# Fileset

[25_AdvSci.pdf](https://mdr.nims.go.jp/filesets/f0e145d3-3b00-4f42-b4d4-69f46ea6df4c/download)

## Creator

Toshimasa Homma, Rie Yamamoto, Lily Zuin Ping Ang, [Alaa Fehaid](https://orcid.org/0000-0001-7759-663X), [Mitsuhiro Ebara](https://orcid.org/0000-0002-7906-0350)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[A Novel Gene Synthesis Platform for Designing Functional Protein Polymers](https://mdr.nims.go.jp/datasets/281a42a8-c016-432c-9a70-d5085a21a53a)

## Fulltext

A Novel Gene Synthesis Platform for Designing Functional Protein PolymersRESEARCH ARTICLEwww.advancedscience.comA Novel Gene Synthesis Platform for Designing FunctionalProtein PolymersToshimasa Homma,* Rie Yamamoto, Lily Zuin Ping Ang, Alaa Fehaid,and Mitsuhiro EbaraRecombinant protein polymers with repeat sequences of specific amino acidscan be regarded as sustainable functional materials that can be designed usinggenetic engineering. However, synthesizing genes encoding these proteins issignificantly time-consuming and labor-intensive owing to the difficulty of usingcommon gene synthesis tools, such as restriction enzymes and PCR primers.To overcome these obstacles, a novel method is proposed herein: seamlesscloning of rolling-circle amplicons (SCRCA). This method involves one-potpreparation of repetitive-sequence genes with overlapping ends for cloning,facilitating the easy construction of the desired recombinants. SCRCA is used tosynthesize 10 genes encoding hydrophilic resilin-like and hydrophobic elastin-like repeat units that induce liquid-liquid phase separation. SCRCA showshigher transformation efficiency and better workability than conventionalmethods, and the time and budget required for SCRCA are comparable to thoserequired for non-repetitive-sequence gene synthesis. Additionally, SCRCAfacilitates the construction of a repeat unit library at a low cost. The libraryshows considerably higher diversity than that of the current state-of-the-artmethod. By combining this library construction with the directed evolutionconcept, an elastin-like protein polymer with the desired functions can berapidly developed. SCRCA can greatly accelerate research on protein polymers.1. IntroductionDiverse repetitive amino acid sequences exist in nature, eachpossessing distinctive functions.[1–5] Recombinant protein poly-mers inspired by these characteristics have attracted attentionas sustainable and biocompatible materials synthesized usingenvironment-friendly processes.[6,7] The applications of theseT. Homma, R. Yamamoto, L. Z. P. Ang, A. FehaidDivision of Chemical Engineering and BiotechnologyNational Institute of TechnologyIchinoseki College, TakanashiHagisho, Ichinoseki, Iwate 021-8511, JapanE-mail: hommatsh@ichinoseki.ac.jpThe ORCID identification number(s) for the author(s) of this articlecan be found under https://doi.org/10.1002/advs.202410903© 2025 The Author(s). Advanced Science published by Wiley-VCHGmbH. This is an open access article under the terms of the CreativeCommons Attribution License, which permits use, distribution andreproduction in any medium, provided the original work is properly cited.DOI: 10.1002/advs.202410903polymers are expanding to high-toughness fibers,[8,9] ultra-high elasticmaterials,[10,11] cell culture substrates,[12,13]bioseparation,[14,15] drug delivery,[16,17]proton-conducting membranes,[18] andsmart cells.[19] The ability to freely designrepeat units and chain lengths via geneticrecombination technology represents amajor advantage during the developmentof new protein-polymer materials.[3,20–22]To create new functional protein polymers,copolymers, and block copolymers con-sisting of two repeat units with differentproperties have been rationally desig-ned.[23–26] Moreover, polymer libraries ofdifferent lengths and repetitive sequenceshave also been constructed for identifyingmutants with novel functions throughhigh-throughput screening.[27–29]However, the synthesis of repetitive-sequence genes encoding protein polymersrequires time-consuming methods due tothe low number of unique parts whose se-quences can be recognized using restrictionenzymes or PCR primers.[30,31] Althougha codon-scrambling algorithm has beendeveloped to improve gene complexity,[32]screening for useful sequences is costly because it requiresmany custom oligo DNAs. Rolling-circle amplification (RCA) ispromising because it can simultaneously synthesize multiplerepetitive-sequence genes with different repeat numbers.[33,34]However, the previously reported RCA methods have sev-eral shortcomings, including the difficulty of isolating geneT. Homma, R. Yamamoto, A. Fehaid, M. EbaraResearch Center for Macromolecules and BiomaterialsNational Institute for Materials Science (NIMS)1-1 Namiki, Tsukuba, Ibaraki 305-0044, JapanA. FehaidForensic Medicine and Toxicology DepartmentFaculty of Veterinary MedicineMansoura UniversityDakahlia, Mansoura 35516, EgyptM. EbaraGraduate School of Pure and Applied SciencesUniversity of Tsukuba1-1-1 Tennodai, Tsukuba, Ibaraki 305-8577, JapanM. EbaraGraduate School of Advanced EngineeringTokyo University of Science6-3-1 Katsushika-ku, Shinjuku, Tokyo 125-8585, JapanAdv. Sci. 2025, 2410903 2410903 (1 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbHhttp://www.advancedscience.commailto:hommatsh@ichinoseki.ac.jphttps://doi.org/10.1002/advs.202410903http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/http://crossmark.crossref.org/dialog/?doi=10.1002%2Fadvs.202410903&domain=pdf&date_stamp=2025-02-23www.advancedsciencenews.com www.advancedscience.comfragments at a desired length from a library, inability to controlthe direction of insertion,[33] and requirement for a specificnucleotide sequence,[34] leaving room for improvement inoperability and versatility.We propose a new gene synthesis method, called seamlesscloning of rolling-circle amplicons (SCRCA). It consists of foursteps: preparation of ssDNA ring templates, RCA reaction, se-lection of DNA fragment size, and seamless cloning (Figure 1a).Seamless cloning simplifies the introduction of DNA fragmentsinto a plasmid vector at the desired insertion position and in aspecific direction, although overlapping sequences are requiredat both ends of the DNA fragments.[35] In SCRCA, these over-lapping sequences are added at each end of the gene whenthe repeat sequence is synthesized, improving the success rateand significantly reducing the time and labor required for genesynthesis. Moreover, through innovative strategies, this methodcan be used to construct copolymer and block copolymer genes,as well as polymer gene libraries with different repeat units(Figure 1b–d).In this study, we first synthesized repetitive-sequence genesencoding repeat units of a resilin-like polymer (RLP) and anelastin-like polymer (ELP), the major protein-polymer backbone.Cost analysis revealed that the SCRCA method is more econom-ical than the other methods in terms of repetitive gene synthesisand can be achieved with the same time and cost as the synthesisof non-repetitive sequence genes via a common method. In fact,copolymer sets designed by SCRCA can easily provide sequenceand functional insights that can be beneficial for rational poly-mer designing. Next, we proved that repeat unit libraries withrandom mutations can be constructed at a low cost via SCRCA.By performing the directed evolutionary experiment using theSCRCA-constructed libraries, we successfully developed a multi-responsive ELP in several months, a task that would have typicallyrequired several years. To the best of our knowledge, there areno published reports describing the application of directed evo-lution for developing protein polymers. Our results confirm thatSCRCA is an important platform for the advancement of protein–polymer research.2. Results and Discussion2.1. Development of the Gene Amplification ProcessFor the SCRCA, we devised a new RCA reaction using forwardand reverse primers with overlapping sequences at their 5′ ends.The Bst DNA polymerase large fragment was selected as thestrand displacement enzyme for RCA because it lacks 5′→3′ ex-onuclease activity, and therefore, the overlapping sequence ofthe primer would not be digested.[36,37] Further, its high opti-mum temperature (60–70 °C)[38,39] is appropriate for amplifyingGC-rich sequences, which are often present in major repetitivepolypeptides. A cyclized ssDNA with a sequence encoding the de-sired repeat unit was used as an RCA template. During isother-mal DNA amplification using these materials, DNA elongationand strand displacement reactions are expected to occur contin-uously (Figure 2a). Consequently, a mixture of genes with repet-itive sequences with different repeat numbers and overlappingsequences at both ends of the repetitive sequences can be pre-pared.To verify whether the reaction occurred as expected, isothermalamplification was performed using three ssDNA rings, namelyR, E, and RE, that have nucleotide sequences encoding thehydrophilic RLP repeat unit[40] [(GRGDSPYS)4]n, hydrophobicELP repeat unit[41] [(VGVPG)6]n, and RLP-ELP tandem repeatunit [GRGD-(SPYSGRGD)3-(GVPGVGVPGV)6-SPYS]n, respec-tively. RLP and ELP induce phase separation below an upper andabove a lower critical solution temperature (UCST and LCST),respectively. We considered that the advanced design of thesepolymers could contribute to the elucidation and application ofthe phase separation phenomenon.[42,43] Therefore, we adoptedthese polymers in our experimental models. Forward and reverseprimers with 15-base overlapping sequences at their 5′ ends wereused in reactions. All reactions generated a mixture of ampli-fied products of different lengths (Figure 2b–d). The positionsof the bands matched the corresponding theoretical values (re-peat number × ring size + overlapping sequence bases), indi-cating that the repetitive-sequence genes were successfully pre-pared (Figure 2e–g). As amplification products were obtainedeven when the large RE ring was used, SCRCA may also be suit-able for gene synthesis of polymers with long repeat units andcopolymers.Gene length can be selected by performing agarose gel elec-trophoresis, fractionating the sequences of different lengths, andcutting the desired ones from the gel. However, the conventionalRCA products exhibit smear-like bands due to the presence of in-complete double-stranded DNA, suggesting the presence of sev-eral sequences other than the targeted gene length.[29,33] In con-trast, the RCA product of the method described herein has clearlyseparated bands. This makes it possible to easily isolate the genewith the target length (Figure 2h). By increasing the run time ofagarose gel electrophoresis, the length of the genes isolated couldbe increased (Figure 2i,j).2.2. Repetitive-Sequence Gene ConstructionTo evaluate the probability of constructing a transformant withthe desired repetitive-sequence gene, E. coli transformation wasperformed using the In-Fusion seamless cloning system.[44] Ho-mopolymer and copolymer genes R6, E6, (RE)3, R12, E12, and(RE)6 were prepared via RCA reactions using R, E, and RE rings(the repetitive-sequence genes are written in italics accordingto their repeat unit and number; for example, R6 represents agene encoding a polypeptide with six repeats of the R ring se-quence). Purified DNA fragments were subjected to fusion withthe linear vector designed for expression and sequence confir-mation (Figure S1, Supporting Information). Block copolymergenes were synthesized via the simultaneous seamless cloningof two DNA fragments (Figure 1b). Here, four block copolymergenes, R3E3, R8E4, R6E6, and R4E8, were constructed in parallelby combining R3, R4, R6, R8, E3, E4, E6, and E8 blocks isolatedfrom two RCA reactions using R and E as templates accordingly.Colony PCR results showed that most transformants had thedesired gene length (Figure S2, Supporting Information). SangerDNA sequencing revealed that 81.4% of the transformants con-tained the desired gene when the gene length was ≈550 bp(R6, E6, (RE)3, and R3E3); even for genes with lengths >1000bp (R12, E12, (RE)6, R8E4, R6E6, and R4E8), 41.5% of all theAdv. Sci. 2025, 2410903 2410903 (2 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.comFigure 1. Seamless cloning of rolling-circle amplicons (SCRCA). a) Synthesis of a homopolymer gene using the SCRCA method: 1) ssDNA oligo with arepeat unit sequence is circularized; 2) A repetitive-sequence gene library with different repeat numbers is prepared via RCA (blue line: repetitive sequence;purple and yellow lines: overlapping sequences); 3) Amplified products are separated based on their repeat numbers via agarose gel electrophoresis,and the gene of the desired length is excised; 4) This gene is inserted into an expression vector via seamless cloning. Copolymer and block copolymergenes composed of two different repeat units are synthesized via RCA using a large-ring template b) and simultaneous seamless cloning c), respectively.d) A repeat unit library is constructed using an ssDNA oligo containing mixed bases.transformants exhibited the desired gene (Figure 3). Of the 32sequences analyzed, there was one large deletion resulting fromcloning failure and two deletions resulting from amplification ofincomplete templates (Figures S3–S7, Supporting Information).Base substitution or deletion errors were less than 0.02% of>29,000 bases, compared to those of the conventional RCAmethod.[33]2.3. Cost AnalysisTo evaluate the cost-effectiveness of SCRCA, we calculated thecost of synthesizing the homopolymer gene E12 alone and threeblock polymer genes R8E4, R6E6, and R4E8 together. Assumingthat all conventional methods can be used for synthesis, gene syn-thesis roadmaps were prepared based on data from previouslyAdv. Sci. 2025, 2410903 2410903 (3 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.comFigure 2. Isothermal rolling-circle amplification (RCA) for seamless cloning of rolling-circle amplicons (SCRCA). a) Schematic of the isothermal amplifi-cation. Blue, gray, and red lines: repetitive sequences; purple and yellow lines: overlapping sequences. The 3′ ends of the forward and reverse primers arehomologous to the gray and red lines, respectively. I) A DNA chain, in which the complementary sequence of the ssDNA ring is repeated, is elongatedfrom the reverse primer. II) Forward primer attaches to the elongated DNA chain, forming a dsDNA section. III) The nascent chain is dissociated bya subsequent elongation reaction. IV) Primers with complementary sequences attach to the dissociated nascent chain to form a dsDNA fragment. V)These reactions continue to occur to construct a library of repetitive genes with different repeat numbers. The 5′ and 3′ ends of the genes in the libraryhave overlapping sequences derived from the 5′ ends of the forward and reverse primers, respectively. (b–d) Amplicons obtained by RCA using the b)R, c) E, and d) RE rings were separated using 1.5% agarose gel electrophoresis. e–g) Band positions of the amplicons obtained by RCA using the (e) R,(f) E, and (g) RE rings, and the linear function (gray line) of their theoretical lengths (repeat number × ring size + overlapping sequence bases) of theseamplicons showed excellent correlation (R > 0.999). Each value represents the mean of three replicates, and the standard errors were less than 10 bpfor all means. h–j) Two DNA fragments equivalent to the 6 (h) and 12 repeats (i) of R ring sequences were cut from the gels electrophoresed for 70 and110 min, respectively. Panel (j) is an enlarged view of the cut gel, showing that the gel contained the 12-repeat DNA fragment.published papers[31–33] (Figures S8 and S9, Supporting Informa-tion). Costs for oligo DNA were determined based on the infor-mation provided on the websites of a contract oligo DNA syn-thesis company (Tables S1 and S2, Supporting Information). Forreagent costs, we calculated the required costs based on the quan-tities and concentrations listed in the Materials and Methods sec-tion of each paper (Tables S3 and S4, Supporting Information).For example, a reaction solution of overlap-extension rolling cir-cle amplification (OERCA) contains 50 μL of 25 mM dNTPs andrequires two reactions per gene synthesis; based on this, weAdv. Sci. 2025, 2410903 2410903 (4 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.comFigure 3. Success rates and costs of various repetitive-sequence gene syntheses. a) The success rates of gene syntheses were calculated from thepositivity ratios obtained via colony PCR and Sanger DNA sequencing. b) Comparison of synthesis cost of SCRCA with that of conventional methods.Assuming that synthesis can be performed according to the roadmap, we calculated the cost of synthesizing i) a homopolymer gene (E12) aloneand ii) block polymer genes (R8E4, R6E6, and R4E8) simultaneously. For the synthesis of the E12 gene, the combination of concatemerization andrecursive directional ligation by plasmid reconstruction (PRe-RDL),[31] OERCA,[33] and combinatorial codon scrambling (CCS)[32] were examined. Forthe synthesis of three block polymer genes, PRe-RDL combined with concatemerization, PRe-RDL combined with OERCA, and CCS were examined. Thecosts are expressed separately for oligo DNA, reaction reagents, and screening. For comparison, costs were calculated for the outsourced synthesis ofnon-repeat sequence genes of the same length.estimated the cost of dNTPs to be $16.7 (although 70% G/CdNTPs were actually used, to simplify the calculation, it wasassumed that equal amounts of mixed dNTPs were used).[33]In contrast, a reaction solution of SCRCA comprises 50 μL of1.5 mM dNTPs and requires only one reaction per gene synthe-sis; therefore, the cost of dNTPs is only $1.0.As the process of acquiring the desired constructs is critical ingenetic recombination experiments, we also calculated the cost ofscreening (Tables S6 and S7, Supporting Information). The num-ber of colonies required for screening was determined based onthe detailed comparative results of Amiram et al.[33] For exam-ple, they reported that using the OERCA method required exam-ining ≈200 colonies of which, only 2 colonies had >1 kb longgenes. They first used colony PCR to ensure that the directionof insertion was correct. Although the authors did not reveal thepercentage of colonies with the correct insertion orientation, wecan safely assume it to be less than 50% based on the resultsof this colony PCR (Figure 3b in the reference) (This is a rea-sonable result since they cloned the gene at the blunt end). For40% of the 200 colonies (80 colonies), we performed an additionalcolony PCR to determine the length. Candidate transformantswere selected from these 40% colonies, and their sequences wereconfirmed by outsourced DNA sequencing. Assuming that thetransformants carrying the E12 gene were isolated by these op-erations, the number of colonies needed to be examined wouldbe 280. In contrast, the method combining concatemerizationand PRe-RDL was estimated to require 100 colonies to obtainthe E4 gene from the concatemerization product and 8 colonieseach to construct the E8 and E12 genes from the obtained E4gene, thereby requiring a reagent cost of 116 colonies for theentire process (Figure S8, Supporting Information). Consider-ing the results of the aforementioned previous studies[33] andprinciples,[31] these numbers of colonies would be appropriate. Incontrast, with SCRCA, transformants with the desired sequencecan be obtained from ≈8 colonies, greatly reducing the numberof colonies to be tested (Figure S2, Supporting Information). Thecosts of the other methods were calculated in the same manner.The SCRCA cost is estimated to be half of the conventionalmethod cost and is equivalent to the cost of outsourced synthe-sis of non-repetitive genes (Figure 3b; Tables S1–S7, SupportingInformation). Concatemerization and OERCA, like SCRCA, cansimultaneously synthesize a variety of long and short repetitivegenes,[31,33] but the screening cost is high due to the large num-ber of colonies with genes of non-target length or those insertedin the reverse direction. In other words, the cost-effectiveness ofSCRCA can be attributed to precise control of gene length and ahigh cloning success rate. Notably, only two weeks were requiredfrom ordering the oligo DNA to preparing the 10 repetitive-sequence genes. Since the outsourced synthesis of a gene of sim-ilar length takes 8–13 business days (Economy Gene SynthesisService of Eurofins Genomics, Inc.), the workability of SCRCA isalso comparable to that of common synthesis methods for non-repeat sequence genes.2.4. Utilization of Copolymer Set Prepared by SCRCACopolymerization or block copolymerization using two repeatunits with different functions is a powerful method to developnew functional protein polymers;[25,26,45–47] however, to createcopolymers with the desired functionality, the effects of copoly-merization efficacy and composition ratio must be investigatedAdv. Sci. 2025, 2410903 2410903 (5 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.com00.20.40.60.81.020 40 60 80AbsorbanceTemperature (°C) Temperature (°C) Temperature (°C)Temperature (°C)Temperature (°C)Temperature (°C)R1200.20.40.60.81.020 40 60 80Absorbance(RE)600.20.40.60.81.020 40 60 80AbsorbanceE1200.20.40.60.81.020 40 60 80AbsorbanceR8E400.20.40.60.81.020 40 60 80AbsorbanceR6E600.20.40.60.81.01.21.420 40 60 80AbsorbanceR4E8R12 (RE)6 E12R8E4 R6E6 R4E8ab c20406080ELP Fraction0% 32% 48% 65% 100%Two phasesOne phaseOne phaseTemperature (°C)Figure 4. Temperature responses of the polymers prepared by seamless cloning of rolling-circle amplicons. a) Phase separation behaviors of polymersR12, E12, (RE)6, R8E4, R6E6, and R4E8 with the increase in temperature. The samples were prepared in PBS (pH 7.4) containing 2 μM polypeptide and8 mM residual urea. Each value represents the mean of three replicates, with error bars showing the standard errors. b) Appearance of the polymersolution during absorbance measurements. Left: at 25 °C, right: at 80 °C. c) Effect of the ELP fraction on temperature-sensitive phase separation. TheELP fraction indicates the ratio of the ELP portion to the total repetitive sequence.in advance. Copolymer sets that can be constructed using SCRCAcan accelerate such investigations. To test this idea, we purifiedthe constructed ELP-RLP copolymer set (Figure S10 (SupportingInformation), the polymers corresponding to genes are writtenin Roman type) and investigated the effect of the block ratio ontemperature-sensitive phase separation (UCST and LCST). Asshown in Figure 4a,b, E12, R4E8, and R6E6 showed cloudinessupon heating, R8E4 showed cloudiness over a wide temperaturerange of 15–85 °C, whereas R12 lost cloudiness upon heating.Microscopic observation confirmed that this cloudiness wasattributed to phase separation (Figure S11, Supporting Informa-tion). These results indicate that to exhibit the phase separationphenomenon over a wide temperature range, the ELP fractionshould be adjusted to ≈30% (Figure 4c). Considering that thefunction of a polymer is affected by its concentration and thesurrounding environment,[40,48–52] it will be highly beneficial toprepare copolymer sets and evaluate their function in the actualenvironment where they will be used. Notably, for the conven-tional PRe-RDL, extra constructs, and multiple transformationsare required when preparing copolymer sets,[26,45,53–55] whereasSCRCA does not require extra constructs and facilitates thepreparation of copolymer sets in a single transformation (FigureS9, Supporting Information). As the simultaneous cloning ofvarious block genes facilitates the construction of block polymersof different lengths and compositions at once, SCRCA can bean ideal gene synthesis method to comprehensively investigatethe effects of sequence and length on the function of blockpolymers.Interestingly, copolymer (RE)6 and block copolymer R6E6,which have the same amino acid composition, showed similarAdv. Sci. 2025, 2410903 2410903 (6 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.comFigure 5. Construction of the repetitive-sequence gene library. a) Nucleotide sequence of the E-3NDT ring template library. b) At each NDT codonsite, one of 12 types of amino acids is encoded by the combination of bases. c) Alignment result of the nucleotide repeat units found in the libraryconstructed by RCA using the E-3NDT ring template library. The height of each nucleotide corresponds to its frequency. N and D represent the mixedbase introduction position in E-3NDT ssDNA. d) Ratios of mutants and errors among all sequences. Each value represents the mean of two replicates.The corresponding standard deviations are 0.7 and 1.4, respectively. e) Abundance ratios of the most abundant mutant, other mutants, and errorsequences. Each value represents the mean of two replicates. The corresponding standard deviations are 0.083%, 0.088%, and 0.005%, respectively. f)The frequency distribution of mutant sequences. Gray and red lines indicate theoretical and experimental medians, respectively. Values represent themean of two measurements, and error bars represent standard errors.temperature responses. When the ELP fraction was greater than48%, the ELP moiety determined the temperature responsive-ness of the polymer. As the protein-induced phase separation incells is sequence-dependent,[56,57] the results of the present studymay be specific to ELPs that undergo phase separation at hightemperatures. Indeed, block and pseudorandom polymers of ELPmutants reportedly exhibit similar LCSTs.[58] Due to the smallnumber of reported cases, future validation experiments usingSCRCA on a diverse range of sequences are warranted.2.5. Constructing the Repeat Unit LibraryDirected evolution is a promising technique to develop highlyfunctional enzymes and antibodies.[59] In this technique, the fol-lowing four steps are repeated: 1) Construction of a gene libraryby mutagenesis; 2) Expression of genetic information in the formof protein; 3) Selection of the most suitable mutant using a sim-ple evaluation method; 4) Designing next-generation libraries us-ing the genes of the selected mutants as templates. To the bestof our knowledge, directed evolution has not been previouslyused to develop protein polymers. This may be due to the factthat mutagenesis methods suitable for low-complexity proteinssuch as protein polymers have not yet been explored. We con-sidered that a repetitive-sequence gene library with repeat unitscontaining mutations would be useful for the directed evolutionof protein polymers and investigated an approach for the prepa-ration of a ring-template mixture using ssDNA containing mixedbases (Figure 1d). ELP was selected as the model sequence; the Vsites of VGVPGVGVPGVGVPGVGVPGVGVPGVGVPG can bereplaced with a guest amino acid, which changes the tempera-ture responsiveness of the ELPs[60] (Figure 5a). We designed anE-3NDT ssDNA oligo in which the three codons (X1, X2, and X3)encoding the V sites were replaced with NDT codons.[61] Thiscodon encodes one of 12 diverse amino acids (Figure 5b).To confirm the construction of the library, RCA products,which were presumed to have five repeats of the E-3NDTunits, were analyzed using a next-generation sequencer. Thenucleotide repeat units present at each end of the gene (the firstat the 5′-end and the fifth at the 3′-end) could be read accurately.However, the sequence information at the second to fourthnucleotide repeat units was unreliable probably owing to poorAdv. Sci. 2025, 2410903 2410903 (7 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.comcluster formation due to the repetitive sequences. The counts ofeach sequence were almost the same in the first (at the 5′-end)and fifth (at the 3′-end) repeat units (R = 0.991), indicatingthat single nucleotide repeat units were repeated five times in amajority of the genes. Multiple alignments of the observed unitsconfirmed that the E unit backbone was maintained (Figure 5c).In contrast, high complexity was observed at sites where mixedbases were introduced. Although this library contained all the1728 expected sequences (all combinations of the mixed bases),it contained few error sequences (Figure 5d). Notably, even themost abundant sequences represented only <0.9% of the totalnumber of sequences, and the median abundance ratio of allvariants was 0.035%± 0.01%, close to the ideal value (100%/1728sequences = 0.058%) (Figure 5e,f). These results indicate thatthis library possessed excellent diversity. The minor deviationfrom the ideal value is believed to be due to a bias in the basetypes during the introduction of mixed bases into the oligo DNA.Genes having six repeats of the E-3NDT unit were introducedinto the E. coli expression system mentioned above. As E-3NDTis based on the E unit, this can be termed an E6 mutant li-brary. We examined 42 transformed colonies and selected 28colonies with the correct gene lengths (Figure S12, SupportingInformation). Sanger sequencing revealed that the majority ofthe selected transformants carried E6 mutant genes with differ-ent amino acids (Figure 6a,b; the mutants are denoted by “m”and the colony number). Interestingly, block copolymer mutantswith two or more different repeat units were obtained in addi-tion to E6 mutants with the same unit repeated six times (asexpected) (Figure 6b). We ascribe this to seamless cloning be-tween the amplicons with ssDNA regions (Figure 6c). This phe-nomenon, which may also have occurred when constructing thegenes comprising only one repeat unit, might facilitate the syn-thesis of long repetitive-sequence genes via SCRCA. Consideringthat polypeptide aggregation is affected by mean hydropathy andnet charge,[62] the presence of block copolymer genes diversifiedthe ELP library (Figure 6d).Recently, a library construction technique applicable to low-complexity sequences, such as ELP, has been reported; however,this technique presents challenges, such as transformation effi-ciency of less than 25% and low library diversity (the occupancyof the most abundant sequences is more than 30%).[29] In con-trast, libraries constructed using SCRCA have a transformationefficiency of ≈70% and excellent diversity. In addition, the costassociated with SCRCA is only one-tenth that of the previouslyreported method because the introduction of mixed bases is free.These advantages considerably reduce the hurdles in applyingevolutionary molecular engineering to the development of func-tional protein polymers.2.6. Development of the Desired Functional Protein Polymer byDirected EvolutionFinally, we tested whether the repeat unit library could be usedto rapidly develop the desired protein polymers. As a develop-ment target, we selected a multi-response ELP that is soluble atlow temperatures (4 °C, pH 7.4), insoluble near body temper-ature (37 °C, pH 7.4), and soluble in the environment aroundcancer cells (37 °C, pH 6.5). Such protein polymers are usefulvehicles for drug delivery to cancer cells.[24,63] To develop poly-mers with these complex properties, molecules with LCSTs inthe range of 4–37 °C at pH 7.4 and >37 °C at pH 6.5 must bedesigned (Figure 7a). The rational design of such polymers fromthe basic ELP sequence requires several years as factors, such asamino acid substitution, chain length, and influence of the op-erating environment, require careful examination to determinetheir effect on polymer responsiveness to temperature and pHchanges.[64–67]In the directed evolution experiment, we aimed to develop thedesired polymer by optimizing the V positions of E6, which hasan LCST higher than body temperature and does not respond topH changes (Figure 7b). A problem encountered during screen-ing is that the temperature responsiveness of ELP is easily af-fected by its concentration and the presence of contaminants.[64]Therefore, prior to screening, we predicted the polymer functionsfrom the amino acid sequences and then selected some mutantsto reduce the number of samples. By evaluating the temperatureresponsiveness of the selected mutants after purification, the im-provement in the function can be compared. We also conductedat least two rounds of experiments to confirm that the librariesconstructed via SCRCA are applicable to the directed evolution.The screening of the first directed evolution round involvedsearching for protein polymers that exhibited temperature andpH responsiveness in the E6 mutant library. Cysteine-containingmutants (m9, m33, and m40) were excluded owing to their poten-tial to form irreversible aggregates. Block copolymers were alsoexcluded because of the complexity of the next-generation librarydesign. Consequently, three mutants, m8, m10, and m19, whosenet charge changed significantly in response to a small changein pH, were selected and purified (Figure 7c; Figure S13, Sup-porting Information). The mutants m8 and m10 exhibited bothtemperature and pH responsiveness (Figure 7d). However, theywere water-soluble near body temperature.The first-round results (Figures 6b and 7c,d) revealed that thesubstitution of histidine results in a response to the small pHchange. Lowering LCST while maintaining pH responsivenesswould require the introduction of functional groups that interactwith histidine, such as tyrosine, phenylalanine, and histidine.[68]Therefore, we designed the m10-derived sequence m10-3YWY,which has YWY codons encoding one phenylalanine, tyrosine,leucine, and histidine in the remaining three V sites (positionsX4, X5, and X6) (Figure 8a,b). Among the 37 transformants exam-ined using colony PCR and Sanger sequencing (Figure S14, Sup-porting Information), 27 mutants possessed the m10 backbonewith different amino acids at positions X4, X5, and X6 (Figure 8c;Figure S15, Supporting Information). This indicates that a tem-plate design can be used to construct a protein polymer librarythat inherits the characteristics of the useful sequences identifiedin the previous round.To reduce the library size, we selected seven unique mutantsbased on their amino acid composition (Figure 8d) and purifiedthem (Figure S16, Supporting Information). We measured tur-bidity in low-temperature (4 °C, pH 7.4), biological (37 °C, pH7.4), and slightly acidic (37 °C, pH 6.5) environments on a mi-croplate scale. Based on the result (Figure 8d), we selected m76,which agglutinated only at 37 °C, pH 7.4, as a strong candidate; itstemperature and pH responses were close to the target properties(Figure 8e,f). This result indicates that the SCRCA-assisted di-Adv. Sci. 2025, 2410903 2410903 (8 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.comFigure 6. Composition of the E6 mutant library. a) Alignment result of the repetitive amino-acid-sequence parts (between 3 and 183 aa) of 27 E6 mutantsfabricated by SCRCA using the E-3NDT ring template. The height of each amino acid corresponds to its frequency. The X1, X2, and X3 positions representNDT codons in the E-3NDT ring template. b) For 27 E6 mutants, the amino acids at X1, X2, and X3 are summarized. The mutants highlighted in redhave a block copolymer sequence consisting of two or more repeat units. c) Principle of block copolymer formation during library construction. Duringisothermal amplification (described in Figure 2a), dsDNAs I) with an ssDNA region at the 3′ end and II) with an ssDNA region at the 5′ end are generated,respectively (blue, gray, and red lines: repetitive sequences; purple and yellow lines: overlapping sequences). III) They are annealed, and IV) the nick isrepaired during the In-Fusion reaction to generate a block copolymer gene. The dark blue line indicates a repeat unit that is partly different from thatindicated by the blue line. d) Mean hydropathy and net charges are plotted for wild-type E6 (blue), 18 mutants with a homopolymer gene (gray), and 9mutants with a block copolymer gene (red).Adv. Sci. 2025, 2410903 2410903 (9 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.coma4 37 7000.20.40.60.81.0AbsorbanceTemperature (℃)c4 37 70AbsorbanceTemperature (℃)pH 7.4 pH 6.54 37 7000.20.40.60.81.01.2AbsorbanceTemperature (℃)m84 37 7000.20.40.60.8AbsorbanceTemperature (℃)m104 37 7000.20.40.60.8AbsorbanceTemperature (℃)m192468E6m1m4m5m8m10m12m19m20m21m22m24m26m2 9m32m39Δ ne t char gebdE6Figure 7. The first round of the directed evolution. a) The development target must have a transition temperature between 4 and 37 °C in a biologicalenvironment (pH 7.4) and >37 °C in a slightly acidic environment (pH 6.5). b) Temperature and pH responses of E6. The sample was prepared inphosphate-buffered saline (PBS) solutions containing 5 μM E6 (blue line: pH 7.4; red line: pH 6.5). Each value represents the mean of three replicates,with error bars showing the standard errors. c) Net charges at pH 7.4 and pH 6.5 are calculated based on the respective amino acid sequences. d)Temperature and pH responses of the three selected E6 mutants, i.e., m8, m10, and m19. The samples were prepared in PBS solutions containing5 μM polymer (blue line: pH 7.4, red line: pH 6.5). Each value represents the mean of three replicates, with error bars showing the standard errors. Thetemperature response of m19 at pH 6.5 was not evaluated because this mutant did not show a lower critical solution temperature below 95 °C.rected evolution strategy could facilitate the development of a de-sired protein polymer with a few transformation operations andwithin several months.3. ConclusionWe developed a repetitive-gene synthesis method that com-bines RCA and seamless cloning and demonstrated that variousgenes encoding protein polymers can be easily synthesized. TheSCRCA method has a high success rate of gene synthesis and ex-cellent workability, and block copolymer genes, which normallyrequire multiple cloning steps, can be constructed in a singlecloning operation. As the operation days and cost of SCRCA arecomparable to those of standard gene synthesis, SCRCA is ex-pected to be a fundamental technology that will accelerate therational design of functional protein polymers.We also demonstrated that protein polymer libraries with dif-ferent repetitive units can be easily constructed by introducingmixed bases into the template ssDNA. The library constructionusing SCRCA achieved superior transformation efficiency anddiversity in relation to that using the state-of-the-art method. Fur-thermore, by combining this library construction technique withthe directed evolution concept, we proved that the developmentof a highly functional protein polymer can be easily achievedwithout the need for diligent research. This development strat-egy may be suitable for further functionalization of conventionalprotein polymers and for the search for polymers with unknownfunctions. Considering its suitability for the directed evolution oflow-complexity sequences, this strategy may also be used to un-derstand diseases involving intrinsically disordered proteins[69,70]and to develop bioproduction processes using intracellular phaseseparation.[71] Future research will focus on efficiently screen-ing desired protein materials and transformants. As proteinsand polypeptides have a wide range of applications, a screeningmethod that is suitable for development purposes must be pro-posed. Further studies will advance SCRCA to a technology thatcontributes substantially to multiple fields.4. Experimental SectionSynthesis of the ssDNA Ring: DNA was synthesized by Eurofins Ge-nomics, Inc. (Tokyo, Japan). The ssDNA ring template was designed basedon the repeat-unit sequence and synthesized as 5′-phosphorylated ssDNA.The forward and reverse primers were designed based on the beginningand end sequences of the repeat unit, respectively. By designing the re-verse primer to anneal to both ends of the 5′-phosphorylated ssDNA, thereverse primer was also used as a split oligo for ssDNA cyclization. Thecombinations of the ssDNA and primers used in this study are summa-rized in Table S8 (Supporting Information).A mixture of 2 μL 50 μM 5′-phosphorylated ssDNA, 4 μL 50 μM reverseprimer, and 29 μL milli-Q water was heated at 95 °C for 2 min. After coolingat 4 °C, 1 μL T4 DNA ligase (400 000 cohesive end units/mL, New EnglandBiolabs) and 4 μL 10× T4 DNA ligase buffer were added, followed by incu-bation at 20 °C overnight. The product was purified using a GenElute PCRClean-up Kit (Sigma-Aldrich, St Louis, MO, USA). The purity and concen-tration were determined using a Nanodrop 2000 device (Thermo FisherScientific, Waltham, MA, USA).To prepare the RE ring template, a split DNA (ACCAGGAACACCATCAC-CACGACC) was designed based on both ends of the 5′-phosphorylatedssDNAs of the R and E rings. A mixture of 2 μL of each of the 50 μM 5′-phosphorylated ssDNAs, 4 μL of the 50 μM split DNA, 4 μL of the 50 μMreverse primer, and 23 μL milli-Q water was prepared. Cyclization was per-formed as described above.Adv. Sci. 2025, 2410903 2410903 (10 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.comFigure 8. The second round of the directed evolution. a) Nucleotide sequence of the m10-3YWY ring template. b) At each YWY codon site, one of fourtypes of amino acids is encoded by the combination of bases. c) Alignment result of repetitive amino-acid-sequence parts (in the range of 3–183 aa) of27 m10 mutants prepared by SCRCA using the m10-3YWY ring template. The height of each amino acid corresponds to its frequency. The X4, X5, andX6 positions represent YWY codons in the m10-3YWY ring template. d) Results of simple turbidity measurements. Seven m10 mutants, i.e., m48, m60,m62, m64, m65, m71, and m76, were selected based on their unique amino acid compositions (insert). The samples were prepared in PBS (pH 7.4)containing 5 μM polymer and <270 mM residual urea. Turbidity was measured at 4 and 37 °C, and then, the pH was adjusted to 6.5 by adding lactic acid.Each value represents the mean of three replicates, with error bars showing the standard errors. Statistical differences were determined using one-wayANOVA with Dunnett’s multiple comparisons post-test. Differences in turbidity (at pH 7.4 and 4 °C) relative to the PBS blank: no asterisk p > 0.1;*p < 0.1; ****p < 0.0001. e) Temperature and pH responses of m76. The samples were prepared in PBS solutions containing 5 μM polymer and 80 mMresidual urea (blue line: pH 7.4, red line: pH 6.5). f) Appearance of m76 polymer solution. Left: 37 °C, pH 7.4; right: 37 °C, pH 6.5.Adv. Sci. 2025, 2410903 2410903 (11 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comwww.advancedsciencenews.com www.advancedscience.comRCA: A reaction solution of 100 μL containing 0.32 units μL−1 BstDNA polymerase large fragment (New England Biolabs), 0.5 μM forwardprimer, 0.5 μM reverse primer, 0.2 ng μL−1 ssDNA ring template, 1.5 mMdNTP mix, 1× Thermopol buffer, and 2 mM magnesium sulfate was pre-pared. Isothermal amplification was performed at 60 °C for 12 h. A 1.5%TBE agarose gel was used to confirm the amplification products. Theagarose gels were stained with ethidium bromide after electrophoresisand photographed using the Dolphin-Doc imaging system (Wealtec Corp.,Sparks, NE, USA). Band positions were calculated using CS Analyzer 3.0(ATTO Corporation, Tokyo, Japan).Cloning and Transformation: For size selection of the DNA fragments,2.5% TBE agarose gels containing 0.01% (v/v) SYBR Safe DNA were used.After electrophoresis, the band positions were checked using a blue-lightilluminator, and the band corresponding to the desired length was cut us-ing a cutter. The DNA fragments were purified using the FastGene Gel/PCRExtraction Kit (Nippon Genetics, Tokyo, Japan). The purity and concentra-tion were determined using the Nanodrop 2000 device.Expression vectors were constructed using an In-Fusion cloning kit(Takara Bio Inc., Shiga, Japan) according to the manufacturer’s in-structions. A linear vector was constructed via inverse PCR using thepET22b vector and two primers (TGGCCGACTCATCATCACCACCACCACand TTCATATGTATATCTCCTTCTTAAAGTTAAAC). The sequences of theseprimers were designed to add the amino acid sequences MK to the N-terminus and WPTHHHHHH to the C-terminus of protein polymers. Es-cherichia coli BLR (DE3)-competent cells prepared in the laboratory wereused for the transformation.Colony PCR, Sanger DNA Sequencing, and Mass Spectrometry: The T7-promoter primer (TAATACGACTCACTATAGG) and T7-terminator primer(GCTAGTTATTGCTCAGCGG) were used for colony PCR and Sanger DNAsequencing. Colony PCR was performed according to the instructions forthe Go Taq Green Mix (Promega, Madison, WI, USA), except that the an-nealing temperature was set to 50 °C, and the elongation time was set to2 min for short genes (≈550 bp) and 3 min for long genes (≈1100 bp).Sanger DNA sequencing was performed by Fasmac Co. Ltd (At-sugi, Japan). BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) was usedto analyze the sequence results. For genes with repetitive sequences of>1000 bp, the results of decoding in the 5′→3′ and 3′→5′ directions werecombined (duplicated portions were confirmed to be >60 bp). The posi-tivity rate (%) was calculated as [number of positive results from colonyPCR / number of tests for colony PCR] × [number of positive results fromSanger DNA sequencing/number of tests for DNA sequencing].For protein polymers whose repeat counts could not be confirmedby Sanger sequencing alone [R12, E12, and (RE)6], their masses weremeasured using matrix-assisted laser desorption ionization time-of-flightmass spectrometry. The measurements were performed by Apro Science(Tokushima, Japan).Preparation of Block Copolymer Genes: For simultaneous seamlesscloning, RLP-block genes were prepared via RCA with a forward primer(GATATACATATGAAAGGGCGCGGTGACTCTCC) and a reverse primer(GACACCTACTGAGTAAGGTGAATCACCACGACC). These genes have anoverlapping sequence at their 3′-end that recognizes the 5′-end of ELP-block genes. The ELP-block genes were prepared using RCA with aforward primer (TACTCAGTAGGTGTCCCAGGTGTCGG) and a reverseprimer (ATGATGAGTCGGCCAACCAGGAACACCAACACCAGGTAC). Thedesired band of the block gene was cut out from the gel and purified. Twoblock genes and a linear vector were linked via an In-Fusion reaction.[44]Cost Calculation: SCRCA synthesis costs were compared with thosefor PRe-RDL,[31] OERCA,[33] CCS,[32] and a common method ofnon-repeat gene synthesis. The costs were calculated separately foroligo DNA synthesis, reaction reagents, and screening. The costsfor oligo DNA and non-repeat gene synthesis were determinedbased on the data provided on the websites of Eurofins GenomicsInc. (https://eurofinsgenomics.jp/jp/product/oligo-dna/standard-oligo-overview.aspx and https://eurofinsgenomics.jp/jp/service/gsy/overview.aspx). Reagent and screening costs were calculated by multiplying the costper sample by the number of samples. Reagent prices were checked onlinein April 2024 and calculated at a currency conversion rate of ¥150 to $1.OERCA and concatemerization yield several mutants with non-purposefulrepeat numbers, necessitating the screening of a large number of colonies.Therefore, based on the number of tests in the literature,[33] we deter-mined that 280 and 100 colonies per transformation were required to eval-uate OERCA and concatemerization, respectively.Expression and Purification: Transformants were inoculated into a200 mL ZYP-5052 medium and incubated at 25 °C for 2–3 days. Afterharvesting by centrifugation (9000×g, 5 min), the bacteria were stored at−30 °C. Protein polymers R12, E12, (RE6), R8E4, R6E6, and R4E8 werepurified using His-tag affinity resin (His GraviTrap or Ni Sepharose 6FastFlow, GE Healthcare, Chicago, IL, USA). Initially, 10 mL binding buffer(20 mM phosphate buffer (pH 7.4) containing 4 M urea, 0.5 m NaCl, and40 mM imidazole) was added to the defrosted bacteria, and the bacteriawere then sonicated on ice. After insoluble matter was removed via cen-trifugation, the supernatant was incubated at 37 °C for 10 min. Centrifu-gation was again performed to remove the generated insoluble matter,and the supernatant was passed through a 0.2 μm filter for sterilization.Next, affinity purification was performed according to the manufacturer’sinstructions, except that 0.04 and 4 m urea were added to the binding andelution buffers, respectively. The eluate was concentrated using an AmiconUltra 10 kDa column (Millipore, Burlington, MA, USA). Protein polymersE6, m8, m10, and m19 were purified via the inverse transition cycling (ITC)method.[15] The purified polypeptides were dissolved in milli-Q water orPBS. Protein polymers m48, m60, m62, m64, m65, m71, and m76 werepurified using the His-tag affinity resin and then further purified using theITC method. The purified products were dissolved in 20 mM phosphatebuffer (pH 7.4) containing 4 M urea to maintain their high concentrationsand stored at −20 °C.Verification of Protein Polymer Purity: Purified protein polymer concen-trations were calculated by measuring the absorbance at 280 nm. The mo-lar absorption coefficient was determined using the method reported byPace et al.[72] To check for contamination by other proteins, sodium do-decyl sulfate-polyacrylamide gel electrophoresis was performed, and thegels were stained with Coomassie brilliant blue.Evaluation of the Temperature Response: Absorbance was measuredat 350 nm using a V-650 spectrophotometer (Jasco, Tokyo, Japan)connected to a temperature controller. PBS at pH 7.4 was used as theblank. To facilitate the adjustment of the polymer concentration, themeasurement solution was prepared by diluting a concentrated solution.Some concentrated solutions contained 4 m urea to prevent phaseseparation at room temperature. The residual urea concentration in themeasurement solution is presented in the figure legends. The temper-ature was increased by 1 °C per min. PBS at pH 6.5 was prepared byadding 22 μL of 50-fold diluted lactic acid to 1000 μL of the measurementsolution.Microscopic Observations: Images of phase separation were capturedat 400 × magnification using an Olympus CKX53 microscope and Olym-pus CellSens software (Standard Version 4.1.1, Tokyo, Japan). A samplesolution of 100 μL was placed in a transparent flat-bottomed 96-well plate,imaged at 25 °C, heated to 45 °C on a hot plate, and imaged again. Forcomparison, the same imaging was performed using PBS without the poly-mer.Next-generation Sequencing Analysis: MiSeq Illumina sequencing(Illumina, San Diego, CA, USA) and data analysis were performed byHokkaido System Science Co., Ltd (Hokkaido, Japan). The sequencingsamples were synthesized via RCA using a set of primers (TCGTCG-GCAGCGTCAGATGTGTATAAGAGACAGTGTAGGTGTCCCAGGTGTCGGand GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGACCAGGAACAC-CAACACCAGGTAC) with linker sequences at their 5′-ends. For sizeselection, agarose gel electrophoresis was conducted, and fragmentscorresponding to five repeats were excised from the gel. After purification,the samples were submitted to the company. After sequencing, thecompany removed the adapter sequence using Cutadapt[73] and trimmedthe low QV region using Trimmomatic.[74] Cutadapt and Trimmomaticparameters are shown in the Supporting Information. The analysis re-vealed that the middle portions of the gene sequences were less reliable;therefore, only the sequences that were 90 bp from the 5′-ends and 90 bpfrom the 3′-ends were included in the analysis. However, since sequenceswith low counts lack reliability, only those with six or more counts wereAdv. Sci. 2025, 2410903 2410903 (12 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comhttps://blast.ncbi.nlm.nih.gov/Blast.cgihttps://eurofinsgenomics.jp/jp/product/oligo-dna/standard-oligo-overview.aspxhttps://eurofinsgenomics.jp/jp/product/oligo-dna/standard-oligo-overview.aspxhttps://eurofinsgenomics.jp/jp/service/gsy/overview.aspxhttps://eurofinsgenomics.jp/jp/service/gsy/overview.aspxwww.advancedsciencenews.com www.advancedscience.comanalyzed. All sequences in a library were aligned in parallel using Weblogo(weblogo.berkeley.edu/logo.cgi).Mean Hydropathy and Net Charge: The mean hydropathy and netcharge were calculated using Expasy (https://web.expasy.org/protscale/)and PROTEIN CALCULATOR 3.4 (https://protcalc.sourceforge.net/), re-spectively. Calculations were performed using default settings.Simple Turbidity Test: Polymer solutions of 150 μL (5 μM polypeptidein PBS, pH 7.4) were prepared in the cells of 96-well microplates (clearbottom, half area). As the measurement solutions were prepared by di-luting the polymer concentrates with 4 m urea, the solutions contained40–270 mM urea, which should have little effect on the transition temper-ature of ELPs.[75] The solutions were incubated at 4 °C for 30 min, and ab-sorbance was measured at 350 nm. The solutions were further incubatedat 37 °C for 30 min, and absorbance was measured again. Then, 3.3 μL of50-fold diluted lactic acid was added to reduce the pH to 6.5. The solutionwas again incubated at 37 °C for 30 min, and absorbance was measuredat 350 nm. PBS was used as a blank. The experiments were performed onthree cells under identical conditions.Statistical Analysis: The correlation coefficients were calculated, andstatistical tests were performed using KaleidaGraph Version 4.5.3 (Syn-ergy Software, Eden Prairie, MN, USA). One-way ANOVA with Dunnett’stest was used for multiple comparisons. Differences with p < 0.1 wereconsidered statistically significant for improving screening accuracy.Supporting InformationSupporting Information is available from the Wiley Online Library or fromthe author.AcknowledgementsThis study was partly supported by a Grant-in-Aid for Early-Career Scien-tists (No. 22K14549) and a Grant-in-Aid for Transformative Research Areas(No. 24H01378) from the Japan Society for the Promotion Science (JSPS),a project (No. JPNP20004) subsidized by the New Energy and IndustrialTechnology Development Organization (NEDO), the project “Researchproject for sericultural bio-industry” (No. JP22680575) commissioned bythe Ministry of Agriculture, Forestry and Fisheries (MAFF), and the Re-search Collaboration Program (No. AM2100) of the National Institute forMaterials Science (NIMS).Conflict of InterestThe National Institute of Technology has filed a patent application on therepetitive-sequence gene library construction (PCT/JP2023/014202). TheNational Institute of Technology and NIMS have filed another patent ap-plication on the gene synthesis method. All other authors declare no com-peting interests.Data Availability StatementThe data sets in this study are available within the source data.Keywordsdirected evolution, gene synthesis, protein polymers, repeat sequences,rolling-circle amplificationReceived: October 30, 2024Revised: January 15, 2025Published online:[1] A. Miserez, J. Yu, P. Mohammadi, Chem. Rev. 2023, 123, 2049.[2] X. Zhang, J. Li, C. Ma, H. Zhang, K. Liu, Acc. Chem. Res. 2023, 56,2664.[3] M. P. Chang, W. Huang, D. J. Mai, J. Polym. Sci. 2021, 59, 2644.[4] D. Chatterjee, F. J. Lewis, H. J. Sutton, J. A. Kaczmarski, X. Gao, Y.Cai, H. A. McNamara, C. J. Jackson, I. A. Cockburn, Cell Rep. 2021,35, 108996.[5] J. Forbes, A. Bissoyi, L. Eickhoff, N. Reicher, T. Hansen, C. G. Bon, V. K.Walker, T. Koop, Y. Rudich, I. Braslavsky, P. L. Davies, Nat. Commun.2022, 13, 5019.[6] T. Collins, J. Azevedo-Silva, A. da Costa, F. Branca, R. Machado, M.Casal, Microb. Cell Fact. 2013, 12, 21.[7] T. Homma, S. Terui, F. Yokoyama, S. Okino, S. Ohta, C. Kato, N.Haraguchi, I. Fujisawa, S. Itsuno, L. Z. P. Ang, Biotechnol. Bioeng.2023, 120, 194.[8] D. Qin, M. Wang, W. Cheng, J. Chen, F. Wang, J. Sun, C. Ma, Y.Zhang, H. Zhang, H. Li, K. Liu, J. Li, Angew. Chem., Int. Ed. 2024,63, e202400595.[9] J. Li, B. Jiang, X. Chang, H. Yu, Y. Han, F. Zhang, Nat. Commun. 2023,14, 2127.[10] X. Hu, X. X. Xia, S. C. Huang, Z. G. Qian, Biomacromolecules 2019,20, 3283.[11] S. Lv, D. M. Dudek, Y. Cao, M. M. Balamurali, J. Gosline, H. Li, Nature2010, 465, 69.[12] D. Kam, A. Olender, A. Rudich, Y. Kan-Tor, A. Buxboim, O. Shoseyov,S. Magdassi, Adv. Funct. Mater. 2023, 33, 2210993.[13] D. Ding, P. A. Guerette, J. Fu, L. Zhang, S. A. Irvine, A. Miserez, Adv.Mater. 2015, 27, 3953.[14] M. Kim, W. G. Chen, J. W. Kang, M. J. Glassman, K. Ribbeck, B. D.Olsen, Adv. Mater. 2015, 27, 4207.[15] S. R. MacEwan, W. Hassouneh, A. Chilkoti, J. Vis. Exp. 2014, e51583.[16] A. G. Goncalves, E. J. Hartzell, M. O. Sullivan, W. Chen, Adv. DrugDelivery Rev. 2022, 191, 114570.[17] R. S. C. Su, R. J. Galas, C. Y. Lin, J. C. Liu, Macromol. Biosci. 2019, 19,1900122.[18] C. Ma, J. Dong, M. Viviani, I. Tulini, N. Pontillo, S. Maity, Y. Zhou, W.H. Roos, K. Liu, A. Herrmann, G. Portale, Sci. Adv. 2020, 6, eabc0810.[19] Y. Dai, M. Farag, D. Lee, X. Zeng, K. Kim, H. Son, X. Guo, J. Su,N. Peterson, J. Mohammed, M. Ney, D. M. Shapiro, R. V. Pappu, A.Chilkoti, L. You, Nat. Chem. Biol. 2023, 19, 518.[20] Z. G. Qian, F. Pan, X. X. Xia, Curr. Opin. Biotechnol. 2020, 65, 197.[21] R. Price, A. Poursaid, H. Ghandehari, J. Control Rel. 2014, 190, 304.[22] A. D. Malay, T. Suzuki, T. Katashima, N. Kono, K. Arakawa, K. Numata,Sci. Adv. 2020, 6, eabb6030.[23] E. G. Roberts, N. G. Rim, W. Huang, A. Tarakanova, J. Yeo, M. J.Buehler, D. L. Kaplan, J. Y. Wong, Macromol. Biosci. 2018, 18, 1800265.[24] M. Abdelghani, J. Shao, D. H. T. Le, H. Wu, J. C. M. van Hest, Macro-mol. Biosci. 2021, 21, 2100081.[25] S. C. Huang, Z. G. Qian, A. H. Dan, X. Hu, M. L. Zhou, X. X. Xia, ACSBiomater. Sci. Eng. 2017, 3, 1576.[26] A. Basheer, S. Shahid, M. J. Kang, J. H. Lee, J. S. Lee, D. W. Lim, ACSAppl. Mater. Interfaces 2021, 13, 24385.[27] Q. Wang, X. Xia, W. Huang, Y. Lin, Q. Xu, D. L. Kaplan, Adv. Funct.Mater. 2014, 24, 4303.[28] H. K. Binz, P. Amstutz, A. Kohl, M. T. Stumpp, C. Briand, P. Forrer, M.G. Grütter, A. Plückthun, Nat. Biotechnol. 2004, 22, 575.[29] N. C. Tang, J. C. Su, Y. Shmidov, G. Kelly, S. Deshpande, P. Sirohi, N.Peterson, A. Chilkoti, Nat. Commun. 2024, 15, 3727.[30] F. Teulé, A. R. Cooper, W. A. Furin, D. Bittencourt, E. L. Rech, A.Brooks, R. V Lewis, Nat. Protoc. 2009, 4, 341.[31] J. R. McDaniel, J. A. MacKay, F. G. Quiroz, A. Chilkoti, Biomacro-molecules 2010, 11, 944.[32] N. C. Tang, A. Chilkoti, Nat. Mater. 2016, 15, 419.Adv. Sci. 2025, 2410903 2410903 (13 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.comhttps://web.expasy.org/protscale/https://protcalc.sourceforge.net/www.advancedsciencenews.com www.advancedscience.com[33] M. Amiram, F. G. Quiroz, D. J. Callahan, A. Chilkoti, Nat. Mater. 2011,10, 141.[34] H. Jung, A. Pena-Francesch, A. Saadat, A. Sebastian, D. H. Kim, R. F.Hamilton, I. Albert, B. D. Allen, M. C. Demirel, Proc. Natl. Acad. Sci.U.S.A. 2016, 113, 6478.[35] P. H. N. Celie, A. H. A. Parret, A. Perrakis, Curr. Opin. Struct. Biol.2016, 38, 145.[36] J. M. Aliotta, J. J. Pelletier, J. L. Ware, L. S. Moran, J. S. Benner, H.Kong, Genet. Anal. Biomol. Eng. 1996, 12, 185.[37] X. Li, Y. Huang, Y. Guan, M. Zhao, Y. Li, Anal. Chim. Acta. 2007, 584,12.[38] K. B. Ignatov, E. V. Barsova, A. F. Fradkov, K. A. Blagodatskikh, T. V.Kramarova, V. M. Kramarov, BioTechniques 2014, 57, 81.[39] T. Yoshimura, T. Suzuki, S. Mineki, S. Ohuchi, PLoS One 2015, 10,e0136532.[40] M. Dzuricky, B. A. Rogers, A. Shahid, P. S. Cremer, A. Chilkoti, Nat.Chem. 2020, 12, 814.[41] J. R. Simon, N. J. Carroll, M. Rubinstein, A. Chilkoti, G. P. López, Nat.Chem. 2017, 9, 509.[42] S. Rekhi, C. G. Garcia, M. Barai, A. Rizuan, B. S. Schuster, K. L. Kiick,J. Mittal, Nat. Chem. 2024, 16, 1113.[43] S. E. Reichheld, L. D. Muiznieks, F. W. Keeley, S. Sharpe, Proc. Natl.Acad. Sci. U.S.A. 2017, 114, E4408.[44] B. Zhu, G. Cai, E. O. Hall, G. J. Freeman, BioTechniques 2007, 43, 354.[45] P. Weber, M. Dzuricky, J. Min, I. Jenkins, A. Chilkoti, Biomacro-molecules 2021, 22, 4347.[46] X. X. Xia, Q. Xu, X. Hu, G. Qin, D. L. Kaplan, Biomacromolecules 2011,12, 3844.[47] B. Dai, C. J. Sargent, X. Gui, C. Liu, F. Zhang, Biomacromolecules 2019,20, 2015.[48] W. Hassouneh, M. L. Nunalee, M. C. Shelton, A. Chilkoti, Biomacro-molecules 2013, 14, 2347.[49] R. Balu, N. K. Dutta, N. R. Choudhury, C. M. Elvin, R. E. Lyons, R.Knott, A. J. Hill, Acta Biomater. 2014, 10, 4768.[50] L. Li, T. Luo, K. L. Kiick, Macromol. Rapid Commun. 2015, 36, 90.[51] M. R. Banki, L. Feng, D. W. Wood, Nat. Methods 2005, 2, 659.[52] D. W. Lim, K. Trabbic-Carlson, J. A. MacKay, A. Chilkoti, Biomacro-molecules 2007, 8, 1417.[53] I. Weitzhandler, M. Dzuricky, I. Hoffmann, F. Garcia Quiroz, M.Gradzielski, A. Chilkoti, Biomacromolecules 2017, 18, 2419.[54] S. Roberts, T. S. Harmon, J. L. Schaal, V. Miao, K. Li, A. Hunt, Y. Wen,T. G. Oas, J. H. Collier, R. V. Pappu, A. Chilkoti, Nat. Mater. 2018, 17,1154.[55] M. W. T. Werten, G. Eggink, M. A. Cohen Stuart, F. A. de Wolf, Biotech-nol. Adv. 2019, 37, 642.[56] E. W. Martin, A. S. Holehouse, I. Peran, M. Farag, J. J. Incicco, A.Bremer, C. R. Grace, A. Soranno, R. V. Pappu, T. Mittag, Science 2020,367, 694.[57] K. M. Ruff, Y. H. Choi, D. Cox, A. R. Ormsby, Y. Myung, D. B. Ascher,S. E. Radford, R. V. Pappu, D. M. Hatters, Mol. Cell 2022, 82, 3193.[58] A. Chilkoti, M. R. Dreher, D. E. Meyer, Adv. Drug Delivery Rev. 2002,54, 1093.[59] Y. Wang, P. Xue, M. Cao, T. Yu, S. T. Lane, H. Zhao, Chem. Rev. 2021,121, 12384.[60] D. W. Urry, Angew. Chemie Int. Ed. Eng. 1993, 32, 819.[61] M. T. Reetz, D. Kahakeaw, R. Lohmer, ChemBioChem 2008, 9, 1797.[62] S. Roberts, M. Dzuricky, A. Chilkoti, FEBS Lett. 2015, 589, 2477.[63] D. J. Callahan, W. Liu, X. Li, M. R. Dreher, W. Hassouneh, M. Kim, P.Marszalek, A. Chilkoti, Nano Lett. 2012, 12, 2165.[64] D. E. Meyer, A. Chilkoti, Biomacromolecules 2004, 5, 846.[65] Y. Cho, Y. Zhang, T. Christensen, L. B. Sagle, A. Chilkoti, P. S. Cremer,J. Phys. Chem. B. 2008, 112, 13765.[66] J. R. McDaniel, D. C. Radford, A. Chilkoti, Biomacromolecules 2013,14, 2866.[67] J. A. MacKay, D. J. Callahan, K. N. FitzGerald, A. Chilkoti, Biomacro-molecules 2010, 11, 2873.[68] S.-M. Liao, Q.-S. Du, J.-Z. Meng, Z.-W. Pang, R.-B. Huang, Chem.Cent. J. 2013, 7, 44.[69] B. Tsang, I. Pritišanac, S. W. Scherer, A. M. Moses, J. D. Forman-Kay,Cell 2020, 183, 1742.[70] J. Wang, J. M. Choi, A. S. Holehouse, H. O. Lee, X. Zhang, M. Jahnel,S. Maharana, R. Lemaitre, A. Pozniakovsky, D. Drechsel, I. Poser, R.V. Pappu, S. Alberti, A. A. Hyman, Cell 2018, 174, 688.[71] S. Lim, D. S. Clark, Trends Biotechnol. 2024, 42, 496.[72] C. N. Pace, F. Vajdos, L. Fee, G. Grimsley, T. Gray, Protein Sci. 1995,4, 2411.[73] M. Martin, EMBnet.journal 2011, 17, 10.[74] A. M. Bolger, M. Lohse, B. Usadel, Bioinformatics 2014, 30, 2114.[75] F. G. Quiroz, N. K. Li, S. Roberts, P. Weber, M. Dzuricky, I.Weitzhandler, Y. G. Yingling, A. Chilkoti, Sci. Adv. 2019, 5, eaax5177.Adv. Sci. 2025, 2410903 2410903 (14 of 14) © 2025 The Author(s). Advanced Science published by Wiley-VCH GmbH 21983844, 0, Downloaded from https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202410903 by National Institute For, Wiley Online Library on [02/04/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttp://www.advancedsciencenews.comhttp://www.advancedscience.com