Human pluripotent stem cells (hPSCs) are a valuable tool for the study of the cellular and molecular mechanisms that underlie different types of diabetes. However, one pitfall of hPSCs is that genomic aberrations can develop during the reprogramming process as a result of gene editing or simply during extended cell culture (1). Some of these aberrations, such as trisomy 1, hamper embryonic development and lead to elimination of the fetus (2). Such detrimental aberrations acquired in hPSCs can cause genetic imbalance, thus affecting cellular identity and function. Therefore, given the fatal phenotypes observed in their in vivo manifestations, the use of hPSC lines that acquire such fatal genomic aberrations in disease modeling studies must be prevented.
In the article by Carrasco et al. (3), the authors used induced pluripotent stem cells (iPSCs) that carry abnormal chromosomal content to investigate molecular mechanisms underlying maturity-onset diabetes of the young type 1 (MODY1). They generated iPSCs from a patient with MODY1 (here termed MODY1-iPSCs) carrying a mutation in the HNF4A gene. They also corrected this mutation in MODY1-iPSCs, using gene editing, to serve as isogenic controls and then differentiated them into pancreatic β-cells to evaluate the effect of mutation on β-cell development and function (3). The differences observed by the authors using both two-dimensional and three-dimensional differentiation approaches varied based on the effect of the cellular environment on biological processes being switched on. Prior to differentiation, they used karyotyping analysis to examine the integrity of the generated iPSC lines. However, the results of the karyotyping were not explained in the text. Rather, the details of karyotyping analysis were mentioned in the supplementary material (Supplementary Fig. 7), which showed major chromosomal abnormalities in all cell lines examined. In their Supplementary Fig. 7, the karyotyping images for patient-derived iPSCs (trisomy 1 in two clones for all 20 fields/cells examined and isochromosome 1q [break p10] in addition to trisomy 1 in the third clone). This indicates that all generated clones commonly had full trisomy of chromosome 1 in the first two clones generated from the MODY1 patient and a mosaic trisomy 1 in the third one. In contrast, trisomy 1 was not observed when these patient-derived iPSCs underwent mutation correction by CRISPR-Cas9. Rather, the mutation-corrected clones showed mosaicism of deletion in chromosome 18. The authors correctly point out that trisomy 1 is identified as a recurring chromosomal abnormality in hPSCs (partial-chromosome or whole-chromosome gain or loss), as mentioned in their supplementary material results. However, the authors failed to address and investigate the effect of trisomy 1 on cellular and molecular profiles of the hPSCs used and how that may affect pancreatic development.
Chromosome 1, the largest of the human chromosomes, is known to contain around 8% of all human genetic data. More than 350 human diseases are associated with defects in the sequence of chromosome 1, and the gain of 1q in chromosome 1 is one of the most common chromosomal aberrations in human cancer (4). Trisomy 1 is one of the most rare trisomies to be recorded in humans, with no clinical report of survival beyond the first trimester of pregnancy (2). Trisomy 1 embryos were shown to undergo cell division and implantation; however, they failed to present a heartbeat and normal embryonic pole development (2). In rare cases, embryos with partial trisomy 1, where duplication of an arm of chromosome 1 occurs (both mosaic and nonmosaic), have a chance to survive; however, these individuals suffer from a myriad of developmental disorders, such as delayed neuropsychomotor development, facial malformations, intellectual disabilities, and, in most cases, cardiac defects (5,6). These findings indicate that trisomy 1 hinders embryo development. Since iPSCs represent the inner cell mass of the blastocysts, the iPSCs with trisomy 1 would not differentiate to the target cells normally. Since the MODY1-iPSCs with acquired trisomy 1 were compared by Carrasco et al. (3) to their isogenic mutation-corrected controls without trisomy 1 (normal copies of chromosome 1), the results cannot confirm if the defective phenotypes observed in the MODY1-iPSCs are due to the HNF4A mutation or trisomy 1. To circumvent this, the authors must validate their findings in a different hPSC model for HNF4A that has a normal karyotype.
No comprehensive study on the impact of trisomy 1 on hPSC transcriptome and differentiation capacities has been performed. Similar to trisomy 1, trisomy 12 is also identified as an acquired recurrent genomic aberration in hPSCs (1). A recent study highlighted the perils of employing such aneuploid hPSCs for disease modeling studies by performing transcriptome comparisons between hPSCs with trisomy 12 and diploid, normal hPSCs (7). The results showed striking dissimilarities between normal hPSCs and those with trisomy 12, revealing that trisomy 12 confers tumorigenicity to the hPSCs and increases their proliferation. Other studies reported that the pluripotency characteristics may not be changed in PSCs with genetic aberrations gained during culturing; however, these aberrations could have negative effects on their differentiation abilities or the function of the derived cells (8,9). A previous study demonstrated that neural stem cells with abnormalities in chromosome 1 fail to proliferate and integrate after implantation into rat brains, while normal neural stem cells can expand and engraft, highlighting the significance of chromosome 1 integrity (9). Therefore, the lessons learned from in vivo manifestation and consequences of trisomy 1 in human embryos as well as the in vitro studies on other recurrent chromosomal abnormalities strongly indicate that hPSCs with trisomy 1 are genetically unstable and unfit for application in disease modeling studies.
It is noteworthy that the occurrence of trisomy 1 in patient-derived iPSCs is not seen in the same iPSC lines after mutation correction (3). Considering the abovementioned points, these mutation-corrected iPSCs do not serve as appropriate isogenic controls for the MODY1-iPSCs due to chromosomal imbalance between the two groups.
It must be noted that a previous article published by Hua et al. (10), wherein similar experiments were performed for iPSCs generated from patients with maturity-onset diabetes of the young type 2, was retracted due to misrepresentation of the karyotyping results. This signifies the importance of ensuring normal chromosomal profiles for hPSC lines used in the experiments.
Importantly, comprehensive studies that assess the impact on transcriptome as well as functional properties of pancreatic β-cells that are derived from hPSCs with chromosomal gain or loss are warranted. Additionally, we emphasize that the in vivo manifestations of the acquired chromosomal abnormalities in hPSCs must be considered before their application, irrespective of their being recurrent genetic abnormalities in hPSCs.
See accompanying article, p. e3.
Article Information
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
