The RNA and protein landscapes of mouse brain organoids

This study demonstrates that three-week-old mouse brain organoids serve as a rapidly maturing and relevant model of neonatal brain development, exhibiting transcriptomic and proteomic profiles, including complex synaptic networks and RNA processing events, that closely mirror those of an in vivo neonatal mouse brain.

The Gist

Imagine you want to study how a human brain grows and works, but you can't just peek inside a living person's skull. For a long time, scientists have been trying to build tiny, 3D "mini-brains" in a dish using stem cells. These are called organoids.

While scientists have been great at building human mini-brains, they take a long time to grow (months) and are hard to make in large numbers. This is where mouse brain organoids come in. They are like the "fast-food" version of brain research: they grow incredibly fast (in just three weeks) and are easier to produce. But, until now, nobody was 100% sure if these fast-growing mouse mini-brains were actually "real" or just a messy imitation.

This paper is like a massive quality control inspection. The researchers asked: "Do these mouse mini-brains actually look and act like a real newborn mouse brain, or are they just a fake?"

Here is what they found, explained with some everyday analogies:

1. The "Fast-Forward" Button

The researchers grew mouse stem cells into brain organoids and checked them at three stages: Day 7, Day 14, and Day 21.

• The Finding: By Day 21, the mini-brain's "instruction manual" (its genetic code) looked almost identical to the instruction manual of a real newborn mouse brain.
• The Analogy: Imagine you are teaching a robot to drive. You start with a robot that knows nothing (the stem cell). By week 3, the robot isn't just driving; it's driving exactly like a newborn baby driver, knowing all the same traffic rules and road signs as a real human baby.

2. The "Spelling Bee" of Genes (Alternative Splicing)

Genes aren't just simple on/off switches. Often, the cell takes a gene and "edits" it, like a director cutting scenes out of a movie script to make different versions for different characters. This is called alternative splicing.

• The Finding: The mouse organoids didn't just turn genes on; they edited the genetic scripts in the exact same complex ways a real brain does. They skipped the right "exons" (parts of the script) to create the right proteins for neurons.
• The Analogy: It's like a baker who doesn't just bake a cake; they know exactly how to swap out ingredients to make a chocolate cake, a vanilla cake, or a strawberry cake depending on who is eating it. The mini-brain knew the recipe perfectly.

3. The "Tail" Length (Alternative Polyadenylation)

Genes also have "tails" at the end that control how long the message lasts and how strong it is. Sometimes, the cell makes the tail longer or shorter.

• The Finding: The mini-brains were also stretching and shrinking these genetic tails just like a real brain does.
• The Analogy: Think of a radio signal. Sometimes you need a short, punchy signal for a quick message; other times, you need a long, sustained signal to keep a song playing. The mini-brain knew exactly when to switch between short and long signals, just like a real brain.

4. The "Construction Site" (Proteins)

RNA is the blueprint, but proteins are the actual bricks and workers that build the house. Sometimes, a blueprint says "build a wall," but the construction site is empty.

• The Finding: The researchers checked the actual proteins in the mini-brains. They found that when the blueprint changed, the construction site changed right along with it. The mini-brains built a complex network of synapses (the connections between brain cells) and even had the right receptors for neurotransmitters (the chemicals that let brain cells talk to each other, like Glutamate and GABA).
• The Analogy: It's not enough to have a drawing of a house; you need the actual bricks. This study proved that the mouse mini-brain didn't just have the drawings; it actually built the house, complete with working doorbells (receptors) and wiring (synapses).

5. What Was Missing? (The "Gaps")

The study wasn't perfect. The mini-brains were missing a few things that a real brain has:

• Blood vessels: No plumbing.
• Microglia: No "security guards" (immune cells) to clean up trash.
• The Analogy: It's like a fully furnished house that has no electricity, no water pipes, and no security system. It looks like a house and has the furniture, but it's not quite a fully functional home yet.

The Big Takeaway

This paper is a huge thumbs-up for using mouse brain organoids.

• Why it matters: Because they grow so fast (3 weeks vs. months for humans), scientists can now run hundreds of experiments to test drugs for brain diseases, study how the brain develops, and understand genetic disorders much faster than before.
• The Verdict: Even though they are missing a few "plumbing" parts, mouse brain organoids are a highly accurate, rapidly maturing model of a newborn brain. They are a powerful new tool that allows scientists to peek into the "construction phase" of the brain without waiting years for the building to finish.

Technical Summary

1. Problem Statement

While mouse brain organoids are increasingly used as scalable models for neurodevelopmental screening due to their rapid differentiation (3 weeks vs. months for human organoids), they remain poorly defined at the molecular level. Unlike human brain organoids, which have well-characterized transcriptomic and proteomic profiles, mouse brain organoids lack a comprehensive multi-omics characterization. Specifically, it is unknown whether they accurately recapitulate complex gene regulatory mechanisms critical for neural identity, such as alternative splicing (AS) and alternative polyadenylation (APA), or whether changes in mRNA levels effectively translate to the protein landscape, particularly regarding synaptic machinery.

2. Methodology

The study employed a multi-omics approach to characterize mouse brain organoids at three developmental stages (Days 7, 14, and 21) and compared them against mouse embryonic stem cells (mESCs) and neonatal mouse brains (NBB).

• Organoid Generation: Mouse hybrid ESCs (C57BL/6J x JF1) were differentiated into brain organoids using a modified Sasai protocol. This involved aggregating ESCs in ultra-low adhesion wells with neural induction media (DMH1, IWP-2) followed by maturation in N2/B27 media without Vitamin A.
• Transcriptomics (RNA-seq):
  • Samples: mESCs, D7/D14/D21 organoids, and NBB.
  • Analysis: Differential gene expression (DESeq2), Principal Component Analysis (PCA), and Gene Ontology (GO) enrichment.
  • Regulatory Mechanisms:
    • Alternative Splicing: Analyzed using rMATS to detect exon skipping events (inclusion/exclusion).
    • Alternative Polyadenylation (APA): Analyzed using APAlyzer to detect 3'UTR lengthening or shortening.
• Proteomics (LC-MS/MS):
  • Samples: mESCs (n=4), D7 (n=3), D14 (n=4), D21 (n=4).
  • Technique: Trypsin digestion followed by Liquid Chromatography coupled to tandem Mass Spectrometry (Orbitrap Exploris 480).
  • Analysis: Identification of ~5,900 proteins. Quantification via iBAQ (intensity-based absolute quantification) to determine relative protein abundance.
  • Correlation: Pearson correlation analysis between log2-fold changes in RNA and protein levels across developmental transitions.
• Validation: Immunofluorescence and cryosectioning were used to validate protein expression (e.g., REELIN, GRM5) and cell identity markers.

3. Key Contributions

This study provides the first comprehensive transcriptomic and proteomic atlas of mouse brain organoids. Its primary contributions include:

1. Validation of the Model: Establishing that 3-week-old mouse brain organoids closely mimic the transcriptome and proteome of a neonatal mouse brain.
2. Gene Regulation Fidelity: Demonstrating that organoids robustly reproduce complex, developmentally regulated gene processing events (AS and APA) that define neural identity.
3. Proteome-RNA Concordance: Showing a high correlation between mRNA and protein dynamics, confirming that transcriptional changes are effectively translated into the proteome during organoid maturation.
4. Synaptic Maturity: Identifying a complex network of synaptic proteins, including ionotropic and metabotropic receptors for glutamate and GABA, present as early as Day 21.

4. Key Results

Transcriptomic Landscape

• Developmental Trajectory: PCA showed a progressive shift from mESCs toward the NBB profile. D21 organoids clustered closely with NBB samples.
• Gene Expression: D21 organoids shared 69.5% of upregulated and 61.5% of downregulated genes with the NBB (relative to mESCs). Enriched GO terms included "glutamatergic synapse," "dendrite," and "axon."
• Limitations: Organoids lacked non-neural cell types present in the NBB (e.g., blood vessels, microglia, bone cells), evidenced by the absence of angiogenesis and inflammatory response terms.
• Alternative Splicing: Organoids recapitulated 77% of exon skipping events found in the NBB (968/1,258 genes). Events like Map2 exon inclusion/exclusion were identical between organoids and NBB.
• Alternative Polyadenylation: Organoids reproduced 66% of 3'UTR lengthening events found in the NBB. APA events favored 3'UTR lengthening (associated with stability and translation), particularly in genes related to the ubiquitin-proteasome pathway and synaptic proteins (e.g., Elavl1, Homer2).

Proteomic Landscape

• Maturation Stages: Proteomic transitions were sharpest between mESC→D7 and D7→D14. The D14 proteome showed a sharp transition toward maturity with enriched terms like "chemical synaptic transmission."
• Cell Identity: Protein markers appeared in a temporal order consistent with in vivo corticogenesis: PAX6 (D7) → EOMES/TBR2 (D14) → Layer-specific markers (TBR1, REELIN, BCL11B, CUX1) by D21.
• Synaptic Proteins: D21 organoids contained abundant synaptic proteins, including Synapsins, Glutamate receptors (GRM5, GRIA2), and GABA receptors (GABBR1, GABRA3).
• Abundance: iBAQ analysis confirmed that radial glia markers (FABP7) and neuronal cytoskeletal proteins (TUBB3, MAP2) were highly abundant, while oligodendrocyte markers were less abundant, and microglia/blood vessel markers were absent.

RNA-Protein Correlation

• A strong positive correlation (Pearson r > 0.65) was observed between RNA and protein fold-changes across all developmental transitions (ESC→D7, D7→D14, D14→D21). This indicates that the transcriptome is a reliable predictor of the proteome in this model.

5. Significance

• Model Validity: The study confirms that mouse brain organoids are a rapidly maturing, relevant model for studying mammalian neurodevelopment in a neonatal-like environment. They successfully recapitulate not just cell types, but the intricate gene regulatory logic (splicing and polyadenylation) required for neural identity.
• Screening Utility: Due to their short generation time (3 weeks) and genetic tractability (mouse), these organoids offer a scalable alternative to human organoids for high-throughput screening of neurodevelopmental disorders and drug testing.
• Mechanistic Insight: The finding that organoids reproduce specific APA and AS events suggests they are suitable for studying the post-transcriptional regulation of neurodevelopmental diseases.
• Limitations & Future Directions: The model currently lacks non-neural components (microglia, vasculature). Future work may require assembloids or co-culture systems to model neuro-immune or neuro-vascular interactions.

In conclusion, this paper provides compelling evidence that mouse brain organoids are not merely aggregates of neurons but are sophisticated models that faithfully reproduce the molecular and functional landscapes of the developing mammalian brain.