Research Article
Altering CLC stoichiometry by reducing non-polar side-chains at the dimerization interface

https://doi.org/10.1016/j.jmb.2021.166886Get rights and content

Highlights

  • Dimerization of CLC-ec1 requires non-polar residues at the dimerization interface.

  • Alanine substitutions at the CLC-ec1 dimerization interface alter stoichiometry in a solvent dependent manner.

  • L194 on helix H is a molecular hot-spot for CLC-ec1 dimerization.

Abstract

CLC-ec1 is a Cl/H+ antiporter that forms stable homodimers in lipid bilayers, with a free energy of −10.9 kcal/mol in 2:1 POPE/POPG lipid bilayers. The dimerization interface is formed by four transmembrane helices: H, I, P and Q, that are lined by non-polar side-chains that come in close contact, yet it is unclear as to whether their interactions drive dimerization. To investigate whether non-polar side-chains are required for dimer assembly, we designed a series of constructs where side-chain packing in the dimer state is significantly reduced by making 4–5 alanine substitutions along each helix (H-ala, I-ala, P-ala, Q-ala). All constructs are functional and three purify as stable dimers in detergent micelles despite the removal of significant side-chain interactions. On the other hand, H-ala shows the unique behavior of purifying as a mixture of monomers and dimers, followed by a rapid and complete conversion to monomers. In lipid bilayers, all four constructs are monomeric as examined by single-molecule photobleaching analysis. Further study of the H-helix shows that the single mutation L194A is sufficient to yield monomeric CLC-ec1 in detergent micelles and lipid bilayers. X-ray crystal structures of L194A reveal the protein re-assembles to form dimers, with a structure that is identical to wild-type. Altogether, these results demonstrate that non-polar membrane embedded side-chains play an important role in defining dimer stability, but the stoichiometry is highly contextual to the solvent environment. Furthermore, we discovered that L194 is a molecular hot-spot for defining dimerization of CLC-ec1.

Introduction

The cell membrane encapsulates all living things with a structurally defined layer of oil - the lipid bilayer. This hydrocarbon core provides an electrostatic barrier to the passive permeation of ions and charged molecules. With that, biology has evolved a special class of membrane proteins, that stably reside within the lipid bilayer and control the precise transport of chemicals and information across the membrane barrier. Membrane proteins are built with the same amino acids that are found within water soluble proteins. However, they must be lined by hydrophobic amino acids at their lipid exposed surfaces to enable partitioning into the lipid bilayer while not disrupting the integrity of the membrane. In the past, it was proposed that membrane proteins have evolved as inversions of soluble proteins, favoring polar and hydrophilic interactions within the core, while exposing the non-polar surfaces to the surrounding lipids.1

While it is true that membrane proteins display hydrophobic residues on the lipid-facing surfaces, and that they often contain hydrophilic permeation or transport pores within, it is not generally true that membrane proteins interact via polar groups embedded within the membrane. In fact, many membrane protein complexes assemble via non-polar surfaces, between transmembrane helix contacts in multi-helix folds, or along larger surfaces that are involved in oligomerization.2 In these cases, the hydrophobicity of the interaction interfaces appears to be similar to that of lipid-facing interfaces. This raises the question of what drives the assembly of membrane proteins via these greasy interfaces, when they appear to be suitably solvated by the similarly greasy lipids in the membrane?

One possibility is that non-polar side-chains at these interfaces allow specific van der Waals (VDW) interactions to form, which exclusively stabilize the assembled state. While some degree of VDW interactions is also expected between the exposed interfaces and the lipids in the dissociated state, it may be that the protein–protein interactions in the associated complex are optimized leading to a net stabilizing driving force for assembly. Indeed, VDW interactions by side-chains have been shown to play an important role in the dimerization stability of Glycophorin-A,3 however in this system, it has also been shown that other backbone contributions, such as hydrogen bonding also contribute to the overall stability.4, 5 Thus, it remains an open question whether side-chains at non-polar surfaces are indeed responsible for driving the assembled state over the dissociated lipid solvated states.

Recently, we have developed a new type of model system that has the potential to expand our ability to investigate how interfacial non-polar side-chains affect membrane protein assembly. CLC-ec1 is a homodimeric Cl/H+ antiporter that is native to E. coli inner membranes (Figure 1(A)). It is a large protein, with 18 transmembrane helices per subunit (A-R) that associate to form a stable dimer, ΔG° = −10.9 kcal/mol, 1 subunit/lipid standard state, in 2:1 POPE/POPG lipid bilayers measured using single-molecule photobleaching analysis.6, 7, 8 The dimerization interface is large, comprised of four transmembrane helices (H, I, P & Q) with a surface area of ~1200 Å2 lined by ≈ 20 non-polar residues (Figure 1(B)). These surfaces exhibit high shape complementarity placing all of these side-chains within VDW contact distance.9

To investigate whether side-chains are important for defining dimerization stability, we carried out an extreme mutagenesis study where we strip each helix of all interfacial side-chains, making 4–5 simultaneous substitutions to alanine (8–10 in the dimer state). These “helix-ala” constructs are predicted to form cavities within the dimerization interface, resulting in a significant loss of side-chain interactions. By examining monomer–dimer populations in detergent and lipid bilayers, carrying out functional transport assays and examining structural changes by x-ray crystallography, we are able to determine that these side-chains contribute to the overall dimerization stability in lipid bilayers, but that they are not required for dimerization in detergent micelles. Furthermore, we identify that residue L194 participates in a molecular hot-spot for CLC-ec1 dimerization. Altogether, these results demonstrate that interfacial side-chains impact dimer stability in lipid bilayers, and that membrane protein stoichiometry is a highly contextual observable that depends strongly on the solvent environment.

Section snippets

Computational analysis of protein interactions at the dimerization interface of CLC-ec1

The CLC-ec1 dimer exhibits many side-chain contacts between the two subunits, all of which may contribute stabilizing interactions in the dimer state (Figure 1(A)). Visually, these contacts are found along four membrane-embedded helices H, I, P & Q, between the domain-swapped helix A, the C-terminal helices Q & R, and the extracellular loop between helices I & J (Figure 1(B) and (C)). To identify the residues that come into close proximity of one another in the dimer complex, we calculated the

Discussion

In our study, we found that reducing non-polar side-chain interactions at the dimerization interface yields functional, monomeric CLC-ec1 in 2:1 POPE/POPG lipid bilayers. In our first approach, we found that an accumulation of 8–10 subtractive substitutions of phenylalanine, leucine and isoleucine to alanine per dimer, produce monomeric CLC-ec1 that behaves comparably to the monomeric I201W/I422W CLC-ec1 control in the lipid bilayer.7, 9 In our original hypothesis, we proposed that removal of

Conclusion

Here we report that CLC-ec1 dimerization in lipid bilayers is destabilized when non-polar side-chains are mutated to alanine at the membrane-embedded dimerization interface. Despite this, we find that several constructs, where substantial side-chain contacts have been removed, can still be found as dimers upon purification in detergent micelles or during crystallization. In addition, we find that CLC-ec1 can be made monomeric with a single alanine substitution at residue L194, identifying a

In silico van der Waals energy calculations

Atomic coordinates were taken from the following x-ray crystal structures: WT CLC-ec1 - PDB ID: 1OTS, C85A/L194A – PDB ID: 7CVS and C85A/H234C/L194A – PDB ID: 7CVT. The CHARMM-GUI (http://www.charmm-gui.org/)32, 33, 34, 35 webserver was used to build missing hydrogens into the atomic coordinates and to mutate C85 to Ala in the WT structure. All side-chain atoms of the CLC-ec1 structures were then energy-minimized for 8000 steps using NAMD36 and the CHARMM27 force-field. Pairwise van der Waals

CRediT authorship contribution statement

Kacey Mersch: Conceptualization, Methodology, Investigation, Formal analysis, Visualization. Tugba N. Ozturk: Methodology, Formal analysis, Visualization. Kunwoong Park: Investigation, Formal analysis. Hyun-Ho Lim: Supervision, Formal analysis, Funding acquisition. Janice L. Robertson: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Visualization, Supervision, Software, Project administration, Funding acquisition.

Acknowledgements

The Robertson lab is supported by the National Institute of General Medical Science, National Institutes of Health (R01GM120260, R21GM126476). We are grateful to the staff at beamlines 5C at PALII (Pohang Light Source II, Pohang Accelerator Laboratory, Pohang, Republic of Korea) for assistance at the synchrotron. This work was partly supported by the KBRI Basic Research Program through Korea Brain Research Institute funded by the Ministry of Science and ICT, Korea (20-BR-01-05) to H.-H. L. We

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (50)

  • S. Jo et al.

    CHARMM-GUI PDB manipulator for advanced modeling and simulations of proteins containing nonstandard residues

    Adv. Protein Chem. Struct. Biol.

    (2014)
  • S. Granier et al.

    Structure and conformational changes in the C-terminal domain of the beta2-adrenoceptor: insights from fluorescence resonance energy transfer studies

    J. Biol. Chem.

    (2007)
  • L.J. Friedman et al.

    Multi-wavelength single-molecule fluorescence analysis of transcription mechanisms

    Methods

    (2015)
  • D.M. Engelman et al.

    Bacteriorhodopsin is an inside-out protein

    Proc. Natl. Acad. Sci. USA

    (1980)
  • A. Oberai et al.

    Structural imperatives impose diverse evolutionary constraints on helical membrane proteins

    Proc. Natl. Acad. Sci. USA

    (2009)
  • K.R. MacKenzie et al.

    A transmembrane helix dimer: structure and implications

    Science

    (1997)
  • R. Chadda et al.

    A model-free method for measuring dimerization free energies of CLC-ec1 in lipid bilayers

    J. Gen. Physiol.

    (2018)
  • R. Chadda et al.

    The dimerization equilibrium of a ClC Cl(-)/H(+) antiporter in lipid bilayers

    Elife.

    (2016)
  • J.L. Robertson et al.

    Design, function and structure of a monomeric ClC transporter

    Nature

    (2010)
  • R. Dutzler et al.

    Gating the selectivity filter in ClC chloride channels

    Science

    (2003)
  • H.H. Lim et al.

    Intracellular proton access in a Cl(-)/H(+) antiporter

    PLoS Biol.

    (2012)
  • R. Dutzler et al.

    X-ray structure of a ClC chloride channel at 3.0 A reveals the molecular basis of anion selectivity

    Nature

    (2002)
  • R.E. Jefferson et al.

    Membrane proteins can have high kinetic stability

    J. Am. Chem. Soc.

    (2013)
  • A. Stetsenko et al.

    An overview of the top ten detergents used for membrane protein crystallization

    Crystals

    (2017)
  • G.F. Ames

    Lipids of Salmonella typhimurium and Escherichia coli: structure and metabolism

    J. Bacteriol.

    (1968)
  • View full text