Binding of SARS-CoV-2 fusion peptide to host membranes
Summary (3 min read)
Introduction
- During infection, viruses first recognize and then enter their target cells.
- The spike S1 subunit recognizes the human target cell by binding to the ACE2 receptor, and the S2 subunit then facilitates fusion of the viral membrane with host cellular membranes. [1] [2] [3].
- As the membrane composition and pH of the endosome change, structural rearrangements may be induced in the S protein that facilitate membrane fusion.
- Mutation studies by Madu et al. confirmed the importance of the LLF motif for the fusion activity of the FP.
The FP in solution forms two short amphipathic helices
- Formed by consecutive residues, the LLF motif is spread across both faces of the NTH, as the two leucines are part of the predicted hydrophobic face and the phenylalanine is not.
- In their simulation of FP in aqueous solution, the first segment expanded to the beginning of the second segment to form a single contiguous helix (AH2), the remainder of the second segment unfolded, and the third segment retained its helical structure.
- These two distinct short helices (AH2 and CTH) are connected via a short loop.
- Whereas the amino-acid sequence of CTH is highly conserved and shows no strong amphipathic properties, AH2 is less conserved but carries a strong hydrophobic moment .
NTH binds membranes with its amphipathic face
- The authors reasoned that the two amphipathic helices NTH and AH2 may insert into the human membranes to anchor the S protein for membrane fusion.
- In all three cases in which the NTH established the first stable contact, the binding followed a consistent path.
- In the remaining insertion event into the endosomal membrane, the NTH established stable binding without F823 flipping in the simulated time and hence the helix remained on top of the membrane interface.
- In the simulations with the mimetic of the outer leaflet of the plasma membrane, the authors observed one spontaneous NTH binding event, after the CTH and AH2 had already been inserted for more than 5 µs .
- All three helices stayed bound to the membrane for ⇡ 0.5 µs until the NTH lost membrane contact.
The C-terminus of the FP binds via flexible elements
- In all their simulations, the authors observed membrane binding also with the C-terminal end of the FP.
- Binding involved AH2, the CTH and flexible elements flanking AH2 at both ends.
- In some cases, these membrane interactions were relatively short lived , in other cases binding was stable for more than 6 µs.
- The short hydrophobic stretches that inserted most frequently are centered around residues I834, L841 and I844.
- Notably however, the short AH2 and to an even larger extent the CTH, in some cases partially unfolded to flexible amphipathic structures when bound to the membranes .
The inserted NTH can withstand high pulling forces
- This process mimics the forces experienced by the FP during its presumed primary function of pulling host and viral membrane into proximity.
- By pulling the C-terminal end up via a harmonic spring moving at constant velocity, the force applied to the peptide increases more or less linearly in time, until peptide segments, and ultimately the entire peptide, are pulled out of the membrane.
- By pulling the bound NTH out of the endosomal membrane, the authors found that the binding of the NTH alone can withstand pulling forces between 40 and 65 pN.
- Higher forces were needed to pull the NTH out of the deeply bound state with inserted F823 .
- Interestingly, even though F823 was pulled out of the membrane along this path , the shallow state did not appear as a distinct intermediate in the pulling traces.
Additional insertion of AH2 and CTH dramatically stabilize membrane anchoring
- The authors performed additional pulling simulations with the full length FP, starting from a binding mode with all three helices inserted into the membrane interface.
- One might assume that these peaks were the result of the CTH and AH2 detaching individually.
- Visual inspection revealed that this is not the case.
- In 35% of the simulations the entire peptide was detached from the membrane after the second force peak, meaning that all three helices detached nearly simultaneously with the detachment of I844.
- This then resulted in a series of additional force peaks, with rupture forces in the same regime as for the isolated NTH fragment.
Discussion
- Two separate binding regions increase the likelihood to stay bound under load Due to the connection via a long disordered linker, the NTH and the two more C-terminal helices, AH2 and CTH, act relatively independent from one another.
- For SARS-CoV-2, the concept of avidity -with multiple spread-out interactions maintaining a bound statehas emerged at multiple levels: in the ACE2-S interaction, 24 in the virion-host interaction, 14 and here in the FP-membrane interaction.
- This underlines their idea of the two sides of the FP acting cooperatively, promoting each other's membrane binding and stabilizing the anchoring overall.
- 25, 26 Gorgun et al. used a truncated version of the FP where they cut it behind L841, right where the authors observe the AH2.
- They identified three binding modes: one with the loop inserted into the membrane, one with the NTH inserted and a last binding mode where the whole FP acquires helical structure and inserts on top of the membrane.
Di↵erences in lipid density may alter preferred binding mode
- The binding modes the authors described seem to loosely group into the NTH binding to the endosomal membrane and C-terminal regions binding to the outer plasma membrane mimetic.
- By contrast, the high density of the outer plasma membrane may favor the insertion of smaller, often disordered hydrophobic stretches such as the ones around I844.
- This creates an artificial energetic penalty that competes with the binding free energy of the amphipathic peptide.
- Unfortunately, this effect is di cult to correct for, short of performing simulations with prohibitively large boxes or with preemptively removed lipids from one leaflet.
- Furthermore, the lower penalty from the lateral pressure would have likely led to the simultaneous binding of two or all three helices.
Binding of few FPs may be strong enough to facilitate membrane fusion
- 21 Whereas already the bound NTH alone can sustain such forces, the full FP is anchored even more strongly by its three helices.
- The structure with a disulfide bridge thus endows the FP with high anchoring strength that is reminiscent of the catch bonds giving cell-cell contacts high mechanostability.
- The observed stability of the binding raises the question of how many bound FPs are required to be engaged for successful fusion.
- 21 This may ultimately increase the infection success of the virus, as it would reduce one source of failure.
Conclusions
- From atomistic molecular dynamics simulations, the authors gained a detailed view of the interactions between the SARS-CoV-2 FP with lipid bilayers mimicking the endosomal membrane and the outer leaflet of the plasma membrane.
- In all four runs with the more flexible and less packed endosomal membrane, the FP eventually bound into the lipid bilayer with its NTH.
- Insertion of all three helices at the same time was observed rarely in their simulations.
- The Cys-Cys disulfide bond linking the centers of AH2 and CTH emerged as an important stabilizer.
- The authors speculate that by spreading the membrane interaction across multiple distinct elements, with NTH, AH2, CTH and the intervening amphipathic loops all connecting to the membrane, the virus achieves a trade-o↵ between rapid insertion of individually small elements into the membrane and their firm membrane anchoring.
Did you find this useful? Give us your feedback
Citations
5 citations
References
25,256 citations
17,939 citations
16,857 citations
15,362 citations
13,937 citations
Related Papers (5)
Frequently Asked Questions (17)
Q2. How did the authors determine the strength of the inserted NTH?
The inserted NTH can withstand high pulling forcesThe authors determined the strength of the membrane anchoring by subjecting the the C-terminus of the FP to mechanical force.
Q3. What is the role of the S protein in the fusion of the viral membrane?
As the membrane composition and pH of the endosome change, structural rearrangements may be induced in the S protein that facilitate membrane fusion.
Q4. How many times did the peptide detached from the membrane?
In 35% of the simulations the entire peptide was detached from the membrane after the second force peak, meaning that all three helices detached nearly simultaneously with the detachment of I844.
Q5. How many spontaneous insertion events of the NTH into the membrane interface?
In eight independent MD simulations of 10 µs each, the authors observed five spontaneous insertion events of the NTH into the membrane interface.
Q6. How did the authors determine the strength of the binding of the NTH?
By pulling the bound NTH out of the endosomal membrane, the authors found that the binding of the NTH alone can withstand pulling forces between 40 and 65 pN.
Q7. What do you think of the FP binding modes?
26Di↵erences in lipid density may alter preferred binding modeThe binding modes the authors described seem to loosely group into the NTH binding to the endosomal membrane and C-terminal regions binding to the outer plasma membrane mimetic.
Q8. How did the authors stabilize the binding mode of the SARS-CoV-2 virus?
the authors could show that by relieving lateral pressure in the exposed leaflet, the authors could stabilize a binding mode with all three helices inserted fully.
Q9. What is the effect of the deep binding state of the FP?
As the authors showed, the deep binding state associated with F823 insertion increases the pulling force that the NTH can withstand, which in turn supports the fusion activity of the FP.
Q10. What is the force required to completely detach the NTH from the deep state?
the forces required to completely detach the NTH from the deep state are 10-15 pN higher compared to the shallow binding mode.
Q11. How many pulls did the AH2 and the NTH stay attached?
In the remaining 65% of pulls, parts of the long linker and the NTH stayed bound independently of the AH2 and the CTH and only detached later.
Q12. What was the initial structure for the 20 replica pulling simulations?
The initial structure for the 20 replica pulling simulations was taken from an unbiased simulation of FP on the outer plasma membrane (for details see methods).
Q13. What was the effect of the flipped orientation on the NTH?
In three simulations (Figure 2; runs 1, 2, 4), F823 flipped its orientation after being bound to the endosomal membrane for ⇡ 0.7, 2, and 3 µs, respectively, so that its aromatic sidechain became completely buried under the lipid headgroup region.
Q14. What is the role of the LLF motif in the fusion of a viral he?
In addition, the authors can directly link the described crucial role of the LLF motif to their observed NTH binding as it consistently includes the insertion of both leucines (L821 and L822) into the glycerol backbone region of the membrane lipids.
Q15. What is the expected binding of the FP to the membrane?
The authors therefore expect that FP binding will eventually converge toall three helices bound to the membrane in the course of a real infection event.
Q16. How can the authors show that the bound FP can withstand large pulling forces?
The authors could show that the bound FP — even though it is bound only to the interface of the membrane — can withstand large pulling forces exceeding 200 pN.
Q17. What did the authors do to relieve the asymmetry between the two leaflets?
To relieve the lateral-pressure asymmetry between the two leaflets caused by inserting a large structure into only the top leaflet of a finite-size membrane patch, the authors selected this relatively short lived state with all three helices bound, removed 5 lipid molecules of the over-compressed leaflet and restarted the simulation.