Lipid bilayer thinning near a ubiquitin ligase selects ER membrane proteins for degradation

Why this new work is important 

This study provides a clear model for how cells monitor the quality of their membrane proteins. The work reveals a simple, physics-driven mechanism: rather than being recognized by a specific molecular sensor, faulty proteins are identified through basic energetics and membrane dynamics. It also uncovers a key difference between single-pass and multi-pass membrane proteins.

By introducing the idea that local membrane thinning acts as a functional gateway for protein movement, this study reshapes our understanding of protein quality control and may have broader implications for other membrane systems, such as mitochondria or Golgi.

Background

Many proteins end up in the endoplasmic reticulum (ER), where they are either inserted into the membrane or folded inside the lumen with the help of the translocon. Once there, they face strict quality control: only properly folded and assembled proteins can stay. If a protein fails to fold correctly, it is sent for degradation through the ER-associated degradation (ERAD) pathway, a conserved process across all eukaryotes that prevents toxic buildup of misfolded proteins. During ERAD, defective proteins are tagged with ubiquitin, extracted into the cytosol, and degraded by the proteasome.

For misfolded luminal proteins, the process is relatively well understood. In yeast, it involves the Hrd1 complex, where Hrd1, a ubiquitin ligase, and its partner Der1 help move misfolded polypeptides through a locally thinned region of the ER membrane. However, the mechanism for membrane protein degradation has been less clear, and this study sheds light on that process. The authors show, using budding yeast, that misfolded membrane proteins are selected for degradation through a general mechanism in which they partition into a locally thinned region of the ER membrane next to Hrd1. This happens when hydrophilic residues become exposed in the transmembrane segment.

Key findings

• Hydrophilic residues trigger degradation

Introducing charged or hydrophilic amino acids into the transmembrane (TM) segment of single-pass proteins was sufficient to trigger Hrd1-dependent degradation. This was demonstrated through cycloheximide-chase experiments in wild-type yeast and strains lacking key ERAD components (Figure 1 in the preprint). The authors used model single-pass proteins containing an idealized hydrophobic TM and compared it to variants in which a single hydrophobic residue was replaced with a charged or hydrophilic one. These modified constructs were rapidly degraded, whereas substitution with hydrophobic residues had no effect.

• Position of hydrophilic residues is critical for degradation rate

Degradation depended strongly on where the hydrophilic residue was located within the TM. Cycloheximide-chase assays showed that residues in the middle or cytosolic side of the membrane triggered the fastest degradation, whereas luminal-proximal positions had weaker effects. To generalize this observation, the authors created a comprehensive library with randomized amino acids at position 10 and analyzed the abundance of each variant using FACS followed by next-generation sequencing. The resulting TM Stability Score correlated closely with the free energy of transferring each amino acid into the lipid phase (Figures 2 and 3 in the preprint).

Mechanistically, the TMs with exposed hydrophilic residues partition into the locally thinned membrane region adjacent to the lateral gate of Hrd1. This movement allows the proteins to escape the energetic penalty of being in the lipid phase by moving into the aqueous environment of Hrd1’s cytosolic cavity. This increased residence time near Hrd1 leads to polyubiquitination.

• Single-pass and multi-pass proteins behave differently

Cycloheximide-chase experiments on model single-pass substrates and endogenous single-pass proteins demonstrated that their degradation requires both Hrd1 and Der1, consistent with their insertion into the thinned membrane region formed cooperatively by these two proteins (Figure 4 in the preprint). In contrast, multi-pass substrates were degraded in an Hrd1-dependent but Der1-independent manner (Figures 5 and 6 in the preprint).

• Hrd1 dimer displaces Der1 to accommodate multi-pass proteins

To explain how multi-pass proteins bypass the need for Der1, the authors determined a cryo-EM structure of a Hrd1 dimer (Figure 6 in the preprint). This revealed that the second Hrd1 monomer occupies the same position normally taken by Der1. Importantly, each Hrd1 subunit contains a large cytosolic cavity, indicating that Der1 must move aside for multi-pass proteins to enter the cavity-adjacent thinned membrane region and undergo degradation.

Future directions and questions for the authors

1. The D4-R single-pass construct, which positioned a hydrophilic residue near the luminal leaflet, was surprisingly stable and poorly recognized by the ER-localized ligases. Does this mean that multi-pass proteins also tolerate hydrophilic residues in their luminal regions?
2. The authors predict that multi-pass proteins may partition less efficiently and be degraded more slowly than single-pass proteins because they cannot utilize Der1. It would be interesting to quantify this.
3. The authors note that Asi and Doa10 recognize substrates via different mechanisms than the partitioning model proposed for Hrd1. It would be interesting to expand on how the other ER-localized ubiquitin ligases identify their substrates.
4. Given that the Dsc ligase complex in the Golgi contains components predicted to be structurally similar to Hrd1 and Der1, does the partitioning model also apply to membrane protein quality control mechanisms within the Golgi apparatus?
5. To what extent can the partitioning mechanism be extended to higher eukaryotes? Mammalian Hrd1 homologs (gp78, RNF145, Trc8) have distinct substrate specificities. Do the cytosolic cavities of these mammalian homologs provide more specific binding sites that refine the general partitioning mechanism?

 

doi: https://doi.org/10.1242/prelights.42199

Read preprint (No Ratings Yet)