PEX11β controls early neural progenitor fate through regulation of peroxisomal fission and lipid metabolism Authors: C Bodnya, RP Theart, ME Colley, JM Spraggins, V Gama Published: 2026 ABSTRACT SummaryPeroxisomes are metabolic organelles that undergo fission to maintain peroxisome number for lipid homeostasis and cellular function. The peroxisomal fission machinery is shared with mitochondria, and although mitochondrial fission regulates neural cell fate decisions during human neurogenesis, it remains unclear whether peroxisomal fission independently influences human neural development. PEX11β selectively regulates peroxisomal fission by promoting membrane deformation, elongation, and downstream fission factor recruitment. To uncouple mitochondrial and peroxisomal fission in early human neurogenesis, we generated PEX11β knockout iPSCs and tracked differentiation into neural progenitors and neural rosettes. At the cellular level, perturbing PEX11β selectively impaired peroxisomal lipid metabolism, characterized by reduced ether-linked phospholipids. At the tissue level, PEX11β deficiency altered neural rosette architecture, resulting in enlarged central lumens and expanded neural progenitor populations. This human neural model uncouples mitochondrial and peroxisomal fission and positions PEX11β as a distinct regulator of lipid metabolism and neural progenitor fate during human neurogenesis.Keywords: peroxisomes, PEX11β, fission, DRP1, neurogenesis, induced pluripotent stem cells, neural progenitor cells Introduction Peroxisomes are specialized organelles that are essential for lipid metabolism, including the β-oxidation of very-long-chain fatty acids (VLCFAs) and the synthesis of plasmalogens, cholesterol bile acid intermediates, and docosahexaenoic acid (Brites et al., 2004; Ferdinandusse et al., 2009; Honsho and Fujiki, 2023; Krisans, 1992; Lazarow, 1978; Su et al., 2001). These metabolic functions are particularly important in the brain, where peroxisomal processes are key for myelin production and neural membrane integrity (Bottelbergs et al., 2010; Hulshagen et al., 2008; Kassmann, 2014; Kassmann et al., 2007; Teigler et al., 2009). In addition to peroxisome biogenesis, peroxisomes undergo fission to maintain sufficient peroxisome number for these critical metabolic functions. At the molecular level, current models posit that the peroxisomal biogenesis factor 11 beta (PEX11β) functions at multiple steps of the peroxisomal fission process, including peroxisomal membrane elongation, constriction, and fission by recruitment of dynamin-related protein 1 (DRP1) (Delille et al., 2010; Koch et al., 2003, Koch et al., 2004, Koch et al., 2010; Koch and Brocard, 2012). PEX11β initiates the peroxisomal fission process by mediating elongation of the peroxisomal membrane via its N-terminal amphipathic helix (Delille et al., 2010; Koch et al., 2010; Schrader et al., 1998, 2012). PEX11β self-interaction has also been suggested to mediate its ability to deform the peroxisomal membrane (Bonekamp et al., 2013; Yang et al., 2025). Some of the lipids required for peroxisomal membrane expansion are proposed to originate via lipid flow from the endoplasmic reticulum (ER) (Costello et al., 2017a; Hua et al., 2017), and cellular docosahexaenoic acid has also been shown to mediate peroxisomal elongation (Itoyama et al., 2012). Following peroxisomal elongation, the peroxisome undergoes constriction via mechanisms that have not been fully characterized; PEX11β is observed at sites of peroxisomal constriction and can constrict proteo-liposomes in vitro (Delille et al., 2010; Schrader et al., 1998; Yoshida et al., 2015). Next, PEX11β interacts with the downstream components of the peroxisomal fission machinery, the receptors mitochondrial fission factor (MFF) and mitochondrial fission 1 (FIS1), promoting their assembly at peroxisomes (Kobayashi et al., 2007; Koch et al., 2005; Koch and Brocard, 2012). Peroxisomal fission culminates in the recruitment of DRP1, which forms oligomeric rings around the peroxisomal membrane, resulting in final membrane scission in a GTPase-dependent manner (Kamerkar et al., 2018; Koch et al., 2003, 2004). PEX11β may interact with DRP1 to enhance its GTPase activity (Williams et al., 2015). It has further been shown that PEX11β mediates peroxisomal fission independently of MFF in a FIS1- and DRP1-dependent manner (Schrader et al., 2022). Individuals with novel variants in the PEX11β gene, which typically lead to loss of function, have been identified (Ebberink et al., 2012; Henning et al., 2024; Khoddam et al., 2024; Malekzadeh et al., 2021; Taylor et al., 2017; Tian et al., 2019; Tsinopoulou et al., 2023), extending the Zellweger spectrum disorder (ZSD) phenotype (MIM: 603867). Individuals with ZSDs, autosomal recessive disorders, typically present with metabolic dysfunction, such as accumulation of VLCFAs and reduction in plasmalogens, and multi-system phenotypes, including severe neurological symptoms such as developmental delay and seizures (Steinberg et al., 1993). Compared to most ZSDs, PEX11β-associated disease is characterized by a milder overall phenotype and relatively subtle metabolic alterations. Yet, the presence of seizures, neuropathy, and developmental delay in individuals with PEX11β mutations emphasizes the indispensable function of PEX11β during human neurodevelopment. Fibroblasts isolated from individuals with PEX11β mutations display elongated peroxisomal morphology (Ebberink et al., 2012); however, the effects of PEX11β deficiency on peroxisomal morphology and metabolic function in neural cells have not been explored. Given the established link between mitochondrial morphology and neural cell fate during human neurodevelopment (Iwata et al., 2020; Khacho et al., 2016), we sought to investigate whether a similar link exists between peroxisomal morphology and neural cell fate. Mitochondrial morphology has emerged as a driver of neural cell fate decisions. In a critical post-mitotic period during neural differentiation, if mitochondria in daughter cells elongate, the cells maintain a neural progenitor (NPC) identity; if mitochondria remain fragmented, the cells commit to an early-born neuronal cell fate (Iwata et al., 2020; Khacho et al., 2016). These changes in mitochondrial morphology are coupled with shifts in mitochondrial metabolism, which can also regulate the tempo of neural development (Iwata et al., 2023). However, the potential impact of peroxisomal morphology and metabolism have been overlooked in these studies. Often, perturbation of DRP1 is leveraged as a modulator of mitochondrial morphology; however, disrupting DRP1 function also impacts peroxisomal fission and potentially function. Unlike DRP1 perturbation, which simultaneously disrupts mitochondrial and peroxisomal fission, PEX11β loss offers a potential means to selectively impair peroxisomal fission, enabling identification of a peroxisome-specific mechanism regulating NPC fate. The neurodevelopmental effects of depleting Pex11β have been examined in vivo in mouse models, where Pex11β deletion is lethal (Li et al., 2002). Pex11β-deficient mice had severe neural migration defects, despite mild metabolic changes, and died shortly after birth (Li et al., 2002). Deletion of a single allele of the Pex11β gene in mice resulted in differentiation delays, impaired neural migration, and neural cell death (Ahlemeyer et al., 2012). Furthermore, when Pex11β was knocked down in vitro in mouse embryonic stem cells, a reduction in the NPC pool and impaired neural differentiation were observed after 14 days of retinoic acid treatment (Esmaeili et al., 2016). However, mouse and human neurodevelopment differ significantly, particularly in brain regions with abundant NPC pools, such as the outer ventricular zone (Molnár et al., 2019). Although individuals with PEX11β mutations present with neurological symptoms, brain magnetic resonance imaging (MRI) appears generally normal (Ebberink et al., 2012; Taylor et al., 2017), and the effects of PEX11β deficiency on early human neurodevelopment have not been rigorously investigated. In this study, we report the first model of PEX11β deficiency in human induced pluripotent stem cell (iPSC)-derived neural cells. Notably, this system selectively results in elongated peroxisomal morphology without disrupting mitochondrial morphology. At the molecular level, we found that PEX11β was required for the recruitment of the peroxisomal fission machinery in NPCs—the first quantification of this colocalization in a human neural system. Furthermore, we observed that peroxisomal elongation, a prerequisite for fission, is largely independent of the peroxisome-endoplasmic reticulum (ER) tether, contrasting other models where ER lipid flow was posited to lead to peroxisomal elongation, suggesting a distinct mechanism in neural cell types. At the cellular level, perturbation of PEX11β-mediated fission led to a reduction in ether-linked phospholipids, indicating impairment of peroxisomal metabolism. At the tissue level, defective PEX11β-mediated fission and disrupted ether-linked phospholipid metabolism resulted in enlarged lumens and increased NPC numbers, which may be linked to an imbalance in symmetric/asymmetric cell divisions and can result in downstream developmental deficiencies. Ultimately, our study establishes a link between elongated peroxisomal morphology, perturbed peroxisomal metabolism, and altered neural cell fate during human neurogenesis. Results PEX11β KO iPSCs are viable and maintain stem cell identity To model the effects of perturbing PEX11β function during neurogenesis, we knocked out PEX11β using CRISPR-Cas9 in human iPSCs (Figure S1A). Two clones containing single-base pair deletions in the PEX11β gene were selected for downstream analysis after Sanger sequencing. Two isogenic clones that underwent CRISPR-Cas9 but did not receive the desired edit were selected and used as controls. Controls and knockouts (KOs) were confirmed using whole genome sequencing, real-time qPCR, and liquid chromatography-mass spectrometry (LC-MS) to detect presence vs. absence of PEX11β peptides (Figures S1B–1D). Both control and PEX11β KO iPSCs maintained a normal karyotype (Figure S1E). Expression of the two additional PEX11 isoforms, PEX11α and PEX11γ, was unchanged in PEX11β KO iPSCs compared with control iPSCs (Figure S1C). Next, we validated whether knocking out PEX11β altered the stem cell identity of iPSCs. We first evaluated pluripotency by comparing the gene expression profiles of control and PEX11β KO clones with a reference dataset of over 450 genome-wide transcriptional profiles of validated pluripotent cells (PluriTest). Both control and PEX11β KO clones exhibited a transcriptional profile consistent with that of a pluripotent cell (Figure S2A). To complement the pluripotency test, we evaluated levels of the key stem cell identity markers octamer-binding transcription factor 4 (OCT4), NANOG, and sex determining region Y-box 2 (SOX2) using real-time qPCR, and of OCT4, NANOG, and SOX2 by immunofluorescence. There were no significant differences in stem cell identity markers between control and PEX11β KO iPSCs (Figures S2B and S2C). To determine whether control and PEX11β KO iPSCs differentiated into the three germ layers, we performed trilineage differentiation. Commitment to the endoderm lineage was evaluated using SRY-box transcription factor 17 (SOX17) and forkhead box A2 (FOXA2); commitment to the mesoderm lineage was evaluated using brachyury and neural cell adhesion molecule (NCAM); and commitment to the ectoderm lineage was evaluated using paired box 6 (PAX6) and Nestin. There were no significant differences in the ability of control and PEX11β KO iPSCs to differentiate into the three germ layers (Bodnya et al., 2026). These findings indicate that PEX11β KO iPSCs retain stem cell identity. PEX11β deficiency leads to elongated peroxisomal morphology in iPSCs, with no changes in mitochondrial morphology PEX11β deficiency leads to elongated peroxisomal morphology in mouse embryonic fibroblasts and in fibroblasts from individuals with PEX11β mutations (Ebberink et al., 2012; Li and Gould, 2002). To test whether loss of PEX11β similarly alters peroxisome morphology in iPSCs, we stained PEX11β KO iPSCs for PEX14, a core component of the peroxisomal matrix protein import machinery (Yamashita et al., 2020). After acquiring super-resolution images, we quantified peroxisomal major axis length, surface area, and volume by segmenting peroxisomes in 3D and extracting data from the resulting 3D mask (Joshi et al., 2020; Rasmussen et al., 2020; Robertson et al., 2023). PEX11β KO iPSCs displayed elongated peroxisomal morphology compared with controls, consistent with previous observations in other systems (Figures S2D and S2E). Because the mitochondrial and peroxisomal fission machineries and metabolic functions are tightly interconnected (Demarquoy and Le Borgne, 2015; Schrader, 2006; Schrader et al., 2015; Subramani et al., 2023), we examined whether knocking out PEX11β alters mitochondrial morphology. We examined mitochondrial morphology in control and PEX11β KO iPSCs using a pan-mitochondrial antibody and super-resolution imaging, with quantification of major axis length, surface area, and volume as described earlier. PEX11β KO iPSCs exhibited no detectable changes in mitochondrial morphology compared with control iPSCs (Figure S2F), indicating that PEX11β loss selectively affects peroxisomes without altering mitochondrial morphology. This finding is essential, as a key limitation of studies that perturb both mitochondrial and peroxisomal fission proteins is that the unique effects of perturbations on peroxisomal morphology and function are often unexplored. PEX11β deficiency transiently alters neural identity marker expression in NPCs To establish whether PEX11β deficiency affects neural cell fate, we differentiated PEX11β KO iPSCs into NPCs over the course of 8 days using dual SMAD inhibition (Chambers et al., 2009) (Figure 1A). PEX11β KO NPCs had no changes in the expression of the NPC identity markers PAX6, Nestin, or SOX2 and the intermediate progenitor marker t-brain gene-2 (TBR2) compared with control NPCs (Figures 1B–1D). However, the expression of beta-3-tubulin (TUBB3), a marker of early-born neurons, was reduced in PEX11β KO NPCs compared with control NPCs (Figure 1E), indicating a transient delay in neuronal differentiation. A similar phenotype is observed when mitochondrial fission is inhibited (Iwata et al., 2020; Khacho et al., 2016). However, by day 12 of NPC differentiation, there were no longer differences in expression of TUBB3 or other neural identity markers (Figures 1F–1H). Therefore, though PEX11β deficiency has an early effect on TUBB3 expression, PEX11β-deficient NPCs recover from or compensate for this effect as they differentiate. These findings contrast with previous observations in Pex11β-deficient mice, where depletion in the NPC pool and differentiation defects have been reported (Ahlemeyer et al., 2012; Esmaeili et al., 2016; Li et al., 2002), highlighting potential species-specific requirements for PEX11β during neurogenesis. PEX11β KO iPSCs differentiate into neural progenitors, with transient transcriptional changes during early neural differentiation (A) Schematic of neural progenitor (NPC) differentiation using dual-SMAD inhibition. (B) Representative maximum intensity projections and quantification of control and PEX11β NPCs at day 8 of differentiation, stained for the NPC identity marker PAX6. Scale bars, 25 μm (SDC). Each dot represents a normalized count per field of view (7 per n), each triangle represents an independent differentiation, and different colors represent different clones (n = 3). Light green, control clone 1; dark green, control clone 2; light blue, PEX11β KO clone 1; dark blue, PEX11β KO clone 2. Outliers removed using ROUT. Analyzed using Welch’s t test; error bars represent mean ± SEM. (C) Real-time qPCR analysis of NPC identity markers, PAX6, SOX2, and NESTIN, at day 8 of NPC differentiation. (D) Real-time qPCR analysis of intermediate progenitor identity marker, TBR2, at day 8 of NPC differentiation. (E) Real-time qPCR analysis of early born neuron marker, TUBB3, at day 8 of NPC differentiation. (F) Real-time qPCR analysis of NPC identity markers, PAX6, SOX2, and NESTIN, at day 12 of NPC differentiation. (G) Real-time qPCR analysis of intermediate progenitor identity marker, TBR2, at day 12 of NPC differentiation. (H) Real-time qPCR analysis of early-born neuron marker, TUBB3, at day 12 of NPC differentiation. For all real-time qPCRs, each dot represents an independent differentiation, consisting of the average of two technical replicates; different colors represent different clones (n = 3). Light green, control clone 1; dark green, control clone 2; light blue, PEX11β KO clone 1; dark blue = PEX11β KO clone 2. PEX11β KO NPCs were analyzed relative to control and normalized to two housekeeping genes (GPI and GAPDH). Analyzed using Welch’s t test; error bars represent mean ± SEM. PEX11β deficiency results in elongated peroxisomal morphology in NPCs To determine whether PEX11β deficiency leads to elongated peroxisomal morphology in iPSC-derived NPCs, we stained peroxisomes for the 70-kDa peroxisomal membrane protein, PMP70, a major component of peroxisomal membranes in mature, import-competent peroxisomes (Imanaka et al., 2000); PEX14 (Yamashita et al., 2020); and catalase, a key enzyme that breaks down the hydrogen peroxide formed as a byproduct of the first step of peroxisomal β-oxidation (Hashimoto and Hayashi, 1987). PEX11β KO NPCs had an elongated peroxisomal morphology compared with control NPCs when visualized with both peroxisomal membrane and peroxisomal matrix markers (Figures 2A–2F). Catalase-positive peroxisomes showed reduced major axis length relative to those positive for PMP70 or PEX14. This difference likely reflects the localization of catalase to the peroxisomal matrix, potentially underrepresenting peroxisomal dimensions captured by membrane-associated markers. To determine whether the elongated peroxisomal morphology in PEX11β-deficient NPCs altered enzymatic activity, a catalase activity assay was performed (Figure 2G). There were no significant differences in catalase activity between control and PEX11β KO NPCs. Overall, these findings demonstrate that PEX11β deficiency results in peroxisomal elongation that is maintained with neural differentiation, without impacting catalase activity. PEX11β KO NPCs have elongated peroxisomal morphology (A) Representative maximum intensity projections of peroxisomal morphology (PMP70) in control and PEX11β KO NPCs at day 8 of differentiation. Scale bars, 10 μm. Zoom scale bars, 2.5 μm (SoRa). (B) Quantification of peroxisomal major axis length, surface area, and volume in control and PEX11β KO NPCs (PMP70). (C) Representative maximum intensity projections of peroxisomal morphology (PEX14) in control and PEX11β KO NPCs at day 8 of differentiation. Scale bars, 10 μm. Zoom scale bars, 2.5 μm (SoRa). (D) Quantification of peroxisomal major axis length, surface area, and volume in control and PEX11β KO NPCs (PEX14). (E) Representative maximum intensity projections of peroxisomal morphology (catalase) in control and PEX11β KO NPCs at day 8 of differentiation. Scale bars, 10 μm. Zoom scale bars, 2.5 μm (SoRa). (F) Quantification of peroxisomal major axis length, surface area, and volume in control and PEX11β KO NPCs (catalase). For all peroxisomal morphology quantification, each dot represents the average per field of view (7 per n), each triangle represents an independent differentiation, and different colors represent different clones (n = 3). Light green, control clone 1; dark green, control clone 2; light blue, PEX11β KO clone 1; dark blue, PEX11β KO clone 2. Analyzed using Welch’s t test; error bars represent mean ± SEM. (G) Catalase activity (pmol/min) in control and PEX11β KO NPCs at day 8 of differentiation normalized to protein concentration (mg). Each dot represents an independent differentiation, and different colors represent different clones (n = 3). Light green, control clone 1; dark green, control clone 2; light blue, PEX11β KO clone 1; dark blue, PEX11β KO clone 2. Analyzed using Welch’s t test; error bars represent mean ± SEM. Changes in peroxisomal number and morphology are neural differentiation dependent Because peroxisomal fission is a mechanism by which cells maintain peroxisome number, cells deficient in PEX11β have been previously reported to have a reduced number of peroxisomes, in addition to their elongated morphology (Ahlemeyer et al., 2012; Ebberink et al., 2012; Yang et al., 2025). Notably, PEX11β KO iPSCs do not have fewer peroxisomes than control iPSCs (Figures S3A and S3E), while PEX11β KO NPCs have significantly fewer peroxisomes than control NPCs (Figures S3B–S3D; Figures 3F–3H). To our knowledge, this is one of the first reports of a PEX11β-deficient model that shows elongated peroxisomal morphology without a significant reduction in peroxisome number. However, this effect was differentiation-dependent, as peroxisome number was reduced in PEX11β KO NPCs but not in iPSCs. Therefore, we investigated how peroxisomal morphology, number, and PEX11β mRNA expression change in control cells during an 8-day differentiation from iPSCs into NPCs. Peroxisome major axis length, surface area, and volume were significantly increased at day 6 of differentiation and declined by day 8 (Figures 3A and 3B). Peroxisome number decreased by nearly half within the first two days of NPC differentiation and remained significantly lower than in iPSCs throughout the differentiation period (Figure 3C). These dynamic changes in peroxisomal morphology coincided with peaks in PEX11β mRNA expression at days 4 and 6 of differentiation (Figure 3D), and with maximal expression of the early NPC marker PAX6 at days 6 and 8 of differentiation (Figure 3E). Together, these data show that peroxisomal remodeling and PEX11β expression are tightly linked to acquisition of NPC identity, suggesting a requirement for PEX11β in peroxisomal fission and proliferation in neural cells but not in iPSCs. Peroxisome morphology and number change during NPC differentiation (A) Representative maximum intensity projections of peroxisomal morphology (PMP70) in control NPCs imaged every two days of differentiation for 8 days. Scale bars, 10 μm. Zoom scale bars, 2.5 μm (SIM). (B) Quantification of peroxisomal major axis length, surface area, and volume in control NPCs over 8 days of differentiation. Each dot represents the average per field of view (7 per n); each triangle represents the mean of an independent differentiation (n = 3). Analyzed using one-way ANOVA with Dunnett’s multiple comparisons; error bars represent mean ± SEM. (C) Quantification of peroxisomal count normalized to DAPI in control NPCs over 8 days of differentiation. Each dot represents the average per field of view (7 per n); each triangle represents the mean of an independent differentiation (n = 3). Analyzed using one-way ANOVA with Dunnett’s multiple comparisons; error bars represent mean ± SEM. (D) Real-time qPCR analysis of PEX11β gene expression during 8-day NPC differentiation. Analyzed relative to day 0 and normalized to two housekeeping genes (ACTB2 and GUSB). Each dot on the graph represents an independent differentiation. Analyzed using one-way ANOVA with Dunnett’s multiple comparisons; error bars represent mean ± SEM. (E) Real-time qPCR analysis of PAX6 gene expression during 8-day NPC differentiation, relative to day 8 of differentiation and normalized to two housekeeping genes (ACTB2 and GUSB). Each dot on the graph represents an independent differentiation. Analyzed using one-way ANOVA with Dunnett’s multiple comparisons; error bars represent mean ± SEM. PEX11β is required for recruitment of the peroxisomal fission machinery in NPCs Given our observation that NPCs may be more reliant on PEX11β-mediated fission than iPSCs, we tested whether PEX11β deficiency reduced the frequency of peroxisomal fission events in NPCs. To examine this, we live-imaged NPCs after transduction with CellLight Peroxisome-GFP, a BacMam vector that encodes GFP linked to the peroxisomal C-terminal targeting sequence (PTS1). We then analyzed the frequency of peroxisomal fission events in control and PEX11β KO NPCs using the mitochondrial event localizer (MEL) algorithm (Theart et al., 2020), applied here as a peroxisomal event localizer (PEL). MEL/PEL processes time-lapse 3D z stack images by operating on 3D instance-labeled organelle masks to detect topological split and merge events. The algorithm outputs per-frame fission event counts together with 3D event locations indicating where peroxisomal fission events are most likely to have occurred at distinct time points. Importantly, MEL/PEL is organelle-agnostic and does not rely on mitochondria-specific assumptions. Briefly, a fission event was defined when a single labeled peroxisomal structure at time t split into two (or more) distinct daughter structures at t + 1 and remained split at t + 2 (a conservative persistence criterion to suppress transient segmentation artifacts). Using this approach, live-cell imaging and PEL quantification of control and PEX11β KO NPCs showed that peroxisomal fission events were significantly reduced in PEX11β KO NPCs compared with control NPCs (Figures S4A–S4C; Videos S1 and S2). To identify whether PEX11β deficiency alters the recruitment of the downstream components of the peroxisomal fission machinery to peroxisomes, we imaged NPCs at day 8 of differentiation and quantified colocalization between PMP70 and MFF positive (+) or FIS1+ puncta. PEX11β KO NPCs displayed reduced colocalization between MFF and PMP70, as well as between FIS1 and PMP70, when compared with control NPCs (Figures 4A, 4B, 4D, and 4E, Figures S5A and S5B). Because colocalization between PMP70 and MFF and FIS1 (the receptors for DRP1) was decreased, we next tested whether there were changes in colocalization between PMP70 and DRP1+ puncta. PEX11β KO NPCs showed reduced colocalization between PMP70 and DRP1 (Figures 4G and 4H; Figure S5C). These findings support prior work identifying PEX11β as a key factor in recruiting MFF and FIS1 (Kobayashi et al., 2007; Koch and Brocard, 2012; Schrader et al., 2012, 2016). Given that MFF, FIS1, and DRP1 are shared components of the mitochondrial and peroxisomal fission systems (Schrader, 2006; Subramani et al., 2023), colocalization of these proteins with mitochondria was also assessed. No significant changes in the association of MFF, FIS1, or DRP1 with mitochondria were detected (Figures 4C, 4F, and 4I), indicating that the basal mitochondrial fission machinery remains intact in PEX11β-deficient cells. These results align with the absence of changes in mitochondrial network morphology in PEX11β KO iPSCs. Colocalization with downstream peroxisomal fission factors is reduced in PEX11β KO NPCs (A) Representative maximum intensity projections of MFF, mitochondria (MitoTracker), and peroxisomes (PMP70) in control and PEX11β NPCs at day 8 of differentiation. White arrowheads mark colocalization. Scale bars, 10 μm. Zoom scale bars, 2.5 μm (SoRa). (B) Quantification of MFF+ puncta that colocalize with PMP70+ puncta, normalized to total number of MFF+ puncta. (C) Quantification of MFF+ puncta that overlap with MitoTracker. (D) Representative maximum intensity projections of FIS1, mitochondria (MitoTracker), and peroxisomes (PMP70) in control and PEX11β NPCs at day 8 of differentiation. White arrowheads mark colocalization. Scale bars, 10 μm. Zoom scale bars, 2.5 μm (SoRa). (E) Quantification of FIS1+ puncta that colocalize with PMP70+ puncta, normalized to total number of FIS1+ puncta. (F) Quantification of FIS1+ puncta that overlap with MitoTracker. (G) Representative maximum intensity projections of DRP1, mitochondria (MitoTracker), and peroxisomes (PMP70) in control and PEX11β NPCs at day 8 of differentiation. White arrowheads mark colocalization. Scale bars, 10 μm. Zoom scale bars, 2.5 μm (SoRa). (H) Quantification of DRP1+ puncta that colocalize with PMP70+ puncta, normalized to total number of DRP1+ puncta. (I) Quantification of DRP1+ puncta that overlap with MitoTracker. For all graphs, each dot represents the average per field of view (7 per n), each triangle represents an independent differentiation, and different colors represent different clones (n = 3). Light green = control clone 1, dark green = control clone 2; light blue = PEX11β KO clone 1, dark blue = PEX11β KO clone 2. Analyzed using Welch’s t test; error bars represent mean ± SEM. Video S1. Peroxisomal fission events in control NPCs, related to Figure S5 Peroxisomal event localizer detection of peroxisomal fission events (red) in control NPCs. 3D time-lapse acquisition every 5 s for 15 min (final minute excluded due to photobleaching). Scale bars = 10 μm (SDC). Video S2. Peroxisomal fission events in PEX11β KO NPCs, related to Figure S5 Peroxisomal event localizer detection of peroxisomal fission events (red) in PEX11β KO NPCs. 3D time-lapse acquisition every 5 s for 15 min (final minute excluded due to photobleaching). Scale bars = 10 μm (SDC). PEX11β deficiency does not affect mitochondrial morphology or respiration in NPCs Although the mitochondrial association of fission proteins was unchanged in PEX11β-deficient NPCs, the tight metabolic connection between peroxisomes and mitochondria prompted examination of mitochondrial morphology. Furthermore, in overexpression systems, targeting of PEX11β to the mitochondria induced division (Schrader et al., 2022). To evaluate mitochondrial morphology in PEX11β-deficient NPCs, mitochondria were labeled with a pan-mitochondrial marker at day 8 of differentiation, imaged by super-resolution microscopy, and quantified for major axis length, surface area, and volume. No significant differences in mitochondrial morphology were detected between PEX11β KO and control NPCs (Figures S6A and S6B). To assess whether PEX11β deficiency affects mitochondrial respiration, oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured in NPCs at day 8 of differentiation using the Seahorse XF Analyzer. ECAR served as a proxy for glycolysis, whereas OCR reported oxidative phosphorylation. No differences in overall OCR or ECAR were detected between PEX11β KO and control NPCs (Figure S6C). Basal, maximal, and ATP-linked respiration were likewise unchanged in PEX11β KO NPCs, indicating preserved ATP production capacity (Figures S6D–S6G). PEX11β KO NPCs similarly did not have changes in proton leak and coupling efficiency (Figures S6H and S6I), demonstrating effective coupling of substrate oxidation to ATP synthesis. Thus, PEX11β-deficient NPCs exhibit normal mitochondrial morphology and respiration. These findings contrast with reports of secondary mitochondrial defects in Pex11β-deficient mice (Colasante et al., 2024) and mitochondrial fragmentation in PEX11β-deficient HEK293T cells carrying a pathogenic polyQ expansion (Kumar et al., 2025). Perturbing the peroxisome-ER tether leads to elongated peroxisomal morphology in control and PEX11β-deficient NPCs Our examination of the downstream peroxisomal fission machinery revealed impaired fission and elongated peroxisomal morphology in PEX11β-deficient NPCs. Although PEX11β has been proposed to mediate both membrane elongation and division, our findings, together with prior reports, indicate that loss of PEX11β predominantly results in an elongated peroxisomal morphology. Previously, it has been suggested that this phenotype may be due to continual lipid flow from the ER to peroxisomes, which is made possible via a tether between vesicle associated membrane protein-associated protein B (VAPB) at the ER and acyl-CoA binding domain-contacting protein 5 (ACBD5) at the peroxisomes (Costello et al., 2017a; Hua et al., 2017) (Figure 5A). To interrogate whether peroxisome-ER contacts, and consequently, ER lipid flow, contribute to the elongated peroxisomal morphology in PEX11β-deficient cells, we disrupted this tether via RNAi-mediated knockdown of ACBD5 in control and PEX11β KO NPCs at day 4 of differentiation, due to the peak in PEX11β mRNA expression at this time point. We hypothesized that if elongated peroxisomal morphology in PEX11β-deficient NPCs is a consequence of lipid flow from the ER, then ACBD5 knockdown would result in shorter (reduced major axis length) and smaller (reduced surface area and volume) peroxisomes. After confirming efficient ACBD5 knockdown (Figure 5B), peroxisomal morphology was assessed by super-resolution microscopy in regions with markedly reduced ACBD5 signal. In PEX11β KO NPCs, peroxisomal major axis length was unchanged between scramble controls and ACBD5-depleted cells (Figures 5C and 5D). Notably, we found increased peroxisomal surface area and volume in PEX11β KO NPCs in which ACBD5 was knocked down (Figure 5D). Thus, elongated peroxisomal morphology in PEX11β KO NPCs does not largely result from ER lipid flow and contrasts with phenotypes described in other models of impaired peroxisomal morphology, including MFF- and DRP1-deficient cells (Costello et al., 2017a; Darwisch et al., 2020; Hua et al., 2017). Furthermore, in control NPCs, we observed elongated peroxisomal morphology upon ACBD5 knockdown (Figures 5C and 5D), which was opposite to the effects reported in non-neural systems (Costello et al., 2017a; Darwisch et al., 2020; Hua et al., 2017). These findings indicate that elongated peroxisomal morphology in PEX11β-deficient NPCs is largely independent of ACBD5-dependent ER lipid transfer, while the effects of ACBD5 knockdown on control NPCs may reflect metabolic requirements unique to neural cell types. ACBD5 knockdown results in peroxisomal elongation in control and PEX11β KO NPCs (A) Schematic of the interaction between ACBD5 and VAPB at peroxisome-ER contact sites, which is mediated by the FFAT motif of ACBD5 and the MSP domain of VAPB (Kors et al., 2022). (B) Representative western blot of ACBD5 knockdown in control and PEX11β KO NPCs at day 4 of differentiation and quantification of western blot band density of ACBD5 normalized to β-actin loading control. Each dot represents an independent differentiation (n = 4); circles represent siSCR and triangles represent siACBD5. Analyzed using two-way ANOVA with uncorrected Fisher’s least significant difference (LSD) multiple comparisons; error bars represent mean ± SEM. (C) Representative maximum intensity projections of ACBD5 and PMP70 in control and PEX11β KO NPCs treated with either scramble control or ACBD5 siRNA at day 4 of NPC differentiation. Zoom scale bars, 2.5 μm (SoRa). (D) Quantification of peroxisomal major axis length, surface area, and volume in control and PEX11β KO NPCs at day 4 of differentiation, treated with either scramble control or ACBD5 siRNA. Each dot represents the average per field of view (7 per n), and each large circle/triangle represents an independent differentiation (n = 4); circles represent siSCR and triangles represent siACBD5. Light green, control clone 1; light blue, PEX11β KO clone 1. Analyzed using two-way ANOVA with uncorrected Fisher’s LSD multiple comparisons; error bars represent mean ± SEM. PEX11β-dependent peroxisomal fission regulates ether lipid and sphingomyelin metabolism in NPCs To assess the impact of PEX11β deficiency on lipid metabolism, global untargeted lipidomics was performed on NPCs at day 8 of differentiation. PEX11β KO NPCs had reduced levels of ether-linked phosphatidylcholine and phosphatidylserine species (Figures 6A–6E , Figures S7A and S7B). These results extend previous findings in PEX11β-deficient human fibroblasts, where reduced levels of phosphatidylethanolamine plasmalogens were observed (Abe et al., 2023). These data suggest disrupted plasmalogen synthesis in PEX11β-deficient NPCs, potentially indicating impaired peroxisomal lipid metabolism (Brites et al., 2004; Honsho and Fujiki, 2023). Furthermore, sphingomyelin species, potentially containing a very-long-chain side chain, were increased (Figures 6A and 6B, Figure S7A), in PEX11β KO NPCs, suggesting perturbations in peroxisomal β-oxidation of VLCFAs in PEX11β-deficient NPCs. PEX11β-deficient NPCs show reduced ether-linked phospholipids (A) Volcano plot showing downregulated (blue) and upregulated (magenta) lipids between control and PEX11β KO NPCs at day 8 of differentiation in positive ion mode. (B) Fold change and p values of significantly changed lipids between control and PEX11β KO NPCs at day 8 of differentiation in positive ion mode. Decreased lipids shown in blue, increased lipids shown in red. (C) Volcano plot showing downregulated (blue) and upregulated (magenta) lipids between control and PEX11β KO NPCs at day 8 of differentiation in negative ion mode. (D) Fold change and p values of significantly changed lipids between control and PEX11β KO NPCs at day 8 of differentiation in negative ion mode. Decreased lipids shown in blue, increased lipids shown in red. (E) Enrichment analysis of significantly changed lipid classes between control and PEX11β KO NPCs at day 8 of differentiation in positive and negative ion modes. PEX11β-dependent peroxisomal fission regulates NPC fate and neuroepithelial organization Given the lipidomic alterations in PEX11β-deficient NPCs and the established importance of ether-linked phospholipids in neurodevelopment (Teigler et al., 2009), we revisited the effects of PEX11β deficiency on neural differentiation. Although neural identity was unchanged between control and PEX11β-deficient NPCs in two-dimensional culture, this system lacks the polarized architecture of the neuroepithelium. Therefore, we differentiated control and PEX11β KO iPSCs into neural rosettes, which recapitulate the apical-basal organization of neuroepithelial cells during early neural tube formation (Elkabetz et al., 2008; Wilson and Stice, 2006). The tight junction marker zonula occludens-1 (ZO-1) localizes apically and marks the central lumen, with PAX6+ NPCs extending radially from the lumen. The NPC cells, which surround the lumen, have been shown to have similar identity to the cells found in the neural epithelial layer of the neural tube (Elkabetz et al., 2008). After 8 days of differentiation, neural rosettes were analyzed for identity and morphology, using PAX6 to quantify NPC number and ZO-1 to measure lumen area (Figure 7A). PEX11β KO neural rosettes exhibited a significantly increased number of PAX6+ NPCs and increased lumen area compared with control neural rosettes (Figures 7B–7E, Figures S8A–S8C). To assess whether differences in apoptosis or proliferation accounted for this phenotype, neural rosettes were stained for cleaved poly (ADP-ribose) (clPARP), a hallmark of apoptosis, and Kiel 67 (Ki67), a cell proliferation marker found exclusively in dividing cells (late G1, S, G2, and M phases). There were no significant differences in clPARP levels between control and PEX11β KO neural rosettes (Figures 7B and 7F). Furthermore, no differences in Ki67 were observed (Figure 7C). Thus, the enlarged lumen size and increased NPC pool in PEX11β KO neural rosettes are unlikely to be attributable to differences in apoptosis or cell proliferation. Collectively, these findings show that PEX11β deficiency has functional consequences for early neural patterning. PEX11β KO NPCs neural rosettes have increased lumen area and neural progenitor number (A) Schematic of neural rosette differentiation. (B) Representative maximum intensity projections of PAX6 (neural progenitors), clPARP (apoptosis), and ZO-1 (lumen) in control and PEX11β KO neural rosettes at day 8 of differentiation. Scale bars, 50 μm (SDC). (C) Representative maximum intensity projections of Ki67 (cells in phases S, G2, or M) and ZO-1 (lumen) in control and PEX11β KO neural rosettes at day 8 of differentiation. Scale bars, 50 μm (SDC). (D) Quantification of neural rosette lumen area (ZO-1) in control and PEX11β KO neural rosettes. Each dot represents the average per field of view (10 per n); each triangle represents an independent differentiation (n = 5). Light green, control clone 1; light blue, PEX11β KO clone 1. Outliers removed using ROUT. Analyzed using Welch’s t test; error bars represent mean ± SEM. (E) Quantification of the number of PAX6+ neural progenitors per neural rosette in control and PEX11β KO neural rosettes. Each dot represents the average per field of view (10 per n); each triangle represents an independent differentiation (n = 5). Light green, control clone 1; light blue, PEX11β KO clone 1. Analyzed using Welch’s t test; error bars represent mean ± SEM. Outliers removed using ROUT. (F) Quantification of clPARP normalized to number of nuclei in and PEX11β KO neural rosettes. Each dot represents the average per field of view (5–7 per n); each triangle represents an independent differentiation (n = 3). Light green, control clone 1; light blue, PEX11β KO clone 1. Outliers removed using ROUT. Analyzed using Welch’s t test; error bars represent mean ± SEM. Discussion PEX11β is a central regulator of peroxisomal fission, from early membrane elongation to recruitment of MFF and FIS1, the receptors for DRP1, which mediates final membrane scission (Koch et al., 2003, Koch et al., 2004, Koch et al., 2010; Koch and Brocard, 2012). Individuals with PEX11β mutations present with neurological symptoms, including seizures and developmental delay (Ebberink et al., 2012; Henning et al., 2024; Khoddam et al., 2024; Malekzadeh et al., 2021; Taylor et al., 2017; Tian et al., 2019; Tsinopoulou et al., 2023), yet how PEX11β deficiency affects peroxisomal metabolism and cell fate during early human neurogenesis has remained unexplored. Given the established relationship between mitochondrial morphology and neural cell fate decisions during early neurogenesis (Iwata et al., 2020; Khacho et al., 2016) and the fact that peroxisomes are also highly dynamic organelles, we sought to examine whether peroxisomal morphology and metabolism similarly drive neural cell fate decisions during human neurogenesis. This study provides the first analysis of PEX11β deficiency in a human iPSC-derived neural system, with two goals: (1) to define the cell biology of PEX11β loss in human neural cells and (2) to determine how elongated peroxisomal morphology influences metabolism and early neural cell fate. Consistent with prior reports, PEX11β KO iPSCs and NPCs displayed elongated peroxisomes. However, unlike most PEX11β-deficient systems, peroxisome number was unchanged in PEX11β-deficient iPSCs. During differentiation, peroxisomal morphology, abundance, and PEX11β expression were dynamically regulated in parallel with PAX6 expression, suggesting a developmental transition from de novo biogenesis to fission-dependent remodeling. iPSCs contained numerous small peroxisomes, whereas NPCs exhibited fewer, larger organelles, consistent with shifting metabolic demands analogous to the glycolysis-to-OXPHOS transition in mitochondria (Agostini et al., 2016; Iwata et al., 2020; Khacho et al., 2016; Zheng et al., 2016) These findings indicate that peroxisomal morphology is developmentally tuned during neural commitment. PEX11β-deficient NPCs showed reduced peroxisome fission frequency and reduced colocalization between peroxisomes and the downstream components of the peroxisomal fission machinery. These results highlight the critical role of PEX11β in peroxisomal fission. Because PEX11β has been reported to also participate in peroxisomal membrane elongation, the basis of the elongated peroxisomal morphology in PEX11β-deficient cells remains unresolved. A prevailing model attributes this phenotype to lipid flow from the ER to peroxisomes, which is mediated by the VAPB-ACBD5 tether (Costello et al., 2017a; Hua et al., 2017). However, disrupting this tether did not shorten peroxisomes in PEX11β-deficient NPCs; instead, ACBD5 knockdown increased peroxisomal surface area and volume in both PEX11β-deficient and control NPCs. These results point to a mechanism where peroxisomal elongation in PEX11β-deficient NPCs is largely independent of ER lipid flow. This unexpected elongated peroxisomal morphology following ACBD5 knockdown in control NPCs deserves further consideration. In non-neural systems, disrupting the VAPB-ACBD5 tether is thought to reduce ER-to-peroxisome lipid flow and results in shorter peroxisomes (Costello et al., 2017a; Darwisch et al., 2020; Hua et al., 2017), consistent with a model in which lipid delivery drives membrane expansion. The opposite finding in control NPCs after ACBD5 knockdown suggests that NPCs may rely on the VAPB-ACBD5 tether in a fundamentally different way, perhaps to regulate lipid efflux from peroxisomes. Alternatively, ACBD5 deficiency may alter fission factor recruitment, phenocopying PEX11β deficiency in this context. Distinguishing between these possibilities will require direct measurement of lipid flux across the ER-peroxisome interface in neural cell types and represents an important direction for future work. It is important to note, however, that the VAPB-ACBD5 tether is not the only ER-peroxisome contact site. The peroxisomal protein, acyl-CoA binding domain (ACBD4) also interacts with VAPB and may function as an additional tether (Costello et al., 2017b, 2023). Several other ER-peroxisome tethers have been identified in yeast systems (David et al., 2013; Knoblach et al., 2013; Schrader et al., 2015), and although these lack direct structural homologs in mammals, functional homologs could similarly contribute to tethering. Ultimately, we propose that continued import of peroxisome matrix and membrane proteins in the absence of efficient fission deforms the peroxisomal membrane, leading to elongation, and that altered lipid composition following ER-peroxisome tether disruption may further affect recruitment of fission factors. Following characterization of peroxisomal morphology and dynamics, we confirmed that PEX11β deficiency did not result in changes in mitochondrial morphology in iPSCs, nor changes in mitochondrial morphology or respiration in iPSC-derived NPCs. Hallmark studies examining the link between mitochondrial morphology and neural cell fate have often relied on genetic or pharmacological manipulation of DRP1, which is critical for both mitochondrial and peroxisomal fission. However, the effects of perturbing DRP1 on peroxisomal fission and metabolic function have remained largely unexplored in such studies. Therefore, our PEX11β-deficient iPSC and iPSC-derived neural model systems, which do not have changes in mitochondrial morphology or respiration, offer a means to study the independent contributions of elongated peroxisomal morphology on metabolism and cell fate decisions during early human neurogenesis. In vivo studies in mouse models linked complete Pex11β loss to neural migration defects and postnatal lethality, while deletion of a single allele of the Pex11β gene in vivo led to differentiation delays, disrupted neural migration, and neural cell death. (Ahlemeyer et al., 2012; Esmaeili et al., 2016; Li et al., 2002). These phenotypes are more severe than those observed in individuals with PEX11β mutations, who typically show normal brain structure by MRI (Ebberink et al., 2012; Taylor et al., 2017). However, in other neurodevelopmental disorders, it has been shown that alterations in early transcriptional programs during early neurogenesis can, in turn, alter cell fate decisions in NPC cells to change neurodevelopmental trajectories, resulting in consequences later in development (Jourdon et al., 2023). Additionally, human neurodevelopment differs markedly from mouse neurodevelopment, with expanded radial glial pools, greater interneuron diversity, and more complex circuitry (Molnár et al., 2019). Therefore, exploring PEX11β-deficiency during early neurogenesis in human systems is critical, as the neurological symptoms observed in individuals deficient in PEX11β may be a result of early cell fate dysregulation at the NPC stage. PEX11β-deficient NPCs showed reduced TUBB3 expression at day 8 of differentiation, reminiscent of other fission-impaired systems that bias progenitors toward self-renewal (Iwata et al., 2020; Khacho et al., 2016). However, by day 12 of NPC differentiation, there were no longer differences in TUBB3 expression in PEX11β-deficient NPCs, suggesting a transient differentiation delay at the transcriptional level. Lipidomic analysis of PEX11β-deficient NPCs at day 8 of differentiation revealed a reduction in ether-linked phospholipids and an increase in a sphingomyelin species with a VLCFA side chain. This suggests perturbations in peroxisomal metabolic functions in PEX11β-deficient NPCs, potentially pointing to deficiencies in the plasmalogen synthesis and β-oxidation pathways, for which peroxisomes are essential. In polarized neural rosettes, PEX11β deficiency produced enlarged lumens and expanded PAX6+ pools, consistent with altered symmetric vs. asymmetric divisions of the NPC pool (Elkabetz et al., 2008). Progenitor imbalances have also been found in Pex11β-deficient epithelia due to spindle misorientation (Asare et al., 2017), though whether this mechanism operates in PEX11β-deficient human iPSC-derived neural rosettes remains unclear. Our findings support a mechanistic model in which PEX11β mediates recruitment of the MFF-FIS1-DRP1 fission machinery to peroxisomes, enabling peroxisomal division during neural differentiation. Loss of this recruitment impairs peroxisomal fission and metabolic function, resulting in reduced ether-linked lipid synthesis. These metabolic alterations, in turn, may influence NPC abundance and neuroepithelial organization, establishing peroxisomal fission as a mechanistic regulator of NPC fate. Phenotypes were milder than in mouse models, suggesting species-specific sensitivity, yet early progenitor imbalance could still reshape developmental later trajectories without overt malformations. Several key open questions remain to be addressed. The primary driver of the elongated peroxisomal morphology in PEX11β-deficient systems remains poorly characterized, and dissecting this mechanism will be important for understanding the upstream regulation of peroxisomal morphology in neural contexts. Furthermore, it is critical to define the specific peroxisomal metabolic functions, whether plasmalogen synthesis, VLCFA oxidation, or other pathways, that underlie the neurodevelopmental phenotypes observed in our study. Finally, whether the early NPC imbalances we describe have lasting consequences for later neurodevelopmental stages remains an open and clinically relevant question, particularly given the neurological symptoms observed in individuals with mutations in PEX11β. Ultimately, leveraging this PEX11β-deficient iPSC system for differentiation into more developmentally mature human brain models, coupled with metabolic profiling in these systems, provides a powerful platform for elucidating the specific cell-fate changes that may contribute to neurological symptoms in individuals with PEX11β mutations. Eventually, these systems may serve as tools for testing potential therapeutic options. Limitations of the study One of the limitations of this study is that PEX11β deficiency results in reduction in peroxisome number, which creates difficulty when interpreting whether findings are the result of elongated peroxisomal morphology or reduced peroxisome number. We propose that the overall content of peroxisomes (i.e., peroxisome biogenesis proteins and matrix enzymes) in PEX11β-deficient peroxisomes may be similar to those in control peroxisomes, although this requires further validation. We also found that in PEX11β-deficient iPSCs, peroxisome number is not changed, offering a platform for investigating the morphological effects of PEX11β deficiency in a system where peroxisome number is not affected. We were also keen to quantify peroxisomal fission events, as these dynamics have not been explored live in neural systems. An important caveat of our PEL approach is that it relies on segmentation and spatial resolution, and some events scored as fission may not be true fission events but rather peroxisomes in proximity that later move apart. Additionally, to minimize photobleaching, peroxisomes were imaged at 5-s intervals. This system, therefore, primarily allows for reliable tracking of slow-moving peroxisomes (Rapp et al., 1996) and may underestimate the total number of fission events. To mitigate this limitation, peroxisomes were analyzed in 3D and fission events were manually verified, confirming a reduction in fission events in PEX11β KO NPCs compared with control NPCs. Future studies will focus on more thorough peroxisomal single-particle tracking, such as with structured-illumination microscopy in Leap mode, to achieve more optimal temporal and spatial resolution. Another limitation of this study is that the CRISPR-generated PEX11β KO represents a complete loss of function, whereas some affected individuals harbor hypomorphic or missense variants that may retain partial activity. Full KO could, therefore, exaggerate some phenotypes or fail to capture allele-specific mechanisms relevant to human disease. Future work will focus on rescue of peroxisomal fission via optimization of PEX11β overexpression constructs for NPC systems and supplementation of lipids, such as plasmalogens. Finally, the differentiation systems used in this study model early stages of human neurogenesis. In the future, complementing these studies with more developmentally mature model systems, such as iPSC-derived neural organoids, will be important to define the impact of PEX11β deficiency on later developmental processes. Resource availability Lead contact Additional information and resource requests can be directed to the lead contact, Gama V (vivian.gama@vanderbilt.edu). Materials availability This study generated PEX11β KO iPSC lines, which will be made available upon request. Data and code availability •The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium with the dataset identifier PXD072703. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium with the dataset identifier PXD072703. •For the mitochondrial/peroxisomal event localizer, the code is available for download at https://github.com/rensutheart/MEL-Fiji-Plugin. For the mitochondrial/peroxisomal event localizer, the code is available for download at https://github.com/rensutheart/MEL-Fiji-Plugin. Acknowledgments We thank Dr. Gabriella Robertson and Caleb Hayes for their technical expertise and assistance with experimental design and data interpretation. We are grateful to Drs. Matthew Tyska, Kevin Ess, and Elma Zaganjor for their critical feedback throughout the project, and to Dr. Kris Burkewitz for his insight and thoughtful review of the manuscript. We also thank Dr. Kristie Rose (Vanderbilt Mass Spectrometry Research Center) for mass spectrometry analysis of PEX11β peptides and Dr. Sergei Chetyrkin (Vanderbilt Mass Spectrometry Core Lab) for untargeted lipidomics data acquisition. Finally, we appreciate Kari Seedle’s, MS, expertise and guidance in high-resolution microscopy and image analysis. This work was supported by NIH grant 2R35GM128915-06 (to V.G), a Vanderbilt Seeding Success Grant (to V.G.), and NIH grant 1F31HD114431-01A1 (to C.B.). All Nikon SIM and SoRa microscopy was performed using the Vanderbilt Cell Imaging Shared Resource, supported by NIH grants CA68485, DK20593, DK58404, DK59637, EY08126, and 1S10MH130456-01A1. Flow cytometry experiments were performed in the VUMC Flow Cytometry Shared Resource, supported by the Vanderbilt-Ingram Cancer Center (P30 CA68485). Core Services performed through Vanderbilt University Medical Center's Digestive Disease Research Center were supported by NIH grant P30DK058404. Figures 1A, 2A, 6A, and 7A in the manuscript were created using Biorender.com. Author contributions C.B. and V.G. conceived the study, designed the experiments, interpreted the data, and wrote the manuscript. R.P.T. performed the PEL analysis. M.E.C. and J.M.S. conducted the lipidomics analysis. V.G. supervised and funded the research. C.B. designed and carried out all the cell biology experiments. All authors edited the document. Declaration of interests The authors declare that they have no conflict of interest. STAR★Methods Key resources table REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesPrimary Antibodies (Immunofluorescence) Mouse anti-OCT4 (1:200)Cell Signaling TechnologiesCat #75463; RRID: AB_2799870 Rabbit anti-NANOG (1:200)Cell Signaling TechnologiesCat #4903;RRID: AB_10559205 Mouse anti-PMP70 (1:200)Thermo Fisher ScientificCat #MA5-31368; RRID: AB_2787005 Rabbit anti-PEX14 (1:200)ProteintechCat #10594-1-AP; RRID: AB_2252194 Rabbit anti-Catalase (1:200)Cell Signaling TechnologiesCat #12980; RRID: AB_2798079 Mouse anti-mitochondria (1:200)AbcamCat #ab92824; RRID: AB_10562769 Rabbit anti-PAX6 (1:200)Cell Signaling TechnologiesCat #60433; RRID: AB_2797599 Mouse anti-PAX6 (1:200)BD BiosciencesCat #561462; RRID: AB_10715442 Rabbit anti-DRP1 (1:200)Cell Signaling TechnologiesCat #8570; RRID: AB_10950498 Rabbit anti-MFF (1:200)Cell Signaling TechnologiesCat #84580; RRID: AB_2728769 Rabbit anti-FIS1 (1:200)Cell Signaling TechnologiesCat #32525; RRID: AB_2927422 Rabbit anti-ACBD5 (1:200)Sigma-AldrichCat #HPA012145; RRID: AB_1844494 Mouse anti-Nestin (1:100)Stem Cell TechnologiesCat #60091; RRID: AB_2650581 Rabbit anti-SOX17 (1:200)Cell Signaling TechnologiesCat #81778; RRID: AB_2650582 Mouse anti-FOXA2 (1:100)BD BiosciencesCat #561580; RRID: AB_10714080 Rabbit anti-Brachyury (1:200)Cell Signaling TechnologiesCat #81694; RRID: AB_2799983 Mouse anti-NCAM (1:100)Stem Cell TechnologiesCat #60021; RRID: AB_2891082 Mouse anti-cleaved PARP (1:200)Cell Signaling TechnologiesCat #32563; RRID: AB_2799024 Rabbit anti-Ki67 (1:200)AbcamCat #ab92742; RRID: AB_10562976 Rat anti-ZO-1 (1:100)Thermo Fisher ScientificCat #14-9776-82; RRID: AB_2572026Secondary Antibodies (Immunofluorescence) Donkey anti-rabbit Alexa Fluor 488Thermo Fisher ScientificCat #A21206; RRID: AB_2535792 Donkey anti-mouse Alexa Fluor 568Thermo Fisher ScientificCat #A10037; RRID: AB_11180865 Donkey anti-mouse Alexa Fluor 647Thermo Fisher ScientificCat #A31571; RRID: AB_162542 Goat anti-mouse Alexa Flour 647 Fcγ subclass 1Jackson ImmunoResearch LaboratoriesCat #115-605-205; RRID: AB_2338916 Donkey anti-rat Alexa Fluor 647Thermo Fisher ScientificCAT #A78947; RRID: AB_2910635Dyes CellLight Peroxisome-GFP, BacMam 2.0Thermo Fisher ScientificCat #C10604 MitoTracker Red CMXRosThermo Fisher ScientificCat #M46752Primary Antibodies (Western blot) Rabbit anti-ACBD5 (1:1,000)Sigma-AldrichCat #HPA012145; RRID: AB_1844494 Mouse β-Actin (1:100,000)Sigma-AldrichCat #A1978-100uL; RRID:Secondary Antibodies (Western blot) Peroxidase AffiniPure Donkey Anti-Rabbit IgG (H + L) (1:10,000)Jackson ImmunoResearchCat #711-035-152; RRID: Peroxidase AffiniPure Goat Anti-Mouse IgG, Fcγ subclass 1 (1:10,000)Jackson ImmunoResearchCat #115-035-205; RRID:Chemicals, peptides, and recombinant proteinsY-27632 Rho/Rock pathway inhibitorStemCell TechnologiesCat #72307DorsomorphinMillipore SigmaCat #P5499SB431542ReprocellCat # 04-0010-10ACBD5 Silencer Select RNAiThermo Fisher ScientificCat #s40666Silencer Select Negative Control No. 1 siRNAThermo Fisher ScientificCat #4390844Lipofectamine RNAiMAX Transfection ReagentThermo Fisher ScientificCat #13778150Critical commercial assaysPluriTest AssayThermo Fisher ScientificCat #A38154KaryoStat AssayThermo Fisher ScientificCat #A38153Sanger sequencingGenhunterN/AWhole genome sequencingNovogeneN/ASTEMdiff Trilineage Differentiation KitStemCell TechnologiesCat #08581Seahorse Cell Mito Stress TestAgilentCat # 103015-100Seahorse XF DMEM medium pH 7.4AgilentCat #103575-100Seahorse XF 1.0 M glucose solutionAgilentCat#103577-100Seahorse XF 100mM pyruvate solutionAgilentCat#103578-100Seahorse XF 200mM glutamine solutionAgilentCat#103579-100Seahorse XF calibrantAgilentCat#100840-000Seahorse XF96 V3 PS cell culture microplatesAgilentCat#101085-004Catalase Assay KitCayman ChemicalCat #707002Deposited dataPEX11β peptide analysisProteomeXchangeDataset identifier: PXD072703Experimental models: cell linesPGP-1 (GM2338) iPSCsSynthegoRRID: RRID: CVCL_F182OligonucleotidesPEX11BFW: AGAAACAGATTCGACAACTGGAGRV: TGATAGGTGAACAGCTCTTTTGGPEX11AFW: AGCCTTGTCAGGGATCTGTATRV: CCACGCTGAACCAAAGAGGAPEX11GFW: GGGGACACGTCTGTTGGTGRV: ACAGGGGTAGTAGAGCTGGTCOCT4FW: GGGCTCTCCCATGCATTCAAACRV: CACCTTCCCTCCAACCAGTTGCNANOGFW: TGGGATTTACAGGCGTGAGCCACRV: AAGCAAAGCCTCCCAATCCCAAACSOX2FW: CCATGCAGGTTGACACCGTTGRV: TCGGCAGACTGATTCAAATAATACAGPAX6FW: TGGGCAGGTATTACGAGACTGRV: ACTCCCGCTTATACTGGGCTANESTINFW: CTGCTACCCTTGAGACACCTGRV: GGGCTCTGATCTCTGCATCTACTBR2FW: GTGCCCACGTCTACCTGTGRV: CCTGCCCTGTTTCGTAATGATTUBB3FW: GGCCAAGGGTCACTACACGRV: GCAGTCGCAGTTTTCACACTCGAPDHFW: ACAACTTTGGTATCGTGGAAGGRV: GCCATCACGCCACAGTTTCGPIFW: GTGTACCTTCTAGTCCCGCCRV: GGTCAAGCTGAAGTGGTTGAAGCACTB2FW: CATGTACGTTGCTATCCAGGCRV: CTCCTTAATGTCACGCACGATGUSBFW: CACCAGGGACCATCCAATACCRV: GCAGTCCAGCGTAGTTGAAAAASoftware and algorithmsFijiSchindelin et al., 2012Database: https://imagej.net/FijiStardistSchmidt et al., 2018Database: https://imagej.net/plugins/stardistGraphPad Prism v10GraphPadDatabase: https://www.graphpad.com/scientific-software/prism/NIS-Elements v.5.42.06Nikon InstrumentsDatabase:MetaboAnalyst 6.0(Chong and Xia, 2018; Chong et al., 2018, 2019; Xia and Wishart, 2010, 2011a, 2011b; Xia et al., 2009)Database: https://www.metaboanalyst.ca/home.xhtmlMS-DIAL v4.90(Tsugawa et al., 2020; Tsugawa et al., 2019; Lai et al., 2017; Tsugawa et al., 2015)sDatabase: https://systemsomicslab.github.io/compms/msdial/main.htmlAffinity DesignerAffinityDatabase: https://www.affinity.studio/ Experimental model and study participant details Cell lines The PGP-1 (GM2338) iPSC line was used in this study. The GM2338 (RRID: CVCL_F182) line is from a 55-year-old healthy, Caucasian male. The iPSCs were maintained on Matrigel-coated plates (Corning #354277) and cultured in StemFlex medium (Thermo Fisher Scientific #A3349401) at 37°C with 5% CO2. Cells were checked for differentiation and cell culture medium was changed daily. Every 3–4 days, the cells were passaged. For passaging, iPSCs were dissociated with Gentle Cell Dissociation Reagent (Stem Cell Technologies #100–0485) and incubated at room temperature for 4 min. Cells were cryopreserved in StemFlex medium (Thermo Fisher Scientific #A3349401) with 10% DMSO (Sigma-Aldrich #D2650) and supplemented with 10μM of Y-27632 Rho/Rock pathway inhibitor (Stem Cell Technologies #72307). All experiments were performed under the supervision of the Vanderbilt Institutional Human Pluripotent Cell Research Oversight (VIHPCRO) Committee. Cells were checked for contamination periodically, validated for pluripotency using the PluriTest Global Gene Expression Profile Assay (Thermo Fisher Scientific #A38154) and karyotyped every 20 passages. Method details Single-cell flow sorting PEX11β knockout (KO) iPSC pools (passage 33) were obtained from Synthego. Two hours prior to sorting, the media on the PEX11β KO iPSC pool was changed to StemFlex medium (Thermo Fisher Scientific #A3349401) supplemented with 10μM of Y-27632 Rho/Rock pathway inhibitor (Stem Cell Technologies #72307). One hour prior to sorting, 96-well plates (Corning #3596) were coated with Matrigel (Corning #354277). Thirty minutes before sorting, Cloning Medium (StemFlex +10% CloneR [StemCell Technologies #05888] + 1% Penicillin Streptomycin [Thermo Fisher Scientific #15-140-122]) was warmed to 37°C. To harvest cells for sorting, cells were first washed with DBPS (Thermo Fisher Scientific #14190144), followed by dissociation with Gentle Cell Dissociation Reagent (Stem Cell Technologies #100–0485) for 8 min at 37°C. After removal of Gentle Cell Dissociation Reagent, StemFlex +10μM of Y-27632 Rho/Rock pathway inhibitor was added, and the cells were gently scraped for harvesting. The cells were then transferred to a 15mL conical and centrifuged at 200xg for 2 min. After the supernatant was aspirated, the cells were resuspended in StemFlex + CloneR. 100μL/well of Cloning Medium was then added to the Matrigel-coated 96-well plates. Resuspended cells were passed through a 40μm filter (Pluriselect #435004051) immediately before sorting. Cells were single cell sorted using a 100μm nozzle, with 1 cell per well. Once sorting was complete, the plates were spun down at 120xg for 3 min to attach the cells to the bottom of the plate. The cells were incubated for 72 h, at which point a full media change was performed with Cloning Medium. 96 h after sorting, a full media change was performed with StemFlex medium without CloneR. Wells with a detectable iPSC colony were marked, while any wells containing more than 1 colony were avoided. Media changes were continued until the colonies were ready to passage (∼12 days after sorting). Colonies were passaged using the iPSC passaging procedures described above. Sample preparation and LC-MS for analysis of PEX11β peptides Coomassie-stained gel regions were excised and diced into 1mm3 cubes. Gel pieces were equilibrated in 100mM ammonium bicarbonate, reduced with 4.5 mM DTT at 56°C for 20 min, and alkylated with 10mM iodoacetamide for 30 min, protected from light (Tateishi et al., 2024). Gels were destained with 1:1 (v/v) 100% acetonitrile and 50mM ammonium bicarbonate and were dehydrated by addition of 100% acetonitrile. Following removal of acetonitrile, gel pieces were dried and in-gel digested with trypsin. Trypsin Gold (mass spectrometry grade, Promega #V5280) was diluted in 25mM ammonium bicarbonate (pH 8.0) to 10ng/μL and added to cover the dehydrated gel pieces, on ice. After 20 min, 25mM ammonium bicarbonate (10μL) was added to the gel pieces, and digestion was performed by incubation at 37°C overnight. The resulting tryptic peptides were recovered by 2 sequential extractions (15 min each) with 100μL of 60% acetonitrile/0.1% trifluoroacetic acid. Extracts were dried, reconstituted in 0.2% aqueous formic acid, and analyzed by LC-coupled tandem mass spectrometry (LC-MS/MS). Using a Dionex Ultimate 3000 nanoLC and autosampler, peptides were loaded onto a reverse phase analytical column (360 μm O.D. x 100 μm I.D.), packed with 20cm of C18 material as described previously (Tateishi et al., 2024). Peptides were gradient-eluted at a flow rate of 350nL/min, using a 90-min gradient. The mobile phase solvents consisted of 0.1% formic acid/99.9% water (solvent A) and 0.1% formic acid/99.9% acetonitrile (solvent B). The gradient was as follows: 2–38% B in 75 min; 38–90% B in 7 min; 90% B for 1 min; 90–2% B in 1 min; 2% B for 6 min. Peptides were analyzed using a data-dependent method on a Thermo Scientific Orbitrap Exploris 240 mass spectrometer, equipped with a nanoelectrospray source. The instrument method consisted of MS1 using an AGC target value of 3 × 106, followed by up to 20 MS/MS scans using an AGC target of 1 × 105. The intensity threshold for MS/MS was set to 1 × 104, HCD collision energy was 30 normal collision energy (nce), and dynamic exclusion (15 s) was enabled. For identification of peptides, tandem mass spectra were searched with Sequest (Thermo Fisher Scientific) against a Homo sapiens database created from the UniprotKB protein database (www.uniprot.org). Variable modification of +15.9949 on Met (oxidation) and +57.0214 on Cys (carbamidomethylation) were included for database searching. Search results were assembled using Scaffold 5.1.0. (Proteome Software), using a minimum filtering criteria of 95% peptide and protein probability. Peptides were also analyzed via data-independent acquisition (DIA) on a Bruker timsTOF HT mass spectrometer. Using a nanoElute2 LC system (Bruker), samples were loaded onto a reverse phase column, packed with 20cm of C18 material. The column was connected with a 20 μm CaptiveSpray emitter, and peptides were gradient-eluted at 500nL/min using a 40-min gradient (2–35% B in 30 min). diaPASEF data were acquired using 12 windows (ramps), spanning 0.75–1.3 V-s/cm2 1/k0 with an accumulation time of 100 ms. Spectronaut, version 20, (Biognosys) (Bruderer et al., 2015) was used for analysis of DIA data. A directDIA search was performed with default settings for identification, which included a false discovery rate (FDR) of 1% for peptides and protein groups. The precursor Qvalue and the experiment-wide protein Qvalue cutoffs were set to 0.01. Method evaluation was added in the experimental settings for the DIA analysis workflow. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (Perez-Riverol et al., 2025) partner repository with the dataset identifier PXD072703. Trilineage differentiation Trilineage differentiation was performed according to the manufacturer’s protocol (Stem Cell Technologies #05230). Briefly, when iPSCs reached 90% confluency, they were dissociated with Gentle Cell Dissociation Reagent (Stem Cell Technologies #100–0485) and incubated at 37°C with 5% CO2 for 8 min. They were then resuspended in StemFlex medium with 10μM Y-27632 Rho/Rock pathway inhibitor (Stem Cell Technologies #72307), scraped gently with cell lifters, and centrifuged at 300xg for 5 min. Cells were plated onto 12-well plates at 800,000 cells/well for the endoderm and ectoderm differentiations, and 200,000 cells/well for the mesoderm differentiation. 24 h later, media was changed to the lineage-specific media. Cells for mesoderm and endoderm differentiation were maintained at 37°C and 5% CO2 for 5 days, while cells for ectoderm differentiation were maintained at 37°C and 5% CO2 for 7 days. Media was changed daily. Differentiation of iPSCs into neural progenitors Neural progenitor cells (NPCs) were generated from iPSCs using dual SMAD neural induction using 10μM of SB431542 (Reprocell #04-0010-10) and 1μM dorsomorphin (Sigma Aldrich #P5499-5MG). When iPSCs reached 80–90% confluency, they were dissociated with Gentle Cell Dissociation Reagent (Stem Cell Technologies #100–0485) and incubated at 37°C with 5% CO2 for 8 min. DMEM/F12 (Thermo Fisher Scientific #11320033) with 10μM Y-27632 Rho/Rock pathway inhibitor (Stem Cell Technologies #72307) was then added to the cells, which were scraped with cell lifters and centrifuged at 300xg for 5 min. They were then resuspended in essential 6 medium (Lippmann et al., 2014, 2015) containing 10μM Y-27632 Rho/Rock pathway inhibitor (Stem Cell Technologies #72307) and SMAD inhibitors SB4 and dorsomorphin, single-celled further with a P1000 pipette, and plated at 2.0–2.5 × 106 cells/well of a 6-well plate or 1.0 × 106 cells/well of a 12-well plate. NPCs were incubated at 37°C and media was changed daily. At day 8 of differentiation, NPCs were collected for downstream processing. Differentiation of iPSCs into neural rosettes When iPSCs reached 80–90% confluency, they were dissociated using Gentle Cell Dissociation Reagent and incubated at 37°C with 5% CO2 for 8 min. DMEM/F12 (Thermo Fisher Scientific #11320033) with 10μM Y-27632 Rho/Rock pathway inhibitor (Stem Cell Technologies #72307) was next added to the cells, and the cells were detached with cell lifters prior to centrifugation at 300xg for 5 min. Cells were then resuspended in STEMdiff Neural Induction Medium (Stem Cell Technologies #05839) with 10μM Y-27632 Rho/Rock pathway inhibitor, further dissociated with a P1000 pipette, and plated at 3 × 106 cells/well in an AggreWell 800 plate (Stem Cell Technologies #34815) pre-treated with 500μL of Anti-Adherence Rinsing Solution for 5 min (Stem Cell Technologies #07010). Embryoid bodies were incubated at 37°C with 5% CO2, with minimal disruption, for 48 h. A 50% media change was performed with STEMdiff Neural Induction Medium on day 2. On day 4, embryoid bodies were harvested using a 40μm mesh cell strainer (Pluriselect #435004051) and transferred to 2, 35mm imaging dishes (Cellvis #D35-10-1.5-N) coated with Matrigel. Media changes were performed daily, and the cells were collected at day 8 of differentiation. RNAi transfection Commercially available siRNA targeting ACBD5 (Thermo Fisher Scientific #s40666) was used to transiently knock down ACBD5 in control and PEX11β KO NPCs. NPCs were generated as described above, and medium was changed on day 1 of neural differentiation. Cells were transfected per manufacturer protocol using Lipofectamine RNAiMax (Thermo Fisher Scientific #13778150) in Essential 6 medium on days 2 and 3 of neural differentiation. Silencer Select Negative Control No. 1 (Thermo Fisher Scientific #4390844) was used as a control. Cells were collected for downstream analysis at day 4 of differentiation. RNA extraction and cDNA synthesis Cells were scraped into 1000μL of Trizol reagent (Thermo Fisher Scientific #15596018) and flash frozen. After thawing, the samples were allowed to sit for 5 min at room temperature. Then, 200μL of chloroform was added (Sigma-Aldrich #C2432–4X25ML) and the samples incubated at room temperature for 2–3 min. The samples were then centrifuged at 12,000xg for 15 min at 4°C, followed by collection of the aqueous phase. Next, 500μL of isopropanol (Sigma-Aldrich #I9516-500ML) was added to precipitate RNA. The samples were incubated for 25 min at room temperature, followed by centrifugation at 12,000xg for 10 min at 4°C. The RNA pellet was washed with 75% ethanol, semi-dried, and resuspended in 30μL of nuclease-free water (Thermo Fisher Scientific #AM9937). Following quantification and adjustment of the volume of all samples to 1μg/μL, the samples were treated with DNAse (New England Biolabs #M0303), which was inactivated using EDTA (Sigma-Aldrich #324506-100ML). cDNA was generated using 10μL of this volume using the manufacturer’s protocol (Thermo Fisher Scientific #4368814). Quantitative RT-PCR (RT-qPCR) RT-qPCR was run on the QuantStudio 3 Real-Time PCR machine according to manufacturer instructions with 1μg of cDNA sample, SYBR Green Universal Master Mix (Thermo Fisher Scientific #4364346), and the primers outlined in the Key Resources Table. Western blot Cells were collected in lysis buffer containing PhosSTOP (Roche #04906837001), Protease Inhibitor Cocktail (PIC) (Roche #04693116001), and PMSF (RPI #P20270–10.0) diluted in 1% Triton X-100 (Sigma-Aldrich #T9284) in PBS. Cells were vortexed every 5 min for 30 min, followed by sonication for further cell lysis. 50μg of protein were loaded and run on 4–20% Mini-Protean TGX precast protein gels (Bio-Rad #4561094) in Tris-Glycine-SDS buffer. Gels were transferred onto polyvinylidene difluoride membranes (PVDF) (Sigma Aldrich #IEVH00005) at 0.31 amps for 1 h at room temperature. After transfer, the membrane was blocked in 5% milk (RPI #M17200500) in TBST for 1 h at room temperature on a benchtop rocker. The primary antibodies were diluted in 5% milk and incubated overnight at 4°C on a benchtop rocker. Blots were washed 3 times for 5 min with TBST, followed by incubation with HRP-conjugated secondary antibodies in 5% milk for 1 h at room temperature on a benchtop rocker. Blots were developed with ECL Plus Reagent (Thermo Fisher Scientific #32106) or SuperSignal West Femto (#Thermo Fisher Scientific #34095). They were then imaged on a CCD AI600 chemiluminescent imager. Band density was quantified using Fiji (ImageJ.net/software/fiji/). Data were normalized to the loading control, and data were transformed and normalized to the average of the three independent experiments of the scramble control. Catalase activity assay Cells were collected in lysis buffer containing PhosSTOP (Roche #04906837001), Protease Inhibitor Cocktail (PIC) (Roche #04693116001), and PMSF (RPI #P20270–10.0) diluted in 1% Triton X-100 (Sigma-Aldrich #T9284) in PBS. Cells were vortexed every 5 min for 30 min, followed by sonication for further cell lysis. Cell lysate was centrifuged at 10,000xg for 15 min at 4°C and the supernatant was used for the assay. The catalase activity assay was conducted according to manufacturer instructions. 20 μL of the sample was combined with 100 μL of Catalase Assay Buffer (1×) and 30 μL of methanol. The reaction was initiated via addition of 20 μL diluted hydrogen peroxide and the plate was incubated on a shaker for 20 min at room temperature. Next, 30 μL of potassium hydroxide was added to terminate the reaction, followed by addition of 30 μL of catalase purpald. The plate was then incubated on a shaker for 10 min at room temperature. Finally, 10 μL of catalase potassium periodate was added to each well and the plate incubated on a shaker for five minutes at room temperature. The absorbance at 540nm was measured using a plate reader. Catalase activity was calculated according to manufacturer instructions based on the standard curve established with the provided standards. Catalase activity was normalized to the protein concentration of each sample. Seahorse mito stress test Neural progenitor cells were plated onto Seahorse XF97 V3 PS cell culture microplates 2 days before the assay (day 6 of neural differentiation) at 80,000 cells per well. One hour before the assay, cell medium was changed to XF DMEM (Agilent #103575–100) containing 1mM pyruvate (Agilent #103578–100), 2mM L-glutamine (Agilent #103579–100), and 10mM D-glucose (Agilent #103577100). Oxygen consumption rate (OCR) was measured sequentially after addition of 1μM oligomycin, 0.5μM FCCP, and 0.5μM rotenone plus antimycin A (all included in Agilent #103015–100). Liquid chromatography-mass spectrometry for untargeted lipidomics To prepare samples, a one-phase method was used to extract lipids. Briefly, 0.5mL of methanol/methyl tert-butyl ether/chloroform mix (1.3:1:1) was added to a frozen pellet of NPCs (2 × 106 cells total), spiked with 10μL EquiSPLASH-lipidomics internal standard mix (Avanti Research), briefly vortexed and shaken gently for 20 min, and centrifuged at 20,000xg for 15 min at 10°C. The supernatant was transferred to a clean Eppendorf tube, evaporated under a gentle stream of N2 gas, resuspended in 100μL methanol/chloroform (9:1), and 2μL were used for LC-HRMS (high-resolution MS) analysis. Each sample was injected two times—one injection in positive electrospray ionization mode, followed by one in negative mode. Pooled QCs were injected to assess the performance of the LC and MS instruments at the beginning, in the middle and at the end of each sequence. Discovery lipidomics data were acquired using a Vanquish UHPLC (ultrahigh performance liquid chromatography) system interfaced to a Q Exactive HF quadrupole/orbitrap mass spectrometer (Thermo Fisher Scientific). Chromatographic separation was performed with a reverse-phase Acquity BEH C18 column (1.7μm, 2.1 × 150mm, Waters, Milford, MA) at a flow rate of 250μL/min and 50°C. Mobile phases were made up of 10 mM ammonium formate and 0.1% formic acid in (A) water/acetonitrile (40:60) and in (B) acetonitrile/isopropyl alcohol (10:90). Gradient conditions were as follows: 0–1 min, B = 20%; 1–8 min, B = 20–100%; 8–10 min, B = 100%; 10–10.5 min, B = 100–20%; 10.5–15 min, B = 20%. The total chromatographic run time was 15 min. Mass spectra were acquired over a precursor ion scan range of m/z 200 to 1,600 at a resolving power of 60,000 using the following HESI-II source parameters: spray voltage 4kV (3kV in negative mode); capillary temperature 250°C; S-lens RF level 60V; N2 sheath gas 40; N2 auxiliary gas 10; auxiliary gas temperature 350°C. MS/MS spectra were acquired for the most abundant precursor ions (Top7) with an AGC target of 1 × 105, a maximum injection time of 100 ms, and a normalized collision energy of 15, 30, 40. High resolution mass spectrometry data were processed with MS-DIAL version 4.90 in lipidomics mode. MS1 and MS2 tolerances were set to 0.01 and 0.025 Da respectively. Minimum peak height was set to 30,000 to decrease the number of false positive hits. Peaks were aligned with RT tolerance of 0.1 min, and mass tolerance of 0.015 Da. Default lipid library was used, solvent type was set to ammonium formate (HCOONH4) to match the solvent used for separation, and the identification score cut off was set to 80%. All lipid classes were made available for the search. After lipid annotation was completed as previously described (Takeda et al., 2024; Tsugawa et al., 2020), MS-DIAL results were exported into Excel and filtered to exclude lipids with an RSD for QC samples exceeding 20% and a QC to Blank below 10. For species identified in both KO and controls, mean levels (areas under peak curves) and their variance were analyzed in Excel. Normalized data were uploaded to MetaboAnalyst for analysis (https://www.metaboanalyst.ca/). Immunofluorescence iPSCs and NPCs were fixed with 4% paraformaldehyde in 1× PBS for 15 min at room temperature, followed by permeabilization with 1% Triton X-100 (Sigma-Aldrich #T9284) in 1× PBS for 5 min at room temperature. Cells were then washed 3 times with PBS, blocked in 10% BSA in 1× PBS, and incubated overnight at 4°C with primary antibody diluted in 10% BSA (Key Resources Table). Cells were washed 3 times with PBS followed by incubation with secondary antibody (Key Resources Table) diluted in 10% BSA for 1 h at room temperature. Following secondary antibody incubation, cells were washed an additional 3 times with 1× PBS, followed by incubation with Hoechst (1mg/mL) (Thermo Fisher Scientific #H3570) in 1× PBS for 10 min at room temperature. Cells were washed once with 1× PBS and sealed with Fluoromount-G slide mounting medium (Thermo Fisher Scientific #00-4958-02). For MitoTracker experiments, cells were incubated with MitoTracker Red CMXRos (Thermo Fisher Scientific #M46752) at a concentration of 100nm for 30 min at 37°C and 5% CO2 prior to fixation. Neural rosettes were fixed on day 8 of differentiation with 100% ice-cold methanol (Thermo Fisher Scientific #A454-4) for 10 min at −20°C. Next, they were washed with TBS and blocked at room temperature with 5% donkey serum (Jackson ImmunoResearch #017-000-121) + 0.3% Triton X-100 in tris-buffered saline (TBS), followed by overnight incubation at 4°C with primary antibodies diluted in blocking solution (Key Resources Table). After being washed three times with 1× TBS, cells were incubated with secondary antibody in 5% donkey serum +0.3% Triton X-100 in TBS for 2 h. After secondary antibody incubation, cells were washed an additional 3 times with 1× TBS and then incubated with Hoechst (1mg/mL) in TBS for 10 min at room temperature. After 1 additional wash, samples were mounted in Fluoromount-G slide mounting medium. Live imaging of peroxisomal morphology NPCs were generated from iPSCs as described above and plated onto Matrigel-coated 35mm imaging dishes (Cellvis #D35-10-1.5-N). At day 3 of neural progenitor differentiation, CellLight Peroxisome-GFP, BacMam 2.0 (Thermo Fisher Scientific #C10604) was added to cells at 5 particles per cell, and the cells were incubated at 37°C and 5% CO2. After 16 h, Peroxisome Cell Light was removed and the cells washed once with essential 6 medium. Prior to live cell imaging, cells were switched into FluoroBrite DMEM (Thermo Fisher Scientific #A1896701) with NucBlue (Thermo Fisher Scientific #R37605). For live imaging of peroxisomal morphology, cells were imaged on the Nikon W1 Spinning Disc Confocal with a 100× Plan Apo 1.49 NA oil objective and a prime 95B CMOS camera. Cells were imaged in a temperature-controlled chamber maintained at 37°C and 5% CO2. Image acquisition For imaging of stem cell identity, trilineage differentiation, neural progenitor identity, and neural rosettes, samples were imaged on the Nikon W1 Spinning Disc Confocal with a 20× Plan Apo 0.75NA objective, 60× Plan Apo 1.40 NA oil objective, or 100× Plan Apo 1.45 NA oil objective and a prime 95B CMOS camera. Super-resolution images of peroxisomal morphology, mitochondrial morphology, and colocalization of peroxisomal antibodies with antibodies marking peroxisomal fission proteins were acquired using a Nikon Super Resolution by Optical Pixel Reassignment (SoRa) microscope equipped with a 100× Plan Apo 1.45 NA oil objective and a Hamamatsu ORCA-Fusion BT camera. Imaging of peroxisomal morphology across neural differentiation was performed using a Nikon Structured Illumination Microscope (SIM) equipped with a 100× Apo TIRF 1.49 NA WD 0.12 oil objective and an Andor iXon Ultra DU-897 EMCCD monochrome camera. All imaging was done with at least 3 independent experiments and 7 images per replicate. Quantification and statistical analysis Image analysis Quantification of mitochondrial and peroxisomal morphology was performed in NIS-Elements version 5.42.06 (Nikon). Peroxisomes or mitochondria were segmented in 3D and data were extracted from the resulting 3D mask. Major axis length, surface area, volume, and count measurements were exported into Excel. Counts of DAPI, counts of transcription factors, and measurement of neural rosette lumen area were performed manually using Fiji (ImageJ.net/software/fiji/). Quantification of cleaved-PARP in neural rosettes was conducted using thresholding in Fiji with the watershed algorithm. For normalization to determine cleaved-PARP percentage, nuclei were counted using the Stardist plugin in Fiji (Schmidt et al., 2018). Peroxisomal event localizer (PEL) analysis A total of 31 cells were analyzed: n1 = 5 control and 4 PEX11β KO cells; n2 = 6 control and 5 KO cells; and n3 = 5 control and 6 KO cells. Time-lapse 3D z stack imaging was conducted over ∼180 timepoints with 10 z-slices every 5 s at a 0.1 μm step size and 100ms exposure. To assess potential photobleaching effects on measurements, analyses were performed on three temporal windows: (1) full acquisition (all ∼180 timepoints), (2) first 50 timepoints, and (3) first 100 timepoints. Automated event detection was performed using custom ImageJ macros implementing the Mitochondrial Event Localizer (MEL) (Theart et al., 2020), repurposed here as the peroxisomal event localizer (PEL). Importantly, MEL/PEL is organelle-agnostic: it detects topological split and merge events in 3D instance-labeled organelle masks (i.e., uniquely labeled 3D connected components) and does not rely on mitochondria-specific assumptions. For each time point, peroxisomal structures were segmented from the 3D z stack and converted to uniquely labeled 3D connected components (instance labeling) and tracked across sequential frames. Fission events were identified when a single labeled structure at time t split into two (or more) daughter structures at t + 1 and remained split at t + 2 (a conservative persistence criterion to suppress transient segmentation artifacts). Movies were not drift-registered, as peroxisomes are near-diffraction-limited and robust registration was unreliable; observed drift was gradual and MEL/PEL detects events via changes in instance topology rather than absolute position. Peroxisome counts at each time point were recorded for normalization. Fission frequency was calculated as the number of fission events per peroxisome per time point. To account for the hierarchical structure of the data and avoid pseudoreplication, statistical comparisons followed a two-stage approach. First, per-cell metrics were aggregated to compute replicate-level means, calculated separately for control and KO genotypes within each independent differentiation. Second, paired statistical tests compared KO versus control using these replicate means (n = 3 independent differentiations), treating each independent differentiation as the unit of analysis. Paired t-tests assessed normally distributed differences. Statistical analysis Experiments were performed with a minimum of 3 independent experiments (independent differentiations in the case of neural progenitors and neural rosettes) and 2 technical replicates (2 control clones and 2 PEX11β KO clones). Normality of data was determined using the Shapiro-Wilk test. Statistical significance was determined using a Welch’s t test, unless otherwise noted in the figure legend. GraphPad Prism v10 was used for all statistical analysis and data visualization. Error bars in all bar graphs represent the standard error of the mean unless otherwise indicated in the figure. Outliers were removed using the robust regression and outlier removal (ROUT) method where appropriate, which is included in the corresponding figure legends. For all statistical analyses, a statistically significant difference was established when p < 0.05. Published: September 17, 2026 Footnotes Supplemental information can be found online at https://doi.org/10.1016/j.stemcr.2026.103080. Supplemental information Document S1. Figures S1–S8 Document S2. Article plus supplemental information References Associated Data Supplementary Materials Video S1. Peroxisomal fission events in control NPCs, related to Figure S5 Peroxisomal event localizer detection of peroxisomal fission events (red) in control NPCs. 3D time-lapse acquisition every 5 s for 15 min (final minute excluded due to photobleaching). Scale bars = 10 μm (SDC). Video S2. Peroxisomal fission events in PEX11β KO NPCs, related to Figure S5 Peroxisomal event localizer detection of peroxisomal fission events (red) in PEX11β KO NPCs. 3D time-lapse acquisition every 5 s for 15 min (final minute excluded due to photobleaching). Scale bars = 10 μm (SDC). Document S1. Figures S1–S8 Document S2. Article plus supplemental information Data Availability Statement •The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium with the dataset identifier PXD072703. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium with the dataset identifier PXD072703. •For the mitochondrial/peroxisomal event localizer, the code is available for download at https://github.com/rensutheart/MEL-Fiji-Plugin. For the mitochondrial/peroxisomal event localizer, the code is available for download at https://github.com/rensutheart/MEL-Fiji-Plugin.