Reverse Proteolysis Uncovers a Hidden Dimension of the Peptidome

This study reveals that lysosomal cysteine cathepsins catalyze reverse proteolysis to generate non-genomically templated fusion peptides, including T1D-associated autoantigens, thereby uncovering a previously unrecognized pathway for peptide diversification and immune activation.

The Gist

The Big Idea: The "Glue" That Was Hiding in the Scissors

For a long time, scientists thought of proteases (enzymes that break down proteins) like a pair of scissors. Their only job was to cut things apart. If you gave them a long protein chain, they would snip it into tiny, useless pieces.

This paper discovers that these "scissors" are actually magic scissors with a hidden glue stick.

Not only can they cut proteins, but under the right conditions, they can also stick pieces back together in new, weird combinations. The authors call this process "Reverse Proteolysis." Instead of just making a mess, these enzymes are acting like molecular architects, building new structures out of the scraps.

The Main Characters: The Cathepsins

The specific "scissors" studied here are called Cathepsins. You can think of them as the janitors inside your cells' recycling centers (called lysosomes). Their normal job is to take old, damaged proteins and chop them up so the cell can recycle the parts.

The researchers found that while these janitors are chopping up trash, they are also accidentally (or perhaps intentionally) gluing random pieces together to create brand new "hybrid" proteins that never existed in your DNA.

The Experiment: How They Found the Glue

The researchers set up a few different scenarios to prove this "gluing" happens:

1. The LEGO Analogy: Imagine you have a box of LEGO bricks (proteins). Usually, a janitor just smashes them into dust. But the researchers found that if you give the janitor a specific type of brick (a peptide), they might smash a big block, pick up a loose piece, and glue it onto another loose piece.

   • The Result: They created "multi-generational" hybrids. It's like taking a red brick, gluing it to a blue one, then gluing that to a green one, creating a long, strange chain that wasn't in the original box.

2. The Viral Mix-Up (Host vs. Virus): This is the most exciting part. The researchers mixed human proteins with SARS-CoV-2 (Coronavirus) proteins.

   • The Discovery: The cathepsins glued a piece of the virus to a piece of a human protein (like insulin).
   • Why it matters: This creates a "Frankenstein" protein. The immune system sees the human part and thinks, "That's me!" but then sees the virus part and thinks, "That's an invader!" This confusion might be why some people develop autoimmune diseases (where the body attacks itself) after a virus infection.

3. The "Citrus" Factor (Citrullination): The paper found that if the proteins are "zested" with a chemical called citrullination (common in conditions like Rheumatoid Arthritis), the janitors get really good at gluing. It's like adding a special lubricant that makes the scissors work faster and stick pieces together more often.

The "CT-TRAP" Tool: Catching the Ghosts

One of the hardest things to prove is that this happens inside a living cell, not just in a test tube. To solve this, the team invented a tool called CT-TRAP.

• The Metaphor: Imagine trying to catch a ghost in a dark room. You can't see it, so you sprinkle glow-in-the-dark glitter on the floor. If the ghost walks through, it gets covered in glitter, and you can see it.
• How it works: They gave cells a special "probe" (the glitter) that only gets glued if the cathepsins are active. They then used a magnet to pull out only the pieces that had the glitter on them.
• The Proof: They found these "glittery" hybrid proteins inside the cells, proving that this gluing process is happening in real life, not just in a lab dish.

Why Should We Care?

This changes how we understand the immune system and diseases like Type 1 Diabetes and Rheumatoid Arthritis.

• The "Fake" Enemy: Our immune system is trained to recognize specific proteins. But if our own enzymes (cathepsins) are constantly gluing random pieces of our own proteins together, they might create new, fake enemies that the immune system has never seen before.
• The Autoimmune Trigger: The immune system might get confused, thinking these new "glued" proteins are foreign invaders and start attacking the body. This could explain why some autoimmune diseases flare up after infections or why they are so hard to predict.

The Takeaway

This paper tells us that our cells are not just passive recycling plants. They are active creators. The enzymes that break things down are also secretly building new things.

This "hidden dimension" of protein creation means that our bodies can generate a vast library of new molecules that our DNA never coded for. While this might be a cool biological trick, it also suggests a new way that diseases like diabetes and arthritis might start: our own janitors are accidentally gluing our body parts together in ways that confuse our immune system.

Technical Summary

1. Problem Statement

Proteases are traditionally viewed solely as degradative enzymes responsible for breaking down proteins into smaller fragments. However, their catalytic machinery theoretically allows for reverse proteolysis (peptide ligation), where peptide bonds are formed rather than broken.

• The Gap: While proteasome-mediated ligation has been observed in the context of MHC Class I, the role of lysosomal cysteine cathepsins in generating cis/transpeptides (fusion peptides formed from two distinct fragments) remains poorly characterized.
• The Skepticism: There is significant doubt in the field regarding whether reported fusion peptides are genuine enzymatic products or artifacts of mass spectrometry, mRNA splicing, or genomic mutations.
• The Biological Question: Can lysosomal cathepsins iteratively cycle between hydrolysis and ligation to generate non-genomically templated peptides (hybrids of host and viral proteins) that could act as neoantigens in autoimmune diseases like Type 1 Diabetes (T1D)?

2. Methodology

The authors employed a multi-pronged approach combining in vitro enzymology, quantitative mass spectrometry, and a novel chemical biology strategy to detect these events in cells.

• In Vitro Enzyme Assays:

  • Substrates: Used human cathepsin S (hCatS) as the primary model, along with CatB, L, K, and V. Substrates included synthetic FRET peptides, biotinylated donor peptides, and full-length proteins (Proinsulin, Insulin, SARS-CoV-2 Spike subunits S1/S2, and human Platelet Factor 4/hPF4).
  • Conditions: Reactions were conducted at varying pH levels (4.5–7.5) to mimic lysosomal and endosomal environments. Post-translational modifications (PTMs), specifically citrullination, were tested to see if they influenced ligation efficiency.
  • Quantification: Used LC-MS/MS with synthetic standards to quantify the ratio of ligation products vs. hydrolysis products.

• Novel Detection Strategy (CT-TRAP):

  • Developed Click-based Targeted Transpeptide Retrieval and Purification (CT-TRAP).
  • Probe Design: A synthetic peptide containing a nucleophilic motif (TTSQGGK), a TAT sequence for lysosomal delivery, a fluorophore (5-FAM), and an azide handle (Lys-N3).
  • Workflow: Cells (RAW 264.7 macrophages) internalize the probe $\rightarrow$ Lysosomal processing occurs $\rightarrow$ Ligation products form $\rightarrow$ Cell lysis $\rightarrow$ CuAAC (Click Chemistry) attaches a biotin tag to the azide $\rightarrow$ Streptavidin enrichment $\rightarrow$ LC-MS/MS detection via diagnostic fragment ions.

• Immunological Validation:

  • Selected fusion peptides were synthesized and tested for binding affinity to HLA-DQ8 (a molecule strongly associated with T1D) using a commercial binding assay (ProImmune REVEAL).

3. Key Contributions

1. Quantification of Reverse Proteolysis: Demonstrated that peptide ligation is not a negligible side reaction but can account for up to 4.5% of proteolytic turnover under specific conditions.
2. Iterative Mechanism: Proved that cathepsins do not just perform single ligation events but catalyze multi-generational cycles of hydrolysis and ligation, creating complex higher-order cis/transpeptides.
3. Novel Detection Platform: Established CT-TRAP as the first method to specifically capture and enrich probe-derived transpeptides within living cells, overcoming previous detection limitations.
4. Biochemical Link to Autoimmunity: Provided direct evidence that cathepsins can generate specific hybrid peptides (e.g., Insulin-IAPP fusions) that match known T1D autoantigens and bind HLA-DQ8 with high affinity.

4. Key Results

• Kinetics and Generational Complexity:

  • hCatS catalyzes the formation of 1st, 2nd, and 3rd-generation cis/transpeptides.
  • While 1st-generation transpeptides are transient, higher-order products (e.g., third-generation) accumulate and show resistance to further hydrolysis.
  • Yield: A first-generation transpeptide reached ~4.5% relative to the main hydrolysis product within 10 minutes.

• Modulation by pH and PTMs:

  • pH: Ligation efficiency increases as pH rises from 4.5 to 6.5–7.5. This is critical because lysosomal alkalinization is a hallmark of autoimmune diseases.
  • Citrullination: Substrates containing citrulline (Arg $\to$ Cit) showed a 2.7-fold increase in catalytic efficiency and a 2–10 fold increase in transpeptide formation compared to non-citrullinated substrates.

• Generation of Disease-Relevant Neoantigens:

  • T1D: hCatS generated the exact proinsulin-IAPP fusion peptide (GQVELGGGNAVEVLK) previously identified as a T1D autoantigen. It also generated novel fusions (e.g., Insulin-Spike) containing anchor residues (Glu at P1 and P9) that bind HLA-DQ8 with 41-fold higher affinity than reference peptides.
  • Viral-Host Hybrids: Successfully generated fusion peptides between human IAPP/PF4 and SARS-CoV-2 Spike proteins (S1 and S2). Notably, a host-viral hybrid (EQDKNNAVEGGK-Biotin) with super-agonist binding potential was identified.
  • Thrombotic Syndromes: Generated cis/transpeptides from hPF4 that overlap with epitopes known to trigger Heparin-Induced Thrombocytopenia (HIT) and Vaccine-Induced Immune Thrombotic Thrombocytopenia (VITT).

• Enzyme Specificity:

  • hCatS and hCatV (immune-dominant cathepsins) showed the highest transpeptidase activity.
  • Ubiquitous cathepsins (CatB, CatL) showed significantly lower yields, suggesting tissue-specific roles in antigen diversification.

• Cellular Validation:

  • Using CT-TRAP in RAW 264.7 macrophages, the team detected probe-derived peptides fused to endogenous proteins (e.g., GAPDH). This fusion was abolished by the cathepsin inhibitor E-64d, confirming the enzymatic nature of the event in a cellular context.

5. Significance and Implications

• Redefining Antigen Processing: This study reframes lysosomal cathepsins from simple "degraders" to "dual-function molecular architects" capable of generating a diverse repertoire of non-genomically templated peptides.
• Mechanism for Autoimmunity: It provides a plausible biochemical mechanism for how molecular mimicry and neoantigen formation occur without genomic mutation. Infection or inflammation (altering pH or inducing citrullination) could trigger cathepsins to stitch together self and non-self (viral) sequences, creating novel epitopes that break immune tolerance.
• Therapeutic Targets: The findings suggest that modulating cathepsin transpeptidase activity (without blocking general proteolysis) could be a novel therapeutic strategy to prevent the generation of pathogenic neoantigens in autoimmune diseases.
• Broader Context: The discovery that pH and PTMs (citrullination) drastically enhance ligation suggests that pathological states (e.g., rheumatoid arthritis, viral infections) create a "perfect storm" for the generation of these hybrid peptides, potentially explaining the epidemiological links between infections/vaccinations and autoimmune onset.