Maturation and integration of transplanted human cortical organelles

Thank you for visiting Nature.com. You are using a browser version with limited CSS support. For the best experience, we recommend that you use an updated browser (or disable Compatibility Mode in Internet Explorer). In addition, to ensure ongoing support, we show the site without styles and JavaScript.
Displays a carousel of three slides at once. Use the Previous and Next buttons to move through three slides at a time, or use the slider buttons at the end to move through three slides at a time.
Self-assembling neural organelles represent a promising in vitro platform for modeling human development and disease. However, organoids lack the connectivity that exists in vivo, which limits maturation and prevents integration with other circuits that control behavior. Here we show that human stem cell-derived cortical organoids transplanted into the somatosensory cortex of neonatal nude rats develop mature cell types that integrate into sensory and motivation-related circuits. MRI revealed post-transplant organoid growth in several stem cell lines and animals, while single-core analysis revealed progression of corticogenesis and the emergence of an activity-dependent transcription program. Indeed, transplanted cortical neurons exhibit more complex morphological, synaptic, and internal membrane properties than their in vitro counterparts, allowing the detection of neuronal defects in patients with Timothy’s syndrome. Anatomical and functional tracing has shown that transplanted organelles receive thalamocortical and corticocortical inputs, and in vivo recordings of neural activity suggest that these inputs can generate sensory responses in human cells. Finally, cortical organoids extend axons throughout the rat brain, and their optogenetic activation leads to reward-seeking behavior. Thus, transplanted human cortex neurons mature and participate in the host’s circuits that control behavior. We expect this approach to facilitate the detection of strand-level phenotypes in patient-derived cells that cannot be detected by other means.
The developing human brain is a remarkable self-organizing process in which cells proliferate, differentiate, migrate, and connect to form functional neuronal circuits that are further refined through sensory experience. A key problem in understanding human brain development, especially in the context of disease, is the lack of access to brain tissue. Self-organizing organelles, including human cortex organoids (hCO; also known as the human cortex sphere), can generate 2,3,4,5,6. However, several limitations limit their wider application to understanding the development and functioning of neural circuits. In particular, it is unclear whether hCO maturation is limited by the absence of certain microenvironmental and sensory inputs present in vivo. In addition, because hCOs are not integrated into circuits that can generate behavioral outcomes, their utility in modeling genetically complex and behavioral neuropsychiatric disorders is currently limited.
The transplantation of hCO into an intact living brain can overcome these limitations. Previous studies have shown that human neurons transplanted into the rodent cortex are able to survive, project, and communicate with rodent cells7,8,9,10,11,12. However, these experiments are usually performed on adult animals, which may limit synaptic and axonal integration. Here, we describe a transplantation paradigm in which we transplanted 3D hCO derived from hiPS cells into the primary somatosensory cortex (S1) of immunodeficient rats at an early stage of plastic development. Transplanted hCO (t-hCO) neurons undergo substantial maturation, receive thalamocortical and cortical-cortical inputs that elicit sensory responses, and extend axonal projections into the rat brain to drive reward-seeking behavior. Extended maturation of t-hCO has revealed neuronal defects in patients with Timothy’s syndrome (TS), a severe genetic disorder caused by mutations in the voltage-sensitive L-type CaV1.2 calcium channel (encoded by CACNA1C).
To study human cortical neurons in circuits in vivo, we stereotactically transplanted intact 3D hCO into S1 of early postnatal athymic rats (days 3-7 postnatally) (Fig. 1a and expanded data of Fig. 1a-c). At this point, the thalamocortical and corticocortical axonal projections have not yet completed their S1 innervation (ref. 13). Thus, this approach is designed to maximize t-hCO integration while minimizing impact on endogenous circuits. To visualize the location of t-hCO in live animals, we performed T2-weighted MRI brain reconstructions of rats 2–3 months after transplantation (Fig. 1b and extended data, Fig. 1d). t-hCO were readily observed and volume measurements of t-hCO were similar to those calculated from fixed slices (Extended Data Fig. 1d,e; P > 0.05). t-hCO were readily observed and volume measurements of t-hCO were similar to those calculated from fixed slices (Extended Data Fig. 1d,e; P > 0.05). t-hCO легко наблюдались, а объемные измерения t-hCO были аналогичны рассчитанным для фиксированных срезов (расширенные данные, рис. 1d, e; P> 0,05). t-hCO were easily observed, and volumetric t-hCO measurements were similar to those calculated for fixed sections (expanded data, Fig. 1d, e; P > 0.05).很容易观察到t-hCO，并且t-hCO 的体积测量值与从固定切片计算的测量值相似（扩展数据图1d、e；P > 0.05）。很容易观察到t-hCO，并且t-hCO t-hCO легко наблюдался, а объемные измерения t-hCO были аналогичны рассчитанным для фиксированных срезов (расширенные данные, рис. 1d, e; P> 0,05). t-hCO was easily observed, and volumetric t-hCO measurements were similar to those calculated for fixed sections (expanded data, Fig. 1d, e; P > 0.05). We determined t-hCO in 81% of transplanted animals approximately 2 months after transplantation (n = 72 animals; hCO from 10 hiPS cell lines; hiPS cell lines in Supplementary Table 1). Of these, 87% were located in the cerebral cortex (Fig. 1c). By performing serial MRI scans at multiple time points in the same transplanted rat, we found a nine-fold increase in t-hCO volume within 3 months (Fig. 1d and expanded data, Fig. 1f). Transplanted animals had a high survival rate (74%) at 12 months post-transplantation (expanded data, Fig. 1g and Supplementary Table 2), and no overt motor or memory impairments, gliosis, or electroencephalogram (EEG) were found. Data Fig. 1g and supplementary table 2). 1h–m and 3e).
a, Schematic of experimental design. hCO derived from hiPS cells was transplanted into S1 of newborn nude rats on days 30-60 of differentiation. b, T2-weighted coronal and horizontal MRI images showing t-hCO in S1 2 months after transplantation. Scale bar, 2 mm. c, Quantification of engraftment success rates shown for each hiPS cell line (n = 108, numbers within bars indicate amount of t-hCO per hIPS cell line) and cortical or subcortical location (n = 88). d, MRI image of a coronary artery (left; scale bar, 3 mm) and corresponding 3D volumetric reconstruction (scale bar, 3 mm) showing an increase in t-hCO over 3 months. e, Review of t-hCO patterns in rat cerebral cortex. Scale bar, 1 mm. f, Representative immunocytochemical images of t-hCO shown from top left to right (during differentiation): PPP1R17 (4 months old), NeuN (8 months old), SOX9 and GFAP (8 months old), PDGFRα; (8 months), MAP2 (8 months) and IBA1 (8 months). Scale bar, 20 µm. Co-expression of HNA indicates cells of human origin. g, snRNA-seq: Unified manifold and projection (UMAP) dimensionality reduction imaging of all high-quality t-hCO nuclei after Seurat integration (n=3 t-hCO samples, n=2 hiPS cell lines). Astrocytes, cells of the astrocyte line; cyc prog, circulating progenitors; GluN DL, deep glutamatergic neurons; GluN DL/SP, deep and sublamellar glutamatergic neurons; GluN UL, upper layer glutamatergic neurons; oligodendrocytes, oligodendrocytes; OPC, oligodendrocyte progenitor cells; RELN, reelin neurons. h, Gene Ontology (GO) term enrichment analysis of genes significantly upregulated (adjusted P < 0.05, fold change > 2, expressed in at least 10% of nuclei) in t-hCO glutamatergic neurons compared with hCO glutamatergic neurons. h, Gene Ontology (GO) term enrichment analysis of genes significantly upregulated (adjusted P < 0.05, fold change > 2, expressed in at least 10% of nuclei) in t-hCO glutamatergic neurons compared with hCO glutamatergic neurons. h, Анализ обогащения терминов Gene Ontology (GO) для генов со значительной активацией (скорректированный P <0,05, кратность изменения > 2, экспрессия по крайней мере в 10% ядер) в глутаматергических нейронах t-hCO по сравнению с глутаматергическими нейронами hCO. h, Gene Ontology (GO) term enrichment analysis for genes with significant activation (adjusted P<0.05, fold change >2, expression in at least 10% nuclei) in t-hCO glutamatergic neurons compared to hCO glutamatergic neurons. h，与hCO 谷氨酸能神经元相比，t-hCO 谷氨酸能神经元中基因显着上调（调整后P < 0.05，倍数变化> 2，在至少10% 的细胞核中表达）的基因本体论(GO) 术语富集分析。 h ， 与 hco 谷氨酸 能 元 相比 ， t-hco 谷氨酸 能 神经 元 基因 显着 上调 （后 后 p <0.05 ， 变化 变化> 2 ， 至少 10% 的 核中 表达） 基因 基因 基因 的 的 的 的 的 的 的 的 的 的 的 的 的 的 的本体论(GO) 术语富集分析。 h, гены значительно активизировались (скорректированный P <0,05, кратность изменения> 2, экспрессируется по крайней мере в 10% ядер) в глутаматергических нейронах t-hCO по сравнению с глутаматергическими нейронами hCO Онтологический (GO) анализ термина обогащения. h, genes were significantly upregulated (adjusted P < 0.05, fold change > 2, expressed in at least 10% of nuclei) in t-hCO glutamatergic neurons compared to hCO glutamatergic neurons Ontological (GO) analysis of the enrichment term. The dotted line indicates a q value of 0.05. i, UMAP imaging of GluN cell types in t-hCO using label transfer from a reference 22 snRNA-seq adult motor cortex dataset. CT — corticothalamic cells, ET — extracerebral cells, IT — internal telencephalic cells, NP — near projection.
We then assessed the cytoarchitecture and overall cellular composition of t-hCO. Antibody staining of rat endothelial cells revealed vascularization with t-hCO, while IBA1 staining revealed the presence of rat microglia throughout the graft (Fig. 1f and expanded data, Fig. 3c,d). Immunostaining revealed human nuclear antigen (HNA) positive cells co-expressing PPP1R17 (cortical progenitors), NeuN (neurons), SOX9 and GFAP (glial-derived cells) or PDGFRα (oligodendrocyte progenitors) (Figure 1f). To study the cellular composition of t-hCO at single cell resolution, we performed single-core RNA sequencing (snRNA-seq) after approximately 8 months of differentiation. Bulk filtration and removal of rat nuclei yielded 21,500 high-quality human mononuclear maps (Fig. 1g and expanded data, Fig. 4a,b). Expression patterns of typical cell-type markers identified clusters of major cortical cell classes, including deep and superficial glutamatergic neurons, circulating progenitors, oligodendrocytes, and astrocyte lineage (Fig. 1g, expanded data, Fig. 4c, and Supplementary Table 3). Immunostaining for SATB2 and CTIP2 showed that despite the presence of cortical subtypes, t-hCO did not show clear anatomical stratification (expanded data, Fig. 3a). stage-matched snRNA-seq hCO produced broadly similar cell classes, with a few exceptions, including the absence of oligodendrocytes and the presence of GABAergic neurons, which may reflect the previously reported favorable in vitro conditions for lateral progenitor cells15 (expanded data, Fig. 4f – i and Supplementary Table 4). Differential gene expression analysis revealed significant differences in glutamatergic neurons between t-hCO and hCO (Supplementary Table 5), including activation of sets of genes associated with neuronal maturation such as synaptic signaling, dendritic localization, and voltage-gated channel activity (Figure 1h and Supplementary Table 5). table 6). Accordingly, cortical glutamatergic t-hCO neurons exhibited accelerated transcriptional maturation.
To elucidate whether these transcriptional changes in t-hCO were related to morphological differences between hCO in vitro and t-hCO in vivo, we reconstructed stage-matched biocytin-filled hCO and hCO in acute sections after 7–8 months of differentiation. hCO neurons (Fig. 2a). t-hCO neurons were significantly larger, had 1.5 times the soma diameter, twice as many dendrites, and an overall six-fold increase in total dendritic length compared to in vitro hCO (Fig. 2b). In addition, we observed a significantly higher density of dendritic spines in t-hCO neurons than in hCO neurons (Fig. 2c). This suggests that t-hCO neurons undergo extensive dendritic elongation and branching, which, in combination with continued cell proliferation, may contribute to the intensive growth of t-hCO after transplantation (Fig. 1d and Extended Data Fig. 1f). This prompted us to investigate the electrophysiological properties. Membrane capacitance was eight times higher (expanded data, Fig. 8d), the resting-state membrane potential was more hyperpolarized (approximately 20 mV), and current injection induced a higher maximum excitation rate in t-hCO neurons than in hCO neurons. in vitro (Fig. 2d), e), which is consistent with the larger and more complex morphological features of t-hCO. In addition, the frequency of spontaneous excitatory postsynaptic current events (EPSC) was significantly higher in t-hCO neurons (Fig. 2f), suggesting that the increased density of dendritic spines observed in t-hCO neurons was associated with functional excitability. sexual synapse. We confirmed the immature character of hCO neurons in vitro by recording labeled glutamatergic neurons (expanded data, Fig. 6a-c).
a, 3D reconstruction of biocytin-filled hCO and t-hCO neurons after 8 months of differentiation. b, Quantification of morphological features (n = 8 hCO neurons, n = 6 t-hCO neurons; **P = 0.0084, *P = 0.0179 and ***P < 0.0001). b, Quantification of morphological features (n = 8 hCO neurons, n = 6 t-hCO neurons; **P = 0.0084, *P = 0.0179 and ***P < 0.0001). б, количественная оценка морфологических признаков (n = 8 нейронов hCO, n = 6 нейронов t-hCO; ** P = 0,0084, * P = 0,0179 и *** P <0,0001). b, quantification of morphological features (n=8 hCO neurons, n=6 t-hCO neurons; **P=0.0084, *P=0.0179, and ***P<0.0001). b，形态学特征的量化（n = 8 个hCO 神经元，n = 6 个t-hCO 神经元；**P = 0.0084，*P = 0.0179 和***P < 0.0001）。 b，形态学特征的量化（n = 8 个hCO 神经元，n = 6 个t-hCO 神经元；**P = 0.0084，*P = 0.0179 和***P < 0.0001）。 б, количественная оценка морфологических признаков (n = 8 нейронов hCO, n = 6 нейронов t-hCO; ** P = 0,0084, * P = 0,0179 и *** P <0,0001). b, quantification of morphological features (n=8 hCO neurons, n=6 t-hCO neurons; **P=0.0084, *P=0.0179, and ***P<0.0001). c, 3D reconstruction of hCO and t-hCO dendritic branches after 8 months of differentiation. Red asterisks indicate putative dendritic spines. Dendritic spine density quantification (n = 8 hCO neurons, n = 6 t-hCO neurons; **P = 0.0092). d, Quantification of the resting membrane potential (n = 25 hCO neurons, n = 16 t-hCO neurons; ***P < 0.0001). d, Quantification of the resting membrane potential (n = 25 hCO neurons, n = 16 t-hCO neurons; ***P < 0.0001). d, количественная оценка мембранного потенциала покоя (n = 25 нейронов hCO, n = 16 нейронов t-hCO; *** P <0,0001). d, resting membrane potential quantification (n = 25 hCO neurons, n = 16 t-hCO neurons; ***P < 0.0001). d，静息膜电位的量化（n = 25 hCO 神经元，n = 16 t-hCO 神经元；***P < 0.0001）。 d，静息膜电位的量化（n = 25 hCO 神经元，n = 16 t-hCO 神经元；***P < 0.0001）。 d, количественная оценка мембранного потенциала покоя (n = 25 нейронов hCO, n = 16 нейронов t-hCO; *** P <0,0001). d, resting membrane potential quantification (n = 25 hCO neurons, n = 16 t-hCO neurons; ***P < 0.0001). e, Repetitive action potential firing in hCO and t-hCO induced by increasing current injections, and quantification of the maximal firing rate (n = 25 hCO neurons, n = 16 t-hCO neurons; ***P < 0.0001). e, Repetitive action potential firing in hCO and t-hCO induced by increasing current injections, and quantification of the maximal firing rate (n = 25 hCO neurons, n = 16 t-hCO neurons; ***P < 0.0001). e, повторное возбуждение потенциала действия в hCO и t-hCO, вызванное увеличением тока, и количественная оценка максимальной скорости возбуждения (n = 25 нейронов hCO, n = 16 нейронов t-hCO; *** P <0,0001). e, action potential re-firing in hCO and t-hCO induced by current increase and quantification of maximum firing rate (n = 25 hCO neurons, n = 16 t-hCO neurons; *** P < 0.0001). e，通过增加电流注入诱导的hCO 和t-hCO 重复动作电位放电，以及最大放电率的量化（n = 25 个hCO 神经元，n = 16 个t-hCO 神经元；***P < 0.0001）。 E ， 通过 增加 电流 注入 的 的 hco 和 t-hco 重复 电位 放电 ， 以及 最 大 的 量化 （（n = 25 个 hco 神经 ， n = 16 个 t-hco 神经 ； *** p <0.0001） 。 e, повторяющееся возбуждение потенциала действия hCO и t-hCO, вызванное увеличением подачи тока, и количественная оценка максимальной скорости возбуждения (n = 25 нейронов hCO, n = 16 нейронов t-hCO; *** P <0,0001). e, repetitive firing of hCO and t-hCO action potentials induced by increased current supply and quantification of maximum firing rate (n = 25 hCO neurons, n = 16 t-hCO neurons; *** P < 0.0001). f, Spontaneous EPSCs (sEPSCs) in hCO and t-hCO neurons at 8 months of differentiation, and quantification of the frequency of synaptic events (n = 25 hCO neurons, n = 17 t-hCO neurons; ***P < 0.0001). f, Spontaneous EPSCs (sEPSCs) in hCO and t-hCO neurons at 8 months of differentiation, and quantification of the frequency of synaptic events (n = 25 hCO neurons, n = 17 t-hCO neurons; ***P < 0.0001) . f, спонтанные EPSC (sEPSC) в нейронах hCO и t-hCO через 8 месяцев дифференцировки и количественная оценка частоты синаптических событий (n = 25 нейронов hCO, n = 17 нейронов t-hCO; *** P <0,0001) . f, Spontaneous EPSCs (sEPSCs) in hCO and t-hCO neurons at 8 months of differentiation and quantification of synaptic event rates (n=25 hCO neurons, n=17 t-hCO neurons; ***P<0.0001). f，分化8 个月时hCO 和t-hCO 神经元中的自发性EPSCs (sEPSCs)，以及突触事件频率的量化（n = 25 hCO 神经元，n = 17 t-hCO 神经元；***P < 0.0001） . f，分化8 个月时hCO 和t-hCO 神经元中的自发性EPSCs (sEPSCs)，以及突触事件频率的量匼（n = 25 hCO 神率的量匼（n = 25 hCO 神率的量匼（n = 25hCO P < 0.0001） . f, спонтанные EPSC (sEPSC) в нейронах hCO и t-hCO через 8 месяцев дифференцировки и количественная оценка частоты синаптических событий (n = 25 нейронов hCO, n = 17 нейронов t-hCO; *** P <0,0001). f, Spontaneous EPSCs (sEPSCs) in hCO and t-hCO neurons at 8 months of differentiation and quantification of synaptic event rates (n = 25 hCO neurons, n = 17 t-hCO neurons; *** P<0.0001). For b-f, hCO and t-hCO in line 1208-2 were taken from the same differentiation batch maintained in parallel. g, Gene set enrichment analysis (one-sided Fisher’s exact test) of genes significantly upregulated (adjusted P < 0.05, fold change > 2, expressed in at least 10% of nuclei) in t-hCO glutamatergic neurons compared with hCO glutamatergic neurons with gene sets of both early-response (ERG) and late-response (LRG) activity-dependent genes identified from an in vivo mouse study16 and human-specific LRGs from in vitro neurons17. g, Gene set enrichment analysis (one-sided Fisher’s exact test) of genes significantly upregulated (adjusted P < 0.05, fold change > 2, expressed in at least 10% of nuclei) in t-hCO glutamatergic neurons compared with hCO glutamatergic neurons with gene sets of both early-response (ERG) and late-response (LRG) activity-dependent genes identified from an in vivo mouse study16 and human-specific LRGs from in vitro neurons17. g, анализ обогащения набора генов (односторонний точный критерий Фишера) генов со значительной активацией (скорректированный P <0,05, кратность изменения > 2, экспрессия по меньшей мере в 10% ядер) в глутаматергических нейронах t-hCO по сравнению с глутаматергическими нейронами hCO наборы генов как раннего (ERG), так и позднего (LRG) генов, зависящих от активности, идентифицированных в исследовании на мышах in vivo16, и специфических для человека LRG из нейронов in vitro17. g, analysis of gene set enrichment (one-tailed Fisher’s exact test) of genes with significant activation (adjusted P<0.05, fold change >2, expression in at least 10% of nuclei) in t-hCO glutamatergic neurons compared to hCO glutamatergic neurons sets of both early (ERG) and late (LRG) activity-dependent genes identified in in vivo mice16 and human-specific LRGs from neurons in vitro17. g，t-hCO谷氨酸能神经元与hCO谷氨酸能神经元相比，t-hCO谷氨酸能神经元显着上调（调整后P<0.05，倍数变化>2，在至少10%的细胞核中表达）的基因集富集分析（单侧Fisher精确检验）从体内小鼠研究中鉴定的早期反应(ERG) 和晚期反应(LRG) 活性依赖性基因的基因组16 和体外神经元17 中的人类特异性LRG。 g ， t-hco 谷氨酸 神经 元 与 hco 谷氨酸 神经 元 相比 ， t-hco 谷氨酸 神经 元 上调 （调整 后 p <0.05 ， 倍数> 2 ， 至少 至少 10%的 细胞 核中 表达） 的 集富集 分析 （单侧 fisher 精确）） 体内 小鼠 研究 中 的 早期 反应 反应 反应 和 晚期 反应 反应 (lrg) 活性 基因 的 基因组 16 和 神经元 神经元 17 中 中 17 中 17的人类特异性LRG。 g, глутаматергические нейроны t-hCO были значительно активизированы по сравнению с глутаматергическими нейронами hCO (скорректированный P<0,05, кратность изменения> 2, не менее 10% Анализ обогащения набора генов (односторонний точный тест Фишера) раннего ответа (ERG) и позднего гены, зависящие от активности ответа (LRG), идентифицированные в исследованиях на мышах in vivo16 и нейронах in vitro17. LRG, специфичные для человека. g, t-hCO glutamatergic neurons were significantly upregulated compared to hCO glutamatergic neurons (adjusted P<0.05, fold change >2, at least 10% Early response (ERG) and late response gene enrichment analysis (one-tailed Fisher’s exact test) response activity dependent genes (LRGs) identified in in vivo mice16 and in vitro neurons.17 Human specific LRGs. The dotted line indicates a Bonferroni-corrected P value of 0.05. h, GluN gene expression (pseudo-package and scaling of each gene) was significantly upregulated in snRNA-seq replicas of LRG genes in t-hCO glutamatergic neurons. i, immunostaining showing SCG2 expression in t-hCO (upper) and hCO (lower) neurons. White arrows point to SCG2+ cells. Scale bar, 25 µm. Data are expressed as mean ± standard deviation.
Based on the increased activity of t-hCO observed in ex vivo slices, snRNA-seq revealed an activity-dependent upregulation of gene transcripts in t-hCO compared to hCO in vitro. Glutamatergic t-hCO neurons expressed higher levels of genes regulating late response activity (Fig. 2g,h), which were found in previous studies in mouse and human neurons16,17. For example, BDNF18, SCG2, and OSTN, a primate-specific activity-regulating gene, showed increased expression in t-hCO neurons compared to hCO neurons (Fig. 2g-i). Thus, t-hCO neurons exhibited enhanced maturation characteristics compared to hCO neurons by transcriptional, morphological, and functional analyses.
To further evaluate the association of t-hCO maturation with human brain development, we performed transcriptomic comparisons of fetal and adult cortical cell types19,20 and adult21,22 as well as extensive data on cortical gene expression23 during development (expanded data, Fig. 5). ). with previous work 24 , the global hCO and t-hCO transcriptome maturation status at 7–8 months of differentiation is broadly consistent with in vivo development time and is most equivalent to late fetal life (Extended Data Fig. 5a). Notably, we observed increased transcriptome maturity in t-hCO compared to age-matched hCO, as well as transcriptome activation associated with synaptogenesis, astrogenesis, and myelination (expanded data, Fig. 5b-d). At the cellular level, we found evidence of a thinner cortex subtype in t-hCO, with clusters of glutamatergic neurons overlapping with adult L2/3, L5, and L6 neuron subtypes (Figure 1i). In contrast, cluster overlap between glutamatergic t-hCO neurons and fetal cortical neurons was more limited in mid-pregnancy (expanded data, Figure 5e-j). To determine whether t-hCO neurons are functionally similar to human postnatal neocortical neurons, we performed electrophysiological recordings and anatomical reconstructions of human L2/3 pyramidal neurons in sharp sections of the human postnatal cortex (expanded data, Fig. 7a). The electrophysiological properties of L2/3 pyramidal neurons were similar to those of t-hCO pyramidal neurons (expanded data, Fig. 7e). Morphologically, L2/3 neurons from postnatal human samples were more similar to t-hCO than to hCO, although L2/3 cells were longer, contained more branches overall, and had a higher spine density (Fig. 3g and expanded data, Fig. 7b-). G).
a, transplantation of hCO produced by control and TS hiPS cell lines into neonatal rats. b, 3D reconstruction of biocytin-filled t-hCO neurons after 8 months of differentiation. c, quantification of mean dendritic length (n = 19 control neurons, n = 21 TS neurons; **P = 0.0041). d, 3D-reconstructed dendritic branches from control and TS t-hCO at 8 months of differentiation, and quantification of dendritic spine density (n = 16 control neurons, n = 21 TS neurons, ***P < 0.0001). d, 3D-reconstructed dendritic branches from control and TS t-hCO at 8 months of differentiation, and quantification of dendritic spine density (n = 16 control neurons, n = 21 TS neurons, ***P < 0.0001). d, 3D-реконструкция дендритных ветвей из контроля и TS t-hCO через 8 месяцев дифференцировки и количественная оценка плотности дендритных шипов (n = 16 контрольных нейронов, n = 21 TS нейронов, *** P <0,0001). d, 3D reconstruction of dendritic branches from control and t-hCO TS at 8 months of differentiation and dendritic spine density quantification (n=16 control neurons, n=21 TS neurons, ***P<0.0001). d，分化8 个月时对照和TS t-hCO 的3D 重建树突分支，以及树突棘密度的量化（n = 16 个对照神经元，n = 21 个TS 神经元，***P < 0.0001）。 d ， 分化 8 个 时 对照 和 ts t-hco 的 3d 重建 分支 分支 以及 树突棘 密度 量化 （n = 16 个 神经 元 ， n = 21 个 ts 神经 ， *** p <0.0001 ）。 d, 3D-реконструкция дендритных ветвей контроля и TS t-hCO через 8 месяцев дифференцировки и количественная оценка плотности дендритных шипов (n = 16 контрольных нейронов, n = 21 TS нейронов, *** P <0,0001). d, 3D reconstruction of control dendritic branches and TS t-hCO at 8 months of differentiation and dendritic spine density quantification (n=16 control neurons, n=21 TS neurons, ***P<0.0001). Red asterisks indicate putative dendritic spines. e, spontaneous EPSCs in control and TS t-hCO neurons after 8 months of differentiation. f, cumulative frequency plot and quantification of frequency and amplitude of synaptic events (n=32 control neurons, n=26 TS neurons; **P=0.0076 and P=0.8102). g, Scholl analysis of TS and control neurons in hCO and t-hCO. Dashed lines show human L2/3 postnatal pyramidal neurons for comparison (n = 24 control t-hCO neurons, n = 21 TS t-hCO neurons, n = 8 control hCO neurons, and n = 7 TS hCO neurons). Data are expressed as the mean ± standard deviation
The ability of t-hCO to replicate the morphological and functional features of human cortex neurons at a high level prompted us to explore whether t-hCO could be used to detect disease phenotypes. We focused on TS, a severe neurodevelopmental disorder caused by gain-of-function mutations in the gene encoding CaV1.2, which initiates activity-dependent gene transcription in neurons. We obtained hCO from three TS patients carrying the most common substitution (p.G406R) and three controls (Fig. 3a). After transplantation, we found that dendritic morphology was altered in TS neurons compared to controls (Fig. 3b and expanded data, Fig. 8a,b), with a two-fold increase in the number of primary dendrites and an overall increase in mean and overall decrease in dendritic length (Fig. 3c and extended data, Fig. 8c). This was associated with an increased density of spines and an increased frequency of spontaneous EPSCs in TS compared to control neurons (Fig. 3d–f and expanded data, Fig. 8g). Further analysis revealed patterns of abnormal dendritic branching in t-hCO TS compared to controls, but not in in vitro TS hCO at a similar stage of differentiation (Fig. 3g). This is consistent with our previous reports of activity-dependent dendritic shrinkage in TS and highlights the ability of this transplant platform to detect disease phenotypes in vivo.
We then asked to what extent t-hCO cells are functionally integrated into rat S1. S1 in rodents receives strong synaptic inputs from the ipsilateral ventral basal and posterior thalamic nuclei, as well as the ipsilateral motor and secondary somatosensory cortices, and contralateral S1 (Fig. 4a). To restore the innervation pattern, we infected hCO with rabies virus-dG-GFP/AAV-G and transplanted hCO into S1 rat 3 days later. We observed dense GFP expression in neurons of the ipsilateral S1 and ventral basal ganglia 7–14 days after transplantation (Fig. 4b, c). In addition, antibody staining of the thalamic marker netrin G1 revealed the presence of thalamic endings in t-hCO (Fig. 4d, e). To evaluate whether these afferent projections could elicit synaptic responses in t-hCO cells, we performed whole cell recordings from human cells in sharp sections of the thalamocortical layer. Electrical stimulation of rat S1, internal capsule, white matter, fibers near t-hCO or optogenetic activation of opsin-expressing thalamic endings in t-hCO induced short-latency EPSCs in t-hCO neurons exposed to the AMPA receptor antagonist NBQX. (Fig. 4f, g and extended data, Fig. 9a–g). These data demonstrate that t-hCO is anatomically integrated into the rat brain and is able to be activated by rat host tissue.
a, Schematic diagram of a rabies tracking experiment. b, GFP and human-specific STEM121 expression between t-hCO and rat cerebral cortex (upper panel). Also shown is GFP expression in the rat ipsilateral ventral basal nucleus (VB) (lower left) and ipsilateral S1 (lower right). Scale bar, 50 µm. The red squares represent the areas of the brain where the images were taken. c, quantification of cells expressing GFP (n = 4 rats). d, e — Netrin G1+ thalamic terminals in t-hCO. d shows a coronal section containing t-hCO and VB nuclei. Scale bar, 2 mm. e shows Netrin G1 and STEM121 expression in t-hCO (left) and VB (right) neurons. Scale bar, 50 µm. The orange dotted line indicates the t-hCO border. f, g, Current traces of t-hCO neurons after electrical stimulation in S1 rat (f) or internal capsule (g), with (purple) or without (black) NBQX (left). EPSC amplitudes with and without NBQX (n = 6 S1 neurons, *P = 0.0119; and n = 6 internal capsule neurons, **P = 0.0022) (center). Percentage of t-hCO neurons showing EPSC in response to electrical stimulation of rat S1 (f) or internal capsule (g) (right). aCSF, artificial cerebrospinal fluid. h, schematic diagram of the 2P imaging experiment (left). Expression of GCaMP6s in t-hCO (middle). Scale bar, 100 µm. Fluorescence time lapse of GCaMP6s (right). i, Z-score of spontaneous activity fluorescence. j, schematic illustration of mustache stimulation. k, z-scored 2P fluorescence trajectories in one trial, aligned with whisker deviation at time zero (dashed line) in example cells. l, population-averaged z-score responses of all cells aligned with whisker deviation at time zero (dashed line) (red) or randomly generated timestamps (grey). m. Schematic diagram of the experiment on optical marking. n, Raw voltage curves from an example t-hCO cell during blue laser stimulation or whisker deflection. Red arrows indicate the first spikes caused by light (top) or caused by whisker deflection (bottom). Gray shading indicates periods of whisker deflection. o, Peak light waveforms and whisker deflection responses. p, spikes of a single attempt, aligned with the deviation of the whiskers in the cells of the example. 0 indicates whisker deviation (dashed line). q, population-averaged z-score firing rate for all photosensitive cells, aligned with whisker deviation at time zero (dashed line) (red) or randomly generated timestamps (grey). r, Proportion of photosensitive units significantly modulated by whisker deviation (n = 3 rats) (left). Peak z-score latency (n = 3 rats; n = 5 (light green), n = 4 (dark green), and n = 4 (cyan) whisker deflection modulation units per rat) (right). Data are expressed as mean ± standard deviation
We then asked if t-hCO could be activated by sensory stimuli in vivo. We transplanted hCO expressing the genetically encoded calcium indicators GCaMP6 into S1 rats. After 150 days, we performed fiber photometry or two-photon calcium imaging (Fig. 4h and expanded data, Fig. 10a). We found that t-hCO cells exhibited synchronized rhythmic activity (Figure 4i, Expanded Data, Figure 10b and Supplementary Video 1). To characterize peak t-hCO activity, we performed extracellular electrophysiological recordings in anesthetized transplant rats (expanded data, Fig. 10c-f). We have generated stereotaxic coordinates from MRI images; thus, these recorded units represent putative human neurons, although electrophysiology alone does not allow a species of origin to be determined. We observed synchronized bursts of activity (expanded data, Fig. 10d). The bursts lasted about 460 ms and were separated by silence periods of about 2 s (expanded data, Fig. 10d, e). Individual units fired an average of about three rounds per burst, which is approximately 73% of registered units per burst. The activities of individual units were highly correlated, and these correlations were higher than those of units identified in unvaccinated animals recorded under the same conditions (expanded data, Fig. 10f). To further characterize the spike responses of identified human-derived neurons, we performed light-tagging experiments on anesthetized rats transplanted with hCO expressing the light-sensitive cation channel rhodopsin 2 (hChR2), through which t-hCO neurons short-latency recognition (less than 10 ms) in response to blue light stimuli (Fig. 4m–o). t-hCO neurons exhibited bursts of spontaneous activity at frequencies similar to those observed in calcium imaging, as well as electrophysiological recordings performed in t-hCO in the absence of light marking (expanded data, Fig. 10c-g). No spontaneous activity was observed in the corresponding stages of hCO recorded in vitro. To assess whether t-hCO could be activated by sensory stimuli, we briefly deflected the rat whiskers away from t-hCO (Fig. 4j,m and extended data, Fig. 10h,k). According to previous studies8,10, a subset of t-hCO cells showed increased activity in response to whisker deflection, which was not observed when the data were compared with random time stamps (Fig. 4k–q and expanded data, Fig. 10h–q). Indeed, about 54% of the opto-labeled single units showed a significantly increased arousal rate after whisker stimulation, peaking at about 650 ms (Fig. 4r). Taken together, these data suggest that t-hCO receives appropriate functional inputs and can be activated by environmental stimuli.
We then investigated whether t-hCO could activate circuits in rats to control behavior. We first investigated whether the axons of t-hCO neurons project into the surrounding tissues of the rat. We infected hCO with a lentivirus encoding hChR2 fused to EYFP (hChR2-EYFP). After 110 days, we observed EYFP expression in ipsilateral cortical regions, including the auditory, motor, and somatosensory cortices, as well as in subcortical regions, including the striatum, hippocampus, and thalamus (Fig. 5a). To assess whether these efferent projections could elicit synaptic responses in rat cells, we optically activated t-hCO cells expressing hChR2-EYFP by recording rat cerebral cortex cells in sharp brain sections. Activation of t-hCO axons with blue light induced short-latency EPSCs in rat pyramidal cortex neurons, which were blocked by NBQX (Figs. 5b–g). In addition, these responses could be blocked by tetrodotoxin (TTX) and restored by 4-aminopyridine (4-AP), suggesting that they were caused by monosynaptic connections (Fig. 5e).
a, Schematic diagram of axon tracking (left). t-hCO EYFP expression (right). Scale bar, 100 µm. A1, auditory cortex, ACC, anterior cingulate cortex, d. striatum, dorsal striatum, HPC, hippocampus; Diaphragm, lateral septum, mPFC, medial prefrontal cortex, piri, piriform cortex, v. striatum, ventral striatum, VPM, ventropostomedial nucleus of the thalamus, VTA, ventral tegmental region. The red squares represent the areas of the brain where the images were taken. b, Schematic diagram of the stimulation experiment. c, d, Examples of the response of blue light-induced photocurrent (top) and voltage (bottom) in human (c) EYFP+ t-hCO or rat (d) EYFP- cells. e, f, Current traces of rat neurons after blue light stimulation of t-hCO axons with TTX and 4-AR (green), TTX (grey) or aCSF (black) (e), with (violet) or without (black) ) ) NBQX (e). g, latency of responses induced by blue light in rat cells (n = 16 cells); horizontal bars indicate average latency (7.13 ms) (left). Amplitude of light-evoked EPSCs recorded with or without NBQX (n = 7 cells; ***P < 0.0001) (middle). Amplitude of light-evoked EPSCs recorded with or without NBQX (n = 7 cells; ***P < 0.0001) (middle). Амплитуда вызванных светом EPSC, зарегистрированных с или без NBQX (n = 7 клеток; ***P <0,0001) (в центре). Amplitude of light-induced EPSCs recorded with or without NBQX (n = 7 cells; ***P < 0.0001) (center).使用或不使用NBQX 记录的光诱发EPSC 的振幅（n = 7 个细胞；***P < 0.0001）（中）。使用或不使用NBQX 记录的光诱发EPSC 的振幅（n = 7 个细胞；***P < 0.0001）（中）。 Амплитуда вызванных светом EPSC, зарегистрированных с или без NBQX (n = 7 клеток; ***P <0,0001) (в центре). Amplitude of light-induced EPSCs recorded with or without NBQX (n = 7 cells; ***P < 0.0001) (center). Percentage of rat cells showing EPSCs that respond to blue light (right). h, Schematic diagram of a behavioral task. d0, day 0. i. Performance of exemplary animals on day 1 (left) or day 15 (right) of training. The mean number of licks performed on day 1 (left) or day 15 (right centre) (n = 150 blue light trials, n = 150 red light trials; ***P < 0.0001). The mean number of licks performed on day 1 (left) or day 15 (right centre) (n = 150 blue light trials, n = 150 red light trials; ***P < 0.0001). Среднее количество облизываний, выполненных в день 1 (слева) или день 15 (в центре справа) (n = 150 испытаний с синим светом, n = 150 испытаний с красным светом; ***P <0,0001). Mean number of licks performed on day 1 (left) or day 15 (center right) (n = 150 blue light trials, n = 150 red light trials; ***P < 0.0001).第1 天（左）或第15 天（右中）执行的平均舔次数（n = 150 次蓝光试验，n = 150 次红光试验；***P < 0.0001）。第1 天（左）或第15 天（右中）执行的平均舔次数（n = 150 次蓝光试验，n = 150 次红光试验；***P < 0.001 Среднее количество облизываний, выполненных в день 1 (слева) или день 15 (в центре справа) (n = 150 испытаний с синим светом, n = 150 испытаний с красным светом; ***P <0,0001). Mean number of licks performed on day 1 (left) or day 15 (center right) (n = 150 blue light trials, n = 150 red light trials; ***P < 0.0001). Cumulative licks for red and blue light trials on day 1 (center left) or day 15 (right). NS, not significant. j,k, Behavioral characteristics of all animals transplanted with t-hCO expressing hChR2-EYFP (j) or control fluorophore (k) on day 1 or 15 (hChR2-EYFP: n = 9 rats, ** P = 0.0049; control: n = 9, P = 0.1497). l, Evolution of preference score (n = 9 hChR2, n = 9 control; **P < 0.001, ***P < 0.0001). l, Evolution of preference score (n = 9 hChR2, n = 9 control; **P < 0.001, ***P < 0.0001). l, Эволюция показателя предпочтения (n = 9 hChR2, n = 9 контрольных; **P <0,001, ***P <0,0001). l, Evolution of preference score (n = 9 hChR2, n = 9 controls; **P < 0.001, ***P < 0.0001). l，偏好评分的演变（n = 9 hChR2，n = 9 对照；**P < 0.001，***P < 0.0001）。 l，偏好评分的演变（n = 9 hChR2，n = 9 对照；**P < 0.001，***P < 0.0001）。 l, Эволюция показателей предпочтения (n = 9 hChR2, n = 9 контролей; **P <0,001, ***P <0,0001). l, Evolution of preference scores (n = 9 hChR2, n = 9 controls; **P < 0.001, ***P < 0.0001). m, FOS expression in response to optogenetic activation of t-hCO in S1. Images of FOS expression (left), and quantification (n = 3 per group; *P < 0.05, **P < 0.01 and ***P < 0.001) (right) are shown. Images of FOS expression (left), and quantification (n = 3 per group; *P < 0.05, **P < 0.01 and ***P < 0.001) (right) are shown. Показаны изображения экспрессии FOS (слева) и количественного определения (n = 3 на группу; * P <0,05, ** P <0,01 и *** P <0,001) (справа). Images of FOS expression (left) and quantification (n = 3 per group; *P<0.05, **P<0.01, and ***P<0.001) are shown (right).显示了FOS 表达（左）和量化（每组n = 3；*P < 0.05、**P < 0.01 和***P < 0.001）（右）的图像。显示了FOS 表达（左）和量化（每组n = 3；*P < 0.05、**P < 0.01 和***P < 0.001）（右）的图像。 Показаны изображения экспрессии FOS (слева) и количественного определения (n = 3 на группу; * P <0,05, ** P <0,01 и *** P <0,001) (справа). Images of FOS expression (left) and quantification (n = 3 per group; *P<0.05, **P<0.01, and ***P<0.001) are shown (right). Scale bar, 100 µm. Data are expressed as mean ± standard error of BLA, basolateral tonsil, MDT, dorsomedial thalamic nucleus, PAG, periaqueductal grey.
Finally, we asked if t-hCO could modulate rat behavior. To test this, we transplanted hChR2-EYFP-expressing hCO into S1, and 90 days later, we implanted optical fibers into t-hCO for light delivery. We then trained the rats with a modified operant conditioning paradigm (Fig. 5h). We placed the animals in a behavioral test chamber and randomly applied 5 second blue (473 nm) and red (635 nm) laser stimuli. Animals received a water reward if they licked during blue light stimulation but did not lick during red light stimulation. On the first day of training, the animals showed no difference in licking when stimulated with blue or red light. However, on day 15, animals transplanted with hCO expressing hChR2-EYFP showed more active licking when stimulated with blue light compared to red light stimulation. These changes in licking behavior were not observed in control animals transplanted with hCO expressing the control fluorophore (learning success rate: hChR2 89%, EYFP 0%, Figure 5i-1 and Supplementary Video 2). These data suggest that t-hCO cells can activate rat neurons to stimulate reward-seeking behavior. To find out which rat t-hCO neural circuits might be involved in these behavioral changes, we optogenetically activated t-hCO in trained animals and harvested tissues 90 minutes later. Immunohistochemistry revealed expression of the activity-dependent FOS protein in several brain regions involved in motivated behavior, including the medial prefrontal cortex, medial thalamus, and periaqueductal gray matter, which was expressed either in unstimulated control animals or in animals. rice. 5m). Taken together, these data suggest that t-hCO can modulate rat neuronal activity to drive behavior.
Neural organoids represent a promising system for studying human development and disease in vitro, but they are limited by the lack of connections between circuits that exist in vivo. We have developed a novel platform in which we transplanted hCO into S1 of immunocompromised early postnatal rats to study human cell development and function in vivo. We have shown that t-hCO develops mature cell types not observed in vitro28 and that t-hCO is anatomically and functionally integrated into the rodent brain. The integration of t-hCO into rodent neural circuits allowed us to establish a link between human cellular activity and studied animal behavior, showing that t-hCO neurons can modulate rat neuronal activity to drive behavioral responses.
The platform we describe has several advantages over previous research on transplanting human cells into rodent brains. First, we transplanted hCO into the developing cortex of early postnatal rats, which may facilitate anatomical and functional integration. Second, t-hCO MRI monitoring allowed us to study graft position and growth in live animals, allowing us to conduct long-term multi-animal studies and establish the reliability of several hiPS cell lines. Finally, we transplanted intact organoids, rather than isolated single cell suspensions, which are less destructive to human cells and can promote the integration and generation of human cortex neurons in rat brains.
We acknowledge that despite advances in this platform, temporal, spatial, and cross-species constraints prevent the formation of human neural circuits with high fidelity, even after transplantation at an early stage of development. For example, it is not clear whether the spontaneous activity observed in t-hCO represents a developmental phenotype similar to the rhythmic activity observed during cortical development, or whether it is due to the absence of suppressive cell types present in t-hCO. Similarly, it is not clear to what extent the absence of lamination in t-hCO affects chain connectivity30. Future work will focus on integrating other cell types such as human microglia, human endothelial cells, and varying proportions of GABAergic interneurons as shown using assembly 6 in vitro, as well as understanding how neural integration and processing can occur in altered t-hCO. transcriptional, synaptic and behavioral levels in cells obtained from patients.
Overall, this in vivo platform represents a powerful resource that can complement in vitro human brain development and disease research. We anticipate that this platform will allow us to discover novel strand-level phenotypes in otherwise elusive patient-derived cells and test new therapeutic strategies.
We generated hCO2.5 from HiPS cells as described previously. To initiate hCO production from hiPS cells cultured on feeder layers, intact colonies of hiPS cells were removed from culture dishes using dispase (0.35 mg/mL) and transferred to ultra-low attachment plastic cultures containing dishes with hiPS cell culture medium. (Corning) supplemented with two SMAD inhibitors dorsomorphine (5 μM; P5499, Sigma-Aldrich) and SB-431542 (10 μM; 1614, Tocris) and ROCK inhibitor Y-27632 (10 μM; S1049, Selleckchem). During the first 5 days, the hiPS cell medium was changed daily and dorsomorphine and SB-431542 were added. On the sixth day in suspension, neural spheroids were transferred to neural medium containing neurobasal-A (10888, Life Technologies), B-27 supplement without vitamin A (12587, Life Technologies), GlutaMax (1:100, Life Technologies), penicillin and streptomycin (1:100, Life Technologies) and supplemented with the epidermal growth factor (EGF; 20 ng ml−1; R&D Systems) and fibroblast growth factor 2 (FGF2; 20 ng ml−1; R&D Systems) until day 24. From day 25 to day 42, the medium was supplemented with brain-derived neurotrophic factor (BDNF; 20 ng ml−1, Peprotech) and neurotrophin 3 (NT3; 20 ng ml−1; Peprotech) with medium changes every other day. On the sixth day in suspension, neural spheroids were transferred to neural medium containing neurobasal-A (10888, Life Technologies), B-27 supplement without vitamin A (12587, Life Technologies), GlutaMax (1:100, Life Technologies), penicillin and streptomycin (1:100, Life Technologies) and supplemented with the epidermal growth factor (EGF; 20 ng ml−1; R&D Systems) and fibroblast growth factor 2 (FGF2; 20 ng ml−1; R&D Systems) until day 24. From day 25 to day 42, the medium was supplemented with brain-derived neurotrophic factor (BDNF; 20 ng ml−1, Peprotech) and neurotrophin 3 (NT3; 20 ng ml−1; Peprotech) with medium changes every other day. On the sixth day in suspension, the neural spheroids were transferred to a neural medium containing Neurobasal-A (10888, Life Technologies), B-27 supplement without vitamin A (12587, Life Technologies), GlutaMax (1:100, Life Technologies), penicillin. и стрептомицин (1:100, Life Technologies) и дополнены эпидермальным фактором роста (EGF; 20 нг/мл; R&D Systems) и фактором роста фибробластов 2 (FGF2; 20 нг/мл; R&D Systems) до 24-го дня. and streptomycin (1:100, Life Technologies) and supplemented with epidermal growth factor (EGF; 20 ng/ml; R&D Systems) and fibroblast growth factor 2 (FGF2; 20 ng/ml; R&D Systems) until day 24. From days 25 to 42, brain-derived neurotrophic factor (BDNF; 20 ng ml-1, Peprotech) and neurotrophin 3 (NT3; 20 ng ml-1, Peprotech) were added to the medium, changing the medium every other day.在悬浮的第6 天，将神经球体转移到含有neurobasal-A（10888，Life Technologies）、不含维生素A 的B-27 补充剂（12587，Life Technologies）、GlutaMax（1:100，Life Technologies）、青霉素的神经培养基中和链霉素（1:100，Life Technologies）并辅以表皮生长因子（EGF；20 ng ml-1；R&D Systems）和成纤维细胞生长因子2（FGF2；20 ng ml-1；R&D Systems）直至第24 天。在 悬浮 的 第 第 6 天 将 神经 球体 转移 含有 含有 neurobasal-a （10888 ， Life Technologies） 不 含 维生素 a 的 b-27 补充剂 （12587 ， Life Technologies） Glutamax （1: 100 ， Life TechNOGIS青霉素 的 神经 培养 基 中 链霉素 （（1: 100 ， Life Technologies） 表皮 生长 因子 （（（20 ng ml-1 ； r & d Systems） 成 纤维 细胞 生长 2 （fgf2 ； 20 ng ml- 1；R&D Systems）直至第24天。 На 6-й день суспензии нейросферы были переведены на добавку, содержащую нейробазал-А (10888, Life Technologies), добавку В-27 без витамина А (12587, Life Technologies), GlutaMax (1:100, Life Technologies), пенициллин- нейтрализованный стрептомицин (1:100, Life Technologies) с добавлением эпидермального фактора роста (EGF; 20 нг мл-1; R&D Systems) и фактора роста фибробластов 2 (FGF2; 20 нг мл-1) 1; On day 6, neurosphere suspensions were switched to a supplement containing neurobasal-A (10888, Life Technologies), B-27 supplement without vitamin A (12587, Life Technologies), GlutaMax (1:100, Life Technologies), penicillin-neutralized streptomycin (1:100, Life Technologies) supplemented with epidermal growth factor (EGF; 20 ng ml-1; R&D Systems) and fibroblast growth factor 2 (FGF2; 20 ng ml-1) 1; R&D Systems) до 24-го дня. R&D Systems) until day 24. From days 25 to 42, brain-derived neurotrophic factor (BDNF; 20 ng ml-1, Peprotech) and neurotrophic factor 3 (NT3; 20 ng ml-1, Peprotech) were added to the culture medium every other day. Medium change once. Starting from day 43, hCO was maintained in unsupplemented neurobasal-A medium (NM; 1088022, Thermo Fisher) with medium change every 4–6 days. To obtain hCO from hiPS cells cultured under feederless conditions, hiPS cells were incubated with Accutase (AT-104, Innovate Cell Technologies) at 37°C for 7 minutes, dissociated into single cells, and plated on AggreWell 800 plates (34815, STEMCELL Technologies) at a density of 3 × 106 single cells per well in Essential 8 medium supplemented with the ROCK inhibitor Y-27632 (10 μM; S1049, Selleckchem). After 24 hours, the media in the wells were pipetted up and down into media containing Essential 6 media (A1516401, Life Technologies) supplemented with dorsomorphine (2.5 μM; P5499, Sigma-Aldrich) and SB-431542 (10 μM; 1614). , Tocrida). From days 2 to 6, Essential 6 medium was replaced daily with dorsomorphine and supplement SB-431542. From the sixth day, the neurosphere suspensions were transferred to the neurobasal medium and maintained as described above.
All animal procedures were performed in accordance with the animal care guidelines approved by the Stanford University Laboratory Animal Care Administrative Committee (APLAC). Pregnant euthymic RNU (rnu/+) rats were purchased (Charles River Laboratories) or housed. Animals were kept on a 12-hour light-dark cycle with food and water ad libitum. Nude (FOXN1–/–) rat pups aged three to seven days were identified by the growth of immature whiskers prior to culling. Puppies (male and female) were anesthetized with 2-3% isoflurane and placed on a stereotaxic frame. A trepanation of the skull with a diameter of approximately 2-3 mm above S1 was performed while maintaining the integrity of the dura mater. Then use a 30-G needle (approximately 0.3 mm) just outside of the craniotomy to pierce the dura. Then apply HCO to a thin 3×3 cm parafilm and remove excess medium. Using a Hamilton syringe attached to a 23 G, 45° needle, gently draw hCO into the most distal end of the needle. Then install the syringe on the syringe pump connected to the stereotaxic device. Then place the tip of the needle over a previously made 0.3 mm wide puncture hole in the dura (z = 0 mm) and narrow the syringe 1–2 mm (z = approximately –1.5 mm) until the needle is between the dura mater A. a dense seal is formed. Then raise the syringe to the center of the cortical surface at z = -0.5 mm and inject hCO at a rate of 1-2 µl per minute. After completion of the hCO injection, the needle is retracted at a rate of 0.2-0.5 mm per minute, the skin is sutured, and the puppy is immediately placed on a warm heating pad until complete recovery. Some animals were transplanted bilaterally.
All animal procedures were performed in accordance with Stanford University APLAC-approved animal care guidelines. Rats (greater than 60 days after transplantation) were induced with 5% isoflurane anesthesia and anesthetized with 1-3% isoflurane during imaging. For visualization, a 7 Tesla actively shielded horizontal borehole scanner Bruker (Bruker Corp.) with an International Electric Company (IECO) gradient drive, a shielded gradient insert with an internal diameter of 120 mm (600 mT/m, 1000 T/m/s) was used using AVANCE. III, eight-channel multi-coil RF and multi-core capabilities, and the accompanying Paravision 6.0.1 platform. Recording was performed using an actively decoupled volumetric RF coil with an internal diameter of 86 mm and a four-channel cryo-cooled RF coil for receive only. Axial 2D Turbo-RARE (repetition time = 2500 ms, echo time = 33 ms, 2 averages) with 16 slice captures, slice thickness 0.6–0.8 mm, containing 256 × 256 samples. The signals were received using a quadrature transceiver volumetric RF coil with an internal diameter of 2 cm (Rapid MR International, LLC). Finally, use the built-in Imaris (BitPlane) surface estimation functions for 3D rendering and volume analysis. A successful transplant was defined as one in which areas of continuous T2-weighted MRI signal were formed in the transplanted hemisphere. Graft rejection was defined as a graft that did not produce areas of continuous T2-weighted MRI signal in the transplanted hemisphere. Subcortical t-hCO was excluded from subsequent analysis.
To stably express GCaMP6s in hCO for two-photon calcium imaging, hiPS cells were infected with pLV[Exp]-EF1a::GcaMP6s-WPRE-Puro followed by selection of antibiotics. Briefly, cells were dissociated with EDTA and suspended in 1 ml of Essential 8 medium at a density of approximately 300,000 cells in the presence of polybrene (5 μg/ml) and 15 μl of virus. The cells were then incubated in suspension for 60 minutes and seeded at a density of 50,000 cells per well. After confluence, cells were treated with 5-10 μg ml-1 puromycin for 5-10 days or until stable colonies appeared. Acute hCO infection was performed as previously described5 with some modifications. Briefly, transfer day 30-45 hCO into 1.5 ml Eppendorf microcentrifuge tubes containing 100 µl of nerve medium. Then approximately 90 µl of the medium is removed, 3-6 µl of high titer lentivirus (from 0.5 x 108 to 1.2 x 109) is added to the tube, and the hCO is transferred to the incubator for 30 minutes. Then add 90–100 µl of medium to each tube and return the tubes to the incubator overnight. The next day, transfer hCO to fresh nerve medium in low attachment plates. After 7 days, hCO was transferred to 24-well glass bottom plates for visualization and evaluation of infection quality. pLV[Exp]-SYN1::EYFP-WPRE and pLV[Exp]-SYN1::hChR2-EYFP-WPRE were generated by VectorBuilder. Lentivirus is used in most experiments because it is integrated into the host genome, allowing reporter gene expression in infected cell lines. For rabies follow-up, day 30-45 hCO was co-infected with rabies-ΔG-eGFP and AAV-DJ-EF1a-CVS-G-WPRE-pGHpA (plasmid #67528, Addgene), washed thoroughly for 3 days, and transplanted into rats in S1 and maintained in vivo for 7-14 days.
For immunocytochemistry, animals were anesthetized and transcardially perfused with PBS followed by 4% paraformaldehyde (PFA in PBS; Electron Microscopy Sciences). Brains were fixed in 4% PFA for 2 hours or overnight at 4°C, cryopreserved in 30% sucrose in PBS for 48-72 hours, and embedded in 1:1, 30% sucrose: OCT (Tissue-Tek OCT Compound 4583 , Sakura Finetek) and coronal sections were made at 30 µm using a cryostat (Leica). For immunohistochemistry of thick sections, animals were perfused with PBS, and the brain was dissected and sectioned coronally at 300–400 µm using a vibratome (Leica) and the sections were fixed with 4% PFA for 30 minutes. Then cryosections or thick sections were washed with PBS, blocked for 1 hour at room temperature (10% normal donkey serum (NDS) and 0.3% Triton X-100 diluted in PBS) and blocked with blocking solution at 4°C. – Incubation Cryosections were incubated overnight and thick sections were incubated for 5 days. Primary antibodies used were: anti-NeuN (mouse, 1:500; ab104224, abcam) anti-CTIP2 (rat, 1:300; ab18465, abcam), anti-GFAP (rabbit, 1:1,000; Z0334, Dako), anti-GFP (chicken, 1:1,000; GTX13970, GeneTex), anti-HNA (mouse, 1:200; ab191181, abcam), anti-NeuN (rabbit, 1:500; ABN78, Millipore), anti-PDGFRA (rabbit, 1:200; sc-338, Santa Cruz), anti-PPP1R17 (rabbit, 1:200; HPA047819, Atlas Antibodies), anti-RECA-1 (mouse, 1:50; ab9774, abcam), anti-SCG2 (rabbit, 1:100; 20357-1-AP, Proteintech), anti-SOX9 (goat, 1:500; AF3075, R&D Systems), Netrin G1a (goat, 1:100; AF1166, R&D Systems), anti-STEM121 (mouse, 1:200; Y40410, Takara Bio), anti-SATB2 (mouse, 1:50; ab51502, abcam), anti-GAD65/67 (rabbit, 1:400; ABN904, Millipore) and anti-IBA1 (goat, 1:100; ab5076, abcam). Primary antibodies used were: anti-NeuN (mouse, 1:500; ab104224, abcam) anti-CTIP2 (rat, 1:300; ab18465, abcam), anti-GFAP (rabbit, 1:1,000; Z0334, Dako), anti -GFP (chicken, 1:1,000; GTX13970, GeneTex), anti-HNA (mouse, 1:200; ab191181, abcam), anti-NeuN (rabbit, 1:500; ABN78, Millipore), anti-PDGFRA (rabbit, 1:200; sc-338, Santa Cruz), anti-PPP1R17 (rabbit, 1:200; HPA047819, Atlas Antibodies), anti-RECA-1 (mouse, 1:50; ab9774, abcam), anti-SCG2 (rabbit , 1:100; 20357-1-AP, Proteintech), anti-SOX9 (goat, 1:500; AF3075, R&D Systems), Netrin G1a (goat, 1:100; AF1166, R&D Systems), anti-STEM121 (mouse , 1:200; Y40410, Takara Bio), anti-SATB2 (mouse, 1:50; ab51502, abcam), anti-GAD65/67 (rabbit, 1:400; ABN904, Millipore) and anti-IBA1 (goat, 1 :100; ab5076, abcam). Использовались следующие первичные антитела: анти-NeuN (мышиные, 1:500; ab104224, abcam), анти-CTIP2 (крысиные, 1:300; ab18465, abcam), анти-GFAP (кроличьи, 1:1000; Z0334, Dako), анти- -GFP (курица, 1:1000; GTX13970, GeneTex), анти-HNA (мышь, 1:200; ab191181, abcam), анти-NeuN (кролик, 1:500; ABN78, Millipore), анти-PDGFRA (кролик, 1:200; sc-338, Санта-Круз), анти-PPP1R17 (кролик, 1:200; HPA047819, Atlas Antibodies), анти-RECA-1 (мышь, 1:50; ab9774, abcam), анти-SCG2 (кролик , 1:100; 20357-1-AP, Proteintech), анти-SOX9 (козий, 1:500; AF3075, R&D Systems), нетрин G1a (козий, 1:100; AF1166, R&D Systems), анти-STEM121 (мышиный , 1:200; Y40410, Takara Bio), анти-SATB2 (мышь, 1:50; ab51502, abcam), анти-GAD65/67 (кролик, 1:400; ABN904, Millipore) и анти-IBA1 (коза, 1 :100; аб5076, абкам). The primary antibodies used were: anti-NeuN (mouse, 1:500; ab104224, abcam), anti-CTIP2 (rat, 1:300; ab18465, abcam), anti-GFAP (rabbit, 1:1000; Z0334, Dako), anti-GFP (chicken, 1:1000; GTX13970, GeneTex), anti-HNA (mouse, 1:200; ab191181, abcam), anti-NeuN (rabbit, 1:500; ABN78, Millipore), anti-PDGFRA ( rabbit, 1:200; sc-338, Santa Cruz), anti-PPP1R17 (rabbit, 1:200; HPA047819, Atlas Antibodies), anti-RECA-1 (mouse, 1:50; ab9774, abcam), anti- SCG2 (rabbit, 1:100; 20357-1-AP, Proteintech), anti-SOX9 (goat, 1:500; AF3075, R&D Systems), netrin G1a (goat, 1:100; AF1166, R&D Systems), anti- STEM121 (mouse, 1:200; Y40410, Takara Bio), anti-SATB2 (mouse, 1:50; ab51502, abcam), anti-GAD65/67 (rabbit, 1:400; ABN904, Millipore) and anti-IBA1 ( goat, 1:100; ab5076, abkam).使用的一抗是：抗NeuN（小鼠，1:500；ab104224，abcam）抗CTIP2（大鼠，1:300；ab18465，abcam），抗GFAP（兔，1:1,000；Z0334，Dako），抗-GFP（鸡，1:1,000；GTX13970，GeneTex），抗HNA（小鼠，1:200；ab191181，abcam），抗NeuN（兔，1:500；ABN78，Millipore），抗PDGFRA（兔， 1:200；sc-338，Santa Cruz），抗PPP1R17（兔，1:200；HPA047819，Atlas 抗体），抗RECA-1（小鼠，1:50；ab9774，abcam），抗SCG2（兔） , 1:100；20357-1-AP，Proteintech），抗SOX9（山羊，1:500；AF3075，R&D Systems），Netrin G1a（山羊，1:100；AF1166，R&D Systems），抗STEM121（小鼠, 1:200;使用的一抗是：抗NeuN（小鼠，1:500；ab104224，abcam）抗CTIP2（大鼠，1:300；ab18465，abcam），抗GFAP（兔，1:1,000；Z0334，Dako），抗-GFP（鸡，1:1,000；GTX13970，GeneTex），抗HNA（小鼠，1:200；ab191181，abcam），抗NeuN（兔，1:500；ABN78，Millipore％弔，抗1: 200；sc-338，Santa Cruz），抗PPP1R17（兔，1:200；HPA047819，Atlas 抗体），抗RECA-1（小鼠，1:50；ab9774，abcam），抗SCG2 100；20357-1-AP，Proteintech），抗SOX9（山羊，1:500；AF3075，R&D Systems），Netrin G1a（山羊，1:100；AF1166，R&D Systems）頏1:20, ; Y40410, Takara Bio), anti-SATB2 (mouse, 1:50; ab51502, abcam), anti-GAD65/67 (rabbit, 1:400; ABN904, Millipore) and anti-IBA1 (goat, 1:100; ab5076, abcam)。 The primary antibodies used were: anti-NeuN (mouse, 1:500; ab104224, abcam), anti-CTIP2 (rat, 1:300; ab18465, abcam), anti-GFAP (rabbit, 1:1000; Z0334, Dako) . , anti-GFP (chicken, 1:1000; GTX13970, GeneTex), anti-HNA (mouse, 1:200; ab191181, abcam), anti-NeuN (rabbit, 1:500; ABN78, Millipore), anti-PDGFRA ( rabbit, 1:200; sc-338, Santa Cruz), anti-PPP1R17 (rabbit, 1:200; HPA047819, Atlas antibody), anti-RECA-1 (mouse, 1:50; ab9774, abcam), anti- SCG2 (rabbit), 1:100; 20357-1-AP, Proteintech), анти-SOX9 (коза, 1:500; AF3075, R&D Systems), Нетрин G1a (коза, 1:100; AF1166, R&D Systems), анти -STEM121 (мышь, 1:200; Y40410, Takara Bio), анти-SATB2 (мышь, 1:50; ab51502, abcam), анти-GAD65/67 (кролик, 1:400; ABN904, Millipore) и анти-IBA1 (коза, 1:100; аб5076, абкам). 20357-1-AP, Proteintech), anti-SOX9 (goat, 1:500; AF3075, R&D Systems), Netrin G1a (goat, 1:100; AF1166, R&D Systems), anti-STEM121 (mouse, 1:200; Y40410, Takara Bio), anti-SATB2 (mouse, 1:50; ab51502, abcam), anti-GAD65/67 (rabbit, 1:400; ABN904, Millipore), and anti-IBA1 (goat, 1:100; ab5076, abkam). Sections were then washed with PBS and incubated with secondary antibody for 1 hour at room temperature (frozen sections) or overnight at 4°C (thick sections). Alexa Fluor secondary antibody (Life Technologies) diluted 1:1000 in blocking solution was used. After washing with PBS, the nuclei were visualized with a Hoechst 33258 (Life Technologies). Finally, the slides were placed on a microscope with coverslips (Fisher Scientific) using an Aquamount (Polysciences) and analyzed on a Keyence fluorescent microscope (BZ-X analyzer) or a Leica TCS SP8 confocal microscope (Las-X) on the image. The images were processed using the ImageJ program (Fiji). To quantify the proportion of human neurons in t-hCO and the rat cortex, 387.5 μm wide rectangular images were taken at the center of the t-hCO, at or near the edge of the rat cortex. Graft margins were determined by assessing changes in tissue transparency, HNA+ nuclei, and/or the presence of tissue autofluorescence. In each image, the total number of NeuN+ and HNA+ cells were divided by the total number of NeuN+ cells in the same area. To ensure that only cells with nuclei in the image plane are counted, only cells that are also Hoechst+ are included in the calculation. Two images separated by at least 1 mm were averaged to reduce the statistical error.
One week prior to sample collection, place hCO transplant animals (approximately 8 months of differentiation) in a dark room with whiskers trimmed to minimize sensory stimulation. Isolation of nuclei was performed as described previously, with some modifications. Briefly, t-hCO and hCO were destroyed using detergent-mechanical cell lysis and a 2 ml glass tissue grinder (D8938, Sigma-Aldrich/KIMBLE). The crude nuclei were then filtered off using a 40 µm filter and centrifuged at 320 g for 10 minutes at 4 °C before performing a sucrose density gradient. After the centrifugation step (320 g for 20 min at 4°C), the samples were resuspended in 0.04% BSA/PBS with the addition of 0.2 units of µl-1 RNase inhibitor (40 u µl-1, AM2682, Ambion) and passed through a 40 µm flow filter. The dissociated nuclei were then resuspended in PBS containing 0.02% BSA and loaded onto a Chromium Single Cell 3′ chip (estimated recovery of 8,000 cells per lane). snRNA-seq libraries were prepared with the Chromium Single cell 3′ GEM, Library & Gel Bead Kit v3 (10x Genomics). snRNA-seq libraries were prepared with the Chromium Single cell 3′ GEM, Library & Gel Bead Kit v3 (10x Genomics). Библиотеки snRNA-seq были приготовлены с помощью Chromium Single cell 3′ GEM, Library & Gel Bead Kit v3 (10x Genomics). The snRNA-seq libraries were prepared using Chromium Single cell 3′ GEM, Library & Gel Bead Kit v3 (10x Genomics). snRNA-seq 文库是使用Chromium Single cell 3′ GEM、Library & Gel Bead Kit v3 (10x Genomics) 制备的。 snRNA-seq 文库是使用Chromium Single cell 3′ GEM、Library & Gel Bead Kit v3 (10x Genomics) 制备的。 Библиотеку snRNA-seq готовили с использованием Chromium Single Cell 3′ GEM, Library & Gel Bead Kit v3 (10x Genomics). The snRNA-seq library was prepared using Chromium Single Cell 3′ GEM, Library & Gel Bead Kit v3 (10x Genomics). Libraries from different samples were pooled and sequenced by Admera Health on NovaSeq S4 (Illumina).
Gene expression levels for each putative nuclear barcode were quantified using the 10x Genomics CellRanger analysis software package (version 6.1.2). Specifically, the reads were matched against a combination of human (GRCh38, Ensemble, version 98) and rat (Rnor_6.0, Ensemble, version 100) reference genomes created with the mkref command and using count with the –include-introns=TRUE command to quantitation include readings mapped to intron regions. For t-hCO samples, human nuclei were identified based on the conservative requirement that at least 95% of all mapped reads match the human genome. All subsequent analyzes were performed on a filtered barcode array output from the CellRanger using the R package (version 4.1.2) Seurat (version 4.1.1)32.
To ensure that only high quality nuclei are included in the subsequent analysis, an iterative filtering process was implemented for each sample. First, low-quality nuclei with less than 1000 unique genes found and more than 20% of the total mitochondria are identified and removed. Subsequently, the raw gene number matrix was normalized by regularized negative binomial regression using the sctransform(vst.flavor=”v2″) function, which also identified the 3000 most variable genes using default parameters. Dimension reduction was performed on the upper variable genes using Principal Component Analysis (PCA) with default parameters using a data set dimension of 30 (dims = 30 was chosen based on visual inspection of knee sites and used for all samples and ensemble analyses). We then performed several rounds of iterative clustering (resolution = 1) to classify genes based on abnormally low gene count (median below 10th percentile), abnormally high mitochondrial gene count (median above 95th percentile) to identify and remove putative cells of low quality. clusters and/or a high proportion of suspected twins identified using the DoubletFinder33 package (mean DoubletFinder score above the 95th percentile). t-hCO samples (n=3) and hCO samples (n=3) were integrated separately using the IntegrateData function with the above parameters. Then another round of qualitative filtering of the integrated data set was performed as described above.
After removing the low quality kernels, the integrated dataset was grouped (resolution = 0.5) and embedded for UMAP34 visualization purposes. Marker genes for each cluster were determined using the FindMarkers function with default parameters calculated from normalized gene expression data. We identify and classify major cell classes by combining fetal and adult cortical reference datasets with marker gene expression 19,20,21,35 and annotation. In particular, circulating precursors were identified by the expression of MKI67 and TOP2A. Progenitor clusters were defined by the absence of mitotic transcripts, high overlap with multipotent glial progenitor clusters described in late metaphase fetal cortex, and EGFR and OLIG1 expression. We use the term astrocyte to encompass several states of astrocyte differentiation, from late radial glia to maturation of astrocytes. Astrocyte clusters express high levels of SLC1A3 and AQP4 and have been shown to map with subtypes of fetal radial glia and/or adult astrocytes. OPCs express PDGFRA and SOX10 while oligodendrocytes express myelination markers (MOG and MYRF). Glutamatergic neurons were identified by the presence of neuronal transcripts (SYT1 and SNAP25), the absence of GABAergic markers (GAD2), and the expression of NEUROD6, SLC17A7, BCL11B, or SATB2. GluN neurons were further divided into upper (SATB2 expression and loss of BCL11B) and deep (BCL11B expression) subclasses. Putative subplate (SP) neurons express known SP18 markers such as ST18 and SORCS1 in addition to deep GluN markers. Choroid plexus-like cells were identified by TTR expression, and meningeal-like cells expressed fibroblast-associated genes and mapped pial/vascular cells of the reference data set.
Differential analysis of gene expression between t-hCO and hCO subclasses was performed using a newly developed pseudo-batch method reproduced in samples implemented using the Libra R package (version 1.0.0). Specifically, edgeR log-likelihood tests (version 3.36.0, package R) were performed for groups by summing the number of genes in cells for a given cell class for each sample replication. For heatmap visualization, normalized per million (CPM) values ​​are computed using edgeR (cpm() function) and scaled (to achieve mean = 0, standard deviation = 1). Gene Ontology (GO) enrichment analysis of significantly upregulated t-hCO GluN genes was performed (Benjamini-Hochberg corrected P value of less than 0.05 expressed in at least 10% of t-hCO GluN cells and a fold increase in change of at least 2 times). performed using ToppGene Suite (https://toppgene.cchmc.org/)37. We use the ToppFun app with default parameters and report Benjamini-Hochberg-corrected P-values ​​calculated from GO-annotated hypergeometric tests.
To match our snRNA-seq clusters with annotated cell clusters from reference studies of primary single-cell RNA-seq or adult snRNA-seq19,20,21,22, we applied a paired dataset integration approach. We used the SCTransform (v2) normalization workflow in Seurat to integrate and compare cluster overlaps between datasets (using the same parameters as above). Individual datasets were randomly subsetted up to 500 cells or cores per original cluster for computational efficiency. Using a similar approach as described previously, cluster overlap was defined as the proportion of cells or nuclei in each pooled cluster that overlapped with the label of the reference cluster. To further classify GluNs, we used Seurat’s TransferData workflow for GluN subset data to assign reference dataset labels to our GluN cells.
To assess the maturation status of the global transcriptome of t-hCO and hCO samples, we compared our pseudo-bulk samples with BrainSpan/psychENCODE23, which consists of a large RNA sequence spanning human brain development. We performed PCA on a combined pattern-normalized gene expression matrix from cortical samples 10 weeks after conception and later, in 5567 genes (together with our data) that were previously identified as active in BrainSpan cortical samples (defined as greater than 50% in developmental variance explained by age using a cubic model)38. In addition, we derived genes associated with major transcriptome signatures of neurodevelopment using non-negative matrix factorization as previously described. The sample weights calculated using the non-negative matrix factorization procedure are plotted in Figs. 5b with expanded data for each of the five signatures described by Zhu et al.38. Again, activity dependent transcriptional markers were derived from previously published studies. In particular, ERG and LRG were significantly upregulated in glutamatergic neurons identified by the mouse visual cortex snRNA-seq collection after visual stimulation from Supplementary Table 3 Hrvatin et al.16. Human-enriched LRGs were obtained from KCl-activated human fetal brain cultures and harvested 6 hours post-stimulation, and the filtered genes were significantly upregulated in humans but not in rodents (Supplementary Table 4). Analysis of gene set enrichment using these gene sets was performed using a one-way Fisher’s exact test.
Anesthetize rats with isoflurane, remove brains and place in cold (approximately 4°C) oxygenated (95% O2 and 5% CO2) sucrose solution for sections containing: 234 mM sucrose, 11 mM glucose, 26 mM NaHCO3, 2.5 mM KCl, 1.25 mM. NaH2PO4, 10 mM MgSO4 and 0.5 mM CaCl2 (about 310 mOsm). Rat brain coronal sections (300–400 µm) containing t-hCO were made using a Leica VT1200 vibratome as described previously39. Sections were then transferred to a sectioning chamber with continuous room temperature oxygenation containing aCSF prepared from: 10 mM glucose, 26 mM NaHCO3, 2.5 mM KCl, 1.25 mM NaHPO4, 1 mM MgSO4, 2 mM CaCl2 and 126 mM NaCl (298 mOsm). at least 45 minutes prior to recording. Sections were recorded in an immersed chamber where they were continuously perfused with aCSF (95% O2 and 5% CO2 vial). All data were recorded at room temperature. t-hCO neurons were terminated with a borosilicate glass pipette filled with a solution containing 127 mM potassium gluconate, 8 mM NaCl, 4 mM magnesium ATP, 0.3 mM sodium GTP, 10 mM HEPES, and 0.6 mM EGTA, pH 7.2, internal solution adjusted with KOH (290 mOsm). In order to recover, biocytin (0.2%) was added to the recording solution.
Data was acquired using a MultiClamp 700B amplifier (Molecular Devices) and a Digidata 1550B digitizer (Molecular Devices), low-pass filtered at 2 kHz, digitized at 20 kHz, and analyzed using Clampfit (Molecular Devices), Origin (OriginPro). 2021b, OriginLab). and custom MATLAB functions (Mathworks). The junction potential was calculated using JPCalc and the entries were adjusted to the calculated value of -14 mV. Operation IV consists of a series of current steps in 10-25 pA steps, from -250 to 750 pA.
The thalamus, white matter, and S1 afferents were electrically stimulated in thalamocortical slices during patch-clamp recording of hCO neurons, as described previously. Briefly, the brain was placed on a 3D printing table tilted at a 10° angle, and the front of the brain was cut at a 35° angle. The brain was then glued to the cut surface and sectioned, preserving the thalamocortical protruding axons. Bipolar tungsten electrodes (0.5 MΩ) were mounted on a second micromanipulator and strategically positioned to stimulate four regions per cell (inner capsule, white matter, S1 and hCO). Record synaptic responses after 300 µA phasic stimulation at 0.03–0.1 Hz.
hChR2-expressing hCO neurons were activated at 480 nm and light pulses generated by an LED (Prizmatix) were applied through a ×40 objective (0.9 NA; Olympus) to record hChR2 expression near the cells. The illuminated field diameter is approximately 0.5 mm and the total power is 10-20 mW. The pulse width was set to 10 ms, which corresponds to the pulse given during the behavioral learning experiment. Various stimulation frequencies were used, from 1 to 20 Hz, but only the first pulse of the series was used for quantification. The intervals between trains are usually longer than 30 s to minimize the effect on synaptic inhibitory or facilitating pathways. To test if the hChR2 response was monosynaptic, we applied TTX (1 μM) to the bath until the EPSC reaction disappeared, and then applied 4-aminopyridine (4-AP; 100 μM). Typically, a response is returned within a few minutes, with a slightly longer delay between LED firing and EPSC generation. NBQX (10 μM) was used to test whether the response is driven by AMPA receptors.
Sharp hCO sections were created as described previously. Briefly, hCO sections were embedded in 4% agarose and transferred to cells containing 126 mM NaCl, 2.5 mM KCl, 1.25 mM NaH2PO4, 1 mM MgSO4, 2 mM CaCl2, 26 mM NaHCO3 and 10 mM d-(+) -glucose into Sections were cut at 200–300 µm at room temperature using a Leica VT1200 vibrator and stored in ASF at room temperature. Then, patch-camp recording of whole cells was performed on hCO sections under a direct SliceScope microscope (Scientifica). Sections were perfused with aCSF (95% O2 and 5% CO2) and cell signals were recorded at room temperature. hCO neurons were applied using a borosilicate glass pipette filled with a solution containing 127 mM potassium gluconate, 8 mM NaCl, 4 mM magnesium ATP, 0.3 mM sodium GTP, 10 mM HEPES, and 0.6 mM EGTA, internal pH 7, 2, adjusted with KOH (osmolarity 290). For recovery purposes, add 0.2% Biocytin to the internal solution.
Data were acquired by Clampex (Clampex 11.1, Molecular Devices) using a MultiClamp 700B amplifier (Molecular Devices) and a Digidata 1550B digitizer (Molecular Devices), low-pass filtered at 2 kHz, digitized at 20 kHz, and analyzed using Clampfit ( version 10.6) for analysis, molecular devices) and custom MATLAB functions (MATLAB 2019b, Mathworks). The junction potential was calculated using JPCalc and the entries were adjusted to the calculated junction potential of -14 mV. Operation IV consists of a series of current steps in 5-10 pA steps from -50 to 250 pA.
For morphological reconstruction of pinched neurons, 0.2% biocytin (Sigma-Aldrich) was added to the internal solution. The cells are primed for at least 15 minutes after hacking. The pipette is then slowly drawn in for 1–2 min until the registered membrane is completely sealed. Following section physiology procedure, sections were fixed overnight at 4° C. in 4% PFA, washed with PBS X3, and diluted 1:1000 with streptavidin-conjugated DyLight 549 or DyLight 405 (Vector Labs). Cells filled with biocytin (2%; Sigma-Aldrich) were labeled during patch clamp recording at room temperature for 2 hours. The sections were then mounted on microscopy slides using an Aquamount (Thermo Scientific) and visualized the next day on a Leica TCS SP8 confocal microscope using an oil immersion objective with a numerical aperture ×40 1.3, magnification ×0.9–1.0, xy. The sampling rate is approximately 7 pixels per micron. Z-stacks at 1 µm intervals were obtained serially, and z-stack mosaics and Leica-based auto-stitching were performed to cover the entire dendritic tree of each neuron. Neurons were then tracked semi-manually using the neuTube 40 interface and SWC files were generated. The files were then uploaded to the SimpleNeuriteTracer41 Fiji plugin (ImageJ, version 2.1.0; NIH).
Human cortical tissue was obtained with informed consent according to a protocol approved by the Institutional Review Board of Stanford University. Two samples of human postpartum tissue (3 and 18 years old) were obtained by resection of the frontal cortex (middle frontal gyrus) as part of surgery for refractory epilepsy. After resection, harvest tissue in ice-cold NMDG-aCSF containing: 92 mM NMDG, 2.5 mM KCl, 1.25 mM NaH2PO4, 30 mM NaHCO3, 20 mM HEPES, 25 mM glucose, 2 mM thiourea, 5 mM sodium ascorbate, 3 mM sodium pyruvate, 0.5 mM CaCl2 4H2O and 10 mM MgSO4 7H2O. Titrate to pH 7.3-7.4 with concentrated hydrochloric acid. Tissues were delivered to the laboratory within 30 minutes and coronal sections were taken according to the procedure described above.
All animal procedures were performed in accordance with Stanford University APLAC-approved animal care guidelines. Rats (greater than 140 days post-transplantation) were induced with 5% isoflurane anesthesia and anesthetized with 1-3% isoflurane intraoperatively. Animals were placed in a stereotaxic frame (Kopf) and sustained release buprenorphine (SR) was injected subcutaneously. The skull is exposed, cleaned and 3-5 bone screws are inserted. To target t-hCO, we generated stereotaxic coordinates from MRI images. A burr hole was drilled at the site of interest and fibers (400 µm diameter, NA 0.48, Doric) were lowered 100 µm below the hCO surface and secured to the skull with UV-curable dental cement (Relyx).
Fiber photometric recordings were performed as previously described42. To record spontaneous activity, rats were placed in a clean cage and a 400 µm diameter fiber optic patch cable (Doric) connected to a fiber optic photometric data acquisition system was connected to the implanted fiber optic cable. During the 10-minute recording of motor activity, the animals were free to explore the cage. To record the evoked activity, rats (more than 140 days after transplantation) were anesthetized with 5% isoflurane for induction and 1-3% isoflurane for maintenance. Place the animal in a stereotactic frame (Kopf) and the whiskers on the opposite side of the t-hCO are trimmed to about 2 cm and passed through a mesh connected to a piezoelectric actuator (PI). A 400 µm fiber optic patch cable (Doric) was connected to the implanted fiber and connected to the data acquisition system. The whiskers on the opposite side of t-hCO were then deflected 50 times (2 mm at 20 Hz, 2 s per presentation) at random times by a piezoelectric drive over a 20 minute recording period. Use the Arduino MATLAB Support Package to control deflection time with custom MATLAB code. Events are synchronized to the data acquisition software using transistor-transistor logic (TTL) pulses.
Rats (greater than 140 days post-transplantation) were induced with 5% isoflurane anesthesia and anesthetized with 1-3% isoflurane intraoperatively. Animals were placed in a stereotaxic frame (Kopf) and buprenorphine SR and dexamethasone were injected subcutaneously. The skull is exposed, cleaned and 3-5 bone screws are inserted. To target t-hCO, we generated stereotaxic coordinates from MRI images. A circular craniotomy (approximately 1 cm in diameter) was performed with a high speed drill directly over the transplanted hCO. Once the bone is as thin as possible, but before drilling through the entire bone, use forceps to remove the remaining intact pelvic disc to reveal underlying t-hCO. The craniotomy was filled with sterile saline, and a coverslip and a special head pin were attached to the skull with UV-cured dental cement (Relyx).
Two-photon imaging was performed using a Bruker multiphoton microscope with a Nikon LWD (×16, 0.8 NA) objective. GCaMP6 imaging was performed at 920 nm with 1.4x single z-plane magnification and 8x average of 7.5 fps. Rats were induced with 5% isoflurane anesthesia and maintained with 1-3% isoflurane. The rats were placed in a custom made head fixture and positioned under the lens. A 3-minute background recording of motor activity was obtained. Over the course of 20 minutes of recording, 50 puffs (each presentation 100 ms long) were randomly delivered to the whisker pad opposite t-hCO using a picospricer. Use the Arduino MATLAB Support Package to control burst time with custom MATLAB code. Synchronize events with data acquisition software (PrairieView 5.5) using TTL pulses. For analysis, the images were corrected for xy motion using affine correction in the MoCo program launched in Fiji. Extraction of fluorescent traces from individual cells using CNMF-E43. Fluorescence was extracted for each region of interest, converted to dF/F curves, and then converted to z-scores.
Rats (greater than 140 days post-transplantation) were induced with 5% isoflurane anesthesia and anesthetized with 1-3% isoflurane intraoperatively. Animals were placed in a stereotaxic frame (Kopf) and buprenorphine SR and dexamethasone were injected subcutaneously. The whiskers on the opposite side of the t-hCO were cut to about 2 cm and threaded through a mesh connected to a piezoelectric actuator. The skull is exposed and cleaned. A stainless steel ground screw is attached to the skull. To target t-hCO, we generated stereotaxic coordinates from MRI images. Perform a circular craniotomy (approximately 1 cm in diameter) with a high speed drill just above the t-hCO. Once the bone is as thin as possible, but before drilling through the entire bone, use forceps to remove the remaining intact pelvic disc to reveal underlying t-hCO. Individual cells were recorded using 32-channel or 64-channel high-density silicon probes (Cambridge Neurotech) grounded to ground screws and pre-amplified with RHD amplifiers (Intan). Use the manipulator to lower the electrodes to the target site through the craniotomy, which is filled with sterile saline. Data collection was performed at a frequency of 30 kHz using the Open Ephys data acquisition system. The recording continued only when we detected highly correlated rhythmic spontaneous activity in more than 10 channels, suggesting that the electrodes were located in the graft (based on two-photon calcium imaging data). A 10-minute background recording of motor activity was obtained. The whiskers on the opposite side of t-hCO were then deflected 50 times (2 mm at 20 Hz, 2 s per presentation) at random times by a piezoelectric drive over a 20 minute recording period. Using the MATLAB Support Package for Arduino (MATLAB 2019b), control deflection time with custom MATLAB code. Use TTL pulses to synchronize events with data acquisition software.
For optical marking experiments, a 200 µm optical patch cord (Doric) connected to a 473 nm laser (Omicron) was connected to a 200 µm optical fiber placed over the craniotomy. Immediately before this, adjust the jumper power to 20 mW. Use the manipulator to lower the electrodes to the target site through the craniotomy, which is filled with sterile saline. At the beginning of the recording, ten pulses of light 473 nm (frequency 2 Hz, pulse duration 10 ms) were emitted. Photosensitive cells were defined as cells that exhibited a spike response within 10 ms of light in 70% or more of the trials.