Human cellular model to study migration patterns in cortical interneurons under hypoxic stress.

A. Schematic illustrating overall experimental design: hiPSCs were used to derive human cortical organoids (hCO) and human subpallial organoids (hSO); for direct visualization of migrating interneurons, at ∼45-55 days in culture hSO were infected with lentivirus Dlxi1/2b::eGFP and then fused with hCO into human forebrain assembloid (hFA); to study interneuron migration during exposure to hypoxia, hFA were imaged 10-14 days post infection using confocal live-imaging setup focused on the hCO part of the hFA; the movement of the same cells was followed for a total of 48 hrs, each condition for 24hrs: 0-24 hrs in control and 24-48 hrs in hypoxia. Created with BioRender.com. B. Schematic of migratory pattern of interneurons, focused on number of saltations, average saltation length and directionality. Created with BioRender.com. C. Example of migration pattern of one cortical interneuron during control and hypoxic conditions; D. Quantification of saltations number/24 hrs in hypoxia-exposed and non-exposed cortical interneurons by individual cells (paired Wilcoxon test, P<0.0001) and by hiPSC line (two-tailed paired t-test, P=0.003); E. Quantification of average saltation length for hypoxia-exposed and non-exposed cortical interneurons by individual cells (paired Wilcoxon test, P=0.29) and by hiPSC line (two-tailed paired t-test, P=0.15); F. Quantification of directionality of migration in hypoxia-exposed and non-exposed cortical interneurons by individual cells (paired Wilcoxon test, P=0.36), and by hiPSC cell line (two-tailed paired t-test, P=0.47); G. Quantification of saltations number/24 hrs in control conditions, in the first 24 hrs (0-24 hrs) versus the subsequent 24 hrs (24-48 hrs) of live imaging by individual cells (paired Wilcoxon test, P=0.09) and hiPSC line (two-tailed paired t-test, P=0.2); H. Quantification of average saltation length in control conditions, in the first 24 hrs (0-24hrs) versus the subsequent 24 hrs (24-48 hrs) of live imaging by individual cells (two-tailed paired t-test, P=0.79) and by hiPSC line (two-tailed paired t-test, P=0.47; I. Quantification of directionality of migration in control conditions, in the first 24 hrs (0-24 hrs) versus the subsequent 24 hrs (24-48 hrs) of live imaging by individual cells (paired Wilcoxon test, P=0.38) and by hiPSC line (two-tailed paired t-test, P=0.62); Bar charts: mean ± s.e.m; scale bar: 50 μm.

Transcriptional changes in hSO exposed to hypoxia.

A. Schematic of hypoxia exposure of hSO and collection of samples for RNA-Sequencing. Created with BioRender.com. B. Heatmap of differentially expressed genes in RNA-Seq data showing clear transcriptional changes in hypoxia-exposed samples. Samples (n = 24) from hSO differentiated from 4 hiPSC lines were collected at 12 hrs and 24 hrs of exposure to hypoxia, as well as 72 hrs after reoxygenation. The union of all differentially expressed genes (n = 1,473) and samples (n = 24) are ordered by hierarchical clustering (complete linkage of Euclidean distance). Z-score normalized expression values are depicted on a continuous scale from lower values (purple) to higher values (orange). Cell lines, treatment and time point are depicted at the top and represented by different colors; C. Volcano plots of differentially expressed genes (DEGs) at 12 hrs, 24 hrs and 72 hrs after reoxygenation. Each dot represents a single gene. DEGs with a padj <0.05 and an absolute fold change >1.5 are shown in red (upregulated) or blue (downregulated) and unchanged genes are shown in gray; D. Bar plot of the top 10 shared enriched gene pathways across hypoxic conditions. Adjusted p-values are depicted as text and colors represent the length of hypoxic exposure; E. Dumbbell plot of the top 6 DEGs with the largest positive difference and the largest negative difference in Log2 fold change between 12 and 24 hrs of hypoxia exposure; the arrow indicates the direction of change from 12 to 24 hrs. F. Transcriptional upregulation (by qPCR) of ADM gene in hSO samples exposed to 24 hrs of hypoxia: ADM (two-tailed paired t-test, P=0.02); G. Quantitative enzyme immunoassay analysis of adrenomedullin (ADM) peptide in media from hSO exposed and non-exposed to hypoxia (unpaired Mann-Whitney test, P<0.0001); for values below the minimal detection range of <0.01 ng/mL, value was approximated to 0 (we had 9 values approximated to 0 in the control samples). Bar charts: mean ± s.e.m.; Different dot colors represent individual hiPSC lines.

Single cell transcriptional profiling in control and hypoxia exposed hSO and hCO.

Schematic of single cell RNA-seq of hCO and hSO from hFA. Created with BioRender.com. B. UMAP visualization of the resolved single cell RNA-seq data of hSO with assignment of main cell clusters, with control and hypoxia exposure shown by condition (n[total] = 16,022 cells, n[control] = 8,691 cells, n[hypoxia] = 7,331 cells); C. UMAP visualization of the resolved single cell RNA-seq data of hCO with assignment of main cell clusters, with control and hypoxia exposure shown by condition (n[total] = 12,567 cells, n[control] = 6,786 cells, n[hypoxia] = 5,781 cells); D. Single cell gene expression level (log) of adrenomedullin (ADM) in main cell clusters of hSO under control and hypoxia conditions; E. Single cell gene expression level (log) of RAMP1 in main cell clusters of hSO under control and hypoxia conditions; Single cell gene expression level (log) of RAMP2 in main cell clusters of hSO under control and hypoxia conditions; Single cell gene expression level (log) of RAMP3 in main cell clusters of hSO under control and hypoxia conditions; F. (Left) Representative blot for RAMP1 protein expression in control conditions, (Right) Representative blot for RAMP2 protein expression in control conditions; normalized to ɑ-Tubulin G. Quantification of RAMP1 and RAMP2 protein expression by (Left) individual hSO sample (two-tailed paired test, P=0.0004) and (Right) by hiPSC line (two-tailed paired test, P=0.0036) in control conditions; H. Representative blots for RAMP2 protein expression changes in control and hypoxia conditions; I. Quantification of RAMP2 protein expression (Left) by individual hSO sample (two-tailed paired test, P=0.0027) and (Right) by hiPSC line (two-tailed paired test, P=0.0235) in control and hypoxia conditions. Bar charts: mean ± s.e.m.; Different dot colors represent individual hiPSC lines.

Exogenous administration of ADM peptide rescues the migration defects in hypoxia-exposed human cortical interneurons in an ex vivo model using human prenatal cerebral cortex at mid-gestation.

A. Schematic of experimental design for pharmacological rescue experiments using ADM; 0.5 μM ADM was added to the media at the beginning of hypoxia exposure. Created with BioRender.com. B. Example of migration pattern for one interneuron in control versus hypoxia + ADM conditions; C. (Left) Quantification of saltation number/24 hrs in control versus hypoxia conditions (paired Wilcoxon test, P<0.0001) and in control versus hypoxia + ADM by individual cells (paired Wilcoxon test, P=0.3); (Right) Quantification of saltation number/24 hrs in control versus hypoxia conditions (two-tailed paired t-test, P=0.02) and in control versus hypoxia + ADM conditions by hiPSC line (two-tailed paired t-test, P=0.90); D. (Left) Quantification of saltation number/24 hrs in control versus control + ADM conditions by individual cells (paired Wilcoxon test, P=0.23); (Right) Quantification of saltation number/24 hrs in control versus control + ADM conditions by hiPSC line (two-tailed paired t-test, P=0.77); E. Schematic of denaturing procedure for ADM, including reduction and alkylation using dithiothreitol (DTT) and iodoacetamide (IAM), resulting in carbamidomethylated (CAM) cysteines at Cys16 and Cys21. Created with BioRender.com. F. (Left) Quantification by individual cells of saltation number/24 hrs in control versus hypoxia (paired Wilcoxon test, P<0.0001), control versus hypoxia + ADM (paired Wilcoxon test, P=0.05) and control versus hypoxia + denatured ADM (paired Wilcoxon test, P<0.0001); (Right) Quantification by hiPSC line of saltation number/24 hrs in control versus hypoxia (two-tailed paired t-test, P=0.002), control versus hypoxia + ADM (two-tailed paired t-test, P=0.07) and control versus hypoxia + denatured ADM (two-tailed paired t-test, P=0.04); G. Schematic of experimental design for pharmacological rescue experiments using ADM22-52 receptor blocker. Created with BioRender.com. H. (Left) Quantification by individual cells of saltation number/24 hrs in control versus hypoxia conditions (paired Wilcoxon test, P<0.0001), control versus hypoxia + ADM (paired Wilcoxon test, P=0.06), and control versus hypoxia + ADM + ADM22-52 conditions (paired Wilcoxon test, P<0.0001); (Right) Quantification of saltation number/24 hrs by hiPSC line in control versus hypoxia conditions (paired t-test test, P=0.0006), control versus hypoxia + ADM (paired t-test, P=0.28), and control versus hypoxia + ADM + ADM22-52 conditions (two-tailed paired t-test, P=0.003). Bar charts: mean±s.e.m.

Migration defect and rescue by ADM in an ex vivo model using human prenatal cerebral cortex at mid-gestation.

A. Schematic illustrating the overall experimental design: sections of ex vivo human cerebral cortex were collected and initially sectioned at ∼3 mm and subsequently at 400 μm thickness; sections were transferred onto cell culture membrane inserts suspended in culture media; for visualization tissue was transfected with Dlxi1/2b::eGFP lentivirus, and imaged 7-10 days post infection directly on inserts; GFP-tagged ex vivo human prenatal cortical interneurons were monitored for 24 hrs in control conditions and 24 hrs in hypoxic conditions in the presence or absence of 0.5 μM ADM. Created with BioRender.com. B. Example of macroscopic view of a 3mm section of fresh ex vivo human prenatal cerebral cortex; scale bar: 1 cm; C. Representative image of fluorescently-tagged cortical interneurons in a section of ex vivo human prenatal cerebral cortex; D. Transcriptional increase of ADM gene following 24 hrs of exposure to hypoxia of ex vivo human prenatal cerebral cortex samples (two-tailed unpaired t-test, P=0.0005); E. Quantification of saltation numbers/24 hrs in control versus hypoxia conditions (paired Wilcoxon test, P<0.0001); F. Quantification of average saltation length for control versus hypoxia conditions (paired Wilcoxon test, P=0.88); G. Quantification of directionality of migration for control versus hypoxia conditions (paired Wilcoxon test, P=0.3); H. Quantification of saltation numbers/24 hrs in control versus hypoxia + ADM conditions by individual cells (paired Wilcoxon test, P=0.09). Bar charts: mean±s.e.m.; scale bars: 1cm, 200 μm and 50 μm.

Molecular mechanism of rescue by ADM

A. Schematic of the previously reported main molecular pathways modulated by ADM. Created with BioRender.com. B.(Left) Quantification of cAMP concentration (pmol/μl) by individual hSOs in control versus hypoxia (one-way ANOVA, P>0.99) and, control versus hypoxia + ADM (one-way ANOVA, P=0.0007), (Right) Quantification of cAMP concentration (pmol/μl) by hiPSC line in control versus hypoxia (one-way ANOVA, P=0.99) and, control versus hypoxia + ADM (one-way ANOVA, P=0.002); C. (Left) Quantification of PKA activity (O.D. 450nm) by individual hSOs in control versus hypoxia (one-way ANOVA test, P=0.98) and control versus hypoxia + ADM (one-way ANOVA, P<0.0001); (Right) Quantification of PKA activity (O.D. 450nm) by hiPSC line in control versus hypoxia (one-way ANOVA, P>0.99) and control versus hypoxia + ADM (one-way ANOVA, P<0.0001); D. (Left) Quantification of pAKT/AKT by individual hSOs in control versus hypoxia (one-way ANOVA, P<0.0001) and control versus hypoxia + ADM (one-way ANOVA, P<0.0001); (Right) Quantification of pAKT/AKT by hiPSC line in control versus hypoxia (Kruskal-Wallis test, P=0.01) and control versus hypoxia + ADM (Kruskal-Wallis test, P=0.03); E. (Left) Quantification of pERK/ERK by individual hSOs in control versus hypoxia (one-way ANOVA test, P=0.001) and control versus hypoxia + ADM (one-way ANOVA test, P=0.004); (Right) Quantification of pERK/ERK by hiPSC line in control versus hypoxia (one-way ANOVA test, P=0.02) and control versus hypoxia + ADM (one-way ANOVA test, P=0.04); F. (Left) Quantification of pCREB/CREB by individual hSOs in control versus hypoxia (one-way ANOVA test, P=0.003) and control versus hypoxia + ADM (one-way ANOVA test, P=0.27); (Right) Quantification of pCREB/CREB by hiPSC line in control versus hypoxia (one-way ANOVA test, P=0.048) and control versus hypoxia + ADM (one-way ANOVA test, P=0.6). G. Schematic of the proposed molecular pathway activation by exogenous ADM in hSOs, including the most common pentameric structure of the GABAA receptor. Created with BioRender.com. H. (Left) Quantification (by q-PCR) of GABRA1 in hSOs samples in control versus hypoxia conditions (one-way ANOVA test, P=0.002) and control versus hypoxia + ADM (one-way ANOVA test, P=0.17); (Center) Quantification (by q-PCR) of GABRB2 in hSOs samples in control versus hypoxia (one-way ANOVA test, P=0.08) and control versus hypoxia + ADM (one-way ANOVA test, P=0.25); (Right) Quantification (by q-PCR) of GABRG2 in hSO samples in control versus hypoxia conditions (one-way ANOVA test, P=0.007) and control versus hypoxia + ADM (one-way ANOVA test, P=0.23); I. (Left) Quantification (by q-PCR) of CXCR4 in hSOs samples in control versus hypoxia conditions (one-way ANOVA test, P=0.0005) and control versus hypoxia + ADM (one-way ANOVA test, P=0.0031); (Center) Quantification (by q-PCR) of CXCR7 in hSOs samples in control versus hypoxia conditions (one-way ANOVA test, P=0.026) and control versus hypoxia + ADM (one-way ANOVA test, P=0.043); (Right) Quantification (by q-PCR) of CXCL12 in hSOs samples in control versus hypoxia conditions (one-way ANOVA test, P=0.024) and control versus hypoxia + ADM (one-way ANOVA test, P=0.0039). Bar charts: mean±s.e.m.; Different dot colors represent individual hiPSC lines.

Schematic of the overall proposed mechanism of interneuron migration defect rescue by ADM upon hypoxia exposure.

Based on our findings and existing data from literature, we propose endogenously produced ADM has decreased biological activity by impaired ability to form the necessary di-sulfide bond in the absence of oxygen in hypoxia. However, exogenous ADM does have biological activity as the di-sulfide bond is present, and thus it binds efficiently to its receptors, especially RAMP2. This binding initiates an activation of the cAMP/PKA/pCREB pathway, which in turn restore the expression of GABAA receptors and rescue the migration. Created with BioRender.com.

Example of hFA, oxygen level measurements, qPCR changes in expression of hypoxia-responsive genes and cell death analyses.

A. Example image of hFA showing cortical interneurons migrated from the hSO side to the hCO side; arrow indicates the direction of migration from hSO to hCO; B. Oxygen levels (PO2 (mmHg)) in cell culture media under control conditions and following exposure to hypoxia for 24 hrs in the confocal microscope environmental chamber (unpaired t-test, P<0.0001); C. Example of Western Blot showing the stabilization of HIF1α protein in hSOs from 4 hiPSC lines upon exposure to hypoxia, and β-actin expression for normalization; D. Quantification of HIF1α protein, normalized to β-actin (two-tailed paired t-test; P=0.01); E. Transcriptional upregulation of hypoxia-responsive genes in hSO exposed to 24 hrs of hypoxia (<1% O2): PFKP (two-tailed paired t-test, P=0.004), PDK1 (two-tailed paired t-test, P=0.001), VEGFA (two-tailed paired t-test, P=0.003); F. Quantification of Annexin V levels in Dlx+ interneurons in hSOs in control and hypoxia conditions (unpaired t-test, P=0.362); G. Representative image of Dlxi1/2b::eGFP and cleaved-CAS3 positive cells in control and hypoxia conditions; H. Quantification (percentage) of Dlxi1/2b::eGFP and cleaved-CAS3 positive cells in control and hypoxia conditions (unpaired t-test, P= 0.125); I. Table showing the number of non-migratory interneurons from 4 hiPSC lines in control and hypoxia. Bar charts: mean ± s.e.m.; scale bar: 100 μm; Different dot colors represent individual hiPSC lines.

Oxygen level measurements for RNA-Sequencing experiments and dendrogram of sample clustering.

A. Oxygen levels (PO2 (mmHg)) in cell culture media under control conditions and following exposure to hypoxia for 24 hrs in the C-chamber hypoxia sub-chamber (Biospherix) (unpaired t-test, P<0.0001); B. Hierarchical clustering, using Ward’s criterion, of the first three principal components of the normalized gene expression profiles of all 17,777 genes passing quality control shows clear separation of 12 and 24 hrs hypoxia-exposed samples versus control and from reoxygenated samples. Bar charts: mean±s.e.m.

Quality control and robustness analyses for scRNA-sequencing data

A. UMAP visualization of main cellular subclusters in hSOs with corresponding gene expression including, astrocytes (SPARC, S100A10), progenitors (SOX9), cycling progenitors (SOX9, TOP2A), glutamatergic neurons (STMN2, NEUROD2), interneuron progenitors (NKX2.1, LHX6, DLX2), and interneurons (STMN2, GAD1, SLC32A1, SCG2); B. UMAP visualization of main cellular subclusters in hCO with corresponding gene expression including, choroid plexus (TTR), progenitors (SOX2), cycling progenitors (SOX2, TOP2A), glutaminergic neurons (STMN2, NEUROD2), interneuron progenitors (NKX2.1), and interneurons (STMN2, GAD1, SLC32A1, SCG2); C. Single cell gene expression level (log) of hypoxia-responsive genes (PDK1 and PFKP) across main cellular subclusters in hSO; D. Single cell gene expression level (log) of hypoxia-responsive genes (PDK1 and PFKP) across main cell clusters in hCO.

A. LC-MS based analysis of the non-modified ADM (top) and the denatured ADM form (bottom), displays characteristic M+4 through M+7 peaks with two CAM modifications (+114 Da) in the bottom panel represented by the peak shift (arrow); B. (Left) Quantification of saltations number/24hr in control versus hypoxia by individual cells (Friedman test, P<0.0001) and control versus reoxygenation by individual cells (Friedman test, P<0.0001) ; (Right) Quantification of saltations/24hr in control versus hypoxia (one-way ANOVA, P=0.003) by hiPSC line and control versus reoxygenation by hiPSC line (one-way ANOVA, P=0.003); C. (Left) Quantification of saltations number/24 hrs in control versus hypoxia + ADM condition by individual cells (Friedman test, P=0.05) and in control versus re-oxygenation after hypoxia + ADM by individual cells (Friedman test, P<0.0001); 35% decrease compared to 65% decrease when ADM not added to hypoxia; (Right) Quantification of saltations number/24 hrs in control versus hypoxia + ADM condition by hiPSC line (one-way ANOVA, P=0.84) and in control versus re-oxygenation after hypoxia + ADM condition by hiPSC line (one-way ANOVA, P=0.12); Bar charts: mean±s.e.m.; Different dot colors represent individual hiPSC lines.

Additional data to support molecular mechanisms of rescue by ADM.

A. (Left) Quantification (by q-PCR) of GABRB3 in individual samples in control versus hypoxia conditions (Kruskal-Wallis test, P=0.02) and control versus hypoxia + ADM (Kruskal-Wallis test, P=0.05); (Center) Quantification (by q-PCR) of GABRG3 in individual samples in control versus hypoxia conditions (one-way ANOVA test, P=0.01) and control versus hypoxia + ADM (one-way ANOVA test, P=0.1); (Right) Quantification (by q-PCR) of GABRA2 in individual samples in control versus hypoxia conditions (one-way ANOVA, P=0.8) and control versus hypoxia + ADM (one-way ANOVA, P=0.75). Bar charts: mean±s.e.m.; Different dot colors represent individual hiPSC lines.