Single Cell Transcriptomics and Surface Protein Expression Reveal Distinct Cellular and Molecular Phenotypes in Human RPESC-RPE and PSC-RPE

This study utilizes CITE-Seq to demonstrate that while both adult RPESC-derived and PSC-derived RPE cells express key markers, they exhibit distinct molecular and surface protein profiles—specifically higher mature retinal function markers and CD24 in RPESC-RPE versus stem cell-related genes and CD57 in PSC-RPE—suggesting significant differences in their functional and immunomodulatory properties that could influence transplantation outcomes for AMD.

The Gist

The Big Picture: Fixing a Broken Camera Lens

Imagine your eye is like a high-end camera. The Retinal Pigment Epithelium (RPE) is the "lens cleaner" and "battery charger" sitting right behind the film (your photoreceptors). Its job is to clean up debris, feed nutrients, and keep the vision sharp.

In diseases like Age-Related Macular Degeneration (AMD), this "lens cleaner" dies off. Without it, the film rots, and you go blind. To fix this, scientists are trying to transplant new RPE cells into the eye.

Currently, there are two main ways to make these new cells:

1. The "Adult" Method (RPESC): Taking stem cells from an adult human eye and turning them into RPE cells.
2. The "Factory" Method (PSC): Taking pluripotent stem cells (the "master" cells that can become anything) from an embryo or skin cell and programming them to become RPE cells.

The Question: Are these two types of "new lens cleaners" actually the same? Or are they different enough that one might work better than the other?

The Study: A High-Definition ID Check

The researchers decided to put these two cell types through a rigorous "ID check" using a super-powerful technology called CITE-Seq.

Think of CITE-Seq as a 2-in-1 security scanner.

• Scanner A (Transcriptomics): Looks at the cell's "instruction manual" (its RNA) to see what it thinks it is.
• Scanner B (Surface Proteins): Looks at the cell's "uniform" (the proteins on its surface) to see what it looks like to the outside world.

They scanned thousands of cells from both the "Adult" group and the "Factory" group to see if they were truly identical twins or just distant cousins.

The Findings: Same Job, Different Personalities

Here is what they discovered, broken down simply:

1. The Uniforms Look Similar (But Not Identical)

At first glance, both groups looked like good RPE cells. They both had the right "badges" (markers) saying, "I am an RPE cell, I can clean and feed." They both formed nice, hexagonal honeycomb patterns, which is how healthy RPE cells sit together.

However, when they looked closer, they saw subtle differences in their "uniforms" (surface proteins).

• The "Adult" Cells wore a badge called CD24. Think of this as a "Do Not Eat Me" sign. It tells the body's immune system (the security guards), "Hey, I'm friendly, don't attack me."
• The "Factory" Cells wore a badge called CD57. This is a different kind of "peacekeeper" badge that also helps them avoid being attacked, but it works in a different way.

2. The Instruction Manuals Tell Different Stories

When they read the "instruction manuals" (the genes) inside the cells, the differences were clearer:

• The "Adult" Cells (RPESC): Their manuals were full of instructions for mature, experienced workers. They were very good at the specific, hard jobs of the eye: recycling light chemicals, managing energy, and handling stress. They felt like veteran employees who have been on the job for years.
• The "Factory" Cells (PSC): Their manuals were full of instructions for growth and development. They were still acting a bit like interns or apprentices. They were focused on dividing, growing, and figuring out their identity. They were more "plastic" (changeable) but less "finished."

3. The Danger of "Wandering"

One major concern with cell transplants is that the new cells might get confused, change their shape, and start wandering around the eye, forming scar tissue (called an Epiretinal Membrane).

• The study found that the "Factory" cells (PSC) had a higher tendency to act like wanderers. Their instructions suggested they were more likely to undergo a transformation that could lead to scar tissue.
• The "Adult" cells (RPESC) seemed more settled and less likely to wander off.

4. The Glue Factor

For a transplant to work, the new cells need to stick to the back of the eye (the Bruch's membrane) like glue.

• The "Adult" cells had a special type of "glue" (a protein called ITGA1) that fits perfectly with the back of the eye.
• The "Factory" cells had a different type of glue (ITGA2) that doesn't stick quite as well to that specific surface.

The Takeaway: Why This Matters

Imagine you are hiring a team to fix a leaking roof.

• Option A (Adult Cells): You hire a crew of retired roofers who have done this exact job for 40 years. They know the materials, they stick to the roof perfectly, and they know exactly how to handle the weather. They are ready to work immediately.
• Option B (Factory Cells): You hire a crew of talented construction students. They are energetic, they can learn quickly, and they are very flexible. But they might need more training, they might wander off the roof, and they might not stick to the shingles as tightly as the veterans.

The Conclusion:
Both types of cells can do the job, but they have different strengths and weaknesses.

• Adult cells might be better at sticking to the eye and surviving the immune system immediately because they are more "mature."
• Factory cells are easier to make in huge numbers, but they might need extra steps to "mature" them before transplanting to ensure they don't wander or get rejected.

This study gives doctors a roadmap. By knowing exactly which "badges" (proteins) and "instructions" (genes) each cell type has, scientists can now try to tweak the "Factory" cells to make them act more like the "Adult" veterans, or select the best "Adult" cells for the job. This could lead to better, safer vision-restoring surgeries in the future.

Technical Summary

1. Problem Statement

Age-related macular degeneration (AMD) is a leading cause of vision loss, characterized by the dysfunction and loss of the Retinal Pigment Epithelium (RPE). Current therapeutic strategies involve cell replacement using either Pluripotent Stem Cell-derived RPE (PSC-RPE) or adult RPE Stem Cell-derived RPE (RPESC-RPE). While both sources are in clinical trials, there is a lack of comprehensive, high-resolution data comparing their intrinsic biological properties. Specifically, it remains unclear how these two sources differ in terms of:

• Cellular maturity and functional specialization.
• Surface protein profiles (surfaceome) critical for adhesion and immune interaction.
• Potential risks such as Epithelial-to-Mesenchymal Transition (EMT) or immune rejection post-transplantation.

2. Methodology

The study employed a multimodal single-cell analysis approach to compare human RPESC-RPE and PSC-RPE cultures derived from three independent donor lines for each source.

• Cell Culture: Both cell types were cultured under identical conditions for 10 weeks post-thaw (Passage 2) to ensure a standardized comparison.
• CITE-Seq (Cellular Indexing of Transcriptomes and Epitopes by Sequencing):
  • Transcriptomics: Single-cell RNA sequencing (scRNA-seq) was performed to profile gene expression.
  • Proteomics: Simultaneously, a panel of 164 DNA-barcoded antibodies (ADTs) was used to quantify surface protein expression.
  • Data Integration: The study utilized the Seurat package for data processing.
    • Reciprocal PCA (RPCA): Used for initial integration of transcriptomic data.
    • Weighted Nearest Neighbor (WNN) Analysis: Integrated RNA and ADT data to create multimodal clusters.
    • DA-seq: Applied to identify origin-enriched subclusters (PSC-specific vs. RPESC-specific).
• Validation: Key findings were validated using flow cytometry and immunohistochemistry on specific surface markers (CD55, CD24, CD57, ITGA1, TNFRSF14).
• Bioinformatic Analysis: Gene Ontology (GO) enrichment, pathway analysis (ClusterProfiler, SCPA), and module scoring were used to characterize functional states (e.g., metabolism, adhesion, immune response).

3. Key Contributions

• First Multimodal Comparison: This is the first study to simultaneously map the transcriptome and surfaceome of RPESC-RPE and PSC-RPE at single-cell resolution, revealing that surface proteins are more effective than transcriptomics alone at distinguishing cell origin.
• Identification of Distinct Surface Markers: The study identified CD24 as a robust marker for adult RPESC-RPE and CD57 for PSC-RPE, providing tools for potential cell enrichment or quality control.
• Functional Heterogeneity Mapping: The research delineated specific subpopulations within both cell types that exhibit distinct functional states (e.g., progenitor-like vs. mature, high adhesion vs. high migration).

4. Key Results

A. Transcriptomic Similarities and Differences

• Global Similarity: Both sources expressed canonical RPE signature genes (e.g., MITF, RPE65, BEST1) and formed similar transcriptional clusters (12 clusters identified), indicating both are bona fide RPE cells.
• Maturity vs. Plasticity:
  • RPESC-RPE: Enriched for genes related to mature retinal functions, including the visual cycle (retinol metabolism), mitochondrial oxidative phosphorylation, and fatty acid metabolism.
  • PSC-RPE: Enriched for developmental and progenitor-related pathways, including chromatin organization, epithelial morphogenesis, and cell growth.
• Stress Response: RPESC-RPE showed enrichment for "response to starvation," while PSC-RPE showed enrichment for "response to hypoxia."
• EMT Risk: PSC-RPE clusters showed significant enrichment for mesenchymal cell migration and EMT-related genes (e.g., C4orf48, TFRC), suggesting a higher risk of forming epiretinal membranes (ERM) compared to RPESC-RPE.

B. Surface Protein (Surfaceome) Divergence

Multimodal clustering (WNN) revealed better separation of cell sources than RNA alone.

• Adhesion Molecules:
  • RPESC-RPE: Highly expressed ITGA1 (integrin alpha-1), which binds preferentially to Type IV collagen (the major component of Bruch's membrane). This suggests superior potential for engraftment to the native Bruch's membrane.
  • PSC-RPE: Highly expressed ITGA2 (binds Type I collagen) and showed higher "cell-cell adhesion" scores but lower "cell-matrix adhesion" scores.
• Immune Modulation ("Don't Eat Me" Signals):
  • CD24: Highly expressed on RPESC-RPE (~45% of cells). CD24 binds Siglec-10 to inhibit macrophage phagocytosis.
  • CD57: Highly expressed on PSC-RPE (~70% of cells). CD57 is associated with immune suppression and graft tolerance.
  • CD55: Higher expression in RPESC-RPE, aiding in complement regulation.
• Developmental Markers: PSC-RPE clusters expressed higher levels of progenitor markers (PAX6, SOX2, NCAM1), while RPESC-RPE uniquely expressed RAX and TNFRSF14.

C. Subpopulation Analysis

• Cluster 6 (Both Sources): Enriched for cell-substrate adhesion and actin filament organization, suggesting a specialized subpopulation capable of strong engraftment.
• Cluster 11 (PSC-biased): Enriched for cell division and DNA replication, indicating a progenitor-like state.

5. Significance and Implications

• Clinical Translation: The findings suggest that RPESC-RPE may be better suited for immediate functional integration due to its mature metabolic profile and strong adhesion to Bruch's membrane (via ITGA1). Conversely, PSC-RPE may require further maturation or selection to reduce EMT risks and enhance adhesion.
• Immune Evasion: The distinct expression of CD24 (RPESC) and CD57 (PSC) indicates that both cell types possess mechanisms to evade immune rejection, but via different pathways. This is crucial for designing allogeneic therapies.
• Quality Control: The identified surface markers (CD24, CD57, ITGA1, CD55) can be used to develop flow cytometry-based assays to enrich specific, high-quality subpopulations before transplantation, potentially improving clinical trial outcomes.
• Future Directions: The study highlights the need to expand CITE-seq panels to include more RPE-specific markers and to investigate how the AMD microenvironment (hypoxia, inflammation) further modulates these distinct phenotypes post-transplantation.

In conclusion, while both PSC-RPE and RPESC-RPE are viable sources for cell therapy, they possess distinct molecular and functional identities. RPESC-RPE appears more "mature" and adhesive, whereas PSC-RPE retains more "plasticity" and developmental potential, carrying different risks and benefits for retinal regeneration.