COMMENTARY

Proteins with weakly funneled energy landscapes challenge the classical structure–function paradigm

Garegin A. Papoian* Department of , University of North Carolina, Chapel Hill, NC 27599-3290

nearthing the origin of incred- ible acceleration of chemical reaction rates by enzymes has occupied the minds of bio- chemistsU and molecular biologists for more than a century (1). The modern paradigm of enzymatic action rests on the ideas presented by Haldane and Pauling, namely, that enzymes selec- tively stabilize transition states in chemi- cal reactions, thus lowering reaction activation energies (2). The widely held microscopic view of enzymatic Fig. 1. The interplay between protein’s folding and functional landscapes. (Left) Energy landscapes of envisions a nearly static enzyme reaction globular proteins are thought to be funneled, such that the native state is both a thermodynamic global chamber carefully sculpted by evolution minimum and is also kinetically accessible (11). It is a common practice to partition the total free energy into protein chain’s structural entropy (horizontal axes) and the remaining part, which is often called to match complementarily the shape and ‘‘effective energy,’’ or simply energy (vertical axes). Note that the latter also includes entropic contribu- electrostatic surface of the reaction tions, mainly from the solvent. (Center) Energy landscapes of many disordered proteins are likely transition state. However, recent NMR organized around a special state, characterized by a weakly funneled landscape (6). However, the driving experiments have suggested that enzyme force for folding is too weak, allowing the protein to remain disordered. Interactions with specific targets active sites are mobile on the micro- create additional favorable contacts, deepening the funnel and driving subsequent folding (6). Transient second to second time scale, commen- population of catalytically competent states, near the funnel’s bottom, may allow for efficient catalysis (9). surate with the time scales for the (Right) A random energy landscape is shown, where search for a specific functionally competent confor- corresponding catalytic reactions (3). mation is extremely inefficient. Single- fluorescent spectro- scopy studies also point to nontrivial structure–function paradigm by using energy landscapes allows molten glob- static and dynamic disorder in enzymatic energy landscape ideas. ules to specifically recognize structural processes, something that is often How can a disordered protein, which partners (6, 10). Still, a free energy pen- masked by ensemble averaging in mac- does not adopt any definite 3D struc- alty needs to be paid during binding, to roscopic kinetic experiments (4). Even ture, carry out specific biological func- overcome the entropy of disordered in the light of these findings, the central tion? In the case of protein–protein states, requiring extra stabilization en- dogma of modern structural biology has interactions, two possibilities come to ergy at the binding interface. Given this remained largely intact; 3D protein mind. First, only a small segment of the penalty, why would proteins evolve to structure determines its function. In- protein is specifically recognized, based become disordered? Various hypotheses deed, this paradigm has worked wonder- on a signal from a local sequence of have been put forward for the ubiquity fully over many decades to explain amino acids. In a way, a protein seg- of the unfolded states: rapid turnover, numerous biological processes, from en- ment behaves as a peptide. Alterna- faster or more specific binding kinetics, zymatic catalysis to signal transduction. tively, and quite commonly, a well and functional promiscuity (7). From a Recent works, however, suggested that defined 3D structure is involved in a practical viewpoint, it is extremely diffi- from one-sixth to one-third of eukary- binding interaction (7). However, with- cult to either predict or experimentally otic proteins are either disordered or out its binding partner, this structure is determine what specific functional role contain large disordered regions (5, 6), well disguised among a multitude of is played by a particular disordered pro- hinting that the traditional interpreta- other structures that the molten globule tein, because molten globule dynamics tion of the structure–function paradigm ensemble explores, in liquid-like dynam- efficiently conceal which particular con- may be too limiting (6). Disordered pro- ics. This particular structure is neverthe- formation is the special one. These diffi- teins regulate many transcriptional and less special. It is low in energy; thus, a culties are clearly exemplified by current signal transduction processes (7). They funnel is protruded in protein’s energy lack of understanding of ␣-synuclein’s even exhibit enzymatic activity, as was landscape because of structural correla- function, a disordered protein that is recently demonstrated experimentally tions, similar to the energy landscape of implicated in Parkinson’s disease (12). (8). How this is accomplished in the ab- globular proteins with well defined na- Could a molten globule also catalyze sence of prearranged 3D structure has tive state (11) (see Fig. 1 Left). In this a chemical reaction? The answer turns not been well understood. In a recent case, however, the funnel is not deep out to be affirmative (8). Does a ran- issue of PNAS, Roca et al. (9) have used enough to render the native state the computer simulations to explain the way thermodynamic global minimum (6) (see a molten globule can accelerate a chem- Fig. 1 Center). Instead, additional ‘‘pull- Author contributions: G.A.P. wrote the paper. ical reaction. Their work, combined with ing down’’ of the funnel’s bottom is The author declares no conflict of interest. recent advances in understanding the needed by specific high-affinity interfa- See companion article on page 13877 in issue 37 of volume way and functional land- cial interactions with protein’s binding 105. scapes are coupled (6, 10), provide es- partner (6). This coupling of folding *E-mail: [email protected]. sential groundwork for generalizing the (3D structure formation) and binding © 2008 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0807977105 PNAS ͉ September 23, 2008 ͉ vol. 105 ͉ no. 38 ͉ 14237–14238 Downloaded by guest on September 25, 2021 dom energy landscape allow structurally ing such a conformation may be so amount about the mechanisms of enzy- specific catalysis? It is very unlikely, be- small that these conformations are virtu- matic reactions. It is clear, however, that cause it would take an astronomically ally invisible to spectroscopic interroga- complementary techniques, both experi- long time for the protein to find the cat- tion. However, if catalytically induced mental and computational, are needed alytically competent conformation, an to study proteins that are not character- argument similar to the necessity of hav- ized by a well defined structure. Above, ing a to overcome exces- Disordered proteins I provided an argument that even for sively long folding times (11) (see Fig. 1 many of these proteins there actually Right). Thus, as in the case of coupling regulate many exists a special conformation at low en- of binding and folding landscapes (6), a ergy, which induces weak funneling in disordered enzyme’s energy landscape transcriptional and the energy landscape. However, if we must be weakly funneled. Pulling down are given a specific disordered protein of the enzyme’s funnel upon transition signal transduction with some function, it might be incredi- state analogue binding, which is mani- bly difficult to determine which confor- fest as nudging the molten globule en- processes. mation is functionally competent among semble toward the native state, has been a myriad of other states that the globu- observed both experimentally and com- lar dynamics explores. For more or- putationally (8, 9). On the other hand, speed-up for the competent conforma- dered, yet still flexible enzymes, it will because enzyme substrates are typically tion is very large, the molten globule be interesting to systematically probe small, this effect might be less pro- will still exhibit enzymatic activity, per- nounced compared with more extensive haps somewhat reduced. In this light, their catalytic landscapes (9), perhaps protein–protein interfaces. Roca et al. would it make sense for nature to use based on a precomputed hierarchy of (9) have elucidated the way a molten molten globules for catalysis if it results native-like states (13), to gain deeper globule acts as a catalyst for a geneti- in a potential slowdown? We can only insights in to the way enzyme folding cally engineered version of chorismate speculate at this point. For example, (14) and dynamics modulates the kinet- mutase from Methanococcus jannaschii, extremely efficient control of the cata- ics of catalysis. In summary, because transforming chorismate to prephenate lytic rate may be achieved by posttrans- disordered proteins play an important (9). Using multiscale modeling, they dis- lational modifications or single point role in eukaryotic biology and human covered that activation barriers are low mutations that effect the native state’s health, controlling numerous transcrip- for several conformational basins within by only 1–2 kcal/mol. Fur- tional and signal transduction processes, the native superbasin, which provide thermore, molten globular proteins are more significant future effort is needed equivalent preorganization of the active quickly degraded by the proteasome, to understand how their energy land- site. The reaction barriers were found to allowing finer temporal control over the scapes determine their function. be high for nonnative conformations. catalytic process. This would be useful if For a weakly funneled protein, reach- only a short burst of catalytic activity is ACKNOWLEDGMENTS. My research is supported ing the catalytically competent confor- desired. by National Science Foundation Award CHE-715225, the Arnold and Mabel Beckman mational manifold may cost some free High-resolution structural biology Foundation Beckman Young Investigator Award, energy, mainly to overcome excess struc- techniques, such as x-ray crystallogra- the Camille and Henry Dreyfus Foundation, and tural entropy. The probability of observ- phy, have taught us a tremendous Petroleum Research Fund Award 47593-G6.

1. Fischer E (1890) U¨ ber die optischen Isomeren des 6. Papoian GA, Wolynes PG (2003) The physics and bioin- sition states using native topology-based energy land- Traubenzuckers, der Gluconsa¨ure und der Zuckersa¨ure formatics of binding and folding: An energy landscape scapes. J Mol Biol 346:1121–1145. [Concerning the optical isomers of the grape sugars, perspective. Biopolymers 68:333–349. 11. Wolynes P, Onuchic J, Thirumalai D (1995) Navigating gluconic acid and saccharic acid]. Ber Dtsch Chem Ges 7. Dyson HJ, Wright PE. (2002) Coupling of folding and the folding routes. Science 267:1619–1620. 23: 2611–2620. binding for unstructured proteins. Curr Opin Struct 12. McNulty BC, Young GB, Pielak GJ (2006) Macromolec- 2. Fersht, A (1999) Structure and Mechanism in Protein Biol 12:54–60. ular crowding in the Escherichia coli periplasm main- Science (Freeman, New York). 8. Pervushin K, Vamvaca K, Vgeli B, Hilvert, D (2007) Struc- tains ␣-synuclein disorder. J Mol Biol 355:893–897. 3. Henzler-Wildman KA, et al. (2007) A hierarchy of time- ture and dynamics of a molten globular enzyme. Nat 13. Materese CK, Goldmon CC, Papoian GA (2008) Hierar- scales in protein dynamics is linked to enzyme catalysis. Struct Mol Biol 14:1202–1206. chical organization of eglin c native state dynamics is Nature 450:913–916. 9. Roca M, Messer B, Hilvert D, Warshel A (2008) On the rela- shaped by competing direct and water-mediated inter- 4. Min W, et al. (2005) Fluctuating enzymes: Lessons from tionship between folding and chemical landscapes in en- actions. Proc Natl Acad Sci USA 105:10659–10664. single-molecule studies. Acc Chem Res 38:923–931. zyme catalysis. Proc Natl Acad Sci USA 105:13877–13882. 14. Osvath S, Sabelko JJ, Gruebele M (2003) Tuning the 5. Romero P, et al. (2001) Sequence complexity of disor- 10. Levy Y, Cho SS, Onuchic JN, Wolynes PG (2005) A survey heterogeneous early folding dynamics of phosphoglyc- dered proteins. Proteins Struct Funct Genet 42:38–48. of flexible protein binding mechanisms and their tran- erate kinase. J Mol Biol 333:187–199.

14238 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0807977105 Papoian Downloaded by guest on September 25, 2021