TL;DR
Membrane proteins are notoriously difficult to predict structurally. This paper challenges the assumption that transmembrane helices settle into their final, thermodynamically favored positions during folding. Instead, they demonstrate that helices can significantly reposition themselves, suggesting a more dynamic folding process and implications for structure prediction.
What Problem Does This Paper Address?
Predicting the three-dimensional structure of membrane proteins is a significant challenge in structural biology. A common approach relies on the "two-stage folding model," which posits that transmembrane (TM) α-helices first insert into the lipid bilayer in their energetically optimal locations and then pack together. This model assumes that the final positions of these helices in the folded protein reflect their initial insertion preferences. However, experimental evidence has begun to suggest a more complex reality, with TM helices exhibiting kinks, reentrant loops, and potentially dynamic behavior. This paper investigates whether TM helices truly maintain their initial insertion positions or if they undergo significant repositioning during the folding and oligomerization process, a phenomenon that would have profound implications for how we predict their structures.
Methods
The researchers employed a combination of experimental and computational approaches:
- Experimental Assay for Membrane Insertion: They utilized a well-established assay involving a reporter protein (Lep) fused to segments of the target protein (GltPh). By monitoring glycosylation patterns of the Lep reporter, they could experimentally determine the apparent free energy of insertion (ΔGapp) for individual TM segments into a lipid bilayer. This assay allows for the quantification of how readily a specific helix segment inserts into the membrane.
- "ΔG Predictor" Program: This computational tool was used to predict the hydrophobicity and insertion efficiency (ΔGapp_pred) of TM helices. It calculates position-dependent contributions to the overall insertion free energy.
- Glycosylation Mapping: This technique was used to precisely map the boundaries of membrane-embedded TM helices. By introducing glycosylation sites at varying distances from the helix, they could determine the minimal distance required for efficient glycosylation, thereby pinpointing the helix's interaction with the membrane.
- Database Analysis: A large database of known membrane protein structures (from the Protein Data Bank, PDB) was analyzed to assess the generality of their findings across different membrane proteins. They categorized helices based on their predicted insertion efficiency and their actual positions in solved structures.
Datasets
- Pyrococcus horikoshii GltPh: This bacterial glutamate transporter homolog served as the primary experimental model.
- Protein Data Bank (PDB): A non-redundant subset of α-helical TM protein structures (601 chains, reduced to 101 chains at 40% sequence identity) was used for the database analysis.
Key Insights and Ideas
- TM Helices Don't Necessarily Stay Put: The most striking finding is that the predicted energetically favorable insertion positions of TM helices often do not match their positions in the final, folded protein structure. This directly challenges the assumption of static insertion.
- Dynamic Folding Process: The results suggest that membrane protein folding is a more dynamic process than previously assumed, with TM helices capable of significant repositioning relative to the membrane during folding and oligomerization.
- Implications for Structure Prediction: Current structure prediction methods that rely on hydrophobicity to determine helix positions may be fundamentally flawed. A more sophisticated approach is needed that accounts for these dynamic shifts.
- Bringing Polar Residues into the Membrane: Repositioning of helices could be a mechanism to strategically place polar or charged residues within the hydrophobic membrane environment, which is crucial for protein function.
- Generality of the Phenomenon: Analysis of a large database of membrane proteins indicates that this repositioning phenomenon is likely widespread, not unique to GltPh.
Key Results
- GltPh TM Helices Reposition: For the GltPh transporter, TM2, TM4, and TM7 helices were found to be dramatically repositioned relative to the membrane during folding and oligomerization.
- Mismatch Between Predicted and Structural Positions: In GltPh, the most hydrophobic segments identified by the "ΔG predictor" did not always coincide with the membrane-buried segments observed in the crystal structure.
- Glycosylation Mapping Confirms Shift: Glycosylation mapping of TM4 in GltPh showed that the initially inserted segment shifted by 12–13 residues relative to the membrane during folding.
- Database Analysis Reveals Widespread Phenomenon: Analysis of PDB structures revealed that a significant fraction of TM helices are not sufficiently hydrophobic to insert optimally on their own, and that "strongly shifted" helices (where predicted and actual positions differ significantly) are common, particularly in transporters.
Limitations and Future Work
- Alternative Explanations: While the authors strongly favor repositioning during folding, they acknowledge the formal possibility that some repositioning might occur within the translocon itself.
- Specific Mechanisms: The precise molecular forces driving these repositioning events are not fully elucidated and warrant further investigation.
- Broader Structural Implications: The findings necessitate a re-evaluation of how membrane protein structures are predicted, suggesting that models need to incorporate the dynamic nature of helix insertion and folding. Future work could focus on developing predictive algorithms that account for these repositioning events.
- Functional Relevance: Understanding how these repositioning events contribute to the protein's function, particularly in dynamic processes like transport, is a key area for future research.
Summary updated