The zinc metalloprotease ZMPSTE24 binds a distinct topological isoform of the tail-anchored protein IFITM3

This study reveals that the zinc metalloprotease ZMPSTE24 interacts with and facilitates the quality control of a normally transient, inverted topological isoform of the antiviral protein IFITM3, which is characterized by a cytosolic C-terminus and hypo-palmitoylation.

The Gist

The Story: The Cell's "Quality Control Inspector" and the "Flipped" Protein

Imagine your cell is a bustling factory. Inside this factory, there are tiny, specialized workers called proteins that do specific jobs. One of these workers is a protein called IFITM3. Think of IFITM3 as a security guard stationed at the factory's front door (the cell membrane). Its job is to stop viruses from sneaking in by stiffening the door so they can't break through.

To do its job correctly, IFITM3 needs to be installed in the door in a very specific way:

• The "Handle" (N-terminus): Must stick out into the factory floor (the cytoplasm) so the cell can control it.
• The "Lock" (C-terminus): Must stick out into the hallway outside (the extracellular space) to block the virus.

This is called the Ncyto-Cexo orientation. It's like a door handle that you pull from the inside to lock the outside.

The Problem: The "Flipped" Guard

However, sometimes things go wrong during the assembly line. Occasionally, a security guard (IFITM3) gets installed upside down.

• The handle is now outside.
• The lock is now inside the factory.

This is the Ncyto-Ccyto orientation (or "flipped" topology). A guard installed this way is useless; it can't lock the door, and it might even cause chaos. Usually, the factory has a "trash crew" (the Ubiquitin-Proteasome system) that quickly finds these flipped guards, drags them out, and throws them in the garbage bin before they can cause trouble. Because they are destroyed so fast, scientists rarely see them.

The Detective: The "Trap" Mutant

Enter the main character of this study: ZMPSTE24. Think of ZMPSTE24 as a senior quality control inspector who walks the factory floor. Its job is to check proteins and make sure they are installed correctly.

The scientists in this study created a special version of this inspector called ZMPSTE24E336A.

• Normal Inspector: Checks a protein, fixes it if it's wrong, or sends it to the trash if it's broken. It works fast and moves on.
• Trap Mutant (E336A): This inspector has a broken "release mechanism." It grabs onto a protein, checks it, but cannot let go. It's like a magnet that gets stuck to a metal object.

The Discovery: Catching the Ghost

When the scientists used the "Trap Mutant" inspector, something amazing happened. They found that the inspector was grabbing onto a specific version of the security guard (IFITM3) that no one had really seen before.

Here is what they learned about this "caught" guard:

1. It was Flipped: The guard was indeed installed upside down (Ncyto-Ccyto). The inspector had caught it in the act of being wrong.
2. It was "Naked": Normal guards wear a special oily coat (palmitoylation) that helps them stick to the door. The flipped guard caught by the inspector was missing most of this coat (hypo-palmitoylated). It was like a guard without a uniform, making it easy to spot and grab.
3. It was Temporary: When the scientists turned off the "trash crew" (using drugs to stop the proteasome), they could see the flipped guards even without the trap mutant. This proved that these flipped guards exist naturally but are usually destroyed so quickly that they are invisible.

The Big Picture: What Does This Mean?

The paper suggests a new theory about how the cell handles mistakes:

1. The Mistake Happens: Sometimes, the factory accidentally installs a guard upside down.
2. The Inspector Intervenes: ZMPSTE24 (the inspector) finds this flipped guard.
   • Possibility A: It tries to flip the guard back to the correct position (a "rescue").
   • Possibility B: If it can't fix it, it marks the guard for the trash crew to destroy.
3. The Trap Reveals the Truth: By using the "Trap Mutant," the scientists essentially froze time. They caught the inspector holding the flipped guard, proving that this "flipped" state is a real, natural part of the cell's quality control process.

Why Should We Care?

IFITM3 is a crucial part of our immune system. If the cell can't manage these flipped guards properly, our defenses against viruses might be weaker. This study shows that ZMPSTE24 is a key player in keeping our viral defenses in top shape. It's not just a trash collector; it's a topological mechanic ensuring that our cellular "security guards" are facing the right way.

In a nutshell: The cell has a quality control inspector (ZMPSTE24) that catches "flipped" security guards (IFITM3) before they can cause trouble. By using a "sticky" version of the inspector, scientists finally caught a glimpse of these rare, upside-down guards, revealing a hidden layer of how our cells maintain their defenses against viruses.

Technical Summary

1. Problem Statement

Integral membrane proteins, particularly tail-anchored (TA) proteins like IFITM3, are critical for cellular function and antiviral defense. IFITM3 typically adopts a specific topology: a cytosolic N-terminus and a lumenal/exocellular C-terminus ($N_{cyto}-C_{exo}$). While the zinc metalloprotease ZMPSTE24 (and its yeast homolog Ste24) is known for processing prelamin A and yeast a-factor, recent studies identified an interaction between ZMPSTE24 and IFITM3. However, the functional significance of this interaction remained unclear. Specifically, it was unknown whether ZMPSTE24 plays a role in the biogenesis, quality control, or topological regulation of IFITM3, and if aberrant topological isoforms of IFITM3 exist and are regulated by cellular machinery.

2. Methodology

The authors employed a multi-faceted approach combining biochemistry, cell biology, and genetic engineering:

• Co-Immunoprecipitation (Co-IP): Used to assess physical interactions between ZMPSTE24 variants (wild-type and catalytic mutants) and IFITM3.
• Catalytic Trapping Mutants: Utilized the catalytically inactive mutant ZMPSTE24$^{E336A}$ (disrupting the HExxH...E active site) to "trap" transient interactions that wild-type enzymes might process and release.
• Palmitoylation Analysis: Employed metabolic labeling with the alkyne-fatty acid analog 17-ODYA followed by click chemistry to determine the palmitoylation status of IFITM3 species bound to ZMPSTE24.
• Disulfide Crosslinking: Analyzed non-reducing SDS-PAGE to detect intermolecular disulfide bonds between ZMPSTE24 and IFITM3, followed by site-directed mutagenesis of cysteine residues to map the interaction interface.
• Split-Fluorescence Topology Reporter: Used a system developed by McKenna et al. where cells express cytosolic GFP(1-10) and ER-luminal mCherry(1-10). IFITM3 constructs were tagged with the $\beta11$ strand at either the N- or C-terminus. Fluorescence (GFP or mCherry) indicates the subcellular location of the tagged terminus.
• Glycosylation Tagging: Inserted an opsin glycosylation tag at the N- or C-terminus of IFITM3 to test for ER lumenal exposure, followed by PNGase F and Endo H treatment.
• Pharmacological Inhibition: Treated cells with inhibitors of the ubiquitin-proteasome system (E1 inhibitor, p97/VCP inhibitor, proteasome inhibitor) to test the stability of topological isoforms.

3. Key Contributions and Results

A. ZMPSTE24$^{E336A}$ Traps a Specific IFITM3 Isoform

• The catalytic mutant ZMPSTE24$^{E336A}$ binds IFITM3 significantly more efficiently than wild-type ZMPSTE24 or other active-site mutants (e.g., H335A, H339A, E415A).
• Crucially, the IFITM3 species trapped by ZMPSTE24$^{E336A}$ migrates faster on SDS-PAGE and is hypo-palmitoylated compared to the IFITM3 found in homo-oligomers.

B. Discovery of an Inverted Topological Isoform ($N_{cyto}-C_{cyto}$)

• Disulfide Mapping: A high-molecular-weight disulfide-linked complex formed between ZMPSTE24$^{E336A}$ and IFITM3 in vitro. Mapping revealed that Cys105 in IFITM3 (predicted to be cytosolic in the canonical topology) forms a bond with Cys359 in ZMPSTE24 (located in the ER lumen).
• Topological Implication: For Cys105 (IFITM3) and Cys359 (ZMPSTE24) to be in proximity within the ER lumen, the IFITM3 transmembrane span must be inverted, placing its C-terminus in the cytosol ($N_{cyto}-C_{cyto}$).
• Split-Fluorescence Confirmation: Using the split-fluorescence reporter, the authors demonstrated that in the presence of ZMPSTE24$^{E336A}$, a significant subpopulation of IFITM3 exhibits a cytosolic C-terminus (GFP signal), whereas wild-type cells show the expected lumenal C-terminus (mCherry signal).
• Biosynthetic Origin: Glycosylation assays showed that the C-terminus of the trapped IFITM3 species was exposed to the ER lumen during biosynthesis (sensitive to PNGase F), suggesting the protein was initially inserted correctly ($N_{cyto}-C_{exo}$) and subsequently underwent a post-insertional transmembrane span inversion.

C. Role of the Ubiquitin-Proteasome System (ERAD)

• In the absence of ZMPSTE24$^{E336A}$, the inverted $N_{cyto}-C_{cyto}$ isoform is normally transient and undetectable.
• However, treatment with inhibitors of ubiquitination (MLN4273), ER extraction (NMS-873/p97), or the proteasome (bortezomib) leads to the accumulation of the inverted IFITM3 isoform.
• This indicates that the inverted isoform is normally recognized as aberrant and rapidly degraded via the ER-associated degradation (ERAD) pathway.

D. Specificity of the Interaction

• The interaction requires specific structural features: Zinc coordination is necessary (mutants disrupting Zn$^{2+}$ binding fail to trap IFITM3), but catalytic activity is not.
• The IFITM3 mutant C71A fails to bind ZMPSTE24 and does not show the inverted topology, suggesting C71 is critical for the initial interaction or stability of the complex.

4. Significance

• Mechanism of Protein Quality Control: The study proposes a novel quality control model where ZMPSTE24 acts as a sensor or "trap" for misfolded or topologically aberrant membrane proteins. It may either attempt to correct the topology (re-inversion) or facilitate the removal of the aberrant isoform via ERAD.
• Plasticity of Membrane Topology: The findings challenge the dogma that transmembrane topology is fixed after insertion, providing evidence for post-translational inversion of tail-anchored proteins in mammalian cells.
• Antiviral Defense: Since IFITM3 is a critical antiviral factor and ZMPSTE24 is required for its full antiviral activity, this study suggests that ZMPSTE24 ensures the functional integrity of IFITM3 by managing its topological landscape. Failure to regulate these isoforms could compromise viral defense mechanisms.
• Broader Implications: The use of catalytic-dead "trap" mutants combined with topology reporters offers a powerful strategy for identifying transient interactions and quality control substrates for other membrane proteases and chaperones.

Conclusion

The paper demonstrates that ZMPSTE24 interacts with a transient, hypo-palmitoylated, and topologically inverted ($N_{cyto}-C_{cyto}$) isoform of IFITM3. This isoform is normally targeted for degradation by the ubiquitin-proteasome system. ZMPSTE24 likely functions as a quality control factor that either corrects this inversion or sequesters the aberrant protein for removal, thereby maintaining the functional pool of IFITM3 required for antiviral immunity.