Tsinghua Science and Technology Tsinghua University, China ISSN: 1007-0212 Vol. 6, Num. 3, 2001, pp. 253-256

Tsinghua Science and Technology, Vol. 6, No. 3, August 2001 pp. 253-256

Designing Stable Antiparallel Coiled Coil Dimers* 

ZENG Xian'gang ** , ZHOU Haimeng †

Department of Chemical Engineering, Wuyi University, Jiangmen, Guangdong 529020, China;
†Department of Biological Sciences and Biotechnology, Tsinghua University, Beijing 100084, China

* Supported  by the “973” Project, the Natural Science Foundation of Guangdong, the Research Foundation for Returned Overseas Scholars of the Ministry of Education, and the State Key Lab of Biomembrane & Membrane Biotechnology  
 ** To  whom correspondence should be addressed, E-mail: xgzeng@letterbox.wyu.edu.cn

Received: 2000-09-18

Code Number: ts01076

Abstract:   

The history of antiparallel coiled coil dimer design is briefly reviewed and the main principles governing the successful designs are explained. They include analysis of the inter-subunit electrostatic repulsion for determining partners for dimerization and of the buried polar interaction for determining the relative orientation of the partners. A theory is proposed to explain the lack of antiparallel coiled coil homodimers in nature.

Key  words:  coiled coil; dimer; antiparallel orientation; hydrogen bond; electrostatic repulsion

Introduction   

The coiled coil[1]  is a universal structure found in a diverse group of proteins, including enzymes, membrane proteins, receptor proteins, structural proteins, transcription factors, etc. It is a bundle of 2, 3, 4, or more strands of interacting a-helices which adopt a geometry of 3.5 residues per helical turn, slightly different from regular a-helices that have 3.6 residues per helical turn. The coiled coil is responsible for the homo-and hetero-oligomerization of many proteins, or in some cases it zips parts of the same oligopeptide chain together.

The two-stranded coiled coil, or a coiled coil dimer, is one of the simplest ways to create a dimerization structure. Actually the beauty of the leucine zipper structures[2, 3] , which are coiled coil dimers, led to a renewed appreciation of the coiled coil. Because of its relative simplicity, the coiled coil dimer has been widely used as a model system for the study of protein assembly, design, folding, interaction, and structural predication. However, de novo design of coiled coils is not just for the love of the basics, but has multiple practical applications[4].

1 Attemped Antiparallel Coiled Coil Dimer Design  

When two helices cross or are next to each other, there are more antiparallel cases (crossing angles greater than 90°) than parallel cases. In coiled coils that have three or more helices, both orientations are found quite often, but the majority of 2-stranded coiled coils studied so far are parallel (Fig.1(a)). Of the very few known antiparallel coiled coil dimers(Fig.1(b)), most are very long or covalently linked. Long coiled coils (e.g., the SMC proteins) have less stringent sequence requirements than short ones, perhaps because longer coils need to provide less stability per heptad (abcdefg, see Fig.1) than shorter coils for the same total stability. The covalently-linked coiled coils may be obligated to form antiparallel coiled coils and when the covalent linkage is broken, the two helices are unlikely to form stable antiparallel coiled coil dimers[5]. Does this mean that antiparallel coiled coil dimers vidate some energy rules that we yet not understand] Can we design peptides that defy the odds]

Hodges and co-workers were the first to systematically study the differences between  parallel and antiparallel coiled coil dimers. They designed two peptides, C2A16 and C33A16 (Fig.2), that formed disulfide-bonded antiparallel heterodimers. But the two peptides did not preferentially form disulfide-bonded antiparallel heterodimers; but also formed disulfide-bonded parallel homodimers. Nevertheless, they studied more coiled coil dimer designs and concluded that interchain electrostatic interactions among the e and g position residues play a major role in controlling the parallel or antiparallel alignment[6, 7].

In  1996, Zeng realized several principles governing coiled coil dimer formation which he used for de novo protein design of peptides X and Z (Fig.2) intended to preferentially and stably form antiparallel coiled coil heterodimers[8]  . Later, Oakley and Kim demonstrated that a pair of designed peptides Acid-a1 and Base-a1 (Fig.2), with many features similar to those of X and Z, preferentially formed antiparallel coiled coil heterodimers without the help of a disultide-bridge[9]. Their results showed that the stable antiparallel coiled coil dimer structure can be engineered, and more and more protein folding fundamentals are understood.

2 Principles  

The structure of a coiled coil dimer is determined by the primary sequences of its two subunits (only one sequence for a coiled coil homodimer). Analysis of sequences, protein folding, protein stability, mutagenesis, and structure, led to several rules related to antiparallel coiled coil dimers.

2.1 b, c, and f positions

Of the different positions in the heptad repeat, the b, c, and f positions generally do not contribute directly to dimerization, since they are occupied by amino acids with good helix-forming propensity and polar amino acids that are soluble without strict sequence requirements. But their energy levels do not favor interaction with the other subunits, so they do not form higher-order oligomers.In peptides X and Z (Fig.2), A, Q, and S are used for all the b, c, and f positions because they are found to be predominate in the b, c, and f positions of 30 leucine zippers in a data bank[10]. In peptides Acid-a1 and Base-a1 (Fig.2), the residues in these positions are from a successful parallel coiled coil dimer design[11].

2.2 g and e positions

The g and e positions flank the dimer interface, so they may be solvent accessible or buried. Studies showed that they are important in determining the oligomerization state and stability, as well as the structural uniqueness of the coiled coil[12]. Charged residues are frequently found in these positions, but the actual role of the inter-subunit side-chain electrostatic interactions between charged residues in these positions is still under debate[13] , even though the electrostatic repulsion is clearly a destabilizing factor in dimer formation[11]. For parallel dimers, these interactions are g-epairs; for antiparallel dimers, these interactions are g-g pairs and e-e pairs (Fig.1).

In peptide X all of these positions are occupied by E. If an XX homodimer is formed, both parallel and antiparallel structures have numerous electrostatic repulsions such as in a ZZ homodimer with all the g and e positions occupied by K. But the formation of an XZ heterodimer will eliminate the unfavorable interactions. This reversed design can be used to obtain heterodimers other than homodimers without considering the orientation. The difference in the pairing of these positions in parallel and antiparallel dimers was not exploited here. Peptides Acid-a1 and Base-a1 were designed using the same strategy in the g and e positions.

2.3 D and A positions

The core of the dimer interface, the D and A positions, prefers hydrophobic amino acids for stability[14]. Polar amino acids in these positions destabilize the dimer, but inter-subunit hydrogen bonding formed between the side chains of aligned polar residues in these positions[2]  keep the two helices in register and determine the specific dimerization at the cost of some stability[15, 16] . For parallel dimers, these interactions are between the d-d and a-a pairs; for antiparallel dimers, these interactions are between the d-a pairs (Fig.3).

For a homodimer with one buried polar residue, the parallel orientation may give rise to a buried inter-subunit hydrogen bond, while the antiparallel orientation will have no buried inter-subunit hydrogen bond but will have two mismatches, each formed between a buried polar residue and a hydrophobic residue. It is reasonable to postulate that the buried polar interaction disfavors the antiparallel orientation in coiled coil homodimers. In a homodimer, the antiparallel orientation would leave the two ends slightly out of register (Fig.3), “wasting” a few interactions that may otherwise provide greater stability, which would be a disadvantage for antiparallel homodimers and may explain why antiparallel coiled coil homodimers are so rare in nature.

In peptides X and Z, most of the D and A positions are occupied by L and I, respectively, except N15 in X and N26 in Z. These two polar residues were designed to give rise to a buried polar interaction only in the antiparallel heterodimer, with two mismatches in the parallel heterodimer. Homodimers are not considered here. Peptides Acid-a1 and Base-a1 used the same design strategy: N21 in Acid-a1 and N10 in Base-a1.

3 Conclusions  

The antiparallel coiled coil dimer can be designed based on the current knowledge of protein folding, stability, and specificity. Successful designs of antiparallel coiled coil heterodimers have been achieved. The main principles are the use of the inter-subunit electrostatic repulsion in determining the peptides for dimerization, and the buried polar interaction in determining the relative orientation of the dimerizing peptides.The next design challenges include: (1) to design shorter but stable antiparallel coiled coil heterodimers, (2) to design antiparallel coiled coil homodimers, and (3) to study the folding of these coiled coils and learn more about the principles of protein folding.Our current research consists of several specific aims: (1) Study the stability and folding of the coiled coil in AraC, which is an antiparallel coiled coil homodimer in the protein structure[17]. (2) Learn from Nature what kinds of inter/intra-helical interactions should be avoid or should be used and identify the determinants of helix orientations in coiled coils. (3) Design peptides using the latest knowledge on coiled coils, with considerations of length, sequence, oligomerization states and, of course,  helix orientation. (4) Test the designs and make adjustments based on the results.

Acknowledgements

We would thank Jim Hu for discussion.

References

1. Crick F H C. The packing of a-helices: Simple coiled coils. Acta Cryst, 1953, 6: 689-697.
2. O'Shea E K, Klemm J D, Kim P S. X-ray structure of the GCN4 leucine zipper, a two-stranded, parallel coiled coil. Science, 1991, 254: 539-544.
3. Harbury P B, Zhang T, Kim P S, Alber T. A switch between two-, three-, and four-stranded coiled coils in GCN4 leucine zipper mutants. Science, 1993, 262: 1401-1407.
4. Hodges R S. De novo design of a-helical proteins: basic research to medical applications. Biochem Cell Biol, 1995, 74: 133-154.
5. Oakley M G, Kim P S. Protein dissection of the antiparallel coiled coil from Escherichia coli Seryl tRNA synthetase. Biochemistry, 1997, 36:  2544-2549.
6. Monera O D, Zhou N E, Kay C M, Hodges R S. Comparison of antiparallel and parallel two-stranded a-helical coiled-coils. Design, synthesis, and characterization. J Biol Chem, 1993, 268:  19[ ]218-19[ ]227.
7. Monera O D, Kay C M, Hodges R S. Electrostatic interactions control the parallel and antiparallel orientation of a-helical chains in two-stranded a-helical coiled-coils. Biochemistry, 1994, 33:  3862-3871.
8. Zeng X. Oligomerization properties of GCN4 leucine zipper mutants: an in vivo study (Ph.D. thesis). Texas A&M University, 1996.
9. Oakley M G, Kim P S. A buried polar interaction can direct the relative orientation of helices in a coiled coil. Biochemistry, 1998, 37: 12 603-12 610.
10. Hu J C, Sauer R T. The basic-region leucine-zipper family of DNA binding proteins. Nucleic Acids and Mol Biol, 1992, 6: 82-101.
11. O'Shea E K, Lumb K J, Kim P S. Peptide “Velcro”: design of a heterodimeric coiled coil. Curr Biol, 1993, 3: 658-667.
12. Zeng X, Zhu H, Lashuel H A, Hu J C. Oligomerization properties of GCN4 leucine zipper e and g position mutants. Protein Sci, 1997, 6:  2218-2226.
13. Lavigne P, Sonnichsen F D, Kay C M, Hodges R S. Interhelical salt bridges, coiled coil stability, and specificity of dimerization. Science, 1996,  271: 1136.
14. Hu J C, O'Shea E K, Kim P S, Saner R T. Sequence requirements for coiled-coils: analysis with lambda repressor-GCN4 leucine zipper fusions. Science, 1990, 250: 1400-1403.
15. Zeng X, Herndon A M, Hu H C. Buried asparagines determine the dimerization specificities of leucine zipper mutants. Proc Natl Acad Sci USA, 1997, 94: 3673-3678.
16. Zhu H, Celinski S A, Scholtz J M, Hu H C. The contribution of buried polar groups to the conformational stability of the GCN4 coiled coil. J Mol Biol, 2000, 300: 1377-1387.
17. Soisson S M, MacDougall-Shaekleton B, Schleif R, Wolberger C. Structural basis for ligand-regulated oligomerization of AraC. Science, 1997, 276:  421-425.