CADGE 2.0, Transcription-Translation-Coupled DNA Replication is Improved in a Chemically Modified Cell-Free System

13 Cell-free gene expression, in vitro transcription-translation (IVTT), in vitro transcription–14 translation–coupled DNA replication (IVTTR), in vitro evolution, cell-free directed evolution, 15 clonal DNA amplification 16 17 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint

18 In vitro directed evolution in synthetic microcompartments can generally support the 19 evolution of genes with functions beyond affinity. The main challenge in the implementation 20 of this strategy is the need to incorporate no more than a single DNA template molecule per 21 microcompartment, thereby establishing a robust genotype-phenotype link age, but which 22

in slow, inconsistent in vitro transcription and translation (IVTT) and poor DNA 23 recovery after selection or screening. To address this challenge, we previously developed 24 CADGE (Clonal Amplification-enhanceD Gene Expression) a strategy that allows the clonal 25 amplification of linear gene-encoding DNA and coupled, in situ transcription-translation of 26 the gene of interest. Here, we show that clonal amplification is highly sensitive to the cell-27 free system’s composition and that robust, highly efficient cell-free DNA amplification via 28 the CADGE platform can be achieved by replacing standard vendor-supplied energy mixes 29 with DNA replication-optimized, homemade counterparts. 30 31 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint

32 Directed evolution is a powerful tool in synthetic biology. 1–3 Yet, despite numerous 33 achievements, it remains largely limited to cell-based systems. Cell-free directed evolution 34 offers an alternative that enables the evolution of toxic or growth-incompatible functions 35 and allows selection to act directly on molecular activity rather than organismal fitness. 4–7 36 Such approaches rely on cell-free gene expression systems that can support in vitro 37 transcription-translation (IVTT) of proteins 8 in an open, chemically tunable environment. 9 A 38 common strategy to establish genotype–phenotype linkage in cell-free evolution is in vitro 39 compartmentalization.6,10,11 Genotype–phenotype linkage can be achieved by using low 40 initial DNA copy number so that each compartment contains, at most, a single variant. 11 41 However, low template concentrations often result in slow and variable IVTT, 12–16 and DNA 42 recovery from microcompartments remains a major bottleneck in both in vitro and cell-43 based systems. 17–19 T o overcome these limitations, various strategies for clonal DNA 44 amplification in microcompartments have been proposed, but these approaches require 45 cumbersome procedures or specialized instrumentation, such as bead display or 46 microfluidic droplet generation.20–27 47 In a previous study, we developed CADGE (Clonal Amplification-enhanceD Gene 48 Expression), a highly streamlined alternative to existing clonal DNA amplification strategies 49 with the potential to accelerate in vitro evolution.28 This strategy combines, within a single 50 compatible environment, isothermal amplification of a template DNA starting from a single 51 copy with in vitro transcription-translation, followed by selection or screening. 28 DNA 52 amplification is carried out by the minimal DNA replication machinery from the B. subtilis 53 phage Φ 29.29 The machinery consists of DNA polymerase (DNAP; p2 gene), terminal 54 protein (TP; p3 gene), double-stranded DNA-binding protein (DSB), and single-stranded 55 DNA-binding protein (SSB). Replication initiates independently at two origins ( ori)30,31 56 located at the termini of the linear Φ 29 genome: DSB binds to ori32,33 and recruits the 57 DNAP–TP complex, 34 TP primes synthesis, 35 and DNAP simultaneously replicates 36 and 58 displaces the non-template strand, 37 which is stabilized by SSB. 38 Two DNAP molecules 59 thus replicate both strands from opposite ends. 39 CADGE repurposes this protein-primed 60 system by placing the gene of interest between two ori sites on a linear template, with DSB 61 and SSB supplied recombinantly and DNAP and TP provided either as purified proteins or 62 expressed in situ .28 Unlike rolling-circle amplification, 40–43 protein-primed replication 44,45 63 regenerates linear DNA identical to the template, minimizing DNA handling between 64 evolution rounds. When encapsulated in liposomes, CADGE boosts phenotypic output 65 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint under low-copy conditions and improves DNA recovery after selection or screening. 28 66 CADGE was established in the recombinant cell-free system, PURE (Protein 67 synthesis Using Recombinant Elements). 46 However, we and others 47 recently observed 68 that commercially available PURE systems currently demonstrate significantly decreased 69 levels of DNA replication compared to previously reported levels, 28 revealing strong 70 sensitivity to modifications in PURE composition. We hypothesized that proprietary 71 changes in reaction composition that optimize transcription and translation can tip the 72 balance against replication. Several studies successfully re-engineered PURE formulations 73 to balance IVTT with Φ 29 DNAP–dependent DNA replication from circular templates, 74 particularly by adjusting NTP and tRNA concentrations and rebalancing Mg 2+ to account for 75 NTP-dependent chelation.42,43,48,49 However, it was unclear whether formulations optimized 76 for nucleic-acid-primed rolling-circle amplification would also support protein-primed 77 replication of linear templates, which requires TP–dAMP initiation and a distinct transition 78 from initiation to elongation. 50,51 These mechanistic and kinetic differences make direct 79 extrapolation nontrivial. Consistent with this concern, despite prior demonstrations of robust 80 Φ 29-mediated rolling-circle replication in extract-based systems, 49 we were unable to 81 implement CADGE in crude E. coli lysate (data not shown). Together, these observations 82 motivated us to test whether previously reported PURE formulations that support rolling-83 circle amplification could be adapted to balance IVTT with protein-primed Φ 29 replication in 84 CADGE. 85 Here, we demonstrate that PURE energy mix (the small-molecule components of 86 the PURE kit) previously optimized 42 for transcription and translation-coupled DNA 87 replication markedly improves protein-primed Φ 29 DNA replication from linear templates 88 across commercial PURE platforms, while maintaining transcription–translation activity. We 89 also observed a modest improvement in autocatalytic self-replication of the Φ 29 DNA 90 polymerase gene. 45,52 This gene circuit can be integrated with diverse genetic modules to 91 enable systems-level evolution and advance the construction of synthetic cells. 53 Together, 92 these results provide a practical foundation for protein-primed in vitro transcription-93 translation-DNA replication (IVTTR) using linear DNA that lower technical barriers to cell-94 free directed evolution of proteins and advance efforts toward building evolvable synthetic 95 cells. 96 97 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint

and Discussion 98 DNA-replication-optimized energy mix restores CADGE activity in PUREfrex 99 To restore CADGE in commercial PURE systems, we tested whether the DNA replication-100 optimized energy mix42 improves DNA replication in CADGE (Fig. 1A, 1B). For easy 101 readout of DNA amplification and coupled transcription and translation, we used the yellow 102 fluorescent protein (YFP) as a reporter. We used the yfp gene encoded on a linear DNA 103 template and flanked by ori sequences at both ends (ori-yfp). The yfp DNA template was 104 introduced in the CADGE reaction at a final concentration of 10 pM. Φ 29 DNAP and TP 105 were expressed by IVTT in situ from a plasmid DNA template. We expected that successful 106 replication would result in increased yfp gene concentration, which would in turn elevate the 107 rate of IVTT and yield of the YFP protein. YFP expression and DNA replication were 108 monitored across three commercially available PURE platforms: PURExpress, PUREfrex 109 1.0, and PUREfrex 2.0. 110 Quantification of yfp DNA by digital PCR (dPCR) before and after the 16-hour 111 incubation of IVTTR reactions at 30°C revealed that reactions containing commercial 112 energy mixes allowed only limited DNA amplification across PURE platforms, with 113 PUREfrex 1.0 showing highest DNA amplification at around two orders of magnitude (Fig. 114 1C, 1D). In contrast, the optimized, homemade, energy mix markedly enhanced DNA 115 amplification in PUREfrex systems, yielding at least three orders of magnitude increase 116 (Fig. 1C, 1D), while PURExpress enabled modest replication around one order of 117 magnitude (Fig. 1C, 1D). Note that PURExpress showed an approximately 10-fold DNA 118 degradation within 16 hours of incubation regardless of the energy mix used (Fig. 1C), 119 possibly due to previously reported nuclease activity in this system, 43 although this 120 degradation appears to be lot-dependent in our experience. A smaller, but still significant 121 degradation of yfp DNA was observed in PUREfrex 2.0 supplemented with the homemade 122 energy mix (Fig. 1C; ~3-fold reduction), but not with the commercial energy mix. In contrast, 123 PUREfrex 1.0 showed no comparable decrease. 124 To verify the integrity of replicated DNA, the ori-yfp DNA was amplified by PCR 125 using primers targeting ori’s, and the amplified DNA was visualized by agarose gel 126 electrophoresis. Across PURE platforms, the full-length ori-yfp amplicon (1.5 kb) was 127 observed in complete CADGE reactions supplemented with the homemade energy mix (Fig. 128 1E, Complete). Moreover, the amount of recovered DNA from IVTTR reactions with the 129 homemade energy mix was considerably higher than from controls lacking dNTPs (Fig. 1E), 130 thus improving DNA recovery from low concentrations of IVTT template. Unexpectedly, in 131 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint IVTTR reactions supplemented with the commercial energy mix, PCR amplification yielded 132 higher apparent yfp DNA levels in control reactions lacking dNTPs than in complete 133 reactions (Fig. 1E). This counterintuitive result likely reflects reduced PCR efficiency of 134 replicated DNA molecules, as Φ 29 DNAP generates products covalently linked to terminal 135 protein (TP) at their 5′ ends, which may sterically hinder the elongating DNA polymerase 136 approaching the 5′ ends of the template. However, the high level of DNA amplification 137 observed in reactions with the homemade energy mix likely overcomes this inhibition. 138 Higher molecular band (~3.2 kb, Fig. 1E) is an artefact that corresponds to the PCR 139 amplicon from the DNAP and TP-expressing plasmid, which similarly contains Φ 29 ori’s. 140 The ori’s on the plasmid DNA encoding DNAP and TP are not required in CADGE, and 141 should be removed prior to the start of directed evolution, which would eliminate the 142 artefact. Together, these results demonstrate that a chemically defined, replication-143 optimized energy mix restores and markedly enhances IVTTR performance across PURE 144 platforms. 145 To validate that DNA replication is compatible with in situ cell-free transcription and 146 translation of the amplified gene, we performed endpoint YFP fluorescence measurements. 147 Across platforms, reactions supplemented with the optimized energy mix generally 148 produced higher YFP fluorescence at 16 hours at 30°C than those containing the 149 corresponding commercial energy mixes, except for PUREfrex 1.0, where fluorescence 150 remained comparable to the commercial formulation. In all three systems, YFP 151 fluorescence increased significantly in the presence of both template DNA and dNTPs, 152 confirming that the observed signal reflects DNA replication-coupled transcription-153 translation (Fig. 1F). Together, these results indicate that while commercial PURE 154 formulations are highly optimized for protein synthesis, the homemade energy mix is more 155 suitable for IVTT-coupled DNA replication in CADGE. 156 157 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint 158 Figure 1. Optimized energy mix restores CADGE across PURE platforms 159 (A) Schematic of CADGE. A linear yfp DNA template flanked by ori from both sides is 160 amplified by in situ-expressed Φ 29 DNAP and TP proteins from plasmid DNA via IVTT in 161 PURE. In the presence of dNTPs (Complete), protein-primed DNA replication increases yfp 162 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint template copy number and enhances YFP expression. Without dNTPs ([−] dNTPs), no 163 replication occurs and YFP is expressed only from the input template. Without DNA 164 template ([−] DNA), no YFP expression is detected. (B) A schematic of chemically defined 165 homemade mix (H) and commercial mixes (C) supplied with PURExpress or PUREfrex 166 systems. (C) Quantification of yfp DNA by dPCR at 0 h and 16 h of CADGE reactions. (D) 167 DNA amplification fold change (16 h / 0 h) calculated from dPCR measurements in Fig. 1C. 168 (E) Estimation of DNA size recovered from CADGE reactions using agarose gel 169 electrophoresis of PCR-amplified DNA. Full-length ori-yfp product: 1486 nt. The 3214-nt 170 band DNA is an artifact originating from the plasmid DNA encoding DNAP and TP . (F) 171 Estimation of in vitro expression of YFP in CADGE reactions by endpoint fluorescence 172 measurements at 16 h. Data are presented as mean + SD (t -test, ** P < 0.01; * P < 0.05; 173 NS, not significant; n = 3). Created in BioRender. Lab, A. (2026) 174 https://BioRender.com/geukhbp. 175 176 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint Optimized energy mix moderately improves protein-primed self-replication 177 We next tested the performance of the optimized energy mix in self-replication of a minimal 178 genetic module encoding Φ 29 DNAP and TP (ori-p2-p3)45,52 (Fig. 2A). In this construct, the 179 p2 and p3 genes are encoded on a single linear DNA template flanked by Φ 29 ori‘s, 180 enabling DNAP and TP to drive autocatalytic replication of the DNA encoding them. Since 181 PUREfrex demonstrated significantly higher replication than PURExpress in the CADGE 182 setup, we proceeded to test self-replication only in PUREfrex. DNA replication from an 183 initial template concentration of 2.3 nM was quantified via dPCR after 16 hours of 184 incubation at 30°C in PUREfrex 1.0 and PUREfrex 2.0. In both platforms, moderate 185 amplification of p2-p3 DNA (10-40-fold) was observed with either energy mix (Fig. 2B). 186 Omission of dNTPs in the IVTTR reactions completely abolished DNA amplification, 187 confirming that the observed increases in DNA concentration were IVTTR-dependent. 188 Comparison of amplification efficiency revealed that the PUREfrex 2.0 189 supplemented with the homemade energy mix produced higher replication yields than the 190 commercial energy mix, whereas no significant difference was observed in PUREfrex 1.0 191 (Fig. 2C). This pattern mirrors the weaker enhancement seen for CADGE in PUREfrex 1.0 192 compared to PUREfrex 2.0 (Fig. 1D), and may reflect the higher translational flux of 193 PUREfrex 2.0. Agarose gel electrophoresis confirmed the accumulation of full-length p2-p3 194 DNA products (3.2 kb) in the presence of dNTPs in PUREfrex 2.0 supplemented with the 195 homemade energy mix (Fig. 2D). 196 These results demonstrate that the PUREfrex 2.0 supplemented with the 197 replication-optimized energy mix enables improved self-replication of a minimal Φ 29 DNA 198 replication module compared to the fully commercial formulation. Overall, we show that a 199 DNA replication-optimized energy formulation originally developed for rolling-circle 200 replication can be directly applied to protein-primed Φ 29 replication from linear DNA 201 templates. Our study provides reaction conditions that could support selection- and 202 evolution-based experiments and offers a starting point for their further development. 203 204 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint 205 Figure 2. Optimized energy mix supports Φ 29 self-replication 206 (A) Schematic of self-replication. p2 (DNAP) and p3 (TP) are expressed via IVTT and form 207 an active DNAP–TP complex that initiates replication at flanking ori sequences in the 208 presence of dNTPs (Complete), driving autocatalytic amplification. In [−] dNTPs, replication 209 is blocked and only expression from the input template occurs. (B) Quantification of p2-p3 210 DNA by dPCR at 0 and 16 h in PUREfrex using homemade (H) or commercial (C) energy 211 mixes. (C) Amplification fold change (16 h / 0 h) derived from Fig. 2B. (D) Agarose gel 212 electrophoresis of PCR-amplified DNA recovered from self-replication reactions. Data are 213 presented as mean + SD (t -test, ** P < 0.01; * P < 0.05; NS, not significant; n = 3). Created 214 in BioRender. Lab, A. (2026) https://BioRender.com/geukhbp. 215 216 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint

217 DNA template preparation 218 Plasmids, primers, and template sequences are listed in Supplementary Tables S1–S3. 219 Linear ori-yfp and ori-p2-p3 DNAs were amplified from plasmids G36528 and G34052 using 220 the same 5′ -phosphorylated primer pair (primers 1 and 2). PCR (50 µL) contained 1x 221 Phusion HF buffer, 200 µM each dNTP , 1 µM each primer, 10 ng template DNA, and 0.5 µL 222 Phusion High-Fidelity DNA Polymerase (Thermo Fisher). Cycling conditions were 98°C for 223 30 s; 25 cycles of 98°C for 10 s, 60°C for 30 s, and 72°C for 15–30 s per kb (1 min 30 s for 224 yfp; 3 min for p2-p3); followed by 72°C for 10 min. PCR products were purified using the 225 QIAquick PCR & Gel Cleanup Kit with two washes in Buffer PE and elution in 30 µL 226 nuclease-free water. 227 Energy mix preparation 228 Our homemade energy mix was based on PURErep 10x energy formulation 42, optimized 229 for rolling circle replication in PURE. The 10x mixture contained the following components: 230 3.6 mM of each of the 20 natural L-amino acids, 700 mM potassium glutamate, 3.75 mM 231 spermidine, 250 mM creatine phosphate potassium salt, 5.18 g/L E. coli tRNA, 1 M 232 HEPES–KOH (pH 8.0), 79 mM hemi-magnesium glutamate, and 60 mM dithiothreitol. tRNA 233 was purified from E. coli strain A19 following the procedure described ipreviously.54,55 234 CADGE reaction 235 The CADGE reaction was adapted using conditions previously optimized for rolling-circle 236 replication.43 All reactions were performed in 10 µL volumes and incubated at 30°C for 16 237 hours in PCR tubes using a thermal cycler, followed by quenching on ice. 238 Reactions using PUREfrex 2.0 (GeneFrontier) supplemented with homemade 239 energy mix consisted of 1 µL 10x homemade energy mix, 0.5 µL solution I, 0.5 µL solution 240 II, 0.5 µL solution III, 18.75 mM ATP, 12.5 mM GTP, 6.25 mM UTP, 6.25 mM CTP, 20 mM 241 ammonium sulfate, 300 µM dNTPs, 375 µg/mL purified Φ 29 SSB, 105 µg/mL purified Φ 29 242 DSB, and 0.6 U/µL SUPERase·In RNase inhibitor (Thermo Fisher), along with 1 nM p2-p3 243 plasmid (G340) and 10 pM ori-yfp linear reporter DNA. Notably, Commercial energy mix (i.e. 244 solution I) was added at 0.1x, primarily to supply the formyl donor (10-formyl-5,6,7,8-245 tetrahydrofolic acid (THF)), which is required for Met-tRNA fMet formylation.43 246 CADGE using the PUREfrex 2.0 with the commercial energy mix differed from the 247 above reaction conditions in that no 10x homemade energy mix was added, and the 248 reaction contained 5 µL solution I, while keeping the rest of the reagents as mentioned 249 above. Click or tap here to enter text. 250 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint In reactions using PUREfrex 1.0, the components corresponding to PUREfrex 2.0 251 solutions I–III were substituted with their respective counterparts, 43 except that 0.25 µL of 252 PUREfrex 1.0 solution III was added, maintaining the same proporti onal adjustment as in 253 the PUREfrex 2.0 reactions. For PURExpress (NEB), solutions A (energy, 2.5x stock, final 254 0.1x) and B (enzymes and ribosomes, 3.33x stock, final 1x) replaced PUREfrex solutions I–255 III.42,43 256 Fluorescence quantification 257 YFP fluorescence (9 µL IVTTR reaction) was measured in 384-well black plates on a 258 BioTek Synergy H1 at 30°C (Ex 513 nm; Em 550 nm to minimize excitation-emission 259 overlap). 260 DNA quantification by dPCR 261 The IVTTR reaction mixtures were diluted 100-fold in TE buffer (10 mM Tris-HCl, 1 mM 262 EDTA, pH 8.0) and stored at −20°C until use. These samples were further diluted in 263 nuclease-free water so that the expected target occupancy ( λ ) was between 0.6 and 1.6 264 copies per microchamber, which keeps a sufficient fraction of partitions negative for 265 accurate calculation of DNA concentration. Target-specific primer-probe sets for yfp and p2-266 p3 were custom-synthesized as FAM-labeled T aqMan Gene Expression Assays (Thermo 267 Fisher) and dPCR was performed using QuantStudio Absolute Q Digital PCR System 268 (Thermo Fisher). 269 IVTTR product size estimation by agarose gel electrophoresis 270 To estimate the size of yfp or p2-p3 DNA amplified in IVTTR reactions, 1 µL of 100-fold 271 diluted reaction mixture was subjected to PCR (100 µL) under the same conditions as for 272 DNA template preparation. PCR products were loaded onto 1% agarose gel without 273 purification and visualized under UV illumination after ethidium bromide staining. 274 Self-amplification reaction 275 Self-replication was conducted under the same bulk CADGE conditions, but without yfp 276 reporter DNA and using 2.3 nM ori-p2-p3 linear DNA (pre-amplified by PCR with 277 phosphorylated primers) instead of G340 plasmid. 278 Statistical analysis 279 All experiments were performed in three independent replicates. Data are presented as 280 mean ± SD. Statistical significance was determined by unpaired two-tailed Student’s t-test 281 (P < 0.01 (**); P < 0.05 (*); not significant (NS)). 282 283 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint Author Information 284 Corresponding Author 285 Zhanar Abil − Department of Biology, University of Florida, Gainesville, Florida 32611, 286 United States; Department of Chemical Engineering, University of Florida, Gainesville, 287 Florida 32611, United States; Department of Interdisciplinary Ecology, University of 288 Florida, Gainesville, Florida 32611, United States; orcid.org/0000-0001-9550-2633; 289 Phone: +1 352 294 6850; Email: [email protected] 290 291 Authors 292 Riku Nagai − Department of Biology, University of Florida, Gainesville, Florida 32611, 293 United States 294 Carlos Chavez Ramirez − Department of Biology, University of Florida, Gainesville, 295 Florida 32611, United States 296 297 Author Contributions 298 Z.A. conceptualized the research and acquired funding. R.N. performed the experiments. 299 R.N. and Z.A. designed the experiments. R.N., C.C.R., and Z.A. analyzed the data and 300 wrote the manuscript. 301 302 Conflict of Interest 303 None declared. 304 305 Acknowledgments 306 We thank members of the Abil laboratory for helpful discussions and feedback on the 307 manuscript. We are grateful to Miguel de Vega (Centro de Biología Molecular Severo 308 Ochoa, Madrid) for kindly providing purified SSB and DSB proteins. We thank Christophe 309 Danelon for the kind sharing of the G365 and G340 plasmids and for helpful discussions. 310 This work was supported by startup funding from the University of Florida College of Liberal 311 Arts and Sciences and the University of Florida Gatorade Award. C.C.R was supported by 312 the University of Florida Biology Department’s 2025 Graduate Student Opportunity Award. 313 Images were created in BioRender Lab, A. (2026) https://BioRender.com/geukhbp. 314 315 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint

316 (1) Arnold, F. H. Innovation by Evolution: Bringing New Chemistry to Life (Nobel Lecture). 317 Angew. Chem. Int. Ed. 2019, 58 (41), 14420–14426. 318 https://doi.org/10.1002/anie.201907729. 319 (2) Arnold, F. H. Engineering Proteins for Nonnatural Environments. FASEB J. 1993, 7 (9), 320 744–749. https://doi.org/10.1096/fasebj.7.9.8330682. 321 (3) Wang, Y.; Xue, P.; Cao, M.; Yu, T.; Lane, S. T.; Zhao, H. Directed Evolution: 322 Methodologies and Applications. Chem. Rev. 2021, 121 (20), 12384–12444. 323 https://doi.org/10.1021/acs.chemrev.1c00260. 324 (4) Seo, K.; Hagino, K.; Ichihashi, N. Progresses in Cell-Free In Vitro Evolution. Adv. 325 Biochem. Eng.Biotechnol. 2023, 186, 121–140. https://doi.org/10.1007/10_2023_219. 326 (5) Lipovsek, D.; Plückthun, A. In-Vitro Protein Evolution by Ribosome Display and mRNA 327 Display. J. Immunol. Methods 2004, 290 (1–2), 51–67. 328 https://doi.org/10.1016/j.jim.2004.04.008. 329 (6) Aharoni, A.; Griffiths, A. D.; Tawfik, D. S. High-Throughput Screens and Selections of 330 Enzyme-Encoding Genes. Curr. Opin. Chem. Biol. 2005, 9 (2), 210–216. 331 https://doi.org/10.1016/j.cbpa.2005.02.002. 332 (7) Dodevski, I.; Markou, G. C.; Sarkar, C. A. Conceptual and Methodological Advances in 333 Cell-Free Directed Evolution. Curr. Opin. Struct. Biol. 2015, 33, 1–7. 334 https://doi.org/10.1016/j.sbi.2015.04.008. 335 (8) Nirenberg, M. W.; Matthaei, J. H. The Dependence of Cell-Free Protein Synthesis in E. 336 Coli upon Naturally Occurring or Synthetic Polyribonucleotides. Proc. Natl. Acad. Sci. 1961, 337 47 (10), 1588–1602. https://doi.org/10.1073/pnas.47.10.1588. 338 (9) Perez, J. G.; Stark, J. C.; Jewett, M. C. Cell-Free Synthetic Biology: Engineering 339 Beyond the Cell. Cold Spring Harb. Perspect. Biol. 2016, 8 (12), a023853. 340 https://doi.org/10.1101/cshperspect.a023853. 341 (10) Griffiths, A. D.; Tawfik, D. S. Directed Evolution of an Extremely Fast 342 Phosphotriesterase by in Vitro Compartmentalization. EMBO J. 2003, 22 (1), 24–35. 343 https://doi.org/10.1093/emboj/cdg014. 344 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint (11) Miller, O. J.; Bernath, K.; Agresti, J. J.; Amitai, G.; Kelly, B. T.; Mastrobattista, E.; Taly, 345 V.; Magdassi, S.; Tawfik, D. S.; Griffiths, A. D. Directed Evolution by in Vitro 346 Compartmentalization. Nat. Methods 2006, 3 (7), 561–570. 347 https://doi.org/10.1038/nmeth897. 348 (12) Noireaux, V.; Libchaber, A. A Vesicle Bioreactor as a Step toward an Artificial Cell 349 Assembly. Proc. Natl. Acad. Sci. 2004, 101 (51), 17669–17674. 350 https://doi.org/10.1073/pnas.0408236101. 351 (13) Stögbauer, T.; Windhager, L.; Zimmer, R.; Rädler, J. O. Experiment and Mathematical 352 Modeling of Gene Expression Dynamics in a Cell-Free System. Integr. Biol. 2012, 4 (5), 353 494–501. https://doi.org/10.1039/c2ib00102k. 354 (14) Chory, J.; Kaplan, S. The in Vitro Transcription-Translation of DNA and RNA 355 Templates by Extracts of Rhodopseudomonas Sphaeroides. Optimization and Comparison 356 of Template Specificity with Escherichia Coli Extracts and in Vivo Synthesis. J. Biol. Chem. 357 1982, 257 (24), 15110–15121. https://doi.org/10.1016/s0021-9258(18)33400-8 . 358 (15) Niederholtmeyer, H.; Xu, L.; Maerkl, S. J. Real-Time mRNA Measurement during an in 359 Vitro Transcription and Translation Reaction Using Binary Probes. ACS Synth. Biol. 2013, 2 360 (8), 411–417. https://doi.org/10.1021/sb300104f. 361 (16) Nourian, Z.; Danelon, C. Linking Genotype and Phenotype in Protein Synthesizing 362 Liposomes with External Supply of Resources. ACS Synth. Biol. 2013, 2 (4), 186–193. 363 https://doi.org/10.1021/sb300125z. 364 (17) Gantz, M.; Neun, S.; Medcalf, E. J.; Vliet, L. D. van; Hollfelder, F. Ultrahigh-Throughput 365 Enzyme Engineering and Discovery in In Vitro Compartments. Chem. Rev. 2023, 123 (9), 366 5571–5611. https://doi.org/10.1021/acs.chemrev.2c00910. 367 (18) Wong, C. F. A.; Vliet, L. van; Bhujbal, S. V.; Guo, C.; Sletmoen, M.; Stokke, B. T.; 368 Hollfelder, F.; Lale, R. A Titratable Cell Lysis-on-Demand System for Droplet-369 Compartmentalized Ultrahigh-Throughput Screening in Functional Metagenomics and 370 Directed Evolution. ACS Synth. Biol. 2021, 10 (8), 1882–1894. 371 https://doi.org/10.1021/acssynbio.1c00084. 372 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint (19) Autour, A.; Merten, C. A. Fluorescence-Activated Droplet Sequencing (FAD-Seq) 373 Directly Provides Sequences of Screening Hits in Antibody Discovery. Proc. Natl. Acad. Sci. 374 United States Am. 2024, 121 (37), e2405342121. https://doi.org/10.1073/pnas.2405342121. 375 (20) Kojima, T.; Takei, Y.; Ohtsuka, M.; Kawarasaki, Y.; Yamane, T.; Nakano, H. PCR 376 Amplification from Single DNA Molecules on Magnetic Beads in Emulsion: Application for 377 High-Throughput Screening of Transcription Factor Targets. Nucleic Acids Res. 2005, 33 378 (17), e150–e150. https://doi.org/10.1093/nar/gni143. 379 (21) Gan, R.; Yamanaka, Y.; Kojima, T.; Nakano, H. Microbeads Display of Proteins Using 380 Emulsion PCR and Cell‐ free Protein Synthesis. Biotechnol. Prog. 2008, 24 (5), 1107–1114. 381 https://doi.org/10.1002/btpr.43. 382 (22) Paul, S.; Stang, A.; Lennartz, K.; Tenbusch, M.; Überla, K. Selection of a T7 Promoter 383 Mutant with Enhanced in Vitro Activity by a Novel Multi-Copy Bead Display Approach for in 384 Vitro Evolution. Nucleic Acids Res. 2013, 41 (1), e29–e29. 385 https://doi.org/10.1093/nar/gks940. 386 (23) Mazutis, L.; Araghi, A. F.; Miller, O. J.; Baret, J.-C.; Frenz, L.; Janoshazi, A.; Taly, V.; 387 Miller, B. J.; Hutchison, J. B.; Link, D.; Griffiths, A. D.; Ryckelynck, M. Droplet-Based 388 Microfluidic Systems for High-Throughput Single DNA Molecule Isothermal Amplification 389 and Analysis. Anal. Chem. 2009, 81 (12), 4813–4821. https://doi.org/10.1021/ac900403z. 390 (24) Holstein, J. M.; Gylstorff, C.; Hollfelder, F. Cell-Free Directed Evolution of a Protease 391 in Microdroplets at Ultrahigh Throughput. ACS Synth. Biol. 2021, 10 (2), 252–257. 392 https://doi.org/10.1021/acssynbio.0c00538. 393 (25) Iizuka, R.; Tahara, K.; Matsueda, A.; Tsuda, S.; Yoon, D. H.; Sekiguchi, T.; Shoji, S.; 394 Funatsu, T. Selection of Green Fluorescent Proteins by in Vitro Compartmentalization 395 Using Microbead-Display Libraries. Biochem. Eng. J. 2022, 187, 108627. 396 https://doi.org/10.1016/j.bej.2022.108627. 397 (26) Diamante, L.; Gatti-Lafranconi, P.; Schaerli, Y.; Hollfelder, F. In Vitro Affinity Screening 398 of Protein and Peptide Binders by Megavalent Bead Surface Display. Protein Eng., Des. 399 Sel. 2013, 26 (10), 713–724. https://doi.org/10.1093/protein/gzt039. 400 (27) Fallah-Araghi, A.; Baret, J.-C.; Ryckelynck, M.; Griffiths, A. D. A Completely in Vitro 401 Ultrahigh-Throughput Droplet-Based Microfluidic Screening System for Protein Engineering 402 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint and Directed Evolution. Lab a Chip 2012, 12 (5), 882–891. 403 https://doi.org/10.1039/c2lc21035e. 404 (28) Abil, Z.; Sierra, A. M. R.; Danelon, C. Clonal Amplification-Enhanced Gene Expression 405 in Synthetic Vesicles. ACS Synth. Biol. 2023, 12 (4), 1187–1203. 406 https://doi.org/10.1021/acssynbio.2c00668. 407 (29) Mencía, M.; Gella, P.; Camacho, A.; Vega, M. de; Salas, M. Terminal Protein-Primed 408 Amplification of Heterologous DNA with a Minimal Replication System Based on Phage 409 Φ 29. Proc. Natl. Acad. Sci. 2011, 108 (46), 18655–18660. 410 https://doi.org/10.1073/pnas.1114397108. 411 (30) Gutiérrez, J.; Garmendia, C.; Salas, M. Characterization of the Origins of Replication 412 of Bacteriophage Ø29 DNA. Nucleic Acids Res. 1988, 16 (13), 5895–5914. 413 https://doi.org/10.1093/nar/16.13.5895. 414 (31) Serrano, M.; Gutiérrez, J.; Prieto, I.; Hermoso, J. M.; Salas, M. Signals at the 415 Bacteriophage Phi 29 DNA Replication Origins Required for Protein P6 Binding and Activity. 416 EMBO J. 1989, 8 (6), 1879–1885. https://doi.org/10.1002/j.1460-2075.1989.tb03584.x . 417 (32) Blanco, L.; Gutièrrez, J.; Làzaro, J. M.; Bernad, A.; Salas, M. Replication of Phage 418 Φ 29 DNA in Vitro: Role of the Viral Protein P6 in Initiation and Elongation. Nucleic Acids 419 Res. 1986, 14 (12), 4923–4937. https://doi.org/10.1093/nar/14.12.4923. 420 (33) Blanco, L.; Lázaro, J. M.; Vega, M. de; Bonnin, A.; Salas, M. Terminal Protein-Primed 421 DNA Amplification. Proc. Natl. Acad. Sci. 1994, 91 (25), 12198–12202. 422 https://doi.org/10.1073/pnas.91.25.12198. 423 (34) Kamtekar, S.; Berman, A. J.; Wang, J.; Lázaro, J. M.; Vega, M. de; Blanco, L.; Salas, 424 M.; Steitz, T. A. The Φ 29 DNA Polymerase:Protein‐ primer Structure Suggests a Model for 425 the Initiation to Elongation Transition. EMBO J. 2006, 25 (6), 1335–1343. 426 https://doi.org/10.1038/sj.emboj.7601027. 427 (35) Peñalva, M. A.; Salas, M. Initiation of Phage Phi 29 DNA Replication in Vitro: 428 Formation of a Covalent Complex between the Terminal Protein, P3, and 5’-dAMP. Proc. 429 Natl. Acad. Sci. 1982, 79 (18), 5522–5526. https://doi.org/10.1073/pnas.79.18.5522. 430 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint (36) Blanco, L.; Salas, M. Replication of Phage Phi 29 DNA with Purified Terminal Protein 431 and DNA Polymerase: Synthesis of Full-Length Phi 29 DNA. Proc. Natl. Acad. Sci. 1985, 432 82 (19), 6404–6408. https://doi.org/10.1073/pnas.82.19.6404. 433 (37) Morin, J. A.; Cao, F. J.; Lázaro, J. M.; Arias-Gonzalez, J. R.; Valpuesta, J. M.; 434 Carrascosa, J. L.; Salas, M.; Ibarra, B. Active DNA Unwinding Dynamics during Processive 435 DNA Replication. Proc. Natl. Acad. Sci. 2012, 109 (21), 8115–8120. 436 https://doi.org/10.1073/pnas.1204759109. 437 (38) Martìin, G.; Làzaro, JosèM.; Mèndez, E.; Salas, M. Characterization of the Phage Π 29 438 Protein P5 as a Single-Stranded DNA Binding Protein. Function in Π 29 DNA-Protein P3 439 Replication. Nucleic Acids Res. 1989, 17 (10), 3663–3672. 440 https://doi.org/10.1093/nar/17.10.3663. 441 (39) Escarmís, C.; Guirao, D.; Salas, M. Replication of Recombinant Φ 29 DNA Molecules 442 in Bacillus Subtilis Protoplasts. Virology 1989, 169 (1), 152–160. 443 https://doi.org/10.1016/0042-6822(89)90051-2 . 444 (40) Sakatani, Y.; Ichihashi, N.; Kazuta, Y.; Yomo, T. A Transcription and Translation-445 Coupled DNA Replication System Using Rolling-Circle Replication. Sci. Rep. 2015, 5 (1), 446 10404. https://doi.org/10.1038/srep10404. 447 (41) Sakatani, Y.; Yomo, T.; Ichihashi, N. Self-Replication of Circular DNA by a Self-448 Encoded DNA Polymerase through Rolling-Circle Replication and Recombination. Sci. Rep. 449 2018, 8 (1), 13089. https://doi.org/10.1038/s41598-018-31585-1. 450 (42) Libicher, K.; Hornberger, R.; Heymann, M.; Mutschler, H. In Vitro Self-Replication and 451 Multicistronic Expression of Large Synthetic Genomes. Nat. Commun. 2020, 11 (1), 904. 452 https://doi.org/10.1038/s41467-020-14694-2. 453 (43) Barthel, S.; Hoffmann-Becking, M.; Karimov, I. G.; Erb, T. J. Genetically Encoded 454 Control of In Vitro Transcription–Translation Coupled DNA Replication. ACS Synth. Biol. 455 2025, 14 (10), 4134–4142. https://doi.org/10.1021/acssynbio.5c00477. 456 (44) Vega, M. de; Salas, M. DNA Replication and Related Cellular Processes. 2011. 457 https://doi.org/10.5772/19309. 458 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint (45) Nies, P. van; Westerlaken, I.; Blanken, D.; Salas, M.; Mencía, M.; Danelon, C. Self-459 Replication of DNA by Its Encoded Proteins in Liposome-Based Synthetic Cells. Nat. 460 Commun. 2018, 9 (1), 1583. https://doi.org/10.1038/s41467-018-03926-1. 461 (46) Shimizu, Y.; Inoue, A.; Tomari, Y.; Suzuki, T.; Yokogawa, T.; Nishikawa, K.; Ueda, T. 462 Cell-Free Translation Reconstituted with Purified Components. Nat. Biotechnol. 2001, 19 463 (8), 751–755. https://doi.org/10.1038/90802. 464 (47) Heras, L. S.; Danelon, C. Autocatalytic Selection of Gene Functions in Vitro. bioRxiv 465 2025, 2025.08.25.672052. https://doi.org/10.1101/2025.08.25.672052 . 466 (48) Seo, K.; Ichihashi, N. Investigation of Compatibility between DNA Replication, 467 Transcription, and Translation for in Vitro Central Dogma. ACS Synth. Biol. 2023, 12 (6), 468 1813–1822. https://doi.org/10.1021/acssynbio.3c00130. 469 (49) Zheng, X.; Gao, W.; Liu, W.-Q.; Zhang, Y.; Huang, S.; Ji, X.; Lu, Y.; Liu, Y.; Ling, S.; Li, 470 J. Reconstituting Transcription–Translation-Coupled DNA Replication within Complex in 471 Vitro Biological Systems. Nat. Commun. 2025, 17 (1), 351. https://doi.org/10.1038/s41467-472 025-67411-2. 473 (50) Blanco, L.; Bernad, A.; Salas, M. Transition from Initiation to Elongation in Protein-474 Primed Phi 29 DNA Replication: Salt-Dependent Stimulation by the Viral Protein P6. J. Virol. 475 1988, 62 (11), 4167–4172. https://doi.org/10.1128/jvi.62.11.4167-4172.1988. 476 (51) Méndez, J.; Blanco, L.; Salas, M. Protein ‐ primed DNA Replication: A Transition 477 between Two Modes of Priming by a Unique DNA Polymerase. EMBO J. 1997, 16 (9), 478 2519–2527. https://doi.org/10.1093/emboj/16.9.2519. 479 (52) Abil, Z.; Sierra, A. M. R.; Stan, A. R.; Châne, A.; Prado, A. del; Vega, M. de; Rondelez, 480 Y.; Danelon, C. Darwinian Evolution of Self-Replicating DNA in a Synthetic Protocell. Nat. 481 Commun. 2024, 15 (1), 9091. https://doi.org/10.1038/s41467-024-53226-0. 482 (53) Abil, Z.; Giaveri, S.; Erb, T. J.; Rothschild, L. J. Integrating Metabolism and Evolution 483 towards the Realization of Synthetic Life. Nat. Rev. Bioeng. 2024, 3 (1), 9–10. 484 https://doi.org/10.1038/s44222-024-00267-1. 485 (54) Kalb, E. M.; Alejo, J. L.; Dias-Fields, L.; Knudson, I.; Davisson, J. A.; Maldonado, E.; 486 Chattrakun, K.; Lin, S.; Schepartz, A.; Zhang, S.; Blanchard, S.; Engelhart, A. E.; Adamala, 487 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint K. P. Purification of Post-Transcriptionally Modified tRNAs for Enhanced Cell-Free 488 Translation Systems. bioRxiv 2025, 2025.06.10.658963. 489 https://doi.org/10.1101/2025.06.10.658963. 490 (55) Zubay, G. The Isolation and Fractionation of Soluble Ribonucleic Acid. J. Mol. Biol. 491 1962, 4 (5), 347–356. https://doi.org/10.1016/s0022-2836(62)80015-1. 492 493 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 1, 2026. ; https://doi.org/10.64898/2026.02.27.708527doi: bioRxiv preprint