Introducing creative destruction as a mechanism in protein evolution

February 2, 2023
120 (6) e2220460120
Research Article
Creative destruction: New protein folds from old
Claudia Alvarez-Carreño, Rohan J. Gupta [...] Loren Dean Williams
To be innovative, molecular evolution needs to screen a gigantic search space, e.g., for a protein of “typcial” length with 300 amino acids, 20300 (equal to a number with 391 zeroes) combinations are possible. It has been a long-standing riddle how evolution has managed to find the presumably small patches in this sheer endless space, which host functional and structured proteins (1). A common theme in many areas of evolution is modularity, which helps to reuse existing parts efficiently and puts them together such that novel traits can be explored with less danger to go astray.
Following the modularity concept, the recombination of small independent units swiftly leads to novelty. By combining or rearranging those modular units, large and more complex ones can be generated enabling the exploration of new functions, without the need to go through the vast sequence space. Modularity is found at different levels in biological systems like the structure of macromolecules (e.g., proteins), protein–protein interaction networks, and gene regulation (2). In proteins, modular rearrangements are possible at the level of domains, motifs, and possibly even smaller modular units. Protein domains, for example, may be duplicated, fused, and rearranged by other processes to form novel proteins featuring new folds and functions (3). One more complex form of rearrangements, called circular permutation, has been described for several proteins and seems to be a universally present phenomenon in protein fold evolution (4). After a duplication event of a multidomain protein, e.g., AB with A and B denoting domains, leading to ABAB, terminal domains can be lost resulting in a new protein with the domains rearranged to BA (5).
In this issue of PNAS, Alvarez-Carreño et al. (6) add to this research area by looking at modularity in protein evolution at the level of single folds which comprise different secondary structure elements as building blocks. These single secondary structure elements can be reshuffled (analogous to circular permutation) in a way which the authors term “creative destruction.” This term, which the authors apply to molecular evolution, is borrowed from the field of economy where Schumpeter (7) used it to describe the process of innovation through the “destruction” of an ancestral product and thereby generating a new functional one. An example of creative destruction of a product from everyday life is the development of an mp3 player from a cassette player. The ancestral product, a cassette player, lost its overall form and the need of a cassette tape to develop into an mp3 player, the new derivative, that is able to play music over headphones from memory space. Form and function of both products are similar while some aspects of the ancestral product have been ‘destroyed’, i.e. cassette, bulky structure and loudspeaker, to create the new features, i.e. memory space, small, headphones.
Applied to protein evolution, this can be illustrated by three so-called -barrel folds which are presumed to be very ancient: SRC homology 3 (SH3), oligonucleotide/oligosaccharide binding (OB), and cradle-loop barrel (CLB) (Fig. 1). Secondary structure elements across all three folds share detectable sequence similarity, indicative of a common evolutionary origin. Furthermore, the overall similarity between the three folds is high, meaning that the spatial orientation of structural elements can be well aligned. Structural conservation was shown to be higher than sequence conservation in many protein families (8), proving useful when studying ancient proteins like ribosomal proteins. By using not only sequence similarity but also structural information, the ancient relationship between the folds could be detected. The three folds (SH3, OB, and CLB) are found in many ribosomal proteins that play a central role in translating RNA to protein sequences serving as a link between the two types of molecules (6). Ribosomal proteins are conserved across the tree of life and have therefore most likely existed long before the mysterious ancestor of all modern cellular organisms (the “last universal common ancestor,” also known as LUCA) spawned the three domains of life as we know them today: Bacteria, Archaea, and Eukarya (9).
“The study by Alvarez-Carreño et al. shines a new light on protein evolution and how new folds evolve.”
—Alvarez-Carreño et al.
Fig. 1.
The three -folds SH3, OB, and CLB were generated through the ‘creative destruction’ mechanism. OB has likely arisen by ‘creative destruction’ of the SH3 fold before existence of LUCA, because both folds are present in the universal ribosomal protein L2. The SH3 sequence was duplicated followed by destruction (symbolized by X) of several -sheets included in the original fold. The CLB fold likely originated by the same mechanism from the OB fold. Two of the -sheets from the OB fold have been remodeled into a loop (L) and a -helix (H). Homologous regions are displayed in the same colors in all folds.
Here, Alvarez-Carreño et al. (6) apply the creative destruction concept to explain the findings of their previous study (10), where they found an ancient homology between OB and SH3 folds, presumably dating back to pre-LUCA times. The corresponding sequences of individual SH3 and OB domains were aligned and superimposed structurally to determine a conserved core structure of the individual folds. The core structure residues were then used to create alignments between pairs of OB and SH3 folds as an intermediary method between sequence and structure alignment. Applying the concept of circular permutation to the OB and SH3 alignments, the authors were able to detect homology between OB and SH3 sequences that had previously been overlooked when using sequence alignments alone (10). The same workflow for homology detection was then applied to the OB fold and the CLB fold, the most recently evolved and most complex one of the three folds. Several pairs of ribosomal proteins containing one of the three β-folds can be used as examples for the creative destruction mechanism as shown in Fig. 1. SH3 and OB fold proteins range from universally conserved sequences (meaning they are present in all forms of life originating from LUCA) to sequences conserved in one domain of life (i.e., Bacteria), while the CLB fold proteins are conserved in single domains of life, indicating CLB origin after LUCA (6). The most probable original fold is SH3 because of simpler overall topology, compared to OB. After duplication of SH3, some of the terminal β-sheets are destroyed because of new structural constraints in the duplicated fold. This scenario is reminiscent of the cassette player being modeled into an mp3 player, where the bulky cassette is destroyed to make way for the much smaller mp3 player. Analogously, the destroyed secondary structure elements of SH3 are subsequently lost resulting in the OB fold. In the universal ribosomal protein L2 (Fig. 1), both ancient folds (SH3 and OB) are present, meaning their ancestor may have existed already before LUCA. The CLB fold evolved somewhat later (i.e., with LUCA or shortly after LUCA) by the same mechanism of creative destruction.’ A duplicated OB fold is put under different structural constraints resulting in the destruction of some secondary structure elements of the original OB fold. Here, some of the secondary structure elements of OB are not completely destroyed but remodeled from the β-sheet into a loop or α-helix (Fig. 1). This would be analogous to the bulky cassette player speakers being remodeled into mp3 headphones.
It is widely assumed that, before LUCA, in the RNA world, polypeptides coevolved with RNA to catalyze functions like replication and provide stability to RNA. The single-stranded RNA is less stable than, for example, double-stranded DNA, especially with the lack of a protecting cellular environment. The ribosome, nowadays comprising RNA as well as proteins, is probably one of the last and most ancient remnants of the RNA world. Ribosomal proteins show strong similarities across all domains of life and, therefore, are assumed to be the oldest protein classes and essential to the evolution of early biotic life. By studying the evolution of ribosomal proteins, we can try to gain a glimpse into the world back at the beginning of the evolution of life as we know it today (9). Since all three folds (SH3, OB, and CLB) can be found in ribosomal proteins and are therefore remnants of the pre-LUCA world, studying their emergence gives crucial mechanistic insights into early protein fold evolution. Indeed, the fold of the SH3 domain is a very frequent fold which provides specific binding interfaces in hundreds of different extant proteins (11). While plenty of research on protein rearrangements exists, it has so far been applied mostly to younger protein folds originating after LUCA. One well-known example is circular permutations of domain combinations that have changed the way we look for homology in proteins. Instead of comparing the proteins in full length, the order of domains is considered more flexible between proteins, resulting in more complex homologous relationships between proteins (5). The early evolution of domains and folds has been analyzed much less so far; however, homology between ancient folds has been detected more and more in recent years (12, 13). According to estimates, only few rudimentary folds have survived the diversification into Archaea, Bacteria, and Eukarya after LUCA (14). The most ancient folds that have survived seem to form simple barrels (like OB, SH3, and CLB) and include the Rossmann and P-loop folds. A recent study by Longo et al. (15) looked at these ancient folds (Rossmann and P-loop) and detected previously unknown homology between the two. Just like the OB and SH3 folds, the Rosmmann and P-loop share ancient homology that most probably predates LUCA. The study by Alvarez-Carreño et al. (6) shines a new light on protein evolution and how new folds evolve. The creative destruction mechanism offers a new explanation of how early evolution works. Such an explanation is particularly insightful because the same genetic processes that have rearranged protein domains over the last millions of years cannot have governed protein evolution at the dawn of times: Genomes as we know them today have not existed back then, enzyme complexes acting on DNA or RNA did not exist, and so forth. The new mechanism for generation of new folds as proposed by the authors complements other described mechanisms in protein evolution (like circular permutation) by taking into account that some elements of the ancestral fold are destroyed in the process to enable novelty. The destruction that takes place can be, for example, the loss of the ancestral secondary structure fold. The destroyed elements can be subsequently lost or remodeled to form the new fold. Researchers will now need to reassess other ancient folds and domains for applicability of the creative destruction mechanism. Examples include the following: Can creative destruction also be applied to explain the evolution of aforementioned Rossmann folds and P-loop folds? What other incidents can be followed by creative destruction, except for the here-described duplication of the fold? Potentially, any occurrence that leads to a new restriction in the protein fold can lead to creative destruction of its original form and give rise to a new fold. Further, it will be of interest to look at other building blocks in evolution and ask whether the creative destruction mechanism may be applied there as well. With this, we could expand our knowledge of different aspects of evolution.


M.A. acknowledges funding by the Volkswagen Stiftung (VWF), grant code 98183 through E.B.-B. We thank Klara Hlouchova for her opinion on the manuscript and Alun Jones for proofreading.

Author contributions

E.B.-B. designed research; and M.A. and E.B.-B. wrote the paper.

Competing interests

The authors declare no competing interest.


J. M. Smith, Natural selection and the concept of a protein space. Nature 225, 563–564 (1970).
G. P. Wagner, M. Pavlicev, J. M. Cheverud, The road to modularity. Nat. Rev. Genet. 8, 921–931 (2007).
E. Dohmen, S. Klasberg, E. Bornberg-Bauer, S. Perrey, C. Kemena, The modular nature of protein evolution: Domain rearrangement rates across eukaryotic life. BMC Ecol. Evol. 20 (year?).
W. C. Lo, C. C. Lee, C. Y. Lee, P. C. Lyu, CPDB: A database of circular permutation in proteins. Nucleic Acids Res. 37, D328–D332 (2009).
J. Weiner III, E. Bornberg-Bauer, Evolution of circular permutations in multidomain proteins. Mol. Biol. Evol. 23, 734–743 (2006).
C. Alvarez-Carreño, R. J. Gupta, A. S. Petrov, L. D. Williams, Creative destruction: New protein folds from old. Proc. Natl. Acad. Sci. U.S.A. 119, e2207897119 (2022).
J. A. Schumpeter, Capitalism, Socialism and Democracy (1942), p. 460.
K. Illergård, D. H. Ardell, A. Elofsson, Structure is three to ten times more conserved than sequence—A study of structural response in protein cores. Proteins: Struct. Funct. Bioinf. 77, 499–508 (2009).
A. N. Lupas, V. Alva, Ribosomal proteins as documents of the transition from unstructured (poly)peptides to folded proteins. J. Struct. Biol. 198, 74–81 (2017).
C. Alvarez-Carreño, P. I. Penev, A. S. Petrov, L. D. Williams, Fold evolution before LUCA: Common ancestry of SH3 domains and OB domains. Mol. Biol. Evol. 38, 5134–5143 (2021).
A. Musacchio, M. Wilmanns, M. Saraster, Structure and function of the SH3 domain. Prog. Biophys. Mol. Biol. 61, 283–297 (1994).
J. A. Farías-Rico, S. Schmidt, B. Hocker, Evolutionary relationship of two ancient protein superfolds. Nat. Chem. Biol. 10, 710–715 (2014).
M. Ammelburg et al., A CTP-dependent archaeal riboflavin kinase forms a bridge in the evolution of cradle-loop barrels. Structure 15, 1577–1590 (2007).
G. Caetano-Anollés, M. Wang, D. Caetano-Anollés, J. Mittenthal, The origin, evolution and structure of the protein world. Biochem. J. 417, 621–637 (2009).
L. M. Longo et al., On the emergence of P-Loop NTPase and Rossmann enzymes from a Beta-Alpha-Beta ancestral fragment. eLife 9, e64415 (2020).

Information & Authors


Published in

Go to Proceedings of the National Academy of Sciences
Proceedings of the National Academy of Sciences
Vol. 120 | No. 6
February 7, 2023
PubMed: 36730205


Submission history

Published online: February 2, 2023
Published in issue: February 7, 2023


M.A. acknowledges funding by the Volkswagen Stiftung (VWF), grant code 98183 through E.B.-B. We thank Klara Hlouchova for her opinion on the manuscript and Alun Jones for proofreading.
Author Contributions
E.B.-B. designed research; and M.A. and E.B.-B. wrote the paper.
Competing Interests
The authors declare no competing interest.


See companion article, “Creative destruction: New protein folds from old,”



Institute for Evolution and Biodiversity, University of Muenster, Muenster 48149, Germany
Erich Bornberg-Bauer2,1 [email protected]
Institute for Evolution and Biodiversity, University of Muenster, Muenster 48149, Germany
Max Planck-Institute for Biology, Tuebingen 72076, Germany


To whom correspondence may be addressed. Email: [email protected].
M.A. and E.B.-B. contributed equally to this work.

Metrics & Citations


Note: The article usage is presented with a three- to four-day delay and will update daily once available. Due to ths delay, usage data will not appear immediately following publication. Citation information is sourced from Crossref Cited-by service.

Citation statements



If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

View Options

View options

PDF format

Download this article as a PDF file


Get Access

Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Personal login Institutional Login

Recommend to a librarian

Recommend PNAS to a Librarian

Purchase options

Purchase this article to access the full text.

Single Article Purchase

Introducing creative destruction as a mechanism in protein evolution
Proceedings of the National Academy of Sciences
  • Vol. 120
  • No. 6







Share article link

Share on social media