Abstract
Huntington’s disease (HD) is a polyglutamine neurodegenerative disease involving pathogenesis within the striatum and cerebral cortex and a neurodevelopmental component, particularly in juvenile HD form (JOHD). We established a fused HD dorsal-ventral system, imitating the cortex and striatum interaction in a single organoid to discover neurodevelopmental impairments at the level of cell populations. We found a range of early pathogenic phenotypes indicating that brain development in HD is affected by impaired neurogenesis. The phenotypes occurred already in early-stage 60-day organoids and the brain of humanized mouse embryos, at time of the beginning of the neurogenesis and choroid plexus development. We demonstrated that HD organoids and HD mouse embryonic brains had gene expression profiles of impaired maturation of neurons and increased expression of genes responsible for proliferation compared to genes responsible for differentiation in control organoids. By using scRNA-seq, the choroid plexus population was highly abundant in HD organoids and embryonic brains. Cortical and choroid plexus cell populations in HD organoids and brains co-expressed genes responsible for HTT function (mitotic spindle and cilia). The impaired maturation and the increased occurrence of the choroid plexus populations were mitigated in our compensatory model, mosaic dorsal/ventral (D/V) or V/D HD/control organoids. Finally, we found that TTR protein, a choroid plexus marker, is elevated in the adult HD mouse serum, indicating that TTR may be a promising marker for detecting HD. In summary, the fused dorso-ventral HD organoids identify a spectrum of neurodevelopmental features, including increased proliferation and delayed cell maturation. We demonstrate that the choroid plexus population is characteristic of aberrant HD neurodevelopment, and contains TTR marker, which can be translated as a blood marker in HD.
1. Introduction
Huntington’s disease (HD) is a fatal, dominantly inherited neurodegenerative disease that affects mainly the central nervous system, causing the loss of neurons in the striatum and cerebral cortex (Cudkowicz and Kowall, 1990; Reiner et al., 1988). Symptoms may occur years or even decades prior to diagnosis, and the disease often worsens as the number of CAG repeats in the huntingtin gene increases (Quarrell et al., 2013). In rare cases, particularly with more than 60 repeats, the disease can occur before the age of 20, known as juvenile HD (JHD) (Fusilli et al., 2018), progresses more rapidly, and has different manifestations than adult-onset HD. Studies suggest that pre-manifest HD patients can develop cognitive, psychiatric, and motor impairments long before diagnosis (Paulsen et al., 2014; Quaid et al., 2017). In addition, decades before adult-onset the gene deregulation involved in neurodevelopment and neuronal degeneration can also occur, in both, mouse models and patients (Hodges et al., 2006; Labadorf et al., 2015; Milnerwood et al., 2006; Schippling et al., 2009). One of such phenotype that occurs years before overt HD symptoms is enlarged ventricular volume in the brain (Hobbs et al., 2010; Jahanshahi et al., 2022). Also, juvenile HD demonstrates an accumulation of various brain defects such as obvious ventricular enlargement. The symptom is characteristic for neurodevelopmental diseases such as autism spectrum disorders (ASD) where it occurs together with abnormalities of the choroid plexus (ChP) and the formation of cysts of ependymal origin in early brain development (Movsas et al., 2013). Similarly, ChP volume was significantly increased in psychosis patients (Lizano et al., 2019). Alzheimer’s disease (AD) is associated with numerous changes in ChP morphology and function, such as changed metabolic activity and reduced toxins clearance (Krzyzanowska and Carro, 2012; Marques et al., 2013; Spector and Johanson, 2013). The ChP is an important physiological barrier and the principal source of cerebrospinal fluid (CSF) responsible for clearance of toxic products from central nervous system (CNS) (Lehtinen et al., 2013; Redzic et al., 2005).
Potential neurodevelopmental conditions such as autism (Birey et al., 2017; Mariani et al., 2015), microcephaly (Tiscornia et al., 2011), HD (Conforti et al., 2018) and BBB (Pellegrini et al., 2020) were already investigated by using small three-dimensional (3D) brain organoids (Lancaster et al., 2013). Recently, the brain organoid approach has been enriched with fused cerebral organoids which recreate the dorsal-ventral forebrain axis (Bagley et al., 2017). In HD, loss of neostriatal volume and medium spiny neurons (MSN) goes along with pyramidal neuron loss in the neocortex (Vonsattel et al., 2011). The combination indicates complex HD pathogenesis, including various cellular and communication impairments between the dorsal (cortex) and ventral (striatum) regions particularly during forebrain development (Rangel-Barajas and Rebec, 2016). Therefore, assembloids of dorso-ventral forebrain origin are an ideal model for studying the interactions between the most important and HD-involved brain regions. Brain organoids originating from juvenile HD cells would be particularly useful for defining neurodevelopmental HD brain pathogenesis at several levels, such as the molecular mechanisms, cellular organization of brain and blood-brain barrier (BBB) and choroid plexus impairments.
In our work, we have established a dorso-ventral HD juvenile fused organoids system to define the complex neurodevelopmental impairments at the levels of molecular biomarkers, cellular differentiation and interactions, and structural changes in HD brain. We showed that assembling the dorso-ventral axis by fusing HD organoids differentiated towards the cerebral cortex (CTX) and striatum (STR) can uncover distinct early brain neurodevelopment in HD. We found changes in the choroid plexus (ChP) in HD brain organoids and embryos of mouse models at the level of composition of cell populations, the presence of ChP pseudo organs in striato-cerebral organoids, and the occurrence of the ChP biomarkers in HD fused organoids and blood serum of HD mice. We demonstrate that ChP abnormalities are an essential neurodevelopmental component of HD pathogenesis and that the markers of ChP in blood serum can be developed as markers of early HD.
2. Results
2.1. Juvenile HD dorso-ventral brain organoids demonstrate markers of early developing cortex and striatum
The generation of our control and HD dorso-ventral forebrain organoids from human juvenile HD iPSC lines is demonstrated on the scheme in Fig. 1A. In brief, neurospheres were primed either by chemical inhibition of SHH by smoothened receptor inhibitor cyclopamine A for obtaining dorsal organoids and by WNT inhibition by IWP2 and priming SHH by agonist SAG to ventral organoids which were later fused and cultivated to model the early pseudo tissue of cerebral cortex (CTX) assembled with the striatum (STR). Since we pursued to model early brain development, we established day 60 as the maturing endpoint for our control and HD organoids. Importantly, we also generated mosaic organoids, which were our compensatory model to assess the contribution of dorsal and ventral parts in the development of HD phenotypes in fused organoids. Fig. 1B presents a representative slice of our HD fused brain organoid stained for TBR1, a marker of the dorsal forebrain and the scheme of the coronal embryonic brain section showing the dorsal-ventral gradients.
First, we used unfused dorsal or ventral 60-day-old brain organoids to confirm the successful proliferation and maturation of various forebrain regions. We assessed the expression of specific cortical and striatal markers (Supp. Fig. 1). The separately primed organoids predominantly differentiated either to dorsal (TBR1 and PAX6) or ventral markers (NKX2.1, DLX2 and GSX2) as demonstrated via qPCR method (Supp. Fig. 1A). Subsequently, to assess the contribution of the cerebral cortex and striatum parts in 60 day fused cortico-striatal assembloids we detected their early markers by immunostaining (Supp. Fig. 1B). We detected immunostaining for TBR1 and PAX6, which are markers of ventricular zone (VZ) and subventricular zone (SVZ) of the developing neocortex (respectively), as dorsal forebrain markers. For examining ventral organoid parts, we detected the immunofluorescence staining of DLX2 and NKX2.1, markers of medial ganglionic eminence (MGE), and GSX2, marker of lateral ganglionic eminence (LGE). We also detected relatively late markers of differentiated neurons such as NEFH and SMI-32, indicating expression of forebrain markers, as well as other neuronal markers (Supp. Fig. 1B). Altogether, these results confirm the successful generation of fused dorso-ventral forebrain organoids, which are suitable to study neurodevelopmental aspects of HD pathology.
2.2. Fused juvenile HD brain organoids are larger and contain prominent PAX6 neuronal stem cell rosettes, which are diminished in control and mosaic organoids
We examined the morphology and quantified the sizes of HD, mosaic, and control organoids (Fig. 2). The 109Q HD organoids were significantly larger and were remarkably more complex in their shape than all control fused organoids. 60-day-old HD organoids were significantly larger in size (pval. < 3.35 e-05) than control organoids (Fig. 2A and B). On average, mosaic 21D/71V organoids were close in size to control organoids, while the opposite configuration of mosaic 71D/21V was close in size to HD organoids (Fig. 2B). All post-hoc test p values are presented in Fig. 2.
In the developing brain, there is an expression of the PAX6 marker, which indicates the presence of a population of early progenitors, and the expression of the TBR1 marker, which indicates the presence of a population of early-born neurons. Our qPCR (Supplement 1A) and RNA-seq DEG results detected the TBR1 [pval = 0.012387; log(FC) = ∼ 2.16] and PAX6 [pval = 0.01252; log(FC) = ∼1.34] markers upregulations in 60 days fused brain HD organoids. To confirm the presence of the PAX6-positive and TBR-positive populations in our organoids, we performed an immunostaining protocol using the same type of organoids as used for scRNA-seq. Surprisingly, the immunostaining results on 60 days fused organoids (Fig.3) demonstrate the occurrence of the large number of PAX6-positive neural rosettes lying close to each other and the occurrence of the separate layers of TBR1 expression in HD fused organoids (71D/71V & 77D/77V). In contrast, the control fused organoids (17D/17V & 21D/21V) demonstrated only several scattered rosettes on organoid slices. Our results demonstrate that the development of the HD organoids is directed towards a more proliferative character of early neural development with the occurrence of a vast number of progenitors in rosettes.
The control organoids present lower expression of PAX6 and TBR1, indicating a more balanced development of structures and their earlier neuron differentiation and maturation. Importantly, in the mosaic organoids, we observed the regression in the HD phenotype represented by reduced neural rosette occurrence, which is consistent with the results from GSEA (Supp. Fig. 2 A; Supp. 1).
2.3. Prominent deregulations of cell population markers of choroid plexus, inhibitory neurons, and glial cells in juvenile HD organoids
We analyzed the publicly available scRNA-seq data in silico to select markers of the most common brain population of neuronal and non-neuronal cells for further analysis of their expression levels in dorso-ventral HD and control organoids (Fig. 4). Another step in the inclusion of markers for analysis was the manual selection, population markers likely affected in the developing HD brain. TTR, H3ST4, PPP1R1B, GPR88, NOS1, NXPH1, CKB, APOE and GRIN3A genes were selected. Additionally, using an in silico bioinformatic decision tree and HTT expression level approach (described in Material and Methods), ATOX1, FOXP1, LIMK2, and SMARCA4 marker genes of cell populations were selected.
Selected genes were general markers associated with all basic types of brain cells, including inhibitory and excitatory neurons, astrocytes and oligodendrocytes, tanycytes, choroid plexus, nitrergic neurons, and neuronal progenitors. Firstly, we compared the expression of several inhibitory neuron markers and showed the deregulated expression of five markers. LIMK2 showed upregulation in all HD organoids compared to 21Q control organoids (ANOVA p.val = ∼ 0.0007). NXPH1 showed highly significant downregulation in all HD lines compared to 21Q and 33Q control organoids (ANOVA p.val = ∼ 3.31e-06). Other markers of early inhibitory neurons, such as FOXP1, GRIN3A, and CKB, have demonstrated inconsistent or non-significant outcomes. Next, we focused on analyzing the expression of selected choroid plexus (ChP) and tanycyte markers. TTR and PPP1R1B showed the most prominent upregulation in all HD lines compared to 21Q and 33Q control organoids. The changes in expression of the TTR mRNA were particularly differentiating across HD lines, with the increasing CAG lengths reaching 11.55 FC for 71D/71V, over 100 FC 77D/77V, and well over 1000 FC all vs 21D/21V control organoids. The 33D/33V organoids showed a 2.37 FC increase in TTR mRNA level vs. 21D/21V. The 33 CAG repeat line is still regarded as the healthy control; however, they contain a number of repeats that may indicate a certain risk of occurring some HD phenotypes (Fig 4. ANOVA p val. ∼ 0.00036; posthoc: T-test * p<0.05, **p<0.01, ***p<0.001). DARPP-32, encoded by the PPP1R1B gene, plays a pivotal role in integrating signal transduction in dopaminoceptive neurons and has a regulatory role in the central nervous system (CNS) (Lin et al., 2021). Another tanycyte marker, GPR88, showed elevated expression levels in the 77Q HD line compared to 21Q and 33Q control organoids (ANOVA p val. = 0.0196). In the group of astrocytes and oligodendrocytes markers, only one gene exhibited upregulation in 109Q HD organoids – SMARCA4 (ANOVA pval = 0.0290). The expression of ATOX1, a marker of neuronal progenitors, and HS3ST4, a marker of excitatory neurons, was also increased in 71Q HD lines in comparison to 21Q and 33Q control organoids (ANOVA pval = 0.0033 and 0.0086, respectively). CCL2 gene expression demonstrated downregulation in 71Q and 77Q organoids but a slight upregulation tendency in 109Q organoids (ANOVA = 0.0005). The opposite result was established for NOS1, a marker of nitrergic neurons, which showed upregulation in 71Q and 77Q organoids and downregulation in the 109Q line (ANOVA < 0.0008). All post-hoc test p values are presented in Fig. 4.
2.4. Pooled cell transcriptomes show early-developmental gene expression and GSEA profiles in Juvenile HD organoids and revers in mosaic juvenile HD/Control
Data from scRNA-seq were first used to identify the total number of DEGs and determine the main processes involved in HD and development in our organoids and mosaic organoids. We pooled cell transcriptomes based on genotypes and compared HD fused organoids (71D/71V, 77D/77V) vs. control (WT) fused organoids (17D/17V, 21D/21V), and mosaic HD/WT fused organoids (71D/21V, 21D/71V) vs. control fused organoids (21D/21V).
Obtained DEGs are presented as volcano plots (Fig. 5 A). We obtained 7815 downregulated genes and 123 upregulated genes for HD (71D/71V, 77D/77V) vs. control (17D/17V, 21D/21V) fused organoids, 9038 downregulated genes and 177 upregulated genes for mosaic HD/WT 71D/21V vs. healthy (21D/21V) organoids, and 8141 downregulated genes and 49 upregulated genes for mosaic WT/HD 21D/71V vs. control 21D/21V organoids.
Due to a certain variability of DEGs among organoids, we performed GSEA (see Materials and Methods section 2.8). The terms that occurred most frequently for all organoids were manually assigned to arbitrary groups and presented on plots based on their p-value in supplementary data (Supp. Fig. 2,3 & Supp. 1,2,3,4,5).
The GSEA demonstrated that HD organoids exhibit upregulation of genes associated with terms tightly connected to early development and brain formation (mostly Dentate gyrus, Forebrain, Telencephalon, and Eye), also connected with epigenetics and Extracellular Matrix (ECM) formation [H3-3A, BIRC5, RPS26, PAX6, LHX2, NR2E1, EMX2, HMGA2, FEZF2, COL3A1, COL1A2, COL4A6, COL21A, VCAM1, TTR, LTBP3]. In contrast, the terms enriched for downregulated genes in HD organoids are linked to processes characteristic of later stages of brain development, such as neuronal and glial cells maturation, migration, and signaling release related to processes of autophagy, mitotic spindle orientation and cilia motility. (Supp. Fig. 2A). Such gen deregulations and GSEA profiles indicate that control organoids are more related to maturation processes than HD organoids. Moreover, the profile of terms indicates that HD organoids delay or even arrest their development, preserving the stemness function of the cells and showing less relation to terms indicating brain structure development compared to healthy organoids. It can also be strongly related to alteration in such processes as mitotic spindle, cell polarity, cilia, cell motility, and autophagy, which play a crucial role in the early stages of healthy brain development. This is reflected by upregulated expression of specific gene markers such as EMX2 [pval = ∼ 1,46E-08; log(FC) = ∼7.02] and PAX6 [pval = ∼ 0,001; log(FC) = ∼1.34], which are early markers indicating a high degree of cell stemness in HD organoids. Moreover, the genes essential for proper cilia function in neurogenesis such as IFT88 [pval = ∼ 4,58E-07; log(FC) = ∼-1.14], ARL13B [pval = ∼ 0.02; log(FC) = ∼-0.43], KIF3A [pval = ∼ 6,62E-06; log(FC) = ∼ -0.65], and mitotic spindle MAPT [pval = ∼ 4,99E-11; log(FC) = ∼ - 1,2] are downregulated.
Furthermore, our GSEA analysis revealed that several processes and terms observed for HD vs control organoids were not found or were reversed for mosaic organoids 71D/21V and 21D/71V vs. control organoids. For instance, terms characteristic for HD-upregulated genes related to initial brain formation were not found, and terms for HD downregulated genes related to neuronal and glial maturation and migration were now identified for upregulated genes in mosaic organoids. (Supp. Fig. 2 B & C; Supp. 1,2,3 for all genes, terms, and statistics). The observation indicates the importance of the interaction of ventral and dorsal regions in brain structure development in HD. Moreover, in our human organoids, both affected ventral and dorsal parts are necessary for the full manifestation of the HD phenotype.
2.5. Embryonic brains of HD mouse contain gene expression and GSEA profiles associated with HTT function
We pooled cell transcriptomes based on genotypes and compared mouse embryonic brains from HD Yac128Q/null vs control Bac21Q/null and HD Hu128Q/21Q vs control Bac21Q (Fig. 5 B). We obtained 328 downregulated and 1531 upregulated DEGs for disease YAC128Q/null vs. control BAC21Q/null, and 1439 downregulated and 985 upregulated DEGs for Hu128Q/21Q vs. control BAC21Q/null embryonic brains.
The analysis of YAC128Q/null vs BAC21Q/null demonstrated downregulated genes mainly related to neuronal and glial cell proliferation, development, and regulation of cell signal release, autophagy, and ECM formation. For upregulated genes, we noticed processes related to embryo development and brain structure formation, regulation of neurogenesis, cell migration, myelination, and cell connection formation and signaling related to processes of mitotic spindle orientation, cell polarity, cilia, and programmed cell death (Supp. Fig. 3 A; Supp. 4).
Similar results were present in Hu128Q/21Q vs Bac21Q/null analysis except that downregulated genes also included terms and processes such as mitotic spindle orientation, cell polarity, and motility, and an increased number of genes assigned to cell maturation and single release terms. The upregulated genes demonstrated similar results to YAC128Q/null vs BAC21Q/null analysis (Supp. Fig. 3 B; Supp. 5).
In addition, the results of both YAC128Q/null and Hu128Q/21Q mouse embryonic brains demonstrated several upregulated genes related to the control of mitotic spindle, potentially indicating the occurrence of altered development and neurogenesis in HD embryos. The upregulated genes included Map1B [YAC128Q/null: pval = ∼ 1,77E-08; log(FC) = ∼ 0.59, Hu128Q/21Q: pval = ∼ 3,47E-11; log(FC) = ∼0.69], Clta [YAC128Q/null: pval = ∼ 4,83E-09; log(FC) = ∼ 1.34, Hu128Q/21Q: pval = ∼ 0,00004; log(FC) = ∼ 1.02].
2.6. Increased TTR mRNA level is a common feature of HD organoids and HD embryonic brains compared to control, mosaic organoids, and control brains
Analyzing all our HD organoid and HD embryonic models, we found that the TTR expression vastly increased in all HD models compared to their control counterparts, demonstrating high statistical significance. Pooled cell transcriptomes of 71D/71V & 77D/77V organoids showed TTR expression log(FC) = ∼8.64 [pval = 3.41E-08] vs Control organoids (Fig. 5 A); pooled cell transcriptomes of YAC128/null embryonic brain showed TTR expression log(FC) = ∼13.15 [pval = 6.03E-23], and Hu128/21 showed TTR expression log(FC) = ∼5.70 [pval = 2.1E-08] all vs control BAC21/null embryonic brain (Fig. 5 B). Surprisingly, we found that our compensatory model, mosaic organoids in both HD ventral/control dorsal and control ventral/HD dorsal configuration compared to the control 21D/21V organoids did not reveal statistically significant changes in TTR expression (Fig. 5 A).
TTR protein often occurs as deposits of fibrils in ECM in neurodegenerative diseases. Interestingly, the ECM formation genes firmly belonged to the downregulated gene pool in the HD embryo brain, and the same term was located in the upregulated gene pool in HD organoids. This could be attributed to the higher demand for extracellular matrix as a scaffold for cell proliferation in organoids, isolated model systems that are not very richly supplied with ECM compared to regular tissue such as the brain. In the tissue, ECM demand is provided by various cell types and is more balanced.
2.7. HD organoids and embryonic brains contain a prominent increase in choroid plexus cells, which are not present or diminished in mosaic organoids
We performed a droplet-based scRNA-seq of 60-day-old fused dorso-ventral JOHD (71Q dorsal / 71Q ventral and 77Q dorsal / 77Q ventral) and control (17Q dorsal / 17Q ventral and 21Q dorsal / 21Q ventral) organoids where we profiled approximately 8.000 single cells, and mosaic organoids (71Q dorsal / 21Q ventral and 21Q dorsal / 71Q ventral) with approximately 3.000 single cells. Additionally, we performed a scRNA-seq sequencing of mouse embryonic brains (e.13.5) from HD disease YAC128Q/null and Hu128Q/21Q, and control BAC21Q/null model where we profiled 9.000 single cells.
We distinguished nine general cell populations from all analyzed organoids and embryonic brains characteristic for STR, CTX, MGE, LGE, caudal ganglionic eminence (CGE), hippocamp (HIP), tectum (TET), MEL, and ChP. Annotation of clusters to reveal population identities was performed as described previously (Kubiś and Figiel, 2023). Thanks to that, we were able to “zoom in” our results and divide general populations into many different subpopulations, and establish the cell lineage composition of the juvenile HD organoids and control organoids (Supp. Fig. 4) and also mouse embryonic HD and normal dorsal and ventral part of brains (Supp. Fig. 5).
The analysis of the single-cell data from organoids (Fig. 6) has shown that there was no ChP population in the 17D/17QV and less than 1% in 21D/21V control organoids. In 71D/71V and 77D/77V fused organoids, the choroid plexus accounted for as much as 10% and 16% of all cells. Another cellular population, present only to HD organoids were melanocytes, which accounted for more than 2% and 4% in 71D/71V and 77D/77V fused organoids, respectively. Our scRNA-seq data show that there is a tendency in HD that more pluripotential cells are directed to develop the dorsal patterns (CTX, HIP, ChP, MEL (Visual Cortex)) in the direction of the cortical cells, than the ventral patterned cells (MGE, LGE, STR). We observed that the presence of ventral and dorsal cell types is imbalanced in mosaic organoids. For the 71D/21V mosaic organoids, where the ventral side is healthy we noticed that the contribution of ventral part cells (MGE [17.19%], CGE [3.33%], LGE [0.6%]) is higher than dorsal cells (CTX [10.63%], ChP [2.24%], HIP [1.49%]). We observed the opposite in 21D/71V mosaic organoids, where the dorsal side is healthy and the ventral was affected. In this case, the share of dorsal part cells (CTX [24.32%], MEL [7.67%], HIP [1.86%]) is higher than ventral cells (MGE [0.88%], LGE [0.77%]). In both mosaic organoids, the share of STR ventral cells was on the same level ∼64%.
scRNA-seq from mouse embryo brains (Fig. 7) has shown that dorsal brain part BAC21Q/null revealed no Choroid plexus populations. Choroid plexus populations were a large fraction of dorsal embryonic brain parts of the YAC128Q/null (49.64%) and Hu128Q/21Q (35.78%) of all cells analyzed. In Hu128Q/21Q we also observed 6.92% of cell occurrence of the melanocyte population. In the ventral part of brains, we were able to see a more balanced population distribution in all mouse embryonic brain models. In YAC128Q/null and Hu128Q/21Q small fractions of melanocyte populations ∼ 1% occurred.
2.8. Juvenile HD organoids and embryonic HD brains express mitotic spindle and ciliary genes in cell subtypes of ChP, CTX and HIP
We selected a set of DEGs involved in mitotic spindle stabilization (EMX2, CLTA, MAP1B, MAPT), cilia (IFT88, ARL13B, KIF3A), and HTT, which is essential for both molecular structures and their functionality. We found that (Fig. 8A) control organoids co-expressed these genes in the cell populations mostly belonging to STR, MGE, CGE and CTX, while HD organoids co-expressed the genes mostly in ChP and CTX cell populations. In mosaic organoids, we have noticed co-expression of these genes in STR, MGE, MEL (melanocytes) cell populations and one cell subtype of ChP. Moreover, HD organoids reveal a high level of EMX2 expression in many cell populations, compared to cell populations in control and mosaic organoids, where expression of EMX2 is marginal.
In the healthy dorsal part of the mouse embryonic brain, we observed spindle- and cilia-related genes co-expression in populations of CTX, MGE, and HIP. In both HD dorsal brains (Fig. 8B), co-expression of these genes is primarily found in CTX, ChP, and HIP cell populations. In the case of the ventral part of the embryonic brain (Fig. 8C), in the control model, co-expression of these genes is predominantly observed in STR, LGE, MGE, CGE, CTX, and HIP. In the dorsal HD brain, there is a higher occurrence of co-expression with genes related to cilia, such as Itf88, Arl13B, and Kif3A, compared to the control model. Additionally, the Htt gene appears to be more frequently co-expressed in cell populations on the ventral brain side, whereas Emx2 is more frequent for the dorsal side of mouse embryonic brain. In this context, the ventral site, as the source of differentiating and migrating interneurons (Bagley et al., 2017; Birey et al., 2017), is crucial for developing the neocortical cell populations in the dorsal part of the brain and may contribute to the developmental HD pathogenesis.
2.9. Choroid plexus structures in HD organoids and increased TTR Protein in blood serum of HD Mouse Model
We performed immunostaining for TTR in 60-days old fused dorso-ventral 21Q/71Q mosaic organoids and 21Q dorso-ventral control organoid (Fig. 9A). The results showed an abundant signal for TTR in 71Q dorsal/21Q ventral mosaic organoid and no signal in 21Q dorso-ventral organoids, which corresponds to the scRNA-seq results. Transthyretin (TTR), also known as prealbumin, is a 55 kDa homotetrameric protein that transports thyroxine and retinol-binding protein (RBP) in the serum and cerebrospinal fluid. Since we found such a significant contribution of choroid plexus populations in HD and mosaic fused organoids, we investigated the levels of TTR protein as a marker candidate in the blood serum of HD mice (YAC128/null) using Western blot assay (Fig. 9B). Our findings show an increase in the level of TTR monomer (16 kDa) and tetramer (55 kDa) in YAC128/null mice compared to control BAC21Q/null mice (Fig. 9). The increased level of TTR protein corresponds directly with the scRNA-seq results of Hu128Q/21Q mice and HD and mosaic fused organoids.
3. Discussion
The malformation or damage of the human brain during its neurodevelopmental phase can contribute to neurodegenerative disorders such as HD. Possible early neurodevelopmental processes that can be affected include aberrant neural stem cell proliferation, neuronal differentiation, and synapse formation (Mariani et al., 2015; McRae et al., 2017).
Brain organoids are rapidly evolving tools for understanding early human brain structure development and neurological diseases in vitro. Modeling of brain inter-regional interactions can be improved by directed differentiation of hiPSCs into distinct, region-specific organoids, which can later be fused, yielding a more faithful structural and functional, in vitro representation of the brain (Bagley et al., 2017; Birey et al., 2017).
An example of structure-dependent healthy brain function includes cerebral cortex layers that can activate striatal neurons through an organized glutamate-driven projection system, tightly regulated by dopamine (DA) and gamma-aminobutyric acid (GABA) (Rangel-Barajas and Rebec, 2016). Such precise interactions ensure appropriate motor and cognitive function of the cerebral cortex and the striatum. In a diseased HD brain, the glutamate, DA, and GABA neurotransmitter systems of the cortex and striatum are compromised. The flow of information is gradually destroyed between the structures, giving rise to both motor and cognitive impairments. The interaction between the striatum and cerebral cortex forms during early neurogenesis, and therefore, its malfunction in HD can be studied in brain organoids, which recapitulate early events in embryonic brain development. For instance, in fused organoids, interneurons generated in the ventral domain can migrate to the dorsal domain (Bagley et al., 2017; Birey et al., 2017), mirroring in vivo tangential migration patterns from the subpallium to the cerebral cortex (Anderson et al., 1997).
Interestingly, the structure that is formed in the vicinity of the cortex and striatum and affects neurogenesis is the choroid plexus. Its formation is a very early event timely correlated with neurogenesis. It begins with the invagination of mesenchymal cells into the neural tube early in gestation, forming a monolayer of cells with motility cilia close to inner cortex layers and striatum in brain ventricles (Guemez-Gamboa et al., 2014; Jain et al., 2010; Pala et al., 2017). The primary function of these ciliary cells is the secretion of CSF and growth factors, which combined influence play a crucial role in brain development by regulating the environment in which neurogenesis occurs (Bitanihirwe et al., 2022; Johansson et al., 2013; Johansson, 2014; Lun et al., 2015b). This impacts neurogenesis, including progenitor survival, proliferation, and gliogenesis, in developing and adult mammalian brains (Moore and Iulianella, 2021).
The important function of the ChP-CSF system is the transport of molecules during critical periods of brain development. Along with choroid plexus, a very early event in development is the production and CSF secretion of transthyretin (TTR), a transporter of thyroid hormone and retinoic acid (Alshehri et al., 2020) (Alshehri et al., 2020; Bitanihirwe et al., 2022; Johansson, 2014; Johansson et al., 2013; Lun et al., 2015b, 2015a). Importantly, the presence of mHTT influences abnormal motility cilia and causes abnormal CSF flow (Keryer et al., 2011). Noteworthy, enlarged volume of lateral ventricles in the brain is an early phenotype of adult and juvenile forms of Huntington’s Disease (HD) (Hobbs et al., 2010; Jahanshahi et al., 2022). This phenotype may also correspond to possible extracellular ionic homeostasis dysregulation since ChP modulates the levels of various ions, such as sodium, potassium, and chloride, for maintaining proper neural stem cell function in new neuron formation (Bitanihirwe et al., 2022; Lun et al., 2015b). Therefore, it is plausible that the early enlargement of the ventricle phenotype may not be the primary result of brain degeneration but specific neurodevelopmental processes.
Our work aimed to discover early brain defects in HD and demonstrate that modeling the dorso-ventral forebrain axis as equivalent to cortico-striatal interactions is suitable for discovering such phenotypes in HD. We focused on early-rising cell populations that may be affected in the early phases of HD brain development using the fused dorsal and ventral forebrain organoids and HD mouse brains. Later, we examined HD phenotypes and cell populations when only one of the fused organoid parts, dorsal or ventral and vice versa, was HD or control (mosaic D/V and V/D fused organoids). Therefore, we have generated the HD and control organoids from iPSC lines by priming to the dorsal/neocortex (TBR1 and PAX6) and ventral/striatum (NKX2.1, DLX2, and GSX2) identities. The dorsal/ventral fused organoids showed cortical and striatal cell markers and mature neurons.
We also noticed that, in general, fused HD organoids were growing significantly larger than the control organoids as measured by their 2D area, indicating the differential composition of cell populations in HD vs. control fused organoids. We hypothesized that the HD organoids have increased proliferative features compared to control fused organoids. Such proliferative responses may indicate increased proliferation of neural stem cells in HD organoids instead of neuronal differentiation. Previous work indicated a transcriptional immaturity in HD organoids relative to controls (Conforti et al., 2018). Indeed, the dorsal part mimicking the developing forebrain of HD organoids demonstrated the abundance of PAX6 positive neural rosettes, indicating the abundance of neural stem cells, surrounded by TBR1 positive surroundings, indicating early-born neurons, while control fused organoids only scarcely contained stem cell rosettes. We also investigated our compensatory mosaic model, which showed few rosettes, a number similar to control organoids, indicating the reversal of the increased-rosette-number phenotype. Importantly, we observed an obvious morphological concrescence zone of cortical and striatal parts in HD organoids. Together, this shows that both parts of organoids may cooperate to exert HD phenotypes. This could indicate a cooperative influence of both cortex and striatum on HD neurodevelopmental deficits in HD patients.
The larger HD organoid size, the prominent occurrence of neural stem cell rosettes, and their concomitant influence on the striatal and dorsal part may be the result of changes in specific cellular populations of HD organoids. We first selected genetic markers of cell populations in the developing brain where HTT expression was present at high levels or completely absent (using public scRNA-seq data). Our findings revealed significant deregulation in several inhibitory neuron markers associated with the neuropathology of AD and autism (Cuberos et al., 2015; Yang and Shcheglovitov, 2020) and downregulation of synaptogenesis and synaptic integrity markers (Saykin et al., 2010).
The most striking result across all HD lines was the 10 to 1000 fold and highly consistent upregulation of markers lying at the interface of cortex and striatum, namely choroid plexus markers, TTR, and high level of dopaminoceptive striatal neuronal cell marker PPP1R1B. Other markers with elevated expression across organoid HD samples were GPR88, which has been linked to various psychiatric disorders (Massart et al., 2016) and astrocytes and oligodendrocytes marker, SMARCA4 deregulated in HD mice (Capizzi et al., 2022).
Next, to better understand the cell composition and cellular phenotypes in HD, we used our fused HD brain organoids for scRNAseq. Further, we aimed to investigate the role of dorsal and ventral brain organoid parts in HD by using our compensatory model, fused mosaic HD/Control organoid, and alone dorsal and ventral brain parts dissected from HD and control embryos. Moreover, we selected the E13 embryo developmental stage, which is just adjacent to the beginning of the accelerating phase of choroid plexus development in mice and associated with expansion with the beginning of neurogenesis (Fame and Lehtinen, 2020).
Cell transcriptomes of scRNAseq were first pooled based on genotypes for the identification of DEGs and the main processes involved in juvenile HD neurodevelopment. Analysis of HD vs control organoids has shown mainly downregulation in crucial processes for brain development regulation, such as mitotic spindle, cell polarity, cilia, cell motility, programmed cell death, and neuronal differentiation and maturation. This indicates that spindle orientation and ciliogenesis, and symmetrical or asymmetrical cell divisions crucial of cortical layers formations which are, in fact, HTT-dependent, can be disturbed in HD organoids (di Pietro et al., 2016; Godin and Humbert, 2011; Guemez-Gamboa et al., 2014; Keryer et al., 2011; Mukhtar and Taylor, 2018; Pala et al., 2017; Park et al., 2019). In turn, the upregulated processes were related to the regulation of neurogenesis, brain structure development, epigenetic regulations, and ECM formation, which are connected to the reprogramming and proliferation of NSC, hence the phase before neuronal differentiation. Again, we observed a reversal in our compensatory mosaic model (HD/control) concerning the main gene deregulations related to biological terms (Supp. Fig. 3 B & C). The processes and gene deregulations are closely related to an upregulation of genes ESCO2, BIRC5, and PAX6, related to mitotic spindle stabilization, and very high expression of EMX2, which in cortical development promotes clonal expansion and symmetric cell divisions (Mukhtar and Taylor, 2018). This demonstrates a strong response of developing HD organoids to the influence of mHTT on these processes and putative cellular attempts of activation of some compensatory pathways, likely activating the cell cycle control system, influencing pluripotency and differentiation (Aponte and Caicedo, 2017; Liu et al., 2019). Again, the DEGs analysis has demonstrated high levels of TTR expression in HD organoids and a lack of upregulated expression in mosaic organoids.
Our flagship analysis of the single-cell transcriptomes defined the populations of HD organoids, mosaic organoids, and embryonic HD brains. We have detected that HD organoids were highly populated by TTR-positive choroid plexus cells, also characterized in our scRNAseq data by the expression of TJP1 (pval = ∼ 2,54E-08; log(FC) = ∼ 7.59) and GATA3 transcription factor (pval = ∼ 1,93E-05; log(FC) = ∼ 4.71), which is prominently active in the development and the S100B (pval = ∼ 0.00048; log(FC) = ∼ 3.13). In contrast, we observed a minimal or complete absence of ChP cells in control fused organoids. In mosaic organoids, we observed the ChP population in the 71D/21V mosaic, whereas it was absent in the 21D/71V mosaic. Similar to HD organoids, dorsal brain parts of YAC128/null and Hu128/21 were also highly populated with choroid plexus cells, while this was not the case for BAC21/null. None of the ventral parts were populated by choroid cells. These results may indicate that the timing of development of choroid plexus cells is different in HD vs control. Moreover, the abundant population of the choroid in HD corresponds to the upregulation of many NSC genes (eg. PAX6 and EMX2, TBR1), indicating overactivated neurogenesis and less differentiation also in embryos. Probably, the overproduction of choroid cells also sustains until adulthood since a high level of TTR protein in CSF was reported in HD patients (Caron et al., 2022). Additionally, in mosaic organoids, when the dorsal part (containing some ChP) was affected, and the ventral part was a control, we observed a higher number of cells with more advanced differentiation characteristics of ventral brain regions (STR, LGE, MGE). Conversely, when the ventral part of the mosaic was affected, and the dorsal part was a control, we observed a higher number of populations of cells with characteristics of dorsal brain regions (CTX, HIP) but still no ChP population. In general, this demonstrates that the HD part is always delayed regarding its differentiation status.
To summarize our findings, we introduce novel organoid models for studying Huntington’s disease and conduct an array of low and high-throughput analyses, revealing changes in HD development closely linked to the HTT protein’s function. We elucidated the neurodevelopmental character of HD and demonstrated a developmental aberration consisting of an increased occurrence of Choroid plexus and TTR expression, which is a very early developmental event in HD. The increased Choroid plexus affected the early developmental deficit in cell differentiation and correlated with increased abundance of NSC. TTR serves as an essential marker of the Choroid plexus. However, its increased levels also occur in the serum of HD animals. Our study highlights the potential of ChP biomarkers in blood serum as non-invasive markers for early detection and monitoring of HD. In addition, the Choroid plexus may represent a potential tissue of interest for novel therapeutic approaches, including detoxification by interference with the blood-brain barrier and regulation of growth factors and cell differentiation within the brain ventricles’ microenvironment.
4. Materials and methods
All experiments were conducted in accordance with the relevant guidelines and established standards.
4.1. Human HD Induced-Pluripotent Stem Cells Culture
Human episomal HD and control iPSC lines were obtained from the NINDS Human Genetics Resource Center DNA and Cell Line Repository (https://catalog.coriell.org/1/ninds), from NINDS Human Cell and Data Repository (NHCDR) and from Cedars-Sinai Induced Pluripotent Stem Cell Core. For the analyses, we used three different HD lines with 71, 77 and 109 CAG repeats, and three control lines with 17/18, 21 and 33 CAG repeats. Human iPSCs were cultured in chemically defined conditions in Essential 8 medium (Life Technologies) and grown on Geltrex™ LDEV-Free, hESC-Qualified, Reduced Growth Factor Basement Membrane Matrix (Life Technologies). Cells were passaged using gentle dissociation with 0.5 mM EDTA in PBS.
4.2. Brain Organoid culture, differentiation and fusion
We developed the protocol for generating dorso-ventral forebrain organoids based on fusion protocol (Bagley et al., 2017; Knoblich et al., 2017) from human control and HD iPSC lines using an intermediate EZ spheres (neurospheres) culture step (Ebert et al., 2013). In brief, hiPSCs were detached with 0.5 mM EDTA in PBS, placed on a polyHEMA [Poly(2-hydroxyethyl methacrylate), Santa Cruz Biotechnology] coated dishes, and then allowed to form EZ spheres (Ebert et al., 2013) the next day. On the first day, EZ spheres were cultured in E8 medium with ROCK inhibitor (Y-2736, Tocris). On the second day, E8 medium were replaced with EZ medium (DMEM, F12, 2% B27 -vit. A, 1% penicillin/streptomycin, 0.5% Glutamax, 0.1% heparin) with growth factors, bFGF and EGF (ORF Genetics). After later mechanical passaging with McIlwain Tissue Chopper (model TC752, Campden Instruments) and obtaining the appropriate shape and diameter (>500 µm), neurospheres were incubated with small molecule inhibitors and activators of signaling pathways for dorsal (cyclopamine A: SHH inhibitor by smoothened inhibition) and ventral forebrain patterning (WNT inhibition by IWP2 and SHH activator by SAG agonist). Subsequently, the dorsal and ventral patterned organoids were fused and differentiated on the orbital shaker. Fused dorsal-ventral HD and control brain organoids, as well as our compensatory models - mosaic fused organoids (dorsal HD/ventral control and ventral HD/dorsal control), were collected on day 60 for scRNA-seq, RT-qPCR and immunohistochemistry analyses. Dorsal and ventral organoids from each cell line, which were cultured without the fusion step, were collected on day 60 for cortical and striatal markers assessment via qPCR.
4.3. Organoids Area Measurement
Pictures of 60-days old fused organoids were captured with Q-scope digital microscope model QS.80200-P (Euromex). Area measurement analysis was performed in ImageJ software using the “Freehand selection” tool. The scale was set based on the photograph of a calibration ruler added to the Q-scope kit, with a known number of pixels equaling the length of 5 mm. Graphs were created using GraphPad Prism Software (Boston, MA, USA).
4.4. HD mouse tissue
The HD (Hu128Q/21Q, Yac128Q/null, Bac21Q/null) mouse model was described previously (Southwell et al., 2017) and was maintained and bread in the Center for Advanced Technology (CAT) animal facility in Poznan. Animal experimentation and handling were approved and monitored by the Local Ethical Commission for Animal Experiments in Poznan (64/2018). For scRNA-seq analyses, pregnant female mice were sacrificed, embryos harvested, and embryonic brains dissected into cortex (dorsal part) and basal ganglia (ventral part). Blood serum for WB analyses was collected from the HD mouse model, from a typical experimental group of three or five mutant mice versus three or five control mice.
4.5. In Silico Analyses of Available Brain Region Transcriptomics Data
Public scRNA-seq datasets were acquired from NCBI GEO and Broad Institute Data Portal (https://singlecell.broadinstitute.org/) to select cell population markers in different brain structures for initial expression analyses. We used mouse brain datasets including hypothalamus [HYP] (7 329 cells; GSM3562050 and GSM3562051); embryonic thalamus (E12.5) [TH] (7 365 cells; GSE122012); motor cortex [MCTX] (13 539 cells; MD704, MD705); auditory cortex [ACTX] (35 594 cells; MD717, MD720, MD21); orbitofrontal cortex [OCTX] (19 531 cells; MD716, MD717); entorhinal cortex [ECTX] (80 653 cells; MD720, MD721, MD698/699); hippocampus [HIP] (6 555 cells; MD718, MD719) and cerebellum [CER] (611 034 cells). The datasets represented both single-cell and single-nucleus approaches; therefore, all mitochondrial and metabolism-related genes were excluded from the analysis based on genes from MitoCarta 2.0. Data from each mouse brain region were analyzed separately using our original JSEQ_scRNAseq pipeline designed to discover the composition of cell populations in high resolution (https://github.com/jkubis96/JSEQ_scRNAseq).
The cell populations from the high-resolution analysis were next analyzed using statistical and machine learning methods, such as a decision tree (rpart), correlation (corr), and statistical cell population comparison using the Shapiro-Wilk, T-Student, and Mann-Whitney tests with differences in the expression level of marker genes on significant level p <= 0.05. Subsequently, the specific cell population (e.g.: GABAergic neurons) was separated using quantiles 1 and 3 for high and low HTT gene expression in particular cells. The populations were then analyzed to obtain high expression of their specific markers and high HTT expression (R software with RStudio IDE). Specific qPCR primers (Table 1) were designed for each marker gene using MEGAX, Primer3, and NCBI primer blast tools.
The preanalyzed data used for this preliminary selection of markers are available and have been described (Kubiś and Figiel, 2023).
4.6. scRNA-seq
We performed scRNA-seq using cells from 60-days-old entire fused brain organoids or from E13.5 mouse embryonic brains dissected into dorsal (containing embryonic CTX, HIP, ChP) and ventral (containing embryonic STR, HYP, TH) parts. The day 60 organoids and dissected dorsal and ventral embryonic brains were incubated with 0.5 ml TrypLE (Thermo Fisher Scientific) for 5 minutes at 37°C with agitation (orbital shaker; 60 RPM), followed by gentle pipette trituration, further incubation for 3 minutes followed by another gentle trituration for 10 times. TrypLE was inactivated by diluting the sample with 1 mL of PBS-EDTA with 0.01% BSA, followed by centrifuging for 3 min at 300 x g. The cells were resuspended in 1 ml of PBS-EDTA with 0.01% BSA, with few pipetting actions, and passed through a 40 µm cell strainer. The cell number was determined using Trypan Blue and manual counting under a light microscope. A suspension of 300.000 cells/mL was prepared from the initial cell suspension using PBS-EDTA with 0.01% BSA.
Encapsulation of cells and further processing were prepared according to the protocol outlined in the Nadia source publication (Davey et al., 2021). In brief, the cell suspension (75,000 cells) was encapsulated with the barcoded beads (150,000 beads) using Nadia Instrument (DO BIO, UK) and disposable microfluidic cartridges, according to the pre-programmed protocol for drop-seq available in the instrument. The droplet beads undergone multiple purification steps using Überstrainer and were washed with 6 x SSC, Tris-EDTA (TE), and Tween solutions to remove the oil remnants from the beads. Subsequently, the beads with captured RNA were reverse transcribed (Maxima RT Fisher Scientific, incubation 30 min RT + 90min 42°C), and the unused barcodes were removed from the beads with exonuclease I (NEB, incubation 45min 37°C), according the Nadia protocol (Davey et al., 2021). For 3’ mRNA profiling, cDNA libraries were amplified using a PCR reaction (2,000 beads per reaction) with Kapa HiFi Mix (Roche) and Smart PCR Primer, employing the previously described protocol (Davey et al., 2021) and (Macosko et al., 2015) with modifications of the PCR cycle extending up to 15 cycles. The amplified cDNA was purified by Ampur XP beads and the concentrations were estimated using Qbit assay. Subsequently, libraries for sequencing (NextSeq550 platform) were prepared using the Nextera XT/Vazyme kits and protocols (Illumina/Vazyme), which was modified after the tagmentation step by adding number of cycles up to 18. For library amplification, custom Index2 (I5) was used to select only tagmentation products that included UMI and BARCODE sequences (Davey et al., 2021) and (Macosko et al., 2015). The library size was adjusted with Ampur XP beads, and its quality and quantity was checked using a quantity control via the Qubit assay and a quality control via the TapeStation System. Sequencing was performed on the Illumina NextSeq550 platform using the NextSeq 500/550 High Output Kit v2.5 (150 Cycles). The sequencing depth was adjusted to around 10,000 reads per cell, which is sufficient for general single-cell characteristics concerning the count of organoids and mouse embryonic brains sequenced.
4.7. Bioinformatic Analysis of scRNA-seq Data for rare cell populations
Raw FASTQ reads collected from NextSeq550 were analyzed by JSEQ_scRNAseq v.2.3.2 bioinformatics pipeline, which was previously described (Kubiś and Figiel, 2023) and is available at https://github.com/jkubis96/JSEQ_scRNAseq. In brief, the pipeline included read quality control (FASTP), annotation to human GRCh38.p13 genome in case of organoids data or mouse GRCm38.p6 genome in case of mouse embryonic brain regions. Further pipeline steps included UMI and BARCODES quality control and repair (UMI-tools, DropSeq), data clustering with markers selection (Seurat), and other additional steps previously described in detail (Kubiś and Figiel, 2023) for final production of high-resolution data including rare cell populations occurring in organoids and embryonic brains. The cell populations were also annotated to their biological functions. Analyses were conducted with default settings included in JSEQ_scRNAseq pipeline and described in Manual available on GitHub and in the publication. The analyses for every single organoid type (HD, control and mosaic) and mouse embryonic brain regions were conducted separately to obtain higher data resolution.
Downstream analysis of obtained cell clusters from JSEQ_scRNAseq results was conducted using a Python script (Supp. 6) with Spyder IDE. The Python scripts were used for joining data from HD vs. normal vs. mosaic organoids and HD vs. normal embryonic mouse brain regions. The analysis included Harmony adjustment, selection of inter-set and intra-set marker genes, visualization of cell populations from different data sets on UMAPs with Harmony adjustment, and visualization of single-cell data composition (pie plot, bar plot).
4.8. Identification of the differentially expressed genes (DEGs) and GSEA processes from pooled cell transcriptomes
The scRNA-seq cell transcriptomes were initially pooled based on genotypes and used to identify the total number of DEGs for HD vs control and mosaic vs control for organoids and YAC128Q/null vs BAC21Q and Hu128Q/21Q vs BAC21Q for embryonic brains. The DEGs were selected by p-value <= 0.05 with Mann-Whitney and minimal log(FC) >= 0.5 and visualized on volcano plot of disease vs. healthy data sets. The identified DEGs were then used for Gene Set Enrichment Analysis (GSEA) performed with GEDSpy (default settings) python library available at https://github.com/jkubis96/GEDSpy with Spyder IDE. The gene enrichment was conducted for GO-TERMs and PATHWAYS (REACTOME, KEGG). The GSEA was performed based on gene overrepresentation in biological terms with Fisher’s Exact test and adjusted by Benjamini and Hochberg FDR correction. In the analysis, we selected only terms with adjusted pval <= 0.05. The GSEA was performed separately for the upregulated pool and downregulated pool of genes. Subsequently, the terms identified by GSEA were analyzed to select the terms that differentially occur between downregulated vs upregulated gene pools by calculating a specific “term fold change”. The “term fold change” is the term normalized by gene occurrence in upregulated or downregulated gene pool calculated as [number of genes related to terms divided by the number of genes inputted to analysis]. We set the term fold change cut off value as FC >= 1.5. The biological terms showing the value of term fold change were then manually selected to only terms related to altered development and assigned to arbitrary groups of terms. The terms with the suitable term fold change and related to development identified from DEGs from organoids (HD, control) were then used as the core terms in other analyses in mosaic organoids and embryonic brains (YAC128Q/null, Hu128Q/21Q, BAC21Q/null). The entire GEDSpy analysis output is available in supplementary data.
4.9. RNA Isolation and RT-qPCR
After medium removal, HD forebrain organoids were washed once with PBS, transferred to 1,5 mL Eppendorf tubes, and then covered with 1 mL of TRIzol reagent (Thermo Fisher Scientific) according to the manufacturer’s instruction and frozen in -80°C. Upon thaw, total RNA isolation was performed according to the manufacturer’s protocol with chloroform, isopropanol, 75% ethanol, and RNase-free water. Reverse transcription was performed using Maxima H Minus Reverse Transcriptase (Thermo Fisher) (200U per reaction) on 1 µg of RNA in 20 µl of total reaction according to the manufacturer’s protocol. For priming a mixture of random hexamers (25 pmol) and oligo(dT) 18 (25 pmol) was used. Additionally RiboLock RNase inhibitor was added to the reaction mix (20U). Before adding the enzyme and the inhibitor, templates were denatured in 65°C for 5 minutes; after mixing all reaction reagents, the reaction was incubated for 10 min at 25°C followed by 15 min at 50°C. The resulting cDNA was further diluted to a final concentration of 10 ng/uL and stored in -20°C. qPCR was performed on CFX96 Real-Time System (Bio-Rad) using HOT FIREPol® EvaGreen® qPCR Mix Plus (ROX) (SOLIS BIODYNE) on 1 µl of cDNA in 10 µl of total reaction volume. The reaction mix included 250/125 nM primers. All primers are listed in Table 1. Thermocycling parameters were as follows: 15 min of initial denaturation at 95°C and 39 two-step cycles with 15 s denaturation at 95°C and 60 s annealing at 60°C, followed by 5 s melt curve from 65 to 95°C, increment 0,5°C. The reaction was run on CFX96 instrument (Bio-Rad). The data were analyzed, and Cq values were determined using a regression model by CFX Manager 3.1 (Bio-Rad). ACTβ was used as a reference gene.
4.10. Immunohistochemistry
After medium removal, organoids were washed with PBS and fixed in cold 4% (vol/vol) paraformaldehyde (PFA) overnight at 4°C. Then, organoids were stained with Ponceau S for 20 minutes, incubated in 30% sucrose solution for 24-48 hours to protect from tissue damage during freezing, and flash-frozen in OCT CompoundTM using dry ice. Flash-frozen organoids were stored in -80°C. Organoids were sectioned using Microtome cryostat (Leica) at -20°C. 25 μm sections were collected at adhesive Superfrost PlusTM glass slides and air-dried at room temperature overnight. Then, the sections were blocked in 4% Normal Goat Serum (NGS) in TBS-T buffer for 1 h at room temperature. Primary antibodies were diluted in solution 4% NGS in TBS-T. All of primary antibodies used in this study are listed in Table 2. Secondary antibodies were used in 1:500 dilution in 4% NGS in TBS-T. Hoechst 33258 was used to visualize nuclei. A series of fluorescent confocal pictures of human brain organoids were acquired using either the TCS SP5 II (Leica Microsystems; Poland) or Opera Phoenix Plus High-Content Screening System with x40 magnification, which required assembling images along the x-, y-, and z-axis. After using the Opera Screening System, the assembly operation was conducted using JIMG v.1.0.0., available at https://github.com/jkubis96/JIMG/tree/v.1.0.0.
4.11. Western Blot Analysis
Fused organoids were harvested from 60-days old cultures. Blood samples were harvested from 6-8-month-old Yac128Q/null and Bac21Q/null mice. All tissues were homogenized in a phosphate buffer (0.1M, pH 7.5) with 2 mM PMSF using pipetting, followed by bath sonication for 10 min. The protein concentration was estimated using a Pierce BCA protein assay kit (Thermo Scientific, Rockford, IL, USA) according to the manufacturer’s instructions. For each analysis, 20 ug of total protein was diluted in a loading buffer containing 2-mercaptoethanol and boiled for 5 min. The proteins were separated by SDS-polyacrylamide gel electrophoresis (4–15% Mini-PROTEAN® TGX™ Precast Protein Gels, BIO-RAD), transferred to nitrocellulose, and stained with Ponceau S solution. The blots were blocked with 5% non-fat milk in PBS/0.05% Tween 20 for 1 h at RT and, subsequently, incubated for 24 h at 4°C with the primary antibodies listed in Table 2. Lamin B1 (Proteintech) protein or bands detected after Ponceau S solution staining were used as a reference for normalization of the TTR monomer (Proteintech) or TTR tetramer (DSHB) levels in blood serum, respectively. For TTR monomer detection, the blots were probed with the respective HRP-conjugated secondary antibody (anti-rabbit or anti-mouse, 1:5000; Jackson ImmunoResearch, Suffolk, UK). The immunoreaction was detected using the ECL substrate (ThermoFisher Scientific, Waltham, MA, USA). For fluorescent Western blot analysis, a secondary antibody directed against mouse IgG, labeled with the Alexa Fluor 488 (Jackson ImmunoResearch; 1:1000) was used.
4.12. Statistics
Statistics was performed using GraphPad Prism 5 (version 5.00) and R v. 4.1.2 in RStudio. In GraphPad, statistical analyses were conducted for Western blots and some qPCR analyses using T-tests when comparing one variable. In R software, the analyses for qPCRs and organoid size were performed. Firstly, the Levene test was calculated to estimate variance. For equal variance, one-way ANOVA was used for multi-group analysis, followed by post-hoc T-tests (qPCR without p-value adjustment; organoid size with Bonferroni adjustment). In case the variance was not equal the, Welch’s ANOVA was used, followed by post-hoc Welch’s T-tests (qPCR without p-value adjustment; organoid size with Bonferroni adjustment).
4.13. Data availability
Data from single-cell sequencing of human fused brain organoids and mouse embryonic brain samples are stored on ZENOD in FASTQ (after QC) and normalized count matrix (after analysis) with metadata formats for organoids [ZENODO link] and embryonic brains [ZENODO link].
Acknowledgements
This work was supported by the grant from the National Science Centre (grant number 2018/31/B/NZ3/03621). We thank Professor R. Kierzek for granting access to real-time quantitative PCR machine. Cell cultures were conducted in Cell and Tissue Culture Laboratory, IBCh, PAS, Poland. Library synthesis was performed in the European Centre for Bioinformatics and Genomics (ECBIG, Poznan).
Additional information
The authors declare no competing or financial interests.
Author contributions
K.Ś-K. generated brain organoids and performed qPCR and immunofluorescence experiments. J.K. processed the embryonic tissue and organoids for scRNAseq, performed scRNAseq library synthesis, designed scRNAseq bioinformatic tools and analyzed the scRNA-seq data; M.F. and J.K. designed the scRNA-seq experiments and collected the embryonic tissue; J.D. and B.K. generated brain organoids, M.S., Ż.K-P and P.P. collected the serum and adult brains; L.H. J.P. M.R. A.S.-C. helped with RNAseq conduction; M.H. and N.C. provided resources, the HD mouse model; K.Ś-K., J.K. and M.F. wrote the article. M.F. was responsible for the concept and obtaining funding.
References
- The Role of Transthyretin in Oligodendrocyte DevelopmentSci Rep 10https://doi.org/10.1038/s41598-020-60699-8
- Interneuron migration from basal forebrain to neocortex: dependence on Dlx genesScience 278:474–476https://doi.org/10.1126/science.278.5337.474
- Stemness in Cancer: Stem Cells, Cancer Stem Cells, and Their MicroenvironmentStem Cells Int 2017https://doi.org/10.1155/2017/5619472
- Fused cerebral organoids model interactions between brain regionsNat Methods 14:743–751https://doi.org/10.1038/nmeth.4304
- Assembly of functionally integrated human forebrain spheroidsNature 545:54–59https://doi.org/10.1038/nature22330
- Deconstructing the functional neuroanatomy of the choroid plexus: an ontogenetic perspective for studying neurodevelopmental and neuropsychiatric disordersMol Psychiatry 27:3573–3582https://doi.org/10.1038/s41380-022-01623-6
- Developmental defects in Huntington’s disease show that axonal growth and microtubule reorganization require NUMA1Neuron 110:36–50https://doi.org/10.1016/j.neuron.2021.10.033
- Cerebrospinal fluid biomarkers for assessing Huntington disease onset and severityBrain Commun 4https://doi.org/10.1093/braincomms/fcac309
- Faulty neuronal determination and cell polarization are reverted by modulating HD early phenotypesProc Natl Acad Sci U S A 115:E762–E771https://doi.org/10.1073/pnas.1715865115
- Roles of LIM kinases in central nervous system function and dysfunctionFEBS Lett 589:3795–3806https://doi.org/10.1016/j.febslet.2015.10.032
- Degeneration of pyramidal projection neurons in Huntington’s disease cortexAnn Neurol 27:200–204https://doi.org/10.1002/ana.410270217
- A flexible microfluidic system for single-cell transcriptome profiling elucidates phased transcriptional regulators of cell cycleSci Rep 11https://doi.org/10.1038/s41598-021-86070-z
- Regulation of mitotic spindle orientation: an integrated viewEMBO Rep 17:1106–1130https://doi.org/10.15252/embr.201642292
- EZ spheres: a stable and expandable culture system for the generation of pre-rosette multipotent stem cells from human ESCs and iPSCsStem Cell Res 10:417–427https://doi.org/10.1016/j.scr.2013.01.009
- Emergence and Developmental Roles of the Cerebrospinal Fluid SystemDev Cell 52:261–275https://doi.org/10.1016/j.devcel.2020.01.027
- Biological and clinical manifestations of juvenile Huntington’s disease: a retrospective analysisLancet Neurol 17:986–993https://doi.org/10.1016/S1474-4422(18)30294-1
- Mitotic spindle: focus on the function of huntingtinInt J Biochem Cell Biol 43:852–856https://doi.org/10.1016/j.biocel.2011.03.009
- Primary Cilia in the Developing and Mature BrainNeuron 82:511–521https://doi.org/10.1016/j.neuron.2014.04.024
- Onset and progression of pathologic atrophy in Huntington disease: a longitudinal MR imaging studyAJNR Am J Neuroradiol 31:1036–1041https://doi.org/10.3174/ajnr.A2018
- Regional and cellular gene expression changes in human Huntington’s disease brainHum Mol Genet 15:965–977https://doi.org/10.1093/hmg/ddl013
- Hidden brain atrophy in ultra-high field MR images in a transgenic rat model of Huntington’s diseaseBrain Disord 6https://doi.org/10.1016/j.dscb.2022.100039
- Temporal relationship between primary and motile ciliogenesis in airway epithelial cellsAm J Respir Cell Mol Biol 43:731–739https://doi.org/10.1165/rcmb.2009-0328OC
- The choroid plexuses and their impact on developmental neurogenesisFront Neurosci 8https://doi.org/10.3389/fnins.2014.00340
- The transcription factor Otx2 regulates choroid plexus development and functionDev Camb Engl 140:1055–1066https://doi.org/10.1242/dev.090860
- Ciliogenesis is regulated by a huntingtin-HAP1-PCM1 pathway and is altered in Huntington diseaseJ Clin Invest 121:4372–4382https://doi.org/10.1172/JCI57552
- Detailed Cerebral Organoid Fusion MethodProtoc Exch https://doi.org/10.1038/protex.2017.064
- Pathological alteration in the choroid plexus of Alzheimer’s disease: implication for new therapy approachesFront Pharmacol 3https://doi.org/10.3389/fphar.2012.00075
- The Cell Subtypes Selection by Genes (CSSG) algorithm for discovering cell populations in high resolutionbioRxiv https://doi.org/10.1101/2023.04.20.537649
- RNA Sequence Analysis of Human Huntington Disease Brain Reveals an Extensive Increase in Inflammatory and Developmental Gene ExpressionPLOS One 10https://doi.org/10.1371/journal.pone.0143563
- Cerebral organoids model human brain development and microcephalyNature 501https://doi.org/10.1038/nature12517
- The Choroid Plexus and Cerebrospinal Fluid: Emerging Roles in Development, Disease, and TherapyJ Neurosci 33:17553–17559https://doi.org/10.1523/JNEUROSCI.3258-13.2013
- Persistently Elevated mTOR Complex 1-S6 Kinase 1 Disrupts DARPP-32-Dependent D1 Dopamine Receptor Signaling and BehaviorsBiol Psychiatry 89:1058–1072https://doi.org/10.1016/j.biopsych.2020.10.012
- The cell cycle in stem cell proliferation, pluripotency and differentiationNat Cell Biol 21:1060–1067https://doi.org/10.1038/s41556-019-0384-4
- Association of Choroid Plexus Enlargement With Cognitive, Inflammatory, and Structural Phenotypes Across the Psychosis SpectrumAm J Psychiatry 176:564–572https://doi.org/10.1176/appi.ajp.2019.18070825
- Spatially heterogeneous choroid plexus transcriptomes encode positional identity and contribute to regional CSF productionJ Neurosci Off J Soc Neurosci 35:4903–4916https://doi.org/10.1523/JNEUROSCI.3081-14.2015
- Development and functions of the choroid plexus-cerebrospinal fluid systemNat Rev Neurosci 16:445–457https://doi.org/10.1038/nrn3921
- Highly Parallel Genome-wide Expression Profiling of Individual Cells Using Nanoliter DropletsCell 161:1202–1214https://doi.org/10.1016/j.cell.2015.05.002
- FOXG1-Dependent Dysregulation of GABA/Glutamate Neuron Differentiation in Autism Spectrum DisordersCell 162:375–390https://doi.org/10.1016/j.cell.2015.06.034
- Blood-brain-barriers in aging and in Alzheimer’s diseaseMol Neurodegener 8https://doi.org/10.1186/1750-1326-8-38
- Developmental and adult expression patterns of the G-protein-coupled receptor GPR88 in the rat: Establishment of a dual nuclear–cytoplasmic localizationJ Comp Neurol 524:2776–2802https://doi.org/10.1002/cne.23991
- Prevalence and architecture of de novo mutations in developmental disordersNature 542:433–438https://doi.org/10.1038/nature21062
- Early development of aberrant synaptic plasticity in a mouse model of Huntington’s diseaseHum Mol Genet 15:1690–1703https://doi.org/10.1093/hmg/ddl092
- Development of the mammalian cortical hem and its derivatives: the choroid plexus, Cajal–Retzius cells and hippocampusOpen Biol 11https://doi.org/10.1098/rsob.210042
- Autism Spectrum Disorder is associated with ventricular enlargement in a Low Birth Weight PopulationJ Pediatr 163:73–78https://doi.org/10.1016/j.jpeds.2012.12.084
- Untangling Cortical Complexity During DevelopmentJ Exp Neurosci 12https://doi.org/10.1177/1179069518759332
- Primary Cilium-Dependent Signaling MechanismsInt J Mol Sci 18https://doi.org/10.3390/ijms18112272
- Roles of Primary Cilia in the Developing BrainFront Cell Neurosci 13https://doi.org/10.3389/fncel.2019.00218
- Clinical and biomarker changes in premanifest Huntington disease show trial feasibility: a decade of the PREDICT-HD studyFront Aging Neurosci 6
- Human CNS barrier-forming organoids with cerebrospinal fluid productionScience https://doi.org/10.1126/science.aaz5626
- Factors related to genetic testing in adults at risk for Huntington disease: the prospective Huntington at-risk observational study (PHAROS)Clin Genet 91:824–831https://doi.org/10.1111/cge.12893
- Managing juvenile Huntington’s diseaseNeurodegener Dis Manag 3https://doi.org/10.2217/nmt.13.18
- Dysregulation of Corticostriatal Connectivity in Huntington’s Disease: A&Role for Dopamine ModulationJ Huntingt Dis 5:303–331https://doi.org/10.3233/JHD-160221
- The choroid plexus-cerebrospinal fluid system: from development to agingCurr Top Dev Biol 71:1–52https://doi.org/10.1016/S0070-2153(05)71001-2
- Differential loss of striatal projection neurons in Huntington diseaseProc Natl Acad Sci U S A 85:5733–5737https://doi.org/10.1073/pnas.85.15.5733
- Alzheimer’s Disease Neuroimaging Initiative biomarkers as quantitative phenotypes: Genetics core aims, progress, and plansAlzheimers Dement J Alzheimers Assoc 6:265–273https://doi.org/10.1016/j.jalz.2010.03.013
- Abnormal Motor Cortex Excitability in Preclinical and Very Early Huntington’s DiseaseBiol Psychiatry, Stress, Aging, and Alzheimer’s Disease 65:959–965https://doi.org/10.1016/j.biopsych.2008.12.026
- A novel humanized mouse model of Huntington disease for preclinical development of therapeutics targeting mutant huntingtin allelesHum Mol Genet 26:1115–1132https://doi.org/10.1093/hmg/ddx021
- Sustained choroid plexus function in human elderly and Alzheimer’s disease patientsFluids Barriers CNS 10https://doi.org/10.1186/2045-8118-10-28
- Diseases in a dish: modeling human genetic disorders using induced pluripotent cellsNat Med 17:1570–1576https://doi.org/10.1038/nm.2504
- Huntington’s disease - neuropathologyHandb Clin Neurol 100:83–100https://doi.org/10.1016/B978-0-444-52014-2.00004-5
- Probing disrupted neurodevelopment in autism using human stem cell-derived neurons and organoids: An outlook into future diagnostics and drug developmentDev Dyn 249:6–33https://doi.org/10.1002/dvdy.100
Article and author information
Author information
Version history
- Sent for peer review:
- Preprint posted:
- Reviewed Preprint version 1:
Copyright
© 2025, Świtońska-Kurkowska et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.
Metrics
- views
- 115
- downloads
- 4
- citations
- 0
Views, downloads and citations are aggregated across all versions of this paper published by eLife.