Divergent consequences of PSEN1 knockout and PSEN2 knockout in stem cell derived models of the brain

Using CRISPR-Cas9-edited iPSC-derived brain cells, this study demonstrates that while PSEN1 and PSEN2 are both catalytic subunits of the {gamma}-secretase complex, they perform distinct and non-redundant functions, with PSEN2 knockout specifically disrupting the endo-lysosomal system without significantly affecting A{beta} generation or Nicastrin maturation, unlike PSEN1 knockout.

The Gist

Imagine your brain is a bustling, high-tech city. Inside the cells of this city, there are specialized factories called γ-secretase complexes. Their job is to act like molecular scissors, snipping long protein chains into smaller, useful pieces.

Two of the most important "scissor blades" in these factories are proteins named PSEN1 and PSEN2.

For a long time, scientists thought these two blades were almost identical twins—interchangeable parts that could do the same job. If one broke, the other would just pick up the slack. This paper, however, reveals a surprising truth: They are not twins; they are specialists with very different jobs.

Here is the story of what happens when you remove one blade versus the other, explained through a few simple analogies.

The Setup: The "City" of the Brain

The researchers used stem cells to build tiny, living models of human brain cells (neurons) and immune cells (microglia). They used a molecular tool called CRISPR (think of it as "molecular scissors" or a "search-and-destroy" command) to completely remove the instructions for making either PSEN1 or PSEN2.

They then asked: What happens to the city when we take away the PSEN1 scissors? What happens when we take away the PSEN2 scissors?

Scenario 1: Removing PSEN1 (The "Main Street" Scissors)

When the scientists removed PSEN1, the city went into chaos, specifically regarding its main traffic routes.

• The Analogy: Imagine PSEN1 is the master mechanic for the city's main highway (the cell membrane). It's responsible for cutting up a specific protein called APP. When PSEN1 is gone, the highway gets clogged with uncut debris (APP fragments), and the production of a specific waste product called Amyloid-beta (Aβ) drops by about 60%.
• The Result: The factory also stopped working properly. A key quality-control worker named Nicastrin couldn't get fully trained (it didn't mature).
• The Takeaway: PSEN1 is the "heavy lifter" for cutting proteins on the surface of the cell. Without it, the main processing line breaks down.

Scenario 2: Removing PSEN2 (The "Warehouse" Scissors)

When the scientists removed PSEN2, the main highway (APP processing) kept running perfectly fine. The traffic flowed, and the waste products (Aβ) were produced at normal levels.

• The Analogy: Instead of the highway, PSEN2 is the manager of the city's underground storage and recycling center (the endo-lysosomal system). This is where cells store things, break down trash, and recycle materials.
• The Result: Without PSEN2, the recycling center got messy.
  • The "Early Endosome" (the intake dock) got piled up with boxes that never got moved.
  • The "Lysosome" (the trash compactor) became empty and weak.
  • Essentially, the city's garbage collection system was failing, even though the main street was clear.
• The Takeaway: PSEN2 isn't the main cutter for the big proteins; it's the specialist for keeping the cell's internal recycling and waste disposal systems running smoothly.

The Microglia (The City's Police)

The researchers also looked at the brain's immune cells (microglia), which act like the city's police force. They wanted to see if removing PSEN2 would stop the police from doing their job.

• The Finding: Removing PSEN1 stopped the police from processing a specific signal (TREM2), leaving them confused. But removing PSEN2? The police kept working perfectly. They didn't even notice PSEN2 was gone.
• The Takeaway: In the immune cells of the brain, PSEN1 is the boss; PSEN2 is barely needed for their main tasks.

The Big Surprise: They Don't Swap Jobs

The most exciting part of the story is what happened when one blade was missing.

• When PSEN1 was gone, the cell tried to compensate by making more PSEN2. It was like a backup generator kicking in.
• But when PSEN2 was gone, the cell did not make more PSEN1. The backup generator didn't turn on.

This proves that PSEN1 and PSEN2 are not redundant. You can't just swap them out. They have distinct, non-interchangeable roles.

Why Does This Matter? (The "So What?")

For years, scientists tried to create drugs to "turn off" γ-secretase to stop Alzheimer's disease (which is caused by too much Amyloid-beta waste). But these drugs failed because they turned off both blades, causing massive side effects. It was like shutting down the entire power grid to fix one flickering lightbulb.

This paper offers a new roadmap:
Since PSEN1 and PSEN2 do different things, we might be able to design drugs that target only PSEN1 (to stop the Amyloid-beta waste) without messing up PSEN2 (which keeps the cell's recycling center clean).

Summary in a Nutshell

• PSEN1 is the Main Street Cutter: It handles the big, important cuts on the cell surface. If it breaks, the main traffic jams, and Alzheimer's-related waste piles up.
• PSEN2 is the Recycling Manager: It handles the internal trash and recycling. If it breaks, the cell gets messy inside, but the main traffic keeps flowing.
• The Lesson: They are not interchangeable. To cure Alzheimer's safely, we need to fix the "Main Street Cutter" without breaking the "Recycling Manager."

This discovery suggests that future treatments can be much more precise, acting like a sniper rather than a sledgehammer, potentially avoiding the side effects that stopped previous treatments.

Technical Summary

1. Problem Statement

$\gamma$-secretase is a multi-subunit enzyme complex responsible for the intramembrane proteolysis of hundreds of substrates. It consists of obligate subunits (PEN-2, Nicastrin, APH1A/B) and a catalytic subunit that can be either Presenilin 1 (PSEN1) or Presenilin 2 (PSEN2).

• Clinical Context: Mutations in PSEN1 and PSEN2 cause familial Alzheimer's Disease (fAD). Conversely, loss-of-function mutations in PSEN1, PSENEN (PEN-2), and NCSTN (Nicastrin) cause familial hidradenitis suppurativa (acne inversa).
• The Gap: Previous clinical trials targeting $\gamma$-secretase failed due to severe side effects, likely caused by a lack of understanding regarding the functional diversity and redundancy of the different $\gamma$-secretase complexes.
• The Question: Do PSEN1 and PSEN2 perform redundant roles, or do they have distinct, non-overlapping functions in human brain cells? Existing knowledge relies heavily on animal models (where Psen1 knockout is embryonic lethal) or cell lines, necessitating a direct comparison in human-derived models.

2. Methodology

The study utilized CRISPR-Cas9 genome engineering to generate isogenic human induced pluripotent stem cell (iPSC) lines to compare the effects of knocking out PSEN1 versus PSEN2.

• Cell Line Generation:
  • Source: Corrected control iPSCs (previously derived from a patient with the int4del PSEN1 mutation) were used as the base.
  • Editing: PSEN2 knockout (KO) was achieved by targeting constitutive Exon 5 using CRISPR-Cas9 ribonucleoprotein (RNP) complexes.
  • Validation: Clones were screened via Western blot and Sanger sequencing. Biallelic frameshift mutations were confirmed. Off-target effects were ruled out via sequencing of the top 5 predicted sites.
• Differentiation:
  • Neurons: iPSCs were differentiated into cortical glutamatergic neurons (Day 100).
  • Microglia: iPSCs were differentiated into microglia-like cells (iMGLCs).
• Comparative Groups: The study compared four genotypes:
  1. Unedited Parental Control.
  2. Isogenic PSEN1-knockout (from previous work).
  3. Isogenic PSEN2-knockout (newly generated).
  4. Unrelated control line (KOLF2.1J).
• Assays:
  • Protein Analysis: Western blotting for PSEN1/2, Nicastrin maturation, APP cleavage fragments (CTFs), TREM2 processing, and endo-lysosomal markers (EEA1, RAB5, LAMP1).
  • Quantification: ELISA/Meso Scale Discovery for Aβ38, Aβ40, and Aβ42 levels in conditioned media.
  • Transcriptomics: qPCR to assess mRNA levels of target genes.
  • Pharmacology: Treatment with the $\gamma$-secretase inhibitor DAPT to assess residual activity.

3. Key Contributions

• First Direct Human Comparison: This is the first study to directly compare PSEN1 and PSEN2 loss-of-function in isogenic human iPSC-derived neurons and microglia, bypassing species-specific differences found in mouse models.
• Functional Divergence: The study definitively demonstrates that PSEN1 and PSEN2 are not functionally redundant in human brain cells. They regulate distinct biological pathways.
• Subcellular Specificity: The findings link specific catalytic subunits to specific subcellular compartments (PSEN1 to plasma membrane cleavage events; PSEN2 to endo-lysosomal trafficking).

4. Key Results

A. Impact on APP Processing and Aβ Generation (Neurons)

• PSEN1-KO: Confirmed previous findings: significant reduction in mature Nicastrin, accumulation of APP C-terminal fragments (CTFs), and a ~60% reduction in all Aβ species (Aβ38, 40, 42).
• PSEN2-KO:
  • No change in APP processing or CTF accumulation compared to controls.
  • No change in Aβ generation (levels of Aβ38, 40, 42 remained normal).
  • Compensation: PSEN1 levels did not increase in PSEN2-KO neurons, suggesting PSEN1 does not compensate for PSEN2 loss. Conversely, PSEN2 levels increased in PSEN1-KO neurons, suggesting a compensatory upregulation of PSEN2 when PSEN1 is lost.

B. Impact on TREM2 Processing (Microglia)

• PSEN1-KO: Significant accumulation of TREM2 C-terminal fragments, indicating impaired cleavage.
• PSEN2-KO: No alteration in TREM2 processing or maturation. This contradicts the hypothesis that PSEN2 is the primary presenilin for microglial $\gamma$-secretase activity regarding TREM2.

C. Impact on the Endo-Lysosomal System (Neurons)

• PSEN1-KO: No significant changes in steady-state levels of endo-lysosomal markers.
• PSEN2-KO:
  • Significant accumulation of early endosome markers (EEA1, RAB5).
  • Significant reduction in the lysosomal marker LAMP1.
  • Mechanism: qPCR showed mRNA levels of LAMP1 and RAB5 were unchanged, indicating the defect is post-transcriptional (likely affecting protein stability or trafficking).

5. Significance and Implications

• Therapeutic Precision: The results suggest that inhibiting PSEN1 and PSEN2 would have vastly different clinical outcomes.
  • Inhibiting PSEN1 would likely drastically reduce Aβ (potentially beneficial for AD) but would severely disrupt Notch signaling and APP processing (causing toxicity).
  • Inhibiting PSEN2 might spare APP/Aβ processing and Notch signaling but could severely disrupt endo-lysosomal function, potentially leading to neurodegeneration via lysosomal failure.
• Disease Mechanisms:
  • fAD: The earlier onset of PSEN1 mutations compared to PSEN2 mutations aligns with PSEN1's dominant role in APP cleavage.
  • Compensation: The upregulation of PSEN2 in PSEN1-KO models may explain why some PSEN1 mutation carriers have variable phenotypes or why PSEN2 mutations can sometimes lead to longer Aβ peptides (Aβ42:40 ratio shifts).
• Future Directions: The study advocates for subtype-specific targeting of $\gamma$-secretase complexes (e.g., stabilizing or modulating specific complexes rather than global inhibition) to avoid the off-target toxicity that plagued previous clinical trials. It also highlights the need to investigate intracellular Aβ pools, as PSEN2 may play a specific role there despite not affecting secreted Aβ levels in this model.

Conclusion: PSEN1 and PSEN2 perform distinct, non-redundant roles in human brain cells. PSEN1 is critical for plasma membrane cleavage events (APP, TREM2), while PSEN2 is essential for endo-lysosomal homeostasis. This divergence provides a critical roadmap for developing safer, more precise therapies for Alzheimer's disease.