This week’s journal club focuses on a study from Diabetologia (2025) showing that the aberrant metabolite trafficking pathways seen in SC-islets are robustly diminished following engraftment in mice.
by Agnes Sandvik
Aim:
The authors start by point out that even though SC-islets have a great potential in cell based T1D therapy they are metabolically immature, most evidently in their deviating reactivity to pyruvate. They aim to investigate whether this immature metabolic phenotype will persist after engraftment into mice.
Methods:
They used H1 human embryonic stem cells through the Barsby et. Al differentiation protocol to produce pancreatic islets which they further implant under the kidney capsule of immunocompromised NOD-scid-gamma mice for 1-4 months. They also did metabolic and morphological assays of the islets before and after implantation, in parallel to cadaveric donor islets.
Results:
SC-islets implanted in mice take up to 3 months to humanize blood glucose levels
After producing the SC-islets they were implanted into the kidney capsules of mice at week 3-4 of the final differentiation stage. By three months the blood glucose levels of the mice reached the human fasting euglycemic level of below 5.6 mmol/L (Fig 1a). At the same time levels of human c-peptide circulating in the blood increased (Fig 1b). The graft function was assessed at 1 and 4 months through Intraperitoneal Glucose Tolerance Tests (IPGTTs) and they found that after four months the glucose clearance was faster (Fig 1c) and the level of c-peptide was greater (Fig 1d). Indicating functional maturation between 1 and 4 months.
Cell composition changes do not explain increased functionality post-engraftment
They wanted to understand how metabolism and/or other aspects of SC-islet biology contributed to the noted functional maturation in vivo as well as in comparison with non-transplanted primary human islets (Fig 1e).
To assess any contribution made by changes in cell composition they studied the percentage of α- and ϐ-cells from the overall endocrine cell pool (Fig 2a). ϐ-cell percentage increased by month 4, reaching 40-60 % as previously reported for human islets (Fig 2b). Alpha cells were already quite similar but had a transient increase around month 1 (Fig 2c) which they speculate could be due to slightly higher proliferation of α- compared to ϐ-cells in SC-islet month 1 grafts (Fig 2d&e). Almost no proliferating α- or β-cells were detected in human islets, and this percentage also decreased in SC-islets between months 1 and 4 (Fig. 2d and e).
Insulin granule morphology matures by 1 month post-engraftment
As they didn’t see any dramatic shifts in the endocrine cell composition after engraftment, they went on to investigate intrinsic factors that could potentially contribute to the functionality of the islets.
They started by studying insulin granules through transmission electron microscopy (TEM), categorizing ‘condensed’ and ‘diffuse’ granules as immature as they have previously been shown in human fetal pancreas at week 8-10.5 and ‘crystallized’ as mature as they did not appear in human fetal pancreas until week 14. Pre-engraftment the crystallized morphology was found in only 14 % of SC-islets, this increased by month 1 to almost 50 %, just like in month 4 and human islets (Fig 2g). Even though the change is great, they state that since it occurred before the large shifts in C-peptide it is unlikely to be a key determinant of functional maturity.
Mitochondrial content increases upon engraftment
To investigate other possible cell intrinsic maturation mechanisms, they looked at the protein levels of ϐ-cell maturation marker MafA and found that, as literature suggested, the proportion of ϐ-cells with high nuclear MafA levels increased after engraftment (Fig 3b). MafA levels were significantly higher in human islets than month 1 SC-islets (Fig 1b) and they thereby state it is a good ϐ-cell maturation marker, even though it is currently unknown how it relates to metabolic maturity.
They followingly moved on to more detailed studies of the mitochondria as they had previously reported that mitochondrial oxidative phosphorylation (OXPHOS) genes were upregulated post engraftment. They noticed that mitochondrial contents gradually increased after engraftment, matching the levels seen in human islets (Fig 3c&d). They also confirmed the increase in mitochondrial level through comparing immunostaining for the mitochondrial outer membrane marker TOMM20 in ϐ- vs non-ϐ cell types and found that the difference increased from 26 % to 57 % in month four grafts vs pre-engraftment.
The increased mitochondrial content was not accompanied by morphological changes in the mitochondria, human islets had smaller mitochondria and the cristae density stayed similar between different timepoints. They also noted significant heterogeneity between ϐ-cells of the same sample in both morphology and insulin granule maturity.
To address the heterogeneity, they estimated degree of ϐ-cell maturation based on the insulin granule crystallization status to compare mitochondrial morphology in differently matured ϐ-cells. Human islets mostly categorized as mature while SC-islets fell in the immature and semi-mature category (Fig 3e). Grafts from month 1 and month 4 fell between the two.
Β-cells with a more mature insulin granular status had an overall smaller mitochondrial size (Fig 3f&h), but the shape did not change much between the maturation categories (Fig 3i) and they saw no significant difference in cristae density (Fig 3j). They thereby conclude that mitochondrial morphological parameters correlate poorly with beta cell maturation.
Utilization of glucose into TCA-cycle intermediates increases upon engraftment
To address if the higher mitochondrial content was connected to increased mitochondrial metabolism they studied 13C labelled glucose in basal and stimulatory glucose concentrations with the most significant changes detected in the TCA-cycle intermediates in high glucose. Citrate, fumarate, α-ketoglutarate, malate and aspartate all increased significantly from in vitro to month 4 with a p-value between 0.005-0.0005. All of which showed increased percentage of labelled carbon incorporation following extended engraftment (Fig 4b-f). This occurred gradually and human-islet-like glucose incorporation levels were reached at the 4 month timepoint when glucose utilization became more concentration-dependent as seen by an increase in glucose incorporation under stimulatory glucose concentrations compared with basal levels (Fig 4b-f).
They concluded that carbon usage into the TCA-cycle increased post engraftment, and therefore they looked into whether this correlated with decreased insulin secretion in response to pyruvate.
Pyruvate reactivity decreases upon engraftment
To test glucose and pyruvate responsive insulin secretion they did dynamic insulin secretion assays in SC-islets, isolated month 1- and 4 grafts and human islets. They saw clear insulin secretion in response to KCl depolarization at the end of the tests indicating viable beta cells throughout the test (Fig 5a-d). They saw that SC-islets response to pyruvate was twice as high as their response to glucose (Fig 5a), the glucose response increased to match pyruvate in month 1 grafts (Fig 5b) and became twice as high at month 4 grafts (Fig 5c) just like for human islets (Fig 5d). They concluded that the relative sensitivity to pyruvate was heavily reduced upon engraftment.
To investigate if this was connected to mitochondrial metabolism, they followed the incorporation of 13C from pyruvate which decreased dramatically by 1 month after engraftment, similar to month 4 and human islets (Fig 5f).
They found that the monocarboxylate transporter MCT1 that transports pyruvate and lactate into the cells (Fig 5g) were expressed in SC-islets (Fig 5h) even though it is a disallowed gene in adult beta cells. Through immunostaining they also found that the MCT1 signal was lost post engraftment (Fig 5 i&j).
Conclusions
They conclude that they have successfully filled a knowledge gap as no previous study had given a detailed analysis of the metabolic maturation of SC-islets upon engraftment.
They state that through the study they have shown a sensitivity shift from pyruvate-responsive to glucose-responsive insulin secretion, which will be useful in clinical trials to avoid dysregulated insulin secretion and hypoglycemia.
Implanted SC-islets humanise mouse blood glucose after 2 months (from Vähäkangas et al., 2025)
Continue your reading here:
Vähäkangas E, Saarimäki-Vire J, Montaser H, Lithovius V, Eurola S, Ibrahim H, Kuuluvainen E, Balboa D, Katajisto P, Barsby T, Otonkoski T. Stem–cell–derived beta cells mature metabolically upon murine engraftment
Diabetologia. 2025 Jul 2;68(9):1997–2010. doi: 10.1007/s00125-025-06474-8