Adult ADHD, PARK2 copy number variations and the blood-brain barrier: Culture duration-dependent metabolic alterations in human induced pluripotent stem cell-derived in vitro models.
Fluids and barriers of the CNS · 2026
Abstract originale
BACKGROUND: Attention-deficit/hyperactivity disorder (ADHD) is a multifactorial neurodevelopmental disorder in which PARK2 copy number variations (CNVs) constitute a rare genetic risk factor. PARK2 encodes parkin, a key regulator of mitochondrial quality control. Mitochondrial dysfunction is implicated in ADHD pathophysiology and can contribute to blood-brain barrier (BBB) disruption. As BBB disruption remains insufficiently understood in ADHD, this study investigated whether adult ADHD and PARK2 CNVs are accompanied by alterations in BBB phenotype, mitochondrial properties, and cellular metabolism. METHODS: Brain capillary endothelial-like cells (BCECs) were generated using 10 donor-specific human induced pluripotent stem cell (hiPSC) lines from adults with ADHD carrying PARK2 CNVs, alongside PARK2 wild type adults with and without ADHD. BCECs were analyzed at the end of differentiation (day 10) and after prolonged culture (day 17) to assess group-associated phenotypes and culture duration-dependent differences. RESULTS: Across all groups, BCECs showed brain endothelial marker expression, high transendothelial electrical resistance (TEER), and low sodium fluorescein (NaFl) permeability at day 10. After prolonged culture, NaFl permeability remained low, whereas TEER-based barrier integrity declined. Prolonged culture of BCECs reduced glycolytic ATP production rates while increasing mitochondrial ATP production and its relative contribution to total ATP production, reflecting a metabolic shift from glycolysis toward oxidative phosphorylation. Despite decreased total ATP production after prolonged culture, cellular energy status was largely preserved, alongside overall lower intracellular reactive oxygen species (ROS) and mitochondrial membrane potential. At day 10, BCECs from ADHD donors carrying PARK2 duplications displayed greater mitochondrial abundance, the highest basal and oxidative stress-induced intracellular ROS levels, and reduced TEER-based barrier integrity, while at day 17, BCECs from ADHD donors with PARK2 deletions exhibited the least glycolysis-dependent ATP production profile, although absolute mitochondrial ATP production remained comparable between groups. CONCLUSIONS: Prolonged culture was associated with broad, ADHD- and PARK2 genotype-independent metabolic remodeling, whereas mitochondrial and metabolic alterations associated with ADHD and PARK2 CNVs were overall small. This study identifies culture duration as a critical parameter for the characterization of hiPSC-derived BBB in vitro models and supports the feasibility of using donor-specific BCECs at the end of differentiation to investigate BBB-associated mechanisms in adult ADHD and PARK2 CNVs.
Parole chiave
Leggi il testo integrale originale
Testo estratto dalla pubblicazione: sono esclusi immagini, media, supplementi e parti soggette a diritti distinti.
Adult ADHD, PARK2 copy number variations and the blood-brain barrier: Culture duration-dependent metabolic alterations in human induced pluripotent stem cell-derived in vitro models
Abstract
Background
Attention-deficit/hyperactivity disorder (ADHD) is a multifactorial neurodevelopmental disorder in which PARK2 copy number variations (CNVs) constitute a rare genetic risk factor. PARK2 encodes parkin, a key regulator of mitochondrial quality control. Mitochondrial dysfunction is implicated in ADHD pathophysiology and can contribute to blood-brain barrier (BBB) disruption. As BBB disruption remains insufficiently understood in ADHD, this study investigated whether adult ADHD and PARK2 CNVs are accompanied by alterations in BBB phenotype, mitochondrial properties, and cellular metabolism.
Methods
Brain capillary endothelial-like cells (BCECs) were generated using 10 donor-specific human induced pluripotent stem cell (hiPSC) lines from adults with ADHD carrying PARK2 CNVs, alongside PARK2 wild type adults with and without ADHD. BCECs were analyzed at the end of differentiation (day 10) and after prolonged culture (day 17) to assess group-associated phenotypes and culture duration-dependent differences.
Results
Across all groups, BCECs showed brain endothelial marker expression, high transendothelial electrical resistance (TEER), and low sodium fluorescein (NaFl) permeability at day 10. After prolonged culture, NaFl permeability remained low, whereas TEER-based barrier integrity declined. Prolonged culture of BCECs reduced glycolytic ATP production rates while increasing mitochondrial ATP production and its relative contribution to total ATP production, reflecting a metabolic shift from glycolysis toward oxidative phosphorylation. Despite decreased total ATP production after prolonged culture, cellular energy status was largely preserved, alongside overall lower intracellular reactive oxygen species (ROS) and mitochondrial membrane potential. At day 10, BCECs from ADHD donors carrying PARK2 duplications displayed greater mitochondrial abundance, the highest basal and oxidative stress-induced intracellular ROS levels, and reduced TEER-based barrier integrity, while at day 17, BCECs from ADHD donors with PARK2 deletions exhibited the least glycolysis-dependent ATP production profile, although absolute mitochondrial ATP production remained comparable between groups.
Conclusions
Prolonged culture was associated with broad, ADHD- and PARK2 genotype-independent metabolic remodeling, whereas mitochondrial and metabolic alterations associated with ADHD and PARK2 CNVs were overall small. This study identifies culture duration as a critical parameter for the characterization of hiPSC-derived BBB in vitro models and supports the feasibility of using donor-specific BCECs at the end of differentiation to investigate BBB-associated mechanisms in adult ADHD and PARK2 CNVs.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12987-026-00892-5.
Keywords: Attention-deficit/hyperactivity disorder, PARK2, Copy number variation, Mitochondrial dysfunction, Blood-brain barrier, Endothelial cells, Induced pluripotent stem cells, Mitochondrial membrane potential, Reactive oxygen species, Cellular bioenergetics
Background
Attention-deficit/hyperactivity disorder (ADHD) is among the most prevalent neurodevelopmental disorders worldwide and is characterized by age-inappropriate levels of inattention, hyperactivity, and impulsivity that interfere with academic, occupational, and social functioning [1–3]. Symptoms usually begin in childhood and may persist into adult life, with epidemiological studies reporting a prevalence of 5.3% in school-aged children and about 2.5% in adults [4, 5]. ADHD is highly heritable, and genetic predisposition accounts for a substantial proportion of interindividual variation in liability, while environmental risk factors across prenatal, perinatal, and psychosocial domains also contribute to its development, supporting a multifactorial etiology [2, 6]. Recent genome-wide association studies have identified multiple common genetic variants linked to ADHD, predominantly single nucleotide polymorphisms, and have enabled follow-up analyses that map these loci to candidate genes and biological pathways [7, 8]. Beyond this overall polygenic background, rare genomic alterations including copy number variations (CNVs) have also been associated with ADHD and partly overlap with loci implicated in other neurodevelopmental and psychiatric disorders [9]. Deletions and duplications in PARK2 are observed at higher frequencies in individuals with ADHD than in controls, supporting PARK2 as a rare structural risk locus for the disorder [10]. While biallelic loss-of-function mutations in PARK2 cause early-onset Parkinson’s disease, PARK2 CNVs are thought to induce more subtle disturbances in parkin-mediated mitochondrial quality control [11, 12].
Encoded by PARK2, the E3 ubiquitin ligase parkin acts together with PTEN-induced kinase 1 (PINK1) to regulate mitophagy and mitochondrial dynamics, thereby maintaining mitochondrial function, including ATP production and redox homeostasis [13, 14]. In fibroblasts and human induced pluripotent stem cell (hiPSC)-derived dopaminergic neurons from adults with ADHD carrying PARK2 CNVs, parkin-related mitochondrial dysfunction was characterized by altered mitochondrial network morphology, lower ATP levels, and reduced oxygen consumption rates [12]. Moreover, accumulating clinical and preclinical evidence suggests that mitochondrial dysfunction is involved in ADHD pathophysiology, with reported perturbations in energy metabolism and redox balance [15, 16]. Blood-brain barrier (BBB) disruption is commonly observed across disorders of the central nervous system and can impair brain homeostasis, while recent findings suggest a potential link between ADHD and BBB disruption, although the underlying mechanisms and pathophysiological implications remain unclear [17–19]. Brain endothelial cells rely on intact mitochondrial function to maintain tight junction-dependent barrier integrity, support active transport across the endothelium, and limit oxidative stress that would otherwise compromise barrier properties [20, 21]. Reduced parkin expression in brain endothelial cells was associated with impaired mitophagy and mitochondrial damage, while mitochondrial dysfunction has been linked to BBB disruption, accompanied by elevated levels of reactive oxygen species (ROS), tight junction disorganization, increased permeability, and altered transport processes [22, 23].
ADHD is most often treated pharmacologically, with stimulant medications as first-line therapy and several non-stimulant agents available as alternatives, all aimed at reducing core symptoms and functional impairment rather than targeting causal pathomechanisms [24, 25]. Adult ADHD frequently remains underdiagnosed, and even when pharmacotherapy is initiated, up to 50% of adult patients respond insufficiently to medication, in part due to comorbid psychiatric and somatic disorders that further limit treatment effectiveness [26–28]. Therefore, adult ADHD patient-specific in vitro models are required to better understand disease mechanisms in individual patients and to provide a translational platform for testing and refining treatment strategies. In addition, such models could enable evaluation of potential contributions of BBB disruption to pathophysiology and therapeutic response.
The aim of this study was to examine whether adult ADHD and PARK2 CNVs are associated with alterations in BBB phenotype, mitochondrial properties, and broader cellular metabolism. For this purpose, BBB in vitro models based on brain capillary endothelial-like cells (BCECs) were generated from hiPSCs derived from adult ADHD patients carrying PARK2 CNVs and control donors. Accordingly, models were analyzed at the end of differentiation and after prolonged in vitro culture to evaluate both group-associated phenotypes and culture duration-dependent alterations.
Methods
Derivation and maintenance of hiPSCs
Donor-specific hiPSC lines were generated and provided by the Department of Psychiatry, Psychosomatics and Psychotherapy, University Hospital Würzburg, using the CytoTune™-iPS 2.0 Sendai Reprogramming Kit (Thermo Fisher Scientific, A16517) according to the manufacturer’s instructions. For quality control, hiPSCs were assessed for pluripotency, trilineage differentiation, and Sendai virus clearance (Fig. S1), and additional mycoplasma and genetic analyses were performed as previously described [12, 29–31]. Donor and hiPSC line characteristics of the 4 WT_CTRL, 2 WT_ADHD, 2 DEL_ADHD, and 2 DUP_ADHD hiPSC lines are summarized in Table S1.
HiPSCs were cultivated on 6-well plates (Thermo Fisher Scientific, 140675) coated with 83.3 µg/mL Matrigel® (Corning, 356231) at 37 °C, 5% CO2, and 95% humidity. Twice per week at around 80% confluency, cells were passaged with ReLeSR™ (STEMCELL Technologies, 100–0483) and reseeded at ratios of 1:8 to 1:20 in StemMACS™ PSC-Brew XF medium (Miltenyi Biotec, 130-127-865) supplemented with 10 µM Y-27632 (Tocris Bioscience, 1254/10). Medium was replaced 24 h after seeding, and the volume was doubled on Fridays to allow weekend-free culture.
Generation of hiPSC-derived BCECs
For differentiation of hiPSCs into BCECs, the co-differentiation protocol of Lippmann et al. was applied with slight modifications [32]. At day −3, hiPSCs were detached with Accutase® (Sigma-Aldrich, A6964-100ML), resuspended in StemMACS™ PSC-Brew XF medium with 10 µM Y-27632, and seeded at a density of 7.8 × 103 cells/cm2 on Matrigel®-coated 6-well plates. Between day 0 and day 5, daily medium changes were performed with unconditioned medium containing 78.5% DMEM/F-12 without glutamine (Thermo Fisher Scientific, 21331020), 20.0% KnockOut™ Serum Replacement (Thermo Fisher Scientific, 10828028), 1.0% MEM non-essential amino acids (Thermo Fisher Scientific, 11140035), 1 mM L-glutamine (Sigma-Aldrich, G7513-100ML), and 0.1 mM 2-mercaptoethanol (Thermo Fisher Scientific, 31350010). At day 6 of differentiation, medium was changed to EC medium containing 99.5% human endothelial SFM (Thermo Fisher Scientific, 11111044) and 0.5% B-27™ without vitamin A (Thermo Fisher Scientific, 12587010), supplemented with 20 ng/mL hbFGF (Thermo Fisher Scientific, 100-18B-500UG) and 10 µM retinoic acid (Miltenyi Biotec, 130-117-339). At day 8, cells were detached with Accutase®, resuspended in EC medium supplemented with 20 ng/mL hbFGF and 10 µM retinoic acid, and seeded at a density of 1.0 × 106 cells/cm2 on 24-well inserts (cellQART®, 9320412), 8-well Nunc™ Lab-Tek™ chamber slides (Thermo Fisher Scientific, 177445PK), 96-well plates (Greiner Bio-One, 655090; Agilent, 103774-100), and 12-well plates (TPP®, 92012), all coated with 200 µg/mL Matrigel®. Between day 9 and day 17, EC medium was replaced daily on the apical side and at day 9, day 10, day 14, and day 17 in the basolateral compartment.
Transendothelial electrical resistance measurement
Transendothelial electrical resistance (TEER) values of BCECs at day 10 and day 17 were determined using a Millicell® ERS-2 V-ohmmeter (Merck) equipped with a STX3 electrode (World Precision Instruments). Prior to usage, the electrode was equilibrated in EC medium for 15 min. Measurements were initiated 45 min after medium change on a heating plate maintained at 37 °C. Blank values for Matrigel®-coated inserts without cells were calculated as the mean of three measurements. For each independent differentiation experiment, electrical resistance of at least 3 inserts containing BCECs was measured in duplicate or triplicate and corrected by subtracting blanks. Results were multiplied by the culture area of a 24-well insert.
Sodium fluorescein permeability assay
Sodium fluorescein (NaFl) permeability of BCECs cultured on 24-well inserts was assessed at day 10 and day 17. For each independent differentiation experiment, three technical replicate inserts were prepared, with Matrigel®-coated cell-free inserts serving as blanks. EC medium supplemented with 10 µM NaFl (Sigma-Aldrich, F6377-100G) was applied apically, while EC medium without NaFl was added to the basolateral compartment. Inserts were incubated on an orbital shaker at 100 rpm for 60 min at 37 °C. Apical and basolateral samples, together with EC medium containing 10 µM NaFl and EC medium without NaFl, were transferred to 96-well plates (Greiner Bio-One, 655076), and fluorescence intensity was measured at 490 nm excitation and 525 nm emission using an Infinite® M200 microplate reader (Tecan). Apical fluorescence values were adjusted for dilution with EC medium. As described previously [33], the permeability coefficient (PC) was calculated from NaFl clearance after 60 min, corrected for NaFl passage through cell-free inserts, and divided by insert surface area.
Immunofluorescence staining
At day 10 and day 17, BCECs cultured on inserts or 8-well chamber slides were fixed with 4.0% paraformaldehyde (AppliChem, A3813.1000) for 10 min, washed with PBS (Sigma-Aldrich, D8537-500ML), and permeabilized with 0.2% Triton® X-100 (Carl Roth, 3051.2) for 5 min. Cells were washed with 0.5% PBST (Sigma-Aldrich, P1379-25ML) and blocked in blocking buffer containing 5.0% donkey serum (Biozol Diagnostica, ECL-ECS0217L), 0.1% Triton® X-100, and 0.02% saponin (Carl Roth, 9622.1) for 20 min. The primary antibodies vascular endothelial cadherin (CDH5, Sigma-Aldrich, V1514-200UL, 1:100), glucose transporter 1 (GLUT1, Abcam, ab115730, 1:100), occludin (OCLN, Thermo Fisher Scientific, 33-1500, 1:100), claudin-5 (CLDN5, Abcam, ab15106, 1:100), and translocase of the outer mitochondrial membrane 20 (TOM20, Proteintech, 66777-1-Ig, 1:400) were diluted in blocking buffer and incubated overnight at 4 °C. Samples were washed with PBST and incubated with Alexa Fluor™ 488-, 555-, or 647-conjugated secondary antibodies (Thermo Fisher Scientific, A-21202; A-31570; A-31573, 1:400) diluted in antibody diluent (DCS Diagnostics, AL120R500) for 60 min. BCECs were washed, membranes were excised from inserts, and mounted on object slides using Fluoromount-G™ Mounting Medium with DAPI (Thermo Fisher Scientific, 00-4959-52). Images were acquired using the Quick Full Focus function on a BZ-X810 fluorescence microscope (Keyence) with a 20× objective (NA 0.75) for qualitative assessment of CDH5, GLUT1, OCLN, and CLDN5.
Quantification of mitochondrial morphology
TOM20- (Alexa Fluor™ 555) and DAPI-stained BCECs were imaged using an LSM 980 confocal microscope (Zeiss) with a 40× water immersion objective (NA 1.20). For each differentiation experiment, 5 fields of view were recorded. Z-stacks were deconvolved in Huygens (Scientific Volume Imaging) using Classic Maximum Likelihood Estimation with a theoretical point spread function. Maximum intensity projections were generated in ImageJ (National Institutes of Health) and batch-processed with the Mitochondria Analyzer plugin [34] to quantify mitochondrial count, area, aspect ratio, and form factor. Pre-processing included a sigma filter plus with radius 2.00, enhancement of local contrast with maximum slope 1.80, and gamma adjustment to 0.80. Local thresholding used the weighted mean method with a block size of 1.250 microns and a C-value of 10. For post-processing, despeckle was applied. Aspect ratio was defined as the ratio of the major to minor axis, and form factor was calculated as perimeter squared divided by the product of 4π and area. Values equal to 1 for both aspect ratio and form factor indicate mitochondria without elongation and a perimeter consistent with a circular shape for the measured area, whereas higher values reflect increased elongation and greater deviation from a circle. Cell counts per image were obtained from DAPI-stained nuclei using Cellpose [35].
Mitochondrial membrane potential measurement
Mitochondrial membrane potential (ΔΨm) of BCECs at day 10 and day 17 was assessed in 96-well plates using tetramethylrhodamine methyl ester (TMRM) staining, based on a publication by Zink et al. with minor adaptations [36]. Briefly, cells were incubated for 30 min in EC medium supplemented with 150 nM TMRM (Thermo Fisher Scientific, T668) and 2 µg/mL Hoechst 33342 (Thermo Fisher Scientific, H3570), together with either dimethyl sulfoxide (Sigma-Aldrich, D8418-500ML) as vehicle control or 20 µM carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP, Sigma-Aldrich, C2920-10MG) plus 20 µM antimycin A (AA, Sigma-Aldrich, A8674-25MG) to induce mitochondrial depolarization. For each differentiation experiment, cells were seeded in triplicate for both conditions, and 3 fields of view were recorded per well. Images were acquired using a CellVoyager™ CQ1 confocal quantitative image cytometer (Yokogawa) with a 20× long working distance objective (NA 0.45), using Hoechst 33342 for nuclear staining. As described by Zink et al. [36], ΔΨm was quantified using CellProfiler [37].
ROS assay
Intracellular ROS levels in BCECs were determined at day 10 and day 17 in 96-well plates. For ROS assays, EC medium contained 0.5% B-27™ without antioxidants (Thermo Fisher Scientific, 10889038) instead of B-27™ without vitamin A. Cells were incubated for 30 min in EC medium with 20 µM 2′,7′-dichlorodihydrofluorescein diacetate (DCFDA, Sigma-Aldrich, D6883-50MG) in the presence or absence of 1 mM tert-butyl hydroperoxide (TBHP, Sigma-Aldrich, 458139-25ML). For each differentiation experiment, BCECs were seeded in triplicate for both conditions. Plates were rinsed with PBS, and fluorescence intensity was measured in PBS at 485 nm excitation and 535 nm emission with an Infinite® M200 microplate reader (Tecan). Fluorescence intensity values were corrected by subtracting blank signals from DCFDA-treated wells, with or without TBHP.
After measurements, BCECs were fixed with 4.0% paraformaldehyde for 10 min, washed with PBS and stained with 0.2 µg/mL DAPI (Thermo Fisher Scientific, 62248) for 10 min. Plates were washed with PBS and imaged on a CellVoyager™ CQ1 confocal quantitative image cytometer with a 20× long working distance objective (NA 0.45), acquiring z-stacks from 5 fields of view per well. Mean cell count per field of view was quantified in CellPathfinder (Yokogawa) and scaled to the total well area.
Seahorse XF analysis
Seahorse XF Real-Time ATP Rate Assays (Agilent, 103592-100) were performed with BCECs at day 10 and day 17 in 96-well plates using a Seahorse XF Pro Analyzer (Agilent). For each differentiation experiment, up to 8 technical replicate wells were prepared. Cells were washed once and incubated with Seahorse XF assay medium containing Seahorse XF DMEM (Agilent, 103575-100), supplemented with 10 mM glucose (Agilent, 103577-100), 1 mM pyruvate (Agilent, 103578-100), and 2 mM glutamine (Agilent, 103579-100) for 60 min at 37 °C in a non-CO2 incubator. According to the manufacturer’s instructions, BCECs were treated with 1.5 µM oligomycin and 0.5 µM rotenone plus 0.5 µM antimycin A through the injection ports of a hydrated Seahorse XF sensor cartridge (Agilent, 103775-100). Results were analyzed using Seahorse Analytics (Agilent).
After measurements, cell counts per well were determined as described for ROS assays, acquiring only 2 fields of view per well due to smaller well area.
Quantification of intracellular metabolites
BCECs were cultured in 12-well plates and processed for intracellular metabolite quantification at day 10 and day 17. Cells were rinsed twice with cold 154 mM ammonium acetate (Sigma-Aldrich, A7330-500G) and harvested by cell scraping in a prechilled mixture of 80% v/v methanol (Sigma-Aldrich, 32213–2.5 L-M) and water. Samples were normalized to DNA content using a NanoPhotometer™ N60 (Implen). Liquid chromatography mass spectrometry analysis was performed using an UltiMate 3000 UHPLC system (Thermo Fisher Scientific) connected to a Q Exactive mass spectrometer (Thermo Fisher Scientific) equipped with a heated electrospray ionization source. Briefly, samples were fortified with external standards, ultrasonicated, centrifuged, dried, and reconstituted in 50% v/v acetonitrile (Merck, 1000292500) and water containing 5 mM ammonium acetate. Metabolite separation was performed on a BEH Amide column using a gradient of acetonitrile and water containing ammonium acetate. Data were acquired in alternating positive and negative full-scan mode, and peaks were manually curated and integrated using TraceFinder (Thermo Fisher Scientific).
Statistical analysis
Analyses were performed using GraphPad Prism (GraphPad Software). Data are reported as mean ± SD, where each independent differentiation experiment constitutes a biological replicate. Outliers were identified with the ROUT method (Q = 1%). Normal distribution of data was assessed by the Shapiro-Wilk test. For comparisons of more than two independent data sets, one-way ANOVA followed by Tukey’s post hoc test was performed when parametric assumptions were met. Otherwise, the Kruskal-Wallis with Dunn’s post hoc test was applied. For analyses involving group and either culture duration or treatment, two-way ANOVA followed by Sidak’s post hoc test was used to compare groups within each culture duration or treatment condition and to assess prolonged culture- or treatment-dependent effects within the same group. For comparisons of two independent data sets, pairwise comparisons were performed by Mann-Whitney or unpaired Welch’s t-tests. Statistical significance was defined as: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
Results
ADHD PARK2 CNV-derived BCECs preserve cellular identity and establish functional barrier properties at day 10
In this study, BCECs were derived from 10 donor-specific hiPSC lines across 4 different groups: (I) PARK2 wild type adults without ADHD (WT_CTRL), (II) PARK2 wild type adults with ADHD (WT_ADHD), (III) PARK2 deletion carriers with adult ADHD (DEL_ADHD), and (IV) PARK2 duplication carriers with adult ADHD (DUP_ADHD). BCECs were analyzed at the end of differentiation (day 10) and after 7 days of additional culture (day 17), with each assay performed at both time points to investigate alterations associated with culture duration.
To evaluate the capacity of each patient-derived hiPSC line to differentiate into BCECs with brain endothelial identity, we performed immunofluorescence staining for characteristic markers. At day 10 of differentiation, BCECs expressed the endothelial cell marker CDH5, showed plasma membrane enrichment of GLUT1, and displayed the tight junction proteins OCLN and CLDN5 at cell-cell boundaries (Fig. 1A). No apparent differences in marker expression or localization were detected across all groups at day 10, suggesting preserved cellular identity of BCECs independent of PARK2 genotype and ADHD. At day 17, BCECs maintained cell type-specific markers (Fig. 1B), indicating that the qualitative marker profile was not altered by the duration of in vitro culture.
Furthermore, TEER and NaFl permeability were assessed to determine whether PARK2 CNVs and ADHD influence barrier properties of BCECs. At day 10, TEER was significantly reduced in DUP_ADHD BCECs compared with WT_CTRL and DEL_ADHD BCECs (Fig. 1C), although all groups established functional barriers with mean TEER values above 1600 Ω cm2. TEER did not differ significantly between groups at day 17, whereas each group showed significantly diminished TEER after prolonged culture, resulting in mean values below 550 Ω cm2. Mean NaFl permeability remained below 0.5 μm/min and comparable across groups at day 10 and day 17 (Fig. 1D), while DUP_ADHD BCECs had the numerically highest mean PC at day 10. After prolonged culture, PCs were similar within WT_CTRL, WT_ADHD, and DEL_ADHD BCECs, whereas DUP_ADHD BCECs displayed a significant decrease, indicating improved paracellular barrier function. Taken together, ADHD PARK2 CNV-derived BCECs retained cellular identity and formed physiologically tight barriers at day 10, whereas prolonged culture reduced TEER-based barrier integrity in all groups without increasing NaFl permeability.
Mitochondrial morphology remains intact in ADHD PARK2 CNV-derived BCECs
Following successful establishment of donor-specific BBB in vitro models and given the role of parkin in mitochondrial quality control, we next analyzed whether mitochondrial morphology in BCECs varied with PARK2 CNVs and ADHD. Immunofluorescence staining for TOM20 was performed in BCECs to quantify mitochondrial abundance and morphology using Mitochondria Analyzer (Fig. 2). Mitochondrial count, area, aspect ratio, and form factor were quantified in BCECs at day 10 and day 17. Figure S2 provides representative images of TOM20-stained BCECs at day 10 and day 17 from the dataset used for quantification. At day 10, mitochondrial count was significantly increased in DUP_ADHD BCECs compared with WT_CTRL, followed by a significant decrease within DUP_ADHD BCECs after prolonged culture (Table 1). Area per mitochondrion was consistent across all groups at day 10, at day 17, and within BCECs after prolonged culture. Mitochondrial aspect ratio and form factor were similar between groups at both day 10 and day 17. After prolonged culture, aspect ratio and form factor were slightly reduced in every group, with significant reductions observed only in WT_CTRL and DEL_ADHD BCECs. In conclusion, ADHD PARK2 CNV-derived BCECs exhibited intact mitochondrial morphology at both day 10 and day 17, while aspect ratio and form factor decreased in WT_CTRL and DEL_ADHD BCECs after prolonged culture.
Table 1ADHD PARK2 CNV-derived BCECs preserved mitochondrial morphology
WT_CTRL WT_ADHD DEL_ADHD DUP_ADHD
Day 10
Mitochondrial count per cell 86.94 ± 26.94a) 105.20 ± 9.42 112.39 ± 35.08 161.62 ± 66.00a), b)
Area per mitochondrion [µm2] 0.092 ± 0.007 0.094 ± 0.005 0.106 ± 0.017 0.100 ± 0.011
Aspect ratio per mitochondrion 1.64 ± 0.13c) 1.66 ± 0.08 1.69 ± 0.11d) 1.68 ± 0.06
Form factor per mitochondrion 1.16 ± 0.07e) 1.16 ± 0.04 1.20 ± 0.07f) 1.18 ± 0.04
Day 17
Mitochondrial count per cell 94.77 ± 48.44 69.38 ± 22.72 99.92 ± 40.75 62.47 ± 20.93b)
Area per mitochondrion [µm2] 0.088 ± 0.003 0.090 ± 0.009 0.096 ± 0.007 0.094 ± 0.006
Aspect ratio per mitochondrion 1.51 ± 0.03c) 1.50 ± 0.07 1.53 ± 0.05d) 1.53 ± 0.08
Form factor per mitochondrion 1.09 ± 0.01e) 1.10 ± 0.03 1.10 ± 0.02f) 1.11 ± 0.03
Image-based quantification of mitochondrial count, area, aspect ratio, and form factor in TOM20-stained BCECs at day 10 and day 17. Data are shown as mean ± SD from WT_CTRL n ≥ 8, WT_ADHD n = 4, DEL_ADHD n = 6, DUP_ADHD n ≥ 5 independent differentiation experiments, each representing a biological replicate. (a–f) Statistical analysis was performed using two-way ANOVA with Sidak’s post hoc test: *p < 0.05 (c, d, e), **p < 0.01 (a, b, f)
Mitochondrial membrane potential is not impaired in ADHD PARK2 CNV-derived BCECs, whereas oxidative stress sensitivity is increased in DUP_ADHD BCECs at day 10
Since parkin has been proposed to contribute to ADHD pathophysiology through impaired mitochondrial quality control, we analyzed whether BCECs derived from ADHD patients carrying PARK2 CNVs displayed alterations in mitochondrial membrane potential (ΔΨm). ΔΨm is a sensitive readout of mitochondrial bioenergetic state and can provide insights into mitochondrial function. Accordingly, we quantified ΔΨm using TMRM under basal conditions and after mitochondrial depolarization with FCCP in combination with AA. Across all groups, basal ΔΨm was comparable in BCECs at day 10 (Fig. 3A) and day 17 (Fig. 3B), whereas FCCP + AA significantly reduced ΔΨm in every group at both time points. Comparing day 17 to day 10 cultures within the same group, ΔΨm showed a significant reduction in WT_CTRL, WT_ADHD, and DUP_ADHD BCECs under basal conditions, and in WT_CTRL BCECs following FCCP + AA treatment (Fig. 3C).
To determine whether PARK2 CNVs and ADHD were associated with altered oxidative stress, which may arise from mitochondrial dysfunction and contribute to cellular damage, we quantified intracellular ROS in BCECs. ROS levels were measured with and without the oxidative stressor TBHP. At day 10, baseline ROS levels in BCECs did not differ significantly between groups, with a tendency toward higher levels in DUP_ADHD BCECs, whereas TBHP treatment significantly increased ROS in every group (Fig. 3D). Notably, TBHP-treated DUP_ADHD BCECs displayed significantly higher ROS levels compared to every other treated group at day 10. Across all groups, day 17 BCECs showed similar basal ROS, while TBHP treatment again significantly elevated ROS (Fig. 3E). Compared with BCECs at day 10, ROS levels of day 17 BCECs were lower in WT_CTRL, WT_ADHD, and DUP_ADHD, yet higher in DEL_ADHD under both basal and TBHP-treated conditions, with only the decrease in untreated DUP_ADHD BCECs achieving statistical significance (Fig. 3F). Overall, ADHD PARK2 CNV-derived BCECs did not exhibit perturbations in ΔΨm at either day 10 and day 17, while DUP_ADHD BCECs showed slightly higher basal ROS levels and a greater TBHP-induced ROS response at day 10, indicating heightened sensitivity to oxidative stressors.
Prolonged culture shifts BCEC metabolism from glycolysis toward oxidative phosphorylation
Given the overall reduction in ΔΨm and ROS after prolonged culture, we next investigated whether culture duration of BCECs was associated with changes in ATP-generating pathways. As ATP is produced primarily through oxidative phosphorylation (OXPHOS) and glycolysis, mitochondrial (mitoATP) and glycolytic (glycoATP) ATP production rates were quantified in BCECs at day 10 and day 17 by performing Seahorse XF analysis. Oxygen consumption rate (OCR) profiles of BCECs were largely comparable across groups at day 10 (Fig. 4A) and day 17 (Fig. 4B), with the expected decrease after oligomycin injection followed by further reduction after rotenone plus AA treatment. Extracellular acidification rate (ECAR) profiles did not reveal apparent group-specific differences at day 10 (Fig. 4C) or day 17 (Fig. 4D), and ECAR increased after BCECs were treated with oligomycin, while ECAR levels were overall lower at day 17.
In comparison with WT_CTRL BCECs, mitoATP production was not impaired in DEL_ADHD and DUP_ADHD BCECs at either day 10 or day 17, with a tendency to increase after prolonged culture within every group, reaching significance only in DUP_ADHD BCECs (Fig. 4E). Compared with WT_CTRL at both day 10 and day 17, DEL_ADHD BCECs showed a lower glycoATP production rate, and prolonged culture induced a significant reduction within every group (Fig. 4F). While total ATP production declined with prolonged culture in every group, a significant decrease was detected only in WT_CTRL and WT_ADHD BCECs (Fig. 4G). After prolonged culture, BCECs exhibited a higher mitoATP and lower glycoATP contribution to total ATP production, reflecting a significant metabolic shift within every group (Fig. 4H). Notably, DEL_ADHD BCECs had significantly increased mitoATP and reduced glycoATP contributions compared with every other group at day 17. Although glycoATP production diminished with prolonged culture, glycolytic reserve was slightly elevated in BCECs at day 17 (Fig. 4I), suggesting that BCECs across all groups retained metabolic capacity to upregulate glycolysis in response to potential impairments in mitochondrial ATP production. In summary, prolonged culture of BCECs promoted a metabolic shift from glycolysis toward OXPHOS, while ADHD PARK2 CNV-derived BCECs preserved mitochondrial ATP production at both day 10 and day 17.
Prolonged culture decreases glycolytic lactate abundance and phosphocreatine buffering reserve in BCECs
Since ATP production profiles of BCECs changed with culture duration, we quantified intracellular lactate, a major end product of glycolytic metabolism, by targeted metabolomic profiling to validate reduced glycolytic ATP production. As expected, after prolonged culture, lactate abundance was significantly lower in day 17 BCECs compared to day 10 within every group (Fig. 5A). Lactate levels were slightly reduced in DEL_ADHD and DUP_ADHD BCECs compared to WT_CTRL at both day 10 (Fig. 5B) and day 17 (Fig. 5C), although these differences were not statistically significant.
Moreover, to assess broader metabolic disturbance in ADHD PARK2 CNV-derived BCECs, energy- and redox-related metabolite ratios were determined. In detail, we focused on the following ratios: ATP/AMP (adenosine triphosphate/adenosine monophosphate), PCr/Cr (phosphocreatine/creatine), NADH/NAD+ (reduced/oxidized nicotinamide adenine dinucleotide), and GSH/GSSG (reduced/oxidized glutathione), which indicate cellular energy state, phosphate buffering, redox state, and oxidative stress, respectively. ATP/AMP ratio of BCECs did not differ significantly between groups at either day 10 or day 17, while it was significantly higher in WT_CTRL BCECs after prolonged culture (Fig. 5D). At day 10, PCr/Cr ratio was significantly increased in DEL_ADHD BCECs compared to WT_CTRL and WT_ADHD BCECs, with prolonged culture leading to a significant decrease in WT_CTRL, DEL_ADHD, and DUP_ADHD BCECs (Fig. 5E). Ratios of NADH/NAD+ (Fig. 5F) and GSH/GSSG (Fig. 5G) were similar across BCECs of all groups at day 10, at day 17, and showed no significant changes after prolonged culture. In conclusion, prolonged culture led to reduced glycolytic lactate abundance and PCr buffering reserve within BCECs of every group, while ADHD PARK2 CNV-derived BCECs maintained overall cellular energy status and redox balance at both day 10 and day 17.
Discussion
In this study, hiPSC-derived BCECs were applied to generate BBB in vitro models from individuals with ADHD carrying PARK2 CNVs, and from PARK2 wild type adults with and without ADHD, using a total of 10 donor-specific hiPSC lines for group-level comparisons. Given the proposed connection between ADHD, PARK2 CNVs, mitochondrial dysfunction, and BBB disruption, BCECs were assessed for BBB phenotype, mitochondrial properties, and cellular metabolism. Analyses were performed both at the end of differentiation and after prolonged culture to characterize alterations associated with culture duration.
Independent of ADHD and PARK2 genotype, BBB in vitro models were successfully established, with BCECs displaying cell type-specific marker expression and physiologically tight barrier function at day 10. In brain endothelial-like cells generated from hiPSCs, extending culture beyond the differentiation endpoint has been associated with a transition toward a more quiescent endothelial state, accompanied by reorganization of junctional marker expression [38, 39], indicating that time in culture is linked to ongoing in vitro phenotypic remodeling. However, in our study, cell type-specific marker expression of BCECs was assessed qualitatively, without apparent culture duration-associated changes. TEER-based barrier integrity of BCECs declined after prolonged culture, with mean values ranging from 1600 to 2500 Ω cm2 at day 10 to below 550 Ω cm2 at day 17. By contrast, mean NaFl permeability of BCECs remained below 0.5 μm/min at both time points, indicating that lower TEER at day 17 was not accompanied by increased paracellular permeability to the small molecule tracer. The impact of prolonged culture on barrier integrity of hiPSC-derived brain endothelial-like cells has been investigated previously. Hollmann et al. demonstrated sustained barrier function, with TEER remaining above 1000 Ω cm2 for at least 8 days [40]. Interestingly, Williams et al. showed that after 7 days of extended culture, TEER changes were heterogeneous in direction across independent differentiation experiments, but mean TEER values were comparable at both time points [38]. Using a closely related differentiation protocol, Choi et al. reported mean TEER of approximately 450 Ω cm2 after an additional 7 days of static culture [39], comparable to that of BCECs at day 17 in our study. Evidence from hiPSC-derived brain endothelial-like cells supports a nonlinear relationship between TEER and NaFl permeability, with NaFl permeability reaching a plateau above 500 Ω cm2 and remaining similar across culture conditions despite large differences in TEER [41, 42]. Although mean TEER values of BCECs were below 550 Ω cm2 at day 17, NaFl permeability was comparable to published mean values of 0.10–0.48 μm/min in other hiPSC-derived brain endothelial-like cells [43–45].
Mitochondrial density in brain endothelial cells is approximately five-fold higher than in peripheral endothelial cells [46]. Nevertheless, it has been reported that brain endothelial cells rely on aerobic glycolysis [47], despite glycolysis being substantially less efficient than OXPHOS in terms of ATP yield per glucose molecule. Glycolysis offers several advantages for endothelial cells, including low oxygen consumption that preserves oxygen availability for perivascular cells, reduced mitochondrial ROS generation while maintaining low-level ROS signaling, and rapid ATP production that supports sprouting and continued vessel growth in avascular hypoxic regions where oxygen becomes limiting sooner than glucose [48, 49]. In this study, hiPSC-derived BCECs generated the majority of ATP via glycolysis at day 10, whereas prolonged culture reduced glycolytic ATP production and increased both absolute mitochondrial ATP production and its relative contribution to total ATP. Despite this redistribution, total ATP production declined after prolonged culture, while glycolytic reserve was slightly higher, suggesting that the reduction in glycolytic ATP production reflected lower utilization of glycolysis rather than diminished glycolytic potential. Consistent with this metabolic shift, intracellular lactate abundance decreased after prolonged culture, further supporting reduced lactate-associated glycolytic metabolism in BCECs at day 17. Similarly, Williams et al. reported lower glycolysis in hiPSC-derived brain endothelial-like cells after 7 days of extended culture [38]. Lactate is not a mere glycolytic end product, as brain endothelial cells actively contribute to lactate homeostasis and monocarboxylate transport across the BBB, and lactate released by the endothelium has been implicated in supporting pericyte function and preserving BBB integrity [50, 51].
Prolonged culture of hiPSC-derived BCECs induced a metabolic shift from glycolysis toward OXPHOS, which coincided with reduced TEER-based barrier integrity. This pattern suggests concurrent, ADHD- and PARK2 genotype-independent metabolic remodeling associated with culture duration. After prolonged culture, total ATP production rate of BCECs decreased, whereas ATP/AMP ratio was generally maintained or increased, indicating that steady-state cellular energy status was preserved. By contrast, phosphocreatine/creatine ratio was lower after prolonged culture, suggesting a reduction in the immediately available phosphocreatine reserve for rapid ATP regeneration, consistent with the role of creatine kinase and phosphocreatine in linking cellular energy supply to ATP utilization [52, 53]. In a study by Sakamuri et al., passage-associated aging of primary human brain microvascular endothelial cells revealed reduced glycolytic and total ATP production rates, whereas mitochondrial ATP production remained stable from early to pre-senescent passages and decreased only in later senescent cultures [54]. In a separate study using a CD34+-derived human BBB model, endothelial senescence coincided with reduced TEER and increased paracellular permeability [55], consistent with broader evidence linking cellular senescence to BBB dysfunction in aging and age-related disease [56]. Given that there are differences between primary human brain microvascular endothelial cells and hiPSC-derived brain endothelial-like cells in barrier function, glycolytic and oxidative metabolism [57], metabolic remodeling with increasing passage in primary cells may reflect different processes than the alterations observed after prolonged culture in differentiated BCECs. Endothelial metabolic remodeling extends beyond glycolysis and OXPHOS and can involve fatty acid, amino acid, and pentose phosphate pathway metabolism, thereby contributing to biosynthesis and redox homeostasis [58]. Although laminar shear stress and blood flow have been shown to maintain oxidative pentose phosphate pathway activity in endothelial cells and thereby support vascular maturation [59], metabolic remodeling was nevertheless observed in BCECs after prolonged static culture, indicating that time in culture can modulate metabolic state even in the absence of flow. After prolonged culture, intracellular ROS levels were overall lower in BCECs, while NADH/NAD+ and GSH/GSSG ratios showed no consistent changes. As the GSH/GSSG redox state has been linked to ROS regulation in primary brain endothelial cells [60, 61], the combination of lower ROS levels and a maintained GSH/GSSG ratio in BCECs suggests that prolonged culture did not disrupt the balance between ROS generation and antioxidant defense. Nevertheless, lower ROS levels should not be considered exclusively beneficial, given the role of physiological ROS signaling in BBB homeostasis [62]. Beyond broader signaling implications, lower ROS levels in BCECs coincided with a decrease in ΔΨm after prolonged culture, consistent with evidence that mitochondrial ROS generation is modulated by ΔΨm and can be reduced upon depolarization of the inner mitochondrial membrane [63, 64]. Notably, lower ΔΨm in BCECs was accompanied by an increased mitochondrial ATP production rate after prolonged culture, arguing against the assumption that decreased ΔΨm indicates impaired mitochondrial function. Rather, reduced ΔΨm may reflect greater utilization of the proton-motive force for ATP synthesis relative to proton pumping by the electron transport chain that maintains ΔΨm [65].
Prolonged culture of BCECs was accompanied by broad metabolic remodeling, whereas mitochondrial and metabolic alterations associated with ADHD and PARK2 CNVs were overall small in comparison. At day 10, mitochondrial count was increased in DUP_ADHD BCECs compared with WT_CTRL, whereas mitochondrial morphology was similar between groups. By contrast, Palladino et al. reported higher aspect ratio and form factor in DUP_ADHD fibroblasts, suggesting a more elongated mitochondrial shape with tubular branching [12]. This discrepancy may reflect cell type-specific differences in mitochondrial morphology, as mitochondrial organization aligns with distinct metabolic and signaling demands across cellular contexts [66]. Although morphology is not a direct measure of mitochondrial function, it remains a useful structural indicator of mitochondrial organization. Higher mitochondrial abundance with stable area per mitochondrion argues against fragmentation and is consistent with enhanced mitochondrial biogenesis [67, 68]. In endothelial cells, peroxisome proliferator-activated receptor γ coactivator 1-α (PGC-1α) links mitochondrial biogenesis to intracellular ROS, as elevated ROS can induce PGC-1α expression while PGC-1α promotes mitochondrial biogenesis and limits ROS accumulation [69, 70]. Accordingly, the greatest mitochondrial abundance in DUP_ADHD BCECs at day 10 may relate to their basal ROS levels, which were also highest among all groups. Enhanced susceptibility to oxidative stress was further indicated by elevated intracellular ROS levels in TBHP-treated DUP_ADHD BCECs at day 10, exceeding those in all other groups. Such a stress-dependent phenotype is consistent with findings by Radtke et al., as the expression of genes related to oxidative stress response, ubiquitin pathways, and mitochondrial quality control showed no baseline differences across donor-specific fibroblasts, hiPSCs, and hiPSC-derived dopaminergic neurons, whereas nutrient deprivation revealed increased expression of the oxidative stress response gene NQO1 in fibroblasts from ADHD patients carrying PARK2 CNVs [29]. Under basal conditions, DUP_ADHD BCECs also exhibited reduced TEER-based barrier integrity compared with WT_CTRL at day 10. In brain endothelial cells, elevated oxidative stress can compromise barrier function through altered tight junction protein expression and localization, lower TEER, and increased paracellular permeability [62, 71]. Accordingly, enhanced oxidative stress susceptibility may have contributed to reduced TEER-based barrier integrity of DUP_ADHD BCECs at day 10. At day 17, a group-specific metabolic phenotype was observed in DEL_ADHD BCECs, with mitochondrial ATP production contributing more and glycolytic ATP production less to the total ATP production rate than in all other groups. Despite comparable total ATP production, absolute glycolytic ATP production was lowest in DEL_ADHD BCECs at both time points, demonstrating the least glycolysis-dependent ATP production after prolonged culture. Nevertheless, absolute mitochondrial ATP production remained comparable between groups at day 17, arguing against mitochondrial bioenergetic impairment in DEL_ADHD BCECs.
Limitations
The inclusion of 10 donor-specific hiPSC lines was a strength of this study, although their division into four different groups resulted in only 2–4 donors per group. Consequently, subtle group-associated effects may have been difficult to distinguish from donor-specific variability and may have remained undetected. Although donor groups were matched for age, sex, and ethnicity, hiPSC lines included in the present study were not isogenic and therefore differed in their broader genetic backgrounds. While genome-editing approaches have been used to generate CNV-corrected isogenic hiPSC lines [72, 73], correction of PARK2 CNVs could have isolated their contribution within the same donor background. However, as ADHD is a multifactorial and highly polygenic disorder [2, 6], an isogenic control without ADHD cannot be generated through correction of a single genetic variant. Intracellular ROS was assessed using DCFDA, which does not distinguish individual ROS species or their subcellular sources [74]. Therefore, observed differences in intracellular ROS cannot be attributed specifically to mitochondrial ROS production. Nevertheless, DCFDA provides an integrated whole-cell readout of oxidative activity in BCECs. While BCECs were assessed for BBB phenotype, mitochondrial properties, and cellular metabolism only at the end of differentiation (day 10) and after prolonged culture (day 17), onset and progression of the prolonged culture-induced metabolic shift could not be characterized. What this metabolic shift represents biologically remains unclear, as it may reflect adaptation to prolonged static monoculture, reduced proliferation-related metabolic demand, a transition toward quiescence, senescence-associated alterations, progressive deviation from the BBB phenotype, or a combination of these processes. In addition, the present study does not establish whether metabolic remodeling in BCECs contributed to decreased TEER-based barrier integrity, occurred in response to this reduction, or developed independently during prolonged culture, and the mechanisms underlying their potential connection are unresolved.
Conclusions
Donor-specific BBB in vitro models were successfully established using 10 hiPSC lines across four different groups, with BCECs displaying brain endothelial marker expression, high TEER-based barrier integrity and low NaFl permeability at day 10, demonstrating a robust BBB-associated phenotype. At the end of differentiation, ATP production in BCECs was predominantly glycolytic, consistent with the characteristic reliance of brain endothelial cells on glycolysis [47]. After prolonged culture, NaFl permeability remained low, whereas TEER-based barrier integrity decreased alongside a metabolic shift from glycolysis toward OXPHOS, indicating adverse effects on the maintenance of BBB characteristics. Prolonged culture of BCECs was further accompanied by broad, ADHD- and PARK2 genotype-independent metabolic remodeling. By comparison, mitochondrial and metabolic alterations associated with ADHD and PARK2 CNVs were overall small. At day 10, DUP_ADHD BCECs showed greater mitochondrial abundance, the highest basal and oxidative stress-induced intracellular ROS levels, and reduced TEER-based barrier integrity, while at day 17, DEL_ADHD BCECs exhibited the least glycolysis-dependent ATP production profile, although absolute mitochondrial ATP production remained comparable between groups. ADHD- and PARK2 CNV-related alterations may remain subtle under baseline conditions and become more apparent in response to environmental stressors. Reported associations include perinatal hypoxic-ischemic adversity [75], prenatal undernutrition [76], and maternal infection during pregnancy [77], supporting the investigation of hypoxia, nutrient deprivation, and inflammation as relevant stressors in future experiments. This study identifies culture duration as a critical parameter for the characterization of hiPSC-derived BBB in vitro models and supports the feasibility of using donor-specific BCECs at the end of differentiation to investigate BBB-associated mechanisms in adult ADHD and PARK2 CNVs.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
This study was supported by the Core Unit Fluorescence Imaging, University of Würzburg and the Metabolomics Core Unit, University of Würzburg. The authors thank Dr. Annika Zink, University Hospital Düsseldorf, for support with the initial optimization of ΔΨm measurements and Nicole Paul, Technical University of Applied Sciences Würzburg-Schweinfurt, for assistance with Cellpose.
Abbreviations
ΔΨmMitochondrial membrane potential
AAAntimycin A
ADHDAttention-deficit/hyperactivity disorder
AMPAdenosine monophosphate
ATPAdenosine triphosphate
BBBBlood-brain barrier
BCECBrain capillary endothelial-like cell
CDH5Vascular endothelial cadherin
CLDN5Claudin-5
CNVCopy number variation
CrCreatine
DCFDA 2',7'-Dichlorodihydrofluorescein diacetate
DEL_ADHDPARK2 deletion with adult ADHD
DUP_ADHDPARK2 duplication with adult ADHD
ECARExtracellular acidification rate
FCCPCarbonyl cyanide-p-trifluoromethoxyphenylhydrazone
GLUT1Glucose transporter 1
GSHReduced glutathione
GSSGOxidized glutathione
hiPSChuman induced pluripotent stem cell
NAD+Oxidized nicotinamide adenine dinucleotide
NADHReduced nicotinamide adenine dinucleotide
NaFlSodium fluorescein
OCLNOccludin
OCROxygen consumption rate
OXPHOSOxidative phosphorylation
PCPermeability coefficient
PCrPhosphocreatine
PGC-1αPeroxisome proliferator-activated receptor γ coactivator 1-α
ROSReactive oxygen species
TBHPTert-butyl hydroperoxide
TEERTransendothelial electrical resistance
TMRMTetramethylrhodamine methyl ester
TOM20Translocase of the outer mitochondrial membrane 20
WT_ADHDPARK2 wild type with adult ADHD
WT_CTRLPARK2 wild type without adult ADHD
Author contributions
AAM, SKS, and RVM contributed to conceptualization, funding acquisition, resources, and supervision. MG, LD, LC, ZS, KH, and WS contributed to methodology. MG, LD, LC, ZS, SG, LS, HO, JR, MF, and WS performed investigation. MG wrote the original draft of the manuscript. AAM, SKS, RVM, ZS, KH, and WS reviewed and edited the manuscript.
Funding
Open Access funding enabled and organized by Projekt DEAL. This project has received funding from the Interdisciplinary Center for Clinical Research (IZKF, N-440) of the University Hospital Würzburg and from the Innovative Medicines Initiative 2 Joint Undertaking (JU) under grant agreement No 807015 (IM2PACT). The JU receives support from the European Union’s Horizon 2020 research and innovation program and EFPIA. The Zeiss LSM 980 was funded by the DFG (INST 93/1022-1).
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
Informed consent was obtained from all donors. Ethical approval was obtained from the ethics committees at the University of Würzburg (#96/11) and University of Frankfurt (#425/14). All methods were performed in accordance with relevant guidelines and regulations.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
1.Cecil CAM, Nigg JT. Epigenetics and ADHD: Reflections on Current Knowledge, Research Priorities and Translational Potential. Mol Diagn Ther. 2022;26(6):581–606.
2.Thapar A, Cooper M. Attention deficit hyperactivity disorder. Lancet. 2016;387(10024):1240–50.
3.Faraone SV, et al. Attention-deficit/hyperactivity disorder. Nat Rev Dis Primers. 2024;10(1):11.
4.Polanczyk G, et al. The worldwide prevalence of ADHD: a systematic review and metaregression analysis. Am J Psychiatry. 2007;164(6):942–8.
5.Song P, et al. The prevalence of adult attention-deficit hyperactivity disorder: A global systematic review and meta-analysis. J Glob Health. 2021;11:04009.
6.Faraone SV, Larsson H. Genetics of attention deficit hyperactivity disorder. Mol Psychiatry. 2019;24(4):562–75.
7.Demontis D, et al. Discovery of the first genome-wide significant risk loci for attention deficit/hyperactivity disorder. Nat Genet. 2019;51(1):63–75.
8.Demontis D, et al. Genome-wide analyses of ADHD identify 27 risk loci, refine the genetic architecture and implicate several cognitive domains. Nat Genet. 2023;55(2):198–208.
9.Gudmundsson OO, et al. Attention-deficit hyperactivity disorder shares copy number variant risk with schizophrenia and autism spectrum disorder. Transl Psychiatry. 2019;9(1):258.
10.Jarick I, et al. Genome-wide analysis of rare copy number variations reveals PARK2 as a candidate gene for attention-deficit/hyperactivity disorder. Mol Psychiatry. 2014;19(1):115–21.
11.Castelo Rueda MP, et al. Molecular phenotypes of mitochondrial dysfunction in clinically non-manifesting heterozygous PRKN variant carriers. NPJ Parkinsons Dis. 2023;9(1):65.
12.Palladino VS et al. Energy metabolism disturbances in cell models of PARK2 CNV carriers with ADHD. J Clin Med. 2020;9(12):4092.
13.Pickrell AM, Youle RJ. The roles of PINK1, parkin, and mitochondrial fidelity in Parkinson’s disease. Neuron. 2015;85(2):257–73.
14.Ma K, et al. Mitophagy, Mitochondrial Homeostasis, and Cell Fate. Front Cell Dev Biol. 2020;8:467.
15.Ogutlu H, Kasak M, Tutku Tabur S. Mitochondrial Dysfunction in Attention Deficit Hyperactivity Disorder. Eurasian J Med. 2022;54(Suppl1):187–95.
16.Corona JC. Role of oxidative stress and neuroinflammation in attention-deficit/hyperactivity disorder. Antioxidants (Basel). 2020;9(11):1039.
17.Archie SR, Shoyaib AA, Cucullo L. Blood-brain barrier dysfunction in CNS disorders and putative therapeutic targets: an overview. Pharmaceutics. 2021;13(11):1779.
18.Ahirwar LK, et al. Reviewing vascular influences on neuronal migration, cortical development, and neurodevelopmental disorders: focus on autism, ADHD and schizophrenia. Mol Psychiatry. 2025;30(12):5953–66.
19.Uzun N, et al. Investigation of tight junction protein alterations in ADHD: the role of claudin-5, beta-catenin and paxillin. J Atten Disord. 2026;30(6):757–64.
20.Shao Y, et al. Endothelial mitochondria in the blood-brain barrier. Fluids Barriers CNS. 2025;22(1):88.
21.Kim S, Jung UJ, Kim SR. Role of oxidative stress in blood-brain barrier disruption and neurodegenerative diseases. Antioxidants (Basel). 2024;13(12):1462.
22.Kou C, et al. Rnf40 Exacerbates Hypertension-Induced Cerebrovascular Endothelial Barrier Dysfunction by Ubiquitination and Degradation of Parkin. CNS Neurosci Ther. 2025;31(1):e70210.
23.Wang Y, et al. Mitochondrial oxidative stress in brain microvascular endothelial cells: Triggering blood-brain barrier disruption. Mitochondrion. 2023;69:71–82.
24.Brown KA, Samuel S, Patel DR. Pharmacologic management of attention deficit hyperactivity disorder in children and adolescents: a review for practitioners. Transl Pediatr. 2018;7(1):36–47.
25.Nazarova VA, et al. Treatment of ADHD: Drugs, psychological therapies, devices, complementary and alternative methods as well as the trends in clinical trials. Front Pharmacol. 2022;13:1066988.
26.Ginsberg Y, Quintero J, Anand E, Casillas M, Upadhyaya HP. Underdiagnosis of attention-deficit/hyperactivity disorder in adult patients: a review of the literature. Prim Care Companion CNS Disord. 2014;16(3):PCC.13r01600.
27.Philipsen A, et al. Effects of Group Psychotherapy, Individual Counseling, Methylphenidate, and Placebo in the Treatment of Adult Attention-Deficit/Hyperactivity Disorder: A Randomized Clinical Trial. JAMA Psychiatry. 2015;72(12):1199–210.
28.Kooij JJS, et al. Updated European Consensus Statement on diagnosis and treatment of adult ADHD. Eur Psychiatry. 2019;56:14–34.
29.Radtke F, et al. ADHD-associated PARK2 copy number variants: A pilot study on gene expression and effects of supplementary deprivation in patient-derived cell lines. Am J Med Genet B Neuropsychiatr Genet. 2022;189(7–8):257–70.
30.McNeill RV, et al. Generation of four human induced pluripotent stem cells derived from ADHD patients carrying different genotypes for the risk SNP rs1397547 in the ADHD-associated gene ADGRL3. Stem Cell Res. 2023;67:103016.
31.McNeill RV, et al. hiPSC-derived cortical neurons from ADHD individuals reveal dysregulated glutamatergic development. Mol Psychiatry. 2026;31(2):1041–50.
32.Lippmann ES, et al. A retinoic acid-enhanced, multicellular human blood-brain barrier model derived from stem cell sources. Sci Rep. 2014;4:4160.
33.Santa-Maria AR, et al. Transport Studies Using Blood-Brain Barrier In Vitro Models: A Critical Review and Guidelines. Handb Exp Pharmacol. 2022;273:187–204.
34.Chaudhry A, Shi R, Luciani DS. A pipeline for multidimensional confocal analysis of mitochondrial morphology, function, and dynamics in pancreatic beta-cells. Am J Physiol Endocrinol Metab. 2020;318(2):E87–101.
35.Stringer C, et al. Cellpose: a generalist algorithm for cellular segmentation. Nat Methods. 2021;18(1):100–6.
36.Zink A, et al. High-content screening of mitochondrial polarization in neural cells derived from human pluripotent stem cells. STAR Protoc. 2022;3(3):101602.
37.Stirling DR, et al. CellProfiler 4: improvements in speed, utility and usability. BMC Bioinformatics. 2021;22(1):433.
38.Williams LM, et al. Prolonged culturing of iPSC-derived brain endothelial-like cells is associated with quiescence, downregulation of glycolysis, and resistance to disruption by an Alzheimer’s brain milieu. Fluids Barriers CNS. 2022;19(1):10.
39.Choi J, Mathew S, Oerter S, Appelt-Menzel A, Hansmann J, Schmitz T. Online measurement system for dynamic flow bioreactors to study barrier integrity of hiPSC-based blood-brain barrier in vitro models. Bioengineering (Basel). 2022;9(1):39.
40.Hollmann EK, et al. Accelerated differentiation of human induced pluripotent stem cells to blood-brain barrier endothelial cells. Fluids Barriers CNS. 2017;14(1):9.
41.Mantle JL, Min L, Lee KH. Minimum Transendothelial Electrical Resistance Thresholds for the Study of Small and Large Molecule Drug Transport in a Human in Vitro Blood-Brain Barrier Model. Mol Pharm. 2016;13(12):4191–8.
42.Neal EH, et al. Influence of basal media composition on barrier fidelity within human pluripotent stem cell-derived blood-brain barrier models. J Neurochem. 2021;159(6):980–91.
43.Canfield SG, et al. An isogenic blood-brain barrier model comprising brain endothelial cells, astrocytes, and neurons derived from human induced pluripotent stem cells. J Neurochem. 2017;140(6):874–88.
44.Danz K, Höcherl T, Wien SL, Wien L, von Briesen H, Wagner S. Experimental comparison of primary and hiPS-based in vitro blood-brain barrier models for pharmacological research. Pharmaceutics. 2022;14(4):737.
45.Koenig L, Ramme AP, Faust D, Mayer M, Flötke T, Gerhartl A, et al. A human stem cell-derived brain-liver chip for assessing blood-brain-barrier permeation of pharmaceutical drugs. Cells. 2022;11(20):3295.
46.Bayir E, Sendemir A. Role of intermediate filaments in blood-brain barrier in health and disease. Cells. 2021;10(6):1400.
47.Kim ES, et al. Brain Endothelial Cells Utilize Glycolysis for the Maintenance of the Transcellular Permeability. Mol Neurobiol. 2022;59(7):4315–33.
48.Leung SWS, Shi Y. The glycolytic process in endothelial cells and its implications. Acta Pharmacol Sin. 2022;43(2):251–9.
49.Li X, Sun X, Carmeliet P. Hallmarks Endothelial Cell Metabolism Health Disease Cell Metab. 2019;30(3):414–33.
50.Wang J, et al. Brain Endothelial Cells Maintain Lactate Homeostasis and Control Adult Hippocampal Neurogenesis. Cell Stem Cell. 2019;25(6):754–e7679.
51.Lee HW, et al. Endothelium-derived lactate is required for pericyte function and blood-brain barrier maintenance. EMBO J. 2022;41(9):e109890.
52.Wallimann T, Tokarska-Schlattner M, Schlattner U. The creatine kinase system and pleiotropic effects of creatine. Amino Acids. 2011;40(5):1271–96.
53.Dringen R, et al. ATP Metabolism of Astrocytes: Consumption, Regeneration and Restoration. Neurochem Res. 2025;50(6):357.
54.Sakamuri S, et al. Glycolytic and Oxidative Phosphorylation Defects Precede the Development of Senescence in Primary Human Brain Microvascular Endothelial Cells. Geroscience. 2022;44(4):1975–94.
55.Novo JP, et al. Blood-brain barrier dysfunction in aging is mediated by brain endothelial senescence. Aging Cell. 2024;23(9):e14270.
56.Knopp RC, et al. Cellular senescence and the blood-brain barrier: Implications for aging and age-related diseases. Exp Biol Med (Maywood). 2023;248(5):399–411.
57.Weber CM, et al. Induced pluripotent stem cell-derived cells model brain microvascular endothelial cell glucose metabolism. Fluids Barriers CNS. 2022;19(1):98.
58.Falkenberg KD, et al. The metabolic engine of endothelial cells. Nat Metab. 2019;1(10):937–46.
59.Facchinello N, et al. Oxidative pentose phosphate pathway controls vascular mural cell coverage by regulating extracellular matrix composition. Nat Metab. 2022;4(1):123–40.
60.Sivandzade F, Alqahtani F, Cucullo L. Impact of chronic smoking on traumatic brain microvascular injury: An in vitro study. J Cell Mol Med. 2021;25(15):7122–34.
61.Gao M, Fu J, Wang Y. The lncRNA FAL1 protects against hypoxia-reoxygenation- induced brain endothelial damages through regulating PAK1. J Bioenerg Biomembr. 2020;52(1):17–25.
62.Shuvalova M, et al. The role of reactive oxygen species in the regulation of the blood-brain barrier. Tissue Barriers. 2025;13(2):2361202.
63.Chenna S, et al. Mechanisms and mathematical modeling of ROS production by the mitochondrial electron transport chain. Am J Physiol Cell Physiol. 2022;323(1):C69–83.
64.Vyssokikh MY, et al. Mild depolarization of the inner mitochondrial membrane is a crucial component of an anti-aging program. Proc Natl Acad Sci U S A. 2020;117(12):6491–501.
65.Tovar-Ferrero O, et al. Measuring mitochondrial membrane potential. EMBO J. 2025;44(24):7334–45.
66.Glancy B, et al. The Functional Impact of Mitochondrial Structure Across Subcellular Scales. Front Physiol. 2020;11:541040.
67.Coronado M, et al. Physiological Mitochondrial Fragmentation Is a Normal Cardiac Adaptation to Increased Energy Demand. Circ Res. 2018;122(2):282–95.
68.Wang C, et al. Serine synthesis controls mitochondrial biogenesis in macrophages. Sci Adv. 2024;10(20):eadn2867.
69.Valle I, et al. PGC-1alpha regulates the mitochondrial antioxidant defense system in vascular endothelial cells. Cardiovasc Res. 2005;66(3):562–73.
70.Sawada N, et al. Endothelial PGC-1alpha mediates vascular dysfunction in diabetes. Cell Metab. 2014;19(2):246–58.
71.Vagvolgyi M, et al. 17-Oxime ethers of oxidized ecdysteroid derivatives modulate oxidative stress in human brain endothelial cells and dose-dependently might protect or damage the blood-brain barrier. PLoS ONE. 2024;19(2):e0290526.
72.Munezane H, et al. Elimination of the extra chromosome of Dup15q syndrome iPSCs for cellular and molecular investigation. Eur J Cell Biol. 2024;103(3):151446.
73.Ye T, et al. Efficient manipulation of gene dosage in human iPSCs using CRISPR/Cas9 nickases. Commun Biol. 2021;4(1):195.
74.Kalyanaraman B, et al. Measuring reactive oxygen and nitrogen species with fluorescent probes: challenges and limitations. Free Radic Biol Med. 2012;52(1):1–6.
75.Zhu T, et al. Association Between Perinatal Hypoxic-Ischemic Conditions and Attention-Deficit/Hyperactivity Disorder: A Meta-Analysis. J Child Neurol. 2016;31(10):1235–44.
76.Cortés-Albornoz MC, García-Guáqueta DP, Velez-van-Meerbeke A, Talero-Gutiérrez C. Maternal nutrition and neurodevelopment: a scoping review. Nutrients. 2021;13(10):3530.
77.Ayubi E, Mansori K. Maternal Infection during Pregnancy and Attention-Deficit Hyperactivity Disorder in Children: A Systematic Review and Meta-Analysis. Iran J Public Health. 2022;51(12):2674–87.
Associated Data
Supplementary Materials
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Come citare questo studio
Markus Glaser, Liane Dieterich, Laura Ceslik, Zora Schickardt, Sabine Gätzner, Luisa Schmitt, Heike Oberwinkler, Jasmin Reining, Mike Friedrich, Katherina Hemmen, Werner Schmitz, Rhiannon Victoria McNeill, Sarah Kittel-Schneider, Antje Appelt-Menzel. (2026). Adult ADHD, PARK2 copy number variations and the blood-brain barrier: Culture duration-dependent metabolic alterations in human induced pluripotent stem cell-derived in vitro models.. Fluids and barriers of the CNS. DOI: 10.1186/s12987-026-00892-5
Consulta la pubblicazione originale ↗