Ethanol Causes Cell Death and Neuronal Differentiation Defect During Initial Neurogenesis of the Neural Retina by Disrupting Calcium Signaling in Human Retinal Organoids

Fetal Alcohol Syndrome (FAS) affects a significant proportion, exceeding 90%, of afflicted children, leading to severe ocular aberrations such as microphthalmia and optic nerve hypoplasia. During the early stages of pregnancy, the commencement of neural retina neurogenesis represents a critical period for human eye development, concurrently exposing the developing retinal structures to the highest risk of prenatal ethanol exposure due to a lack of awareness. Despite the paramount importance of this period, the precise influence and underlying mechanisms of short-term ethanol exposure on the developmental process of the human neural retina have remained largely elusive. In this study, we utilize the human embryonic stem cells derived retinal organoids (hROs) to recapitulate the initial retinal neurogenesis and find that 1% (v/v) ethanol slows the growth of hROs by inducing robust cell death and retinal ganglion cell differentiation defect. Bulk RNA-seq analysis and two-photon microscope live calcium imaging reveal altered calcium signaling dynamics derived from ethanol-induced down-regulation of RYR1 and CACNA1S. Moreover, the calcium-binding protein RET, one of the downstream effector genes of the calcium signaling pathway, synergistically integrates ethanol and calcium signals to abort neuron differentiation and cause cell death. To sum up, our study illustrates the effect and molecular mechanism of ethanol on the initial neurogenesis of the human embryonic neural retina, providing a novel interpretation of the ocular phenotype of FAS and potentially informing preventative measures for susceptible populations.


Introduction
Prenatal alcohol exposure results in a set of severe heterogeneous neurodevelopmental conditions termed fetal alcohol syndrome (FAS), characterized by growth retardation, craniofacial anomalies, and neurodevelopmental disruption [1].FAS is the commonest non-inherited cause of neurological deficit, with a high prevalence rate of up to 55.42 per 1,000 worldwide [2].Eye is the most sensitive organ for teratogen, and ocular abnormalities occur in over 90% of children with FAS, including coloboma, microphthalmia, Yu Gong and Lingling Ge contributed equally to this work.and optic nerve hypoplasia [3].Initial neurogenesis of the embryonic retina around six-week gestation is the most critical period of human eye development but sustains the highest risk of prenatal ethanol (EtOH) exposure because of ignorance of early pregnancy [4].Therefore, it calls for a more profound elaboration of the mechanisms of ethanol on the initial neurogenesis of the human embryonic retina.
Retinal neurogenesis is a complex process that involves the formation of the neural retina in the inner wall of the optic cup and the differentiation of multiple retinal neurons in certain spatial and temporal dimensions to develop into a specific intricate neural circuit.As retinal neurogenesis initiates, retinal progenitor cells at the apical of neural retina (NR) start to differentiate into postmitotic differentiating cells with the earliest-born retinal ganglion cell (RGC), then locate at the inner basal layer of the neural retina [5].The human retina developmental process cannot be recapitulated in vitro until a novel established retinal organoid system based on human embryonic/pluripotent stem cells in recent years.The human retinal organoids (hROs) are three-dimensional self-organized aggregates that resemble human retinas at multiple dimensions from the generation of retinal ganglion cells to functional maturation of the retinal neural circuit [6].Large numbers of publications demonstrate that hROs are reliable models in vitro to investigate the development, disease, toxicology, and drug screening of human retinas [7][8][9].Thus, hROs enable direct exploring the prenatal alcohol exposure on early neurogenesis of human embryonic retinas instead of nonhuman animal retinas [10].
Early neurogenesis is also a matter of calcium.Calcium signals are instrumental for differentiating neural progenitor cells into neurons during nervous system development.Increasing evidence demonstrates that intracellular Ca 2+ dynamics are necessary and sufficient to initiate the proneural gene expression, cell cycle progression, and neuron axonal maturation [11].On the other hand, calcium signaling also responds to intrauterine stress, such as valproic acid [12], cholinomimetic agents [13], and virus infection [14], and contributes to neurodevelopmental defects.Moreover, as organized by the Ca 2+ channel on the cell membrane and endoplasmic reticulum, dynamic Ca 2+ is restricted in seconds [15] and serves as a more sensitive tool to orchestrate the early neurogenesis both in the physiology and pathological conditions [16,17], compared with classic secreted signaling molecules such as Shh, Wnts, BMPs, FGFs, and retinoic acid, which are involved in previous FAS related nonhuman animal model [18,19].Calcium signaling also actively participates in the retinal development by contributing to the formation of the embryonic retinal waves and individual saltatory movements of the nucleus within retinal progenitor cells [20,21].However, the impact of prenatal ethanol exposure on calcium signaling during initial neurogenesis of the human neural retina and the role of calcium signaling in the ethanol-induced human ocular phenotype remain elusive.
In this work, we utilized hROs to decipher the impact and mechanism of ethanol exposure on the initial neurogenesis of human neural retina.Short-term ethanol exposure from D24 to D30 slowed the growth of NR in hROs, reduced the proliferation of retinal progenitor cells, increased the apoptosis in hROs and obstructed the differentiation and axonal maturation of retinal ganglion cells.We highlighted the role of calcium signaling pathway in the ethanol-induced cell death and neuronal differentiation defect.Moreover, we demonstrated that calcium dynamic alteration in ethanol-treated hROs was attributed to down-regulation of Ryanodine Receptor 1 (RYR1) and Calcium Voltage-Gated Channel Subunit Alpha1 S(CACNA1S), and transduced by downstream effector RET to abort neuron differentiation and cause apoptosis via MAPK pathway.

Human Embryonic Stem Cells Derived Retinal Organoids Recapitulate the Initial Neurogenesis of the Human Neural Retina
As the high consistency of early development between hROs and corresponding human fetal retina in vivo [22], to recapitulate the initial neurogenesis of human embryonic neural retina, we determined to explore hROs from D24 to D30 when retinal ganglion cells were just born using a neural retina enriched induction protocol previously provided by Kuwahara et al. [23].Embryonic bodies acquired Rax + optic vesicle with unlaminated epithelium morphology at D20 (Fig. 1A, B) and further developed into SOX2 + CHX10 + neural retina at D24 (Fig. 1C).The hROs at D24 contain ATOH7 + retinal progenitor cells and differentiating p27kip1 + postmitotic cells at the basal layer of the neural retina (Fig. 1D, E).Pax6 +strong retinal ganglion precursor cells first emerged at Sox2 − basal layer, accompanying the primitive lamination of neural retina at D24 (Fig. 1F).Retinal ganglion precursor cells continuously differentiated into Islet1 + Tuj1 + immature RGCs at D30 (Fig. 1G).The immature RGCs turned to a mature phenotype with extending outward Tuj1 + axonal processes after hROs were attached to the gelatin-coated plate (Fig. 1H, I).To detect the molecular dynamic of the initial neurogenesis of neural retina in hROs, we constructed a human embryonic stem cell line expressing GCaMP5G, a genetically encoded calcium indicator suit for neural activity imaging, with pLOV-CMV-GCaMP5G transfection.More than 90% of cells expressed GCaMP5G indicating differential cellular calcium dynamic in human embryonic stem cell clones and sustained standard cell morphology and pluripotency after flow cytometry sorting (Fig. S1; Video 1).After induction for 24 days, we again harvested hROs with GCaMP5G expressed in the neural retina (Fig. 1J).Cells at the apical outer neuroblast layer were clearly labeled with GCaMP5G (Fig. 1K) and exhibited spontaneous robust global calcium transients which are representative of initial neurogenesis of neural retina [24] (Fig. 1L).These data supported that hROs from D24 to D30 well recapitulate the initial neurogenesis of the human neural retina around six-week gestation in vivo at the histological, cellular, and molecular level, distinguishing our work from previous model animal research.

Ethanol Exposure Slows the Growth of the Neural Retina in hROs
Ethanol, a well-known teratogen for human embryonic development, has a role as an antiseptic drug, a polar solvent, a central nervous system depressant, and a disinfectant.Prenatal alcohol exposure during the initial neurogenesis of the neural retina is catastrophic for eye development.To address the impact of ethanol on the initial neurogenesis of the human neural retina, we first assessed the morphological change of hROs after ethanol exposure.The hROs were stochastically divided into four groups which treated ethanol from D24 to D30 with 0% (control, v/v), 1%, 2%, and 3% EtOH, respectively.As hROs continuously grow in the control condition, 3% EtOH directly brought hROs to a dehydrated-like morphology.Treatment with 2% EtOH induced hROs an atrophic phenotype, whereas 1% EtOH almost aborted the growth of neural retinal in hROs (Fig. 2A).The thickness of neural retina at D24 stayed at about 185 ± 23 μm (mean ± SD), grew to 249 ± 24 μm during the development.Treatment of 1% EtOH resulted in a stagnation of thickness at 187 ± 26 μm.In comparison, 2% EtOH resulted in a shrinkage of thickness at 142 ± 40 μm (Fig. 2B).The analysis of variance in neural retina thickness demonstrated a notable growth reduction with 1% EtOH treatment and a detrimental effect with 2% EtOH treatment (Fig. 2D).In line with this, measurement of the maximum sectional area of hROs and its variance between D30 and D24 also indicated a similar trend that hROs slowly ceased to grow with 1% EtOH and atrophied with 2% EtOH (Fig. 2C, E).To find the most significant altered mechanism in the maximum pathophysiological condition of heavy drinking rather than the toxicology of ethanol on the eye development, we selected 1% EtOH as a concentration for further experiment.

Ethanol Exposure Causes Neuronal Differentiation Defect and Cell Death in Neural Retina of hROs
As we found that 1%EtOH robustly reduced the growth rate of the neural retina at the initial neurogenesis stage, we sorted out to investigate the generation of retinal ganglion cells (RGCs), the first-born neuron in the neural retina.Compared with the control group, Pax6 +strong retinal ganglion precursor cells robustly decreased with 1% EtOH (Fig. 3A,  B, I).Moreover, immature Tuj1 + RGCs failed to correctly locate at the basal layer of the neural retina (Fig. 3C, D).When attached to the gelatin-coated dish for a week, RGCs in the neural retina could extend their axons out of hROs for up to 2.4 mm long with dense spines in the control condition (Fig. 3E, G, K).However, 1% EtOH weakened the ability of axonal extension and spine maturation of RGCs and leads to a reduction in the axon density, axonal length, and spine density after hROs attachment (Fig. 3F, H, J ~ L).
As the cell death and cell cycle disruption accounts for most ethanol induced neurodevelopmental defects, we further we performed cell cycle analysis and apoptosis marker -cleaved caspase 3 and TUNEL analysis to investigated the cell death of hROs after ethanol treatment.Immunostaining showed the percentage of cleaved-caspase3 + cells in ethanol-treated hROs significantly elevated to more than 22.1 ± 7.6% (mean ± SD) compared with 4.9 ± 1.5% in control hROs (Fig. 4A ~ C), indicated a burst of regulated cell death after ethanol treatment.TUNEL assay showed a similar result: ethanol increased the percentage of TUNEL + cells in hROs from 3.3 ± 0.7% to 23.8 ± 7.0% (Fig. 4D ~ F).On the other hand, Ki67 + proliferative cells in hROs drop down to 4.5 ± 0.8% after ethanol treatment compared with 20.1 ± 4.8% in control hROs(Fig.4G ~ I).To detect the cell cycle, we dissociated the hROs and analyzed the cell cycle with flow cytometric analysis.More than 59.6 ± 7.3% of cells stayed at G1 phase in the control condition while increasing to 82.5 ± 4.8% after ethanol treatment.Consistently, cells in S and G2/M phase dropped to 10.3 ± 2.5% and 7.2 ± 2.4% after ethanol treatment compared with the control group staying at 24.4 ± 5.4% and 16.0 ± 1.9%, respectively (Fig. 4J, K).This result indicated ethanol-induced cell-cycle arrest in the G1 phase of cells in hROs, which supports previous results from animal models [25].

Transcriptome Profiling Reveals the Calcium Signaling Pathway as a Candidate Mechanism of Cell Death and Neuronal Differentiation Defect in Ethanol-treated hROs
To investigate the possible molecular mechanism beneath the ethanol-induced cell death and neuronal differentiation defect, we applied bulk RNA-seq analysis of both control hROs and ethanol-treated hROs.Bioinformatics analysis showed that there were a total of 434 (173 genes upregulated and 261 downregulated genes) differentially expressed genes (DEGs) between the control and ethanol-treated hROs (Fig. 5A).Cluster heat maps showed DEGs between the control hROs and ethanol-treated hROs (Fig. 5B).Volcano plots presented the distribution of DEGs as gray, red, and green circles (Fig. 5C).Enrichment analysis in the PaGenBase [26] demonstrated that these DEGs of ethanol-treated hROs could be enriched as retina tissue-specific pattern genes.Enrichment analysis in Cell Type Signatures [27] further identified DEGs of ethanol-treated hROs as embryonic neural stem cells of homo sapiens.To interpret these DEGs for specific biological processes, we next conducted the Gene Ontology (GO) annotation analysis.The terms such as chromosome organization, positive regulation of cell death, visual perception, regulation of cell development, regulation of neuron differentiation, regulation of ion transport, eye development, and response to calcium ion were significantly enriched for ethanol-treated hROs based on gene counts and P-value (Fig. 5F).This result aligned with the previous study about chromosome damage in ethanol-related diseases [28], and supported the biological phenomenon shown in Figs. 3  and 4. It is worth noting that ion transport and calcium ion response were also enriched as significant biological process.Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotation further distinguished several pathways such as systemic lupus erythematosus, amyotrophic lateral sclerosis, pathways in cancer, shigellosis, toxoplasmosis, circadian rhythm, and calcium signaling pathway (Fig. 5G).The chromosome related pathway and immune system disruption of teratogenic effects of ethanol exposure has been widely discovered, however the calcium signaling alteration

Ethanol Exposure Alters the Dynamic of Calcium Signaling in hROs
The genetically encoded calcium indicator and twophoton microscope enabled us to capture the living cellular calcium signaling in hROs without chemotoxicity and phototoxicity.As shown, the neural retina of hROs presented multiple spontaneous calcium spiking regions characterized as global and local transients (Fig. 6A, B).These regions spontaneously actively generated calcium transients at 0.55 ± 0.46 events per minute (mean ± SD) in the local transient domain and at 0.46 ± 0.20 events per minute in global transient domain (Fig. 6C, O, R; Video 2).Amplitude and duration of the local calcium transient maintained at 8.5 ± 4.8 Z and 4.8 ± 4.3 s, respectively (Fig. 6D, M, N).Amplitude and duration of the global calcium transient maintained were at 4.0 ± 3.2 Z and 17.9 ± 15.8 s, respectively (Fig. 6E, P, Q).Ethanol exposure produced a profound effect on the number of calcium spiking domains contributed by a robust increase of local calcium transient (Fig. 6F, G, K, L).The spiking frequency of local and global transient reduced to 0.14 ± 0.08 and 0.16 ± 0.10 events per minute respectively after ethanol exposure (Fig. 6H, O, R; Video 3).For local calcium transient, ethanol exposure did not alter the amplitude but prolonged the duration to 16.2 ± 18.2 s (Fig. 6M, N).For global calcium transient, ethanol exposure reduced the amplitude to 2.3 ± 1.8 Z but did not alter the duration (Fig. 6P, Q).These data showed ethanol exposure could significantly increase the numbers of local transient domains, lower local calcium transients' frequency, and prolong its duration.Moreover, ethanol exposure significantly lowered global calcium transients' frequency and reduced amplitude.Together, these results supported that ethanol exposure altered the calcium signaling dynamic in hROs and may induce cell death and neuronal differentiation defect during the initial neurogenesis of the neural retina.

Identification of RYR1 and CACNA1S as Regulators of Calcium ion Transport and RET as Synergistic Effecter in the Calcium Signaling Pathway of Ethanol-treated hROs
To uncover the potential molecular mechanism beneath the ethanol-induced calcium signaling alteration, we next performed the gene set enrichment analysis (GSEA) analysis of DEGs in ethanol-treated hROs to screen target genes of the calcium signaling pathway.As shown, genes involved in voltage-gated calcium channel activity were enriched, and the most significantly changed genes, such as RYR1 and CACNA1S, were selected as candidate target genes (Fig. 7B).RYR1 encodes the ryanodine receptor 1 on the endoplasmic reticulum and acts as a calcium channel to release the calcium into the cytoplasm [29].CACNA1S encodes the α1S subunit of L-type voltage-dependent calcium channel expressed on the plasm membrane to transport extracellular calcium into the cytoplasm [30].Both gene expressions were verified by qRT-PCR (Fig. 7C) and could help explain the calcium signaling alteration observed with the two-photon microscope.Besides, to investigate the effector gene of a downstream cascade of the calcium signaling pathway, we further performed Cnet plotting of their associated genes in DEGs of ethanol-treated hROs.In the network, calcium transporters, such as RYR1 and CAC-NA1S, were included, and the RET gene was highlighted as the most related target genes with only its degree ≥ 6 (Fig. 7A).RET has four cadherin-like domains binding of calcium ions and participating transduction of downstream effect of calcium signaling via MAPK-and AKT-cascade that plays a role in cell differentiation and survival.To further understand the critical nodes of the RET-related pathway in ethanol-treated hROs, we analyzed these DEGs with the STRING online database to predict the functional protein-protein interaction of RET (Fig. S2).Among these results, CTNNB1, VEGFA, KIT, and TH were predicted as candidate interactive proteins which has been demonstrated playing a role in the cell death/survival and neuron differentiation.As a member of the calcium signaling pathway, the Cnet plotting showed RET simultaneously participated communities, such as eye development, positive regulation of cell death, and regulation of neuron differentiation, and regulation of cell development.To verify this prospective relation, we confirmed mRNA levels of selected genes, such as the specific retinal ganglion cell transcription factor POU4F2, the anti-apoptotic gene BCL-2, and its transcriptional repressor, the pro-apoptotic gene BCLAF1, and RET.The result showed that the mRNA level of RET, BCL-2, and POU4F2 were downregulated, but BCLAF1 was upregulated in ethanol-treated hROs (Fig. 7C, D).This result indicated that RET acts as a hub gene of the calcium signaling pathway in ethanol-induced cell death and neuron differentiation in hROs.

Discussion
Ethanol is a teratogen, and the eye is the most sensitive organ to teratogens.As early as 1910, Stockard et al. discovered that prenatal ethanol exposure could cause abnormal eye development of chicks and fish.In 2002, Strömland et al. reviewed the ocular signs of patients with fetal alcohol syndrome, such as microphthalmia and optic nerve hypoplasia, and proposed that the neural retina is the pivotal toxic target tissue of ethanol [31].It is generally recognized that pregnant women should not drink alcohol during pregnancy.However, due to alcohol addiction and recreational drug abuse, the risk of prenatal ethanol exposure continues to exist.The risk and degree of prenatal ethanol exposure reach the highest when the early pregnancy is not informed (less than six-week gestation) and the neural retina/brain begins to generate neurons [32].The initial neurogenesis of the neural retina plays a decisive role in the development of the whole eye, but whether and how ethanol affects this critical process is unclear.Most previous studies are based on animal model and set a long ethanol exposure time in experiments (span across the entire gestation).However, animal experiments of different species for the same question even have contradictory results.It prompted us to study the effect of short-term ethanol exposure on the initial neurogenesis of the human neural retina using the human retinal organoids.
Retinal ganglion cells are the earliest differentiated neurons during retinal neurogenesis, which determine the primary retinal lamination.Its development deficit often leads to irreversible blindness in patients.Previous studies have shown that ethanol can interfere with the formation of optic vesicles before the differentiation of retinal ganglion cells and even neural retina formation [33].After the differentiation of retinal ganglion cells, ethanol exposure will reduce the number of axons of retinal ganglion cells and the thickness of optic nerve sheath, and affect the electrophysiological function of the retina to varying degrees [34].Different time windows of ethanol exposure lead to different cell component abnormalities and morphological changes because of cell-selective and tissue-selective responses to ethanol [35].Our study found that the postmitotic ATOH7 + cell population who mainly give birth to the retinal ganglion cell of human embryonic retina emerged at D24 hROs, advancing the emerging timeline of the first transitional cell population of hROs described by Sridhar's work from D60 to D24 [22].Hence we think D24 hROs recapitulate the initial neurogenesis of human embryonic retina which is the very beginning of RGCs differentiation.Short-term ethanol exposure from D24 to D30 significantly slowed the growth of hROs, reduced the number of RGCs, and disrupted their corrective location to the basal side of the neural retina after the asymmetric division of retinal progenitor cells.As immature retinal ganglion cells must undergo radial migration to correctly locate at the basal destination to eventually mature both in cell architecture and electrophysiological function [5,36] and the Nestin + scaffold has not been significantly affected (Fig. 3C, D), suggesting that Tuj1 + ganglion cell radial migration ability might be damaged.At the same time, we found that ethanol inhibited the axon growth of retinal ganglion cells, similar to the inhibition effect of ethanol on the axonal growth of neurons in human brain organoids [37].Previous animal experimental studies based on zebrafish and chicks have put forward conflicting views on retinal cell death after ethanol exposure [38,39].Our results showed that short-term ethanol exposure caused significant regulated cell death in the neural retina which aligns with the research in chick embryos.At the same time, the cell cycle of the neural retina cells was arrested in G1 phase, which is consistent with the previous animal experimental results [25].In short, short-term ethanol exposure does have a critical impact on the initial neurogenesis of the neural retina by inducing cell death and neuronal differentiation defect.
It is known that ethanol can interfere with chromosome assembly, signal molecule secretion, and transcription factors expression, thus participating in the occurrence of cancer occurrence, neuronal degeneration, immune system diseases, and tissue development abnormalities [28].This study found, for the first time, that the calcium signal of the neural retina was also significantly changed by ethanol exposure, which suggests that the calcium signaling pathway may also be one of the molecular mechanisms of ethanol induced teratogenesis.Calcium signal is a vital intracellular second messenger.Calcium signals can widely participate in many critical cell functions, such as cell proliferation, death, differentiation, and movement in the development of nervous system and served as a common target in neurological disorders and neurogenesis [40].We found that short-term ethanol exposure at the initial neurogenesis of the neural retina would lead to a decreased frequency and amplitude of global calcium transients in hROs.Previous studies have suggested that global calcium transients can directly regulate cell proliferation and gene transcription [41], which can explain the neuronal differentiation defects after ethanol-induced calcium dynamic alteration.RNA-seq data and qRT-PCR verification indicated that these calcium dynamic changes might be mediated by the differential expression of the endoplasmic reticulum calcium channel RYR1 and the α1S subunit of the plasma membrane voltagegated calcium channel CACNA1S protein caused by ethanol exposure.On the other hand, ethanol exposure led to an increase in the number of local calcium transient regions in hROs, a decrease in the frequency of calcium spiking, and an increase in the duration.This characteristic change of calcium dynamic is similar to the over-enhancement of local calcium transients caused by calcium overload in mitochondria associated with apoptosis [42].Mechanically, it is worth mentioning that HSPA8, the most significant up-regulated gene in RNA-seq, can mediate the influx of calcium ions in plant mitochondria [43,44].Meanwhile, HSPA5, another member of the HSPA family is clearly engaged in physical interactions with animal endoplasmic reticulum InsP3R and mitochondrial VDAC1 channel to mediate calcium overload to trigger regulated cell death [45,46].Hence we speculated that the cell death in hROs caused by ethanol exposure might be related to the mitochondrial calcium overload associated with the up-regulation of HSPA8.
Calcium signal transduction often requires the activation of downstream effector proteins.Through bioinformatics analysis, we screened the RET gene as the core target gene of the calcium signaling pathway during ethanol exposure.RET protein contains cadherin-like calcium binding domains, which can directly respond to changes in calcium concentration [47].In addition, it is known that RET is directly involved in the differentiation and maturation of neurons and can also regulate cell death through downstream MAPK or AKT signaling pathway [48].The enrichment analysis of the KEGG pathway found that the MAPK signaling and calcium signaling pathway also produced significant differences in hROs after ethanol exposure (Fig. 5G).The qRT-PCR verification showed that the expression of RET gene and the anti-apoptotic gene BCL2, the downstream target gene of ERK1/2 in MAPK signal [49], were significantly down-regulated.Its transcription-inhibitory pro-apoptotic gene BCLAF1 was significantly up-regulated, supporting that RET-regulated cell death was caused by ethanol exposure through the MAPK signal pathway.At the same time, the specific transcription factor POU4F2 of retinal ganglion cells was also significantly down-regulated (Fig. 7D).These results support that RET gene plays a critical synergetic role in ethanol-induced neuron differentiation defect and cell death (Fig. 8).
To sum up, in this work we explained the causes of cell death and neuron differentiation defects caused by shortterm ethanol exposure at the initial neurogenesis of the human neural retina from the perspective of the calcium signaling pathway and obtain the results closest to human development in vivo using hROs.These results may help us understand the mechanism of optic nerve hypoplasia and microphthalmia in FAS patients.In the future, utilizing organoid assemblies, neural chimeras, and microfluidic chips for a more realistic developmental environment exposure to ethanol with a dynamic moderate physiological concentration, we could elaborate on more detailed principles of ethanolinduced teratogenesis and instruct the rational application of alcohol and ethanol-contained drugs during pregnancy.

Human Embryonic Stem Cell Culture and Transduction of pLOV-CMV-GCaMP5G
All experiments were performed in accordance with the guidelines of the Ethics Committee of Southwest Hospital, Third Military Medical University (Army Medical University).Briefly, hESCs Line H1 (RRID: CVCL_9771) were cultured under feeder-free condition using Essential 8TM Medium (Gibco) and Vitronectin (Gibco).H1 cells express key pluripotency markers such as OCT4, SOX2, NANOG, and SSEA4 which are crucial for maintaining their undifferentiated state (Fig. S1, E ~ H).H1 cells typically display a classic morphology of small, tightly-packed colonies with high nucleus-to-cytoplasm ratio and prominent nucleoli, which has a distinct appearance under a microscope, showing tightly packed cells with defined boundaries (Fig. S1, A and B).For routine culture, hESCs colonies were passaged every 4 days using Versene (Gibco) at 37℃ for 5 min.For lentiviral transduction, hESCs were dissociated as single cell by TrypLE™ Express (Gibco) and suspended in Essential 8TM medium containing pLOV-CMV-GCaMP5G (OBiO Technology, multiplicity of infection equal to 5), Y27632 (Sigma,10 μM), and polybrene (10 μg/ml), and then seeded in a Vitronectin coated plate.Sixteen hours later, virus supernatant was removed and fresh Essential 8TM medium containing Y27632 was added and withdraw Y27632 the day after.To enrich infected cells, the cell population was subjected to fluorescence-activated cell sorting as described later.

Fluorescence-activated Cell Sorting
Fluorescence-activated cell sorting (FACS) was performed as described previously [50].Briefly, cells were sorted by using MoFlo XDP (Beckman Coulter).GCaMP5G positive cells were gated by comparison to samples of wild-type hESCs.FlowJo 10.2 was used for data analysis (Tree Star).

Live Calcium-imaging Recordings of hROs-GCaMP5G Using Two-photon Microscopy
Calcium imaging was performed as we previously published [53].The hROs were incubated in DMEM/F12 without phenol red (Abcam Cat# 11-039-021) at room temperature.We used a moveable objective microscope (Sutter) equipped with a Chameleon titanium-sapphire laser tuned to 915 nm (Coherent), and employed an Olympus LUM-PlanFI 40 × water immersion objective (NA 0.8).Emitted fluorescence was captured by the objective and filtered using an HQ 535/50GFP emission filter (Chroma Technology) before detection using Pho Image v.3.0 software3, run on a computer.Images were acquired at a resolution of 256 × 256 pixels.Image sequences were acquired at 1 ms per line using 256 × 256 per frame and analyzed with Igor Pro v.6.10 or ImageJ v.1.53t(NIH).The regions of interest were defined by the standard deviations of image sequences by semiautomatic analysis described previously [54].Scanning and image acquisition were controlled using scan normalized fluorescence (Z score) calculated as follows: △F/F = (F-F0)/ F0, where F is the instantaneous GCaMP5G fluorescence, and F0 is the baseline GCaMP5G fluorescence.

TUNEL Assays
The TdT-mediated dUTP-X nick end labeling (TUNEL) assays were performed by In Situ Cell Death Detection Kit (Roche, Basel, Switzerland).Buffer 1 and buffer 2 were mixed at a ratio of 1:9 and then diluted at 1:3 with 1 × PBS.The cryosections of hROs were finally incubated with the TUNEL reaction mixture at 37 °C for 30 min in a black humidified box.Nuclei were counterstained with DAPI for 10 min.The sealed slides were photographed by a LSM880 microscopy (Zeiss) and images were then analyzed using ImageJ v.1.46.

Cell cycle Analysis by Flow Cytometry
hROs were subjected to Tryple™ Express (Gibco) digestion for 30 min to obtain a single-cell suspension.The suspension was then centrifuged at 200 × g for 5 min to collect the cell pellet.The cell pellet was washed once with 1 ml of pre-chilled PBS and centrifuged again.The cell pellet was resuspended in pre-chilled 70% ethanol and fixed at 4 °C overnight.After fixing, the cells were centrifuged at 200 × g for 5 min and washed once with pre-chilled PBS.Subsequently, the cell pellet was resuspended in 0.5 ml of staining buffer (≤ 1 × 10 6 cells) and mixed with 10 μl of propidium iodide (PI) and 10 μl of RNase A solution (Takara).The mixture was then incubated at 37 °C in the dark for 30 min for flow cytometry analysis.Flow cytometry was performed using the BD Accuri C6 flow cytometer with excitation at 1 3 488 nm, and cell DNA content was analyzed using FlowJo 7.6.1 software.

Bulk RNA-sequencing Profiling and Bioinformatic Analysis
In this study, Bulk RNA-sequencing and bioinformatic analysis were perfomed as described previously [55].Specifically, human retinal organoids (hROs) were treated with 1% ethanol (EtOH) from D24 to D30, resulting in the ethanoltreated group, while hROs treated with phosphate-buffered saline (PBS) during the same period served as the control group.Each group consisted of three independent biological replicates and each replicate consisted a batch of samples gathering more than fifteen D30 hROs for analysis.Total RNA was extracted from the hROs using TRIzol reagent (Gibco), and its concentration was determined using a Nan-oDrop (DE) spectrophotometer.RNA quality was assessed using an Agilent 2100 bioanalyzer (CA).Reverse transcription was carried out using the SMARTer Stranded Total RNA Sample Prep Kit (Takara, Japan) according to the manufacturer's instructions, with subsequent isolation of cDNA using Oligo (dT)-attached magnetic beads.The resulting cDNA fragments were amplified via PCR to construct the cDNA libraries.Sequencing of the qualified libraries was performed on the Illumina HiSeq platform using a 150-bp paired-end model.Post-sequencing, the reads obtained were subjected to further filtering using Cutadapt.The filtering criteria included the removal of reads containing adapters, polyA and polyG sequences, and reads containing more than 5% of unknown nucleotides.Additionally, low-quality reads containing more than 20% of low-quality bases were eliminated.The quality of the filtered sequences was verified using FastQC.To analyze the expression levels of transcripts, the filtered reads from all samples were aligned to the reference genome of the research species using the HISAT2 package.Subsequently, the mapped reads from each sample were assembled, and a final transcriptome encompassing all samples was generated.For transcript expression quantification, the Fragments Per Kilobase of transcript per Million mapped reads (FPKM) metric was calculated.To identify differentially expressed genes (DEGs), the read count data matrices were subjected to DESeq2 software.DEGs were selected based on fold changes (≥ 2) and adjusted p-values (< 0.05) (|log2FC|≥ 1 and q < 0.05) as the thresholds.Visualization of DEG patterns was accomplished through clustered heat maps and volcano plots.Upregulated genes were represented in red (log2FC ≥ 1 and q < 0.05), while downregulated genes were represented in blue (log2FC ≤ -1 and q < 0.05).To gain insights into the biological processes and enriched pathways associated with the DEGs, Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were conducted using the Metascape platform (metascape.org).Enrichment analysis in PaGenBase [26] and Cell Type Signatures [27] were also performed on the Metascape platform.Additionally, Gene Set Enrichment Analysis (GSEA) was performed using GSEA software (http:// softw are.broad insti tute.org/ gsea/).All raw data have been made transferred via the Gene Expression Omnibus (GEO) database.

Plotting of Cnet Network and Protein-protein Interaction Network
For Cnet plotting depicting genes in those most significant and interested enriched terms, we used the Cytoscape software (Version: 3.9.1) to visualize the linkage of imported genes and correspondent biological processes as a network.The raw disorganized network was arranged in a more visually appealing way via the Cytoscape's built-in layout algorithms (yFiles Radial Layout) and the further beautification of layout was performed via Adobe Illustraor Artwork 24.0.1.
For plotting protein interaction network on the STRING website.We go to the official STRING website at https:// cn.string-db.org, enter the cell death related and eye development related genes in the search bar, and select the homo sapiens organism.To generate the interaction network, we set the network type as full STRING network, adjust the required confidence score as 0.400 and FDR stringency as 5 percent.The network stats were listed as below: number of nodes is 85; number of edges is 135; expected number of edges is 66; average node degree is 3.18; avg.local clustering coefficient is 0.458; PPI enrichment p-value is 4.83e −14 .

Quantitative Real-time Polymerase Chain Reaction (RT-PCR)
Three independent experiments were performed in this study.Briefly, total RNA was extracted from hROs with TRIzol reagent (Gibco), and reverse-transcribed using the PrimeScript® RT Reagent Kit (Takara).The cDNA was amplified with specific gene primers as listed below.Quantitative RT-PCR was conducted using the CFX96 Real-Time PCR System (Bio-Rad) with at least three separate RNA samples.RYR1, For CAC CAA TGG CCT ATA CAA CCAG, Rev GCT CAG GAT AAC GCC CTC G; CACNA1S, For TTG CCT ACG GCT TCT TAT TCCA, Rev GTT CCA GAA TCA CGG TGA AGAC; POU4F2, For CTC GCT CGA AGC CTA CTT TG, Rev GAC GCG CAC CAC GTT TTT C; RET, For GCG ATG TTG TGG AGA CCC AA, Rev AGC ACC GAG ACG ATG AAG GA; BCL2, For GGT CGC CAG GAC CTC GCC GCTG, Rev GGT TGA CGC TCT CCA CAC ACAT; BCLAF1, For AGG TCT GGG TCT GGT TCT GTTG, Rev GAA GCC TCT TTA TCC CTG GTAT.

Quantification of the NR Thickness, Volume and Attached Axons
A Leica DMI3000 was used to acquire bright field images of hROs.The NR thickness, and the maximun sectional areas were measured by ImageJ-win64 software as indicated in Fig. 2A.The hROs at D30 were attached to the gelatin-coated dish for another a week and axons and spines were counted clockwise with ImageJ.

Fig. 1
Fig. 1 Identification of initial neurogenesis of human neural retina in human embryonic stem cells derived retinal organoids (hROs).(A) Bright field view showing the typical epithelium morphology of hROs at D20 with induction protocol.Immunostaining showed hROs with the Rax + optic vesicle at D20 (B) and Sox2 + Chx10 + neural retina at D24 (C).Cells of hROs at D24 showing neurogenic competence (ATOH7 + D) and exit of cell cycle (p27kip1 + E) to ready to differentiate into neurons.Pax6 +strong retinal ganglion precursor cells (F) emerge at the Sox2 − inner basal layer of hROs at D24 and gradually specified into immature Islet1 + Tuj1 + RGC (G) at D30.

Fig. 2
Fig. 2 Neural retina in hROs retards its growth after ethanol exposure.(A) Bright field view showing the morphological change during D24 ~ D30 with different ethanol concentration(v/v).Black arrow heads indicate the laminated neural retina (NR) of hROs.3% ethanol directly induce tissue necrosis of hROs as 2% induce an atrophy phenotype of neural retina of hROs.Only 1% ethanol induce a growth reduction.Measurement of NR thickness and maximum sectional area of hROs exposed to 0%, 1%, and 2% ethanol respectively at D24 and D30.Thickness of NR (B) and maximum sectional area of hROs

Fig. 3
Fig. 3 Ethanol impairs the specification, radial migration, and axonal growth of retinal ganglion cells in hROs.Immunostaining showing Pax6 +strong retinal ganglion precursor cells mainly distributed at the inner basal layer of hROs (A) and robustly decreased after the ethanol treatment at D30 (B).Immunostaining showing radial migration of immature Tuj1 + RGC along the nestin + basal scaffold and located at the inner basal layer of hROs (C) whilst it failed to migrate and as a result dislocated at the apical side of neural retina in ethanol-treated hROs (D).Representatives of the hROs attached to the gelatin-coated

Fig. 4
Fig. 4 Ethanol causes cell death and cell cycle arrest in hROs.(A) Representatives of cleaved-caspase3 + apoptotic cells in D30 hROs without ethanol treatment.(B) Representatives of cleaved-caspase3 + apoptotic cells in D30 hROs treated with 1% EtOH from D24 ~ D30.(C) Quantification of cleaved-caspase3 + apoptotic cells in D30 hROs both in control and ethanol group (n = 9 and 9 for each group).(D) Representatives of TUNEL assay staining in D30 hROs without ethanol treatment.(E) Representatives of TUNEL assay staining in D30 hROs treated with 1% EtOH from D24 ~ D30.(F) Quantification of TUNEL assay staining in D30 hROs both in control and ethanol group (n = 9 and 9 for each group).(G) Representatives of Ki67 +

Fig. 5
Fig. 5 The RNA-Seq analysis of control hROs and ethanol-treated hROs at D30. (A) The number of differentially expressed genes (DEGs) in hROs after treatment with 1% EtOH for D24 ~ D30.(B) Cluster heatmap of hROs between control and ethanol-treated hROs (n = 3, three independent experiments).(C) Volcano plot displaying significant DEGs in hROs after treatment with 1% EtOH for

Fig. 6
Fig. 6 The calcium signaling dynamic was significantly altered in ethanol-treated hROs.(A) Median intensity projection image of 1000 frames from neural retina of hROs-GCaMP5G in control group.Map of spontaneously active calcium spiking regions overlaid on image.(B) Map of calcium spiking regions recorded in control hROs-GCaMP5G (Red, local transient domain; Green, global transient domain).(C) Raster plots displaying time and intensity of calcium transients in the control hROs-GCaMP5G.Intensity versus time traces for calcium spiking regions (corresponding to B), showing characteristics of Ca 2+ transients of local (D) and global transient domains (E). (F) Median intensity projection image of 1000 frames from neural retina of hROs-GCaMP5G in ethanol-treated group.Map of spontaneously active calcium spiking regions overlaid on image.(G) Map of calcium spiking regions recorded in ethanol treated hROs-GCaMP5G (Red, local transient domain; Green, global transient domain).(H) Raster plots displaying time and intensity of calcium transients in ethanol-treated hROs-GCaMP5G.Intensity versus time traces for calcium spiking regions (corresponding to G), showing characteristics of Ca 2+ transients of local (I) and global transient domains (J).Graph of numbers of calcium spiking regions (K) and percentage of local transient domains (L) in hROs.Graph showing amplitude per event (M), duration per event (N), and frequency of events per domain (O) in local transient domains of control and ethanol-treated hROs.Graph showing amplitude per event (P), duration per event (Q), and frequency of events per domain (R) in global transient domains of control and ethanol-treated hROs.Data are shown as mean ± SEM (n = 3 independent experiments).*p < 0.05; **p < 0.01; ***p < 0.005.Scale bars: 25 μm for A,F ◂

Fig. 7
Fig. 7 The enrichment of RYR1/CACNA1S/RET-dependent calcium signaling pathway to uncover the cell death and neuronal differentiation defect in ethanol-treated hROs.(A) Cnet plotting showing the interaction of 7 enriched pathways and genes associated with ethanol induced ocular aberration.Red line and circle represent the main pathway enriched in Cnet plotting.Green line and circle represent the downstream pathways and genes.Blue line and circle repre-