nejdl L. Gumulec J. Zitka O. Masarik M. Eckschlager T. Stiborova M. Adam V. Kizek R. The Role of Metallothionein in Oxidative Stress Int. J. Mol. Sci. 14 2013 6044 6066 10.3390/ijms14036044 23502468
31 Girskis K.M. Stergachis A.B. DeGennaro E.M. Doan R.N. Qian X. Johnson M.B. Wang P.P. Sejourne G.M. Nagy M.A. Pollina E.A. Rewiring of human neurodevelopmental gene regulatory programs by human accelerated regions Neuron 109 2021 3239 3251.e7 10.1016/j.neuron.2021.08.005 34478631
32 Lai Y. Li X. Li T. Nyunoya T. Chen K. Kitsios G.D. Nouraie S.M. Zhang Y. McVerry B.J. Lee J.S. Endotoxin stabilizes protein arginine methyltransferase 4 (PRMT4) protein triggering death of lung epithelia Cell Death Dis. 12 2021 828 10.1038/s41419-021-04115-7 34480022
33 Fernández-Sáiz V. Targosz B.-S. Lemeer S. Eichner R. Langer C. Bullinger L. Reiter C. Slotta-Huspenina J. Schroeder S. Knorn A.-M. SCFFbxo9 and CK2 direct the cellular response to growth factor withdrawal via Tel2/Tti1 degradation and promote survival in multiple myeloma Nat. Cell Biol. 15 2013 72 81 10.1038/ncb2651 23263282
34 Diehl F.F. Miettinen T.P. Elbashir R. Nabel C.S. Darnell A.M. Do B.T. Manalis S.R. Lewis C.A. Vander Heiden M.G. Nucleotide imbalance decouples cell growth from cell proliferation Nat. Cell Biol. 24 2022 1252 1264 10.1038/s41556-022-00965-1 35927450
35 Overton K.W. Spencer S.L. Noderer W.L. Meyer T. Wang C.L. Basal p21 controls population heterogeneity in cycling and quiescent cell cycle states Proc. Natl. Acad. Sci. USA 111 2014 E4386 E4393 10.1073/pnas.1409797111 25267623
36 Rodriguez R. Meuth M. Chk1 and p21 Cooperate to Prevent Apoptosis during DNA Replication Fork Stress Mol. Biol. Cell 17 2006 402 412 10.1091/mbc.e05-07-0594 16280359
37 Saldivar J.C. Hamperl S. Bocek M.J. Chung M. Bass T.E. Cisneros-Soberanis F. Samejima K. Xie L. Paulson J.R. Earnshaw W.C. An intrinsic S/G2 checkpoint enforced by ATR Science 361 2018 806 810 10.1126/science.aap9346 30139873
38 Brock C.K. Wallin S.T. Ruiz O.E. Samms K.M. Mandal A. Sumner E.A. Eisenhoffer G.T. Stem cell proliferation is induced by apoptotic bodies from dying cells during epithelial tissue maintenance Nat. Commun. 10 2019 1044 10.1038/s41467-019-09010-6
39 Concordet J.P. Haeussler M. CRISPOR: Intuitive guide selection for CRISPR/Cas9 genome editing experiments and screens Nucleic Acids Res. 46 2018 W242 W245 10.1093/nar/gky354 29762716
40 Schindelin J. Arganda-Carreras I. Frise E. Kaynig V. Longair M. Pietzsch T. Preibisch S. Rueden C. Saalfeld S. Schmid B. Fiji: an open-source platform for biological-image analysis Nat. Methods 9 2012 676 682 10.1038/nmeth.2019 22743772
41 Lithovius V. Saarimäki-Vire J. Balboa D. Ibrahim H. Montaser H. Barsby T. Otonkoski T. SUR1-mutant iPS cell-derived islets recapitulate the pathophysiology of congenital hyperinsulinism Diabetologia 64 2021 630 640 10.1007/s00125-020-05346-7 33404684
42 Love M.I. Huber W. Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 Genome Biol. 15 2014 550 10.1186/s13059-014-0550-8 25516281
43 Kolberg L. Raudvere U. Kuzmin I. Adler P. Vilo J. Peterson H. g:Profiler—interoperable web service for functional enrichment analysis and gene identifier mapping (2023 update) Nucleic Acids Res. 51 2023 W207 W212 10.1093/nar/gkad347 37144459
44 Ritchie M.E. Phipson B. Wu D. Hu Y. Law C.W. Shi W. Smyth G.K. limma powers differential expression analyses for RNA-sequencing and microarray studies Nucleic Acids Res. 43 2015 e47 10.1093/nar/gkv007 25605792
45 Metaboanalyst 6.0 https://www.metaboanalyst.ca/home.xhtml 2024
46 Karjoo Z. Chen X. Hatefi A. Progress and Problems with the Use of Suicide Genes for Targeted Cancer Therapy Adv. Drug Deliv. Rev. 99 2016 113 128 10.1016/j.addr.2015.05.009 26004498
47 Berger C. Flowers M.E. Warren E.H. Riddell S.R. Analysis of transgene-specific immune responses that limit the in vivo persistence of adoptively transferred HSV-TK–modified donor T cells after allogeneic hematopoietic cell transplantation Blood 107 2006 2294 2302 10.1182/blood-2005-08-3503 16282341
48 Sheikh S. Ernst D. Keating A. Prodrugs and prodrug-activated systems in gene therapy Mol. Ther. 29 2021 1716 1728 10.1016/j.ymthe.2021.04.006 33831557
49 Achón Buil B. Tackenberg C. Rust R. Editing a gateway for cell therapy across the blood–brain barrier Brain 146 2023 823 841 10.1093/brain/awac393 36397727
50 Tiberghien P. Reynolds C.W. Keller J. Spence S. Deschaseaux M. Certoux J.-M. Contassot E. Murphy W.J. Lyons R. Chiang Y. Ganciclovir treatment of herpes simplex thymidine kinase-transduced primary T lymphocytes: an approach for specific in vivo donor T-cell depletion after bone marrow transplantation? Blood 84 1994 1333 1341 10.1182/blood.v84.4.1333.1333 8049449
51 Liang Q. Monetti C. Shutova M.V. Neely E.J. Hacibekiroglu S. Yang H. Kim C. Zhang P. Li C. Nagy K. Linking a cell-division gene and a suicide gene to define and improve cell therapy safety Nature 563 2018 701 704 10.1038/s41586-018-0733-7 30429614
52 Austin E.A. Huber B.E. A first step in the development of gene therapy for colorectal carcinoma: cloning, sequencing, and expression of Escherichia coli cytosine deaminase Mol. Pharmacol. 43 1993 380 387 8450832
53 Felmer R.N. Clark J.A. The gene suicide system NTR/CB1954 causes ablation of differentiated 3T3L1 adipocytes by apoptosis Biol. Res. 37 2004 449 460 10.4067/s0716-97602004000300009 15515969
54 Gadi V.K. Alexander S.D. Kudlow J.E. Allan P. Parker W.B. Sorscher E.J. In vivo sensitization of ovarian tumors to chemotherapy by expression of E. coli purine nucleoside phosphorylase in a small fraction of cells Gene Ther. 7 2000 1738 1743 11083495
55 Sato T. Neschadim A. Konrad M. Fowler D.H. Lavie A. Medin J.A. Engineered Human tmpk/AZT As a Novel Enzyme/Prodrug Axis for Suicide Gene Therapy Mol. Ther. 15 2007 962 970 10.1038/mt.sj.6300122 17375075
56 Pellegrini S. Zamarian V. Sordi V. Strategies to Improve the Safety of iPSC-Derived β Cells for β Cell Replacement in Diabetes Transpl. Int. 35 2022 10575 10.3389/ti.2022.10575
57 Serafini M. Manganini M. Borleri G. Bonamino M. Imberti L. Biondi A. Golay J. Rambaldi A. Introna M. Characterization of CD20-transduced T lymphocytes as an alternative suicide gene therapy approach for the treatment of graft-versus-host disease Hum. Gene Ther. 15 2004 63 76 10.1089/10430340460732463 14965378
58 Wang X. Chang W.C. Wong C.W. Colcher D. Sherman M. Ostberg J.R. Forman S.J. Riddell S.R. Jensen M.C. A transgene-encoded cell surface polypeptide for selection, in vivo tracking, and ablation of engineered cells Blood 118 2011 1255 1263 10.1182/blood-2011-02-337360 21653320
59 Kieback E. Charo J. Sommermeyer D. Blankenstein T. Uckert W. A safeguard eliminates T cell receptor gene-modified autoreactive T cells after adoptive transfer Proc. Natl. Acad. Sci. USA 105 2008 623 628 10.1073/pnas.0710198105 18182487
60 Birrer M.J. Moore K.N. Betella I. Bates R.C. Antibody-Drug Conjugate-Based Therapeutics: State of the Science J. Natl. Cancer Inst. 111 2019 538 549 10.1093/jnci/djz035 30859213
61 Straathof K.C. Pulè M.A. Yotnda P. Dotti G. Vanin E.F. Brenner M.K. Heslop H.E. Spencer D.M. Rooney C.M. An inducible caspase 9 safety switch for T-cell therapy Blood 105 2005 4247 4254 10.1182/blood-2004-11-4564 15728125
62 Liu Y. Yang Y. Suo Y. Li C. Chen M. Zheng S. Li H. Tang C. Fan N. Lan T. Inducible caspase-9 suicide gene under control of endogenous oct4 to safeguard mouse and human pluripotent stem cell therapy Mol. Ther. Methods Clin. Dev. 24 2022 332 341 10.1016/j.omtm.2022.01.014 35229007
63 Wunderlich S. Haase A. Merkert S. Jahn K. Deest M. Frieling H. Glage S. Korte W. Martens A. Kirschning A. Targeted biallelic integration of an inducible Caspase 9 suicide gene in iPSCs for safer therapies Mol. Ther. Methods Clin. Dev. 26 2022 84 94 10.1016/j.omtm.2022.05.011 35795779
