CN116064597B - Directed evolution and darwinian adaptation in mammalian cells by autonomous replication of RNA - Google Patents

Abstract

The present invention relates to an orthogonal alphavirus RNA replication system REPLACE system to evolve RNA-based elements, enabling sustained in vivo evolution of RNA replicase-assisted mammalian cells. The method of the present invention is capable of continuous diversification and selection of over 10 billion autonomously replicating RNA copies, RNA replication by replicase restriction patterns and powerful inducible RNA mutations. The present invention evolves new functions of fluorescent proteins, transcription factors, and mini-Cas proteins (dCasMINI) or improves upon existing functions. The present invention shows that cells equipped with replacement can adapt to challenges outside or inside the cell by continually evolving autonomously replicating RNAs that carry the key genes associated with cancer (i.e., MEK1 and KRAS). The novel RNA-based evolution platform of the invention provides a novel high-performance tool box for mammalian synthetic biology and facilitates the adaptive engineering of mammalian cells and tissues.

Description

Directed evolution and Darwinian adaptation in mammalian cells via autonomously replicating RNA

Technical Field

The present invention relates to the field of bioengineering, and in particular to directed evolution and Darwinian adaptation in mammalian cells achieved through autonomous replication of RNA.

Background Art

Although mammalian synthetic biology holds promise for addressing human health ^{issues7–9,21–24} , it still needs to overcome two major obstacles. First, due to the complexity of the correspondence between gene sequence and function, it is very laborious to optimize the synthetic machinery for a given ^{function7–9} . Second, the pre-set synthetic machinery is not only difficult to adapt to the changing environment indefinitely, but also may undergo ^mutations that lead to loss of function6,8,10,25,26. The directed evolution of synthetic machinery in mammalian cells can help address these challenges: on the one hand, directed evolution can accelerate the construction of synthetic machinery, that is, diversify genes or gene circuits (for example, by inducing mutations) and then screen ^them27–30 ; on the other hand, the ability to evolve synthetic machinery, especially continuous evolution in living cells, will enable cells to evolve and adapt to environmental changes, or reverse harmful mutations in synthetic ^{machinery10,25,26} .

As with the yeast-based directed evolution method OrthoRep, ³¹ an ideal mammalian-based directed evolution method should isolate the synthetic machinery from the host genome to ensure continuous, orthogonal, and long-term evolution. It also needs to overcome the difficulties of inefficient transfection and slow proliferation of mammalian cells1 ^, which make it difficult to generate large mutant libraries in mammalian cells. Furthermore, a relatively broad mutation spectrum is necessary to form a complete sequence-function map of the synthetic machinery, which will also improve the performance of the ^method1,3 . In addition, it is very beneficial to achieve adjustable control over the mutation ^rate3 , because this will enable the mutagenesis time and degree to be adjusted to suit various application scenarios. However, existing ^{methods1,2,11–19} do not meet all of the above requirements.

Summary of the invention

Different from the usual method of evolving DNA-based devices, the inventors chose to evolve RNA in mammalian cells for a number of reasons. First, the prospects for RNA-based therapeutic approaches are very promising: RNA can be evolved orthogonally due to its ease of entry into mammalian cells and tissues and minimal interference with the host ^genome32,33 . Second, mammalian cells can stably carry many copies of foreign ^RNA34 , providing a way to construct and screen large variant libraries. Third, RNA can be replicated in mammalian cells by RNA viral ^{replicases35,36} , and the accuracy of its replication can be regulated. By considering the above characteristics of RNA, the inventors here developed REPLACE ( RNA replicated continuous in vivo evolution ) technology, a directed evolution technology based on autonomously replicating RNA using mammalian cells as chassis cells. It can ensure orthogonality, construct a relatively large mutant library, provide a relatively balanced and broad mutation spectrum, allow control of the mutagenesis rate, and operate in continuous or discontinuous mode. The inventors here demonstrate a series of application scenarios of REPLACE, which will become a general and powerful synthetic biology tool for directly applicable mammalian biomacromolecule engineering and providing synthetic evolution capabilities in mammalian cells and tissues.

In one aspect, the present invention provides an RNA replicase-assisted continuous in vivo evolution (REPLACE) system, which comprises an autonomously replicating RNA and a host cell, and optionally further comprises a mutagen that induces mutations.

In one embodiment, the autonomously replicating RNA lacks RNA replicase, and the host cell constitutively expresses the lacked RNA replicase.

In another embodiment, the autonomously replicating RNA is derived from a positive-strand RNA virus, preferably from an alphavirus, more preferably from a Sindbis virus, more preferably from repRNA-v1, more preferably from repRNA-v2.

In another embodiment, the autonomously replicating RNA is selected from any one of repRNA-v3, repRNA-v4 or is derived from any one of repRNA-v3, repRNA-v4, preferably the autonomously replicating RNA is repRNA-v4 or is derived from repRNA-v4, more preferably the autonomously replicating RNA is repRNA-v4.

In another embodiment, the host cell is a eukaryotic host cell, preferably the host cell is an animal host cell, more preferably the host cell is a mammalian host cell, more preferably the host cell is selected from a hamster fibroblast cell line or derived from a hamster fibroblast cell line, more preferably the host cell is selected from BHK-21 (ATCC) or derived from BHK-21 (ATCC), more preferably the host cell is a replicase-limited eukaryotic host cell, preferably the host cell is a replicase-limited animal host cell, more preferably the host cell is a replicase-limited mammalian host cell, more preferably the host cell is selected from a replicase-limited hamster fibroblast cell line or derived from a replicase-limited hamster fibroblast cell line, more preferably the host cell is selected from a replicase-limited BHK-21 (ATCC) or derived from a replicase-limited BHK-21 (ATCC).

In yet another embodiment, the mutagen is selected from any one of a small molecule mutagen, a nucleoside analog, and optionally the mutagen is selected from any one of favipiravir, molnupiravir.

In another aspect, the present invention provides an RNA replicase-assisted continuous in vivo evolution (REPLACE) method, comprising: providing autonomously replicating RNA; providing a host cell; allowing the autonomously replicating RNA to continuously evolve in vivo; and optionally providing a mutagen that induces mutation.

In one embodiment, the autonomously replicating RNA lacks RNA replicase, and the host cell constitutively expresses the lacked RNA replicase.

In another embodiment, the autonomously replicating RNA is derived from a positive-strand RNA virus, preferably from an alphavirus, more preferably from a Sindbis virus, more preferably from repRNA-v1, more preferably from repRNA-v2.

In another embodiment, the autonomously replicating RNA is selected from any one of repRNA-v3, repRNA-v4 or is derived from any one of repRNA-v3, repRNA-v4, preferably the autonomously replicating RNA is repRNA-v4 or is derived from repRNA-v4, more preferably the autonomously replicating RNA is repRNA-v4.

In another embodiment, the host cell is a eukaryotic host cell, preferably the host cell is an animal host cell, more preferably the host cell is a mammalian host cell, more preferably the host cell is selected from a hamster fibroblast cell line or derived from a hamster fibroblast cell line, more preferably the host cell is selected from BHK-21 (ATCC) or derived from BHK-21 (ATCC), more preferably the host cell is a replicase-limited eukaryotic host cell, preferably the host cell is a replicase-limited animal host cell, more preferably the host cell is a replicase-limited mammalian host cell, more preferably the host cell is selected from a replicase-limited hamster fibroblast cell line or derived from a replicase-limited hamster fibroblast cell line, more preferably the host cell is selected from a replicase-limited BHK-21 (ATCC) or derived from any one of the replicase-limited BHK-21 (ATCC).

In yet another embodiment, the mutagen is selected from any one of a small molecule mutagen, a nucleoside analog, and optionally the mutagen is selected from any one of favipiravir, molnupiravir.

In yet another aspect, the present invention provides the use of the REPLACE system of the present invention in the manufacture of biomacromolecules, vaccines, drugs, in the generation of mutant libraries, or in achieving Darwinian adaptation.

In another aspect, the present invention provides a vector for expressing the autonomously replicating RNA of the present invention.

In yet another aspect, the present invention provides a host cell comprising an expression vector for expressing the autonomously replicating RNA of the present invention.

In another aspect, the invention provides a vaccine composition comprising RNA produced or producible by the REPLACE system of the invention.

In yet another aspect, the invention provides a pharmaceutical composition comprising RNA produced or producible by the REPLACE system of the invention.

In another aspect, the invention provides a delivery vector comprising RNA produced or producible by the REPLACE system of the invention.

Advantages of the present invention include that the present invention can ensure orthogonality, construct a relatively large-scale mutant library, provide a relatively balanced and broad mutation spectrum, allow control of the mutagenesis rate, and operate in a continuous or discontinuous mode.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic diagram of self-replicating RNA in mammalian cells. These two self-replicating RNAs are from the literature ^42,43 , and the candidate RNA variants therein show reduced cytopathic characteristics (see the method of this application for details).

Figure 2 is a quantification of the EGFP signal expressed 24 hours after electroporation by flow cytometry, and the number of cells expressing the indicated autonomously replicating RNAs 10 days after electroporation (under selection with 10 μg/ml puromycin). P values are from two-tailed t-tests, and error bars represent standard deviations (n=3 biological replicates).

Figure 3 is a library screening experiment of self-replicating RNA in mammalian cells. The RNA library with caps and tails containing polyadenylic acid (based on repRNA-v2) was electroporated into wild-type BHK-21 cells. The penultimate amino acid of nsP1 and nsP2 is encoded by the NNK codon. Representative images of EGFP (left) and bright field (right) show cells 7 days after electroporation.

Figure 4 quantifies the relative proportion of amino acids corresponding to each NNK codon in the RNA library before and after screening in Figure 3. Error bars represent standard deviations (n=3 biological replicates).

Figure 5 is a diagram of the construction and characterization of autonomously replicating RNAs with enhanced host compatibility in mammalian cells. a, Library screening experiments of designed viral autonomously replicating RNAs in mammalian cells. RNA containing nsP1-3 and a coding frame driven by a subgenomic (SG) promoter was replicated in cells expressing nsP4. The penultimate two amino acids of nsP1 and nsP2 are encoded in NNK codons. b, Representative EGFP (top) and bright field images (bottom) show cells 7 days after library electroporation.

Figure 6 is a quantification of cell number 7 days after screening of two RNA libraries (library 1 refers to Figures 3-4, and library 2 refers to Figures 5 and 7). P values are from two-tailed t-tests, and error bars represent standard deviations (n=3 biological replicates).

Figure 7 quantifies the relative proportions of individual amino acids at each NNK codon in the mutant library before and after screening. Error bars represent standard deviation (n=4 biological replicates).

FIG8 shows representative images of three versions of self-replicating RNA and their corresponding EGFP (top) and bright field (bottom), taken 1 day and 7 days after electroporation into host cells.

Figure 9 shows the cell numbers quantified by different RNA versions 7 days after electroporation. P values are from two-tailed t-tests, and error bars represent standard deviations (n=3 biological replicates).

FIG. 10 is an experimental design for quantifying and comparing short-term stability of RNA by measuring EGFP expression at early time points after RNA entry into cells.

Figure 11 is a flow cytometric quantitative analysis of EGFP and iRFP fluorescence in cells electroporated with different versions of RNA at the indicated time points. Note that only cells hosting repRNA-v4 expressed iRFP signals (i.e., indicating replicase levels).

Figure 12 is the quantification of EGFP signals (including the proportion of EGFP-positive cells and the average EGFP intensity of EGFP-positive cells) by two versions of autonomously replicating RNA over 3 days (repRNA-v2; repRNA-v3). n=2 biological replicates.

Figure 13 is a version of the self-replicating RNA quantification of EGFP signal over 3 days (including the proportion of EGFP-positive cells and the average EGFP intensity of EGFP-positive cells) (repRNA-v4). Cells with different iRFP signal levels were sorted and quantified to compare their behavior under different replicase levels. n = 2 biological replicates.

FIG14 is a pathway enrichment analysis of genes upregulated in cells carrying repRNA-v3 compared to cells carrying repRNA-v4 (see Methods for details).

Figure 15 shows short-term RNA stability quantified by relative changes in the EGFP-positive fraction at 36 and 72 hours after electroporation. repRNA-v4 was quantified in two hosts, where cells from either host were grouped by replicase level (iRFP) and the proportion of EGFP-positive cells was calculated for each population.

Figure 16 is a schematic diagram of a low replicase host carrying repRNA-v4. Compared with the high replicase host (Figure 5a, Figure 10 and see Methods for details), the low replicase host greatly reduces its expression by inserting a stop codon in front of the nsP4 gene.

Figure 17 is a quantification of the EGFP signal of low replicase host cells carrying repRNA-v4 within 3 days (including the proportion of EGFP-positive cells and the average EGFP intensity of EGFP-positive cells) (as shown in Figure 10). Cells with different iRFP signal levels were grouped and quantified to compare the behavior under different replicase levels, and the data were also used in Figure 15.

Figure 18 is a scatter plot showing the steady-state EGFP levels and the corresponding replicase levels of each cell population in Figure 15. The red line represents the linear fit.

Figure 19 shows the results of quantification of EGFP and iRFP signals of high replicase host cells with and without repRNA-v4 using FACS (3 weeks after electroporation) and the average iRFP signal level of the data (i.e., a representative of the replicase level), with error bars representing standard deviations (n=3 biological replicates).

FIG. 20 is a quantification of low replicase hosts similar to FIG. 19 and the mean iRFP signal levels of the data therein, with error bars representing standard deviation (n=3 biological replicates).

FIG. 21 quantifies long-term RNA stability by changes in the EGFP-positive fraction during one month of culture using a cell population that has stabilized after a period of culture (2 weeks after electroporation) (n=3 biological replicates).

Figure 22 is a flow cytometry quantification of EGFP and iRFP signals in high replicase host cells carrying repRNA-v4 over a one month time scale. The data are quantified and presented in Figure 21.

FIG. 23 is a representative image showing the superposition of dsRNA signals (red, immunostaining signals) and nuclear signals (blue, Hoechst) of host cells with and without repRNA-v4.

FIG. 24 is a quantification of dsRNA signals from the immunostaining experiments in FIG. 23 (see Methods for details) shown by box plots.

FIG. 25 shows the quantification results of the RPKM of Gapdh and nsP1-nsP3 by RNA sequencing of host cells with and without repRNA-v4 (n=2 biological replicates).

Figure 26 is a controlled RNA mutagenesis achieved by regulating the fidelity of RNA replicase using nucleoside analogs. a, Experimental design for mutation analysis of replicating RNA in the absence of perturbations (i.e., basal conditions). b, Total mutation frequency of target regions (nsP1-3) on plasmid DNA or autonomously replicating RNA at indicated time points after electroporation. Error bars represent standard deviations (n = 3 biological replicates).

Figure 27 is a scatter plot showing the relationship between total mutation frequency and the number of days replicated in host cells. Data are from b of Figure 26. Asterisks represent the average of 3 experiments. The red line represents the linear fit to the data.

Figure 28 is the mutation spectrum of IVT RNAs and autonomously replicating RNAs of cells 21 days after electroporation. Error bars represent standard deviation (n=3 biological replicates). p values were calculated by two-tailed t-test (*p<0.05, **p<0.01, and ***p<0.001).

Figure 29 is a characterization of the mutation capacity of the nsP4 C488G mutant. Left panel, comparison of the total mutation frequency of the target region (nsP1-3) 10 days after electroporation of wild-type nsP4 and nsP4 C488G mutant. Right panel, comparison of mutation spectra. P values were calculated by two-tailed t-test, and error bars represent standard deviations (n = 3 biological replicates).

Figure 30 shows the EGFP fluorescence level after treatment with the indicated concentrations of ribavirin and 5-AzaC for 72 hours. Error bars represent standard deviation (n=3 biological replicates).

Figure 31 is a FACS quantification of the fluorescence intensity of EGFP after treating cells with nucleoside analogs molnupiravir and favipiravir at the indicated concentrations for 72 hours. The upper right figure is a representative picture showing the reduction of EGFP fluorescence after molnupiravir treatment. The error bars represent standard deviations (n = 3 biological replicates).

FIG. 32 shows the total mutation frequency of the target region (nsP1-3) on the replicating RNA after treatment with molnupiravir or favipiravir at the indicated concentrations for 72 hours.

FIG. 33 is the distribution of EGFP signals in cells carrying repRNA-v4 after treatment with the indicated concentrations of molnupiravir or favipiravir for 72 hours.

Figure 34 is the distribution and corresponding bar graph of the EGFP signal after the cells carrying repRNA-v4 were processed by molnupiravir or favipiravir in the time course of 72 hours. Error bars represent standard deviation (n=3 biological repeats), and in 72 hours, the total mutation frequency of the target region on the replication RNA processed with molnupiravir or favipiravir. Error bars represent standard deviation (n=3 or 6 biological repeats).

Figure 35 is the distribution of iRFP signals of host cells carrying repRNA-v4 after treatment with the specified concentration of molnupiravir or favipiravir for 72 hours; distribution of iRFP signals of host cells carrying repRNA-v4 treated with molnupiravir or favipiravir on a 72-hour time scale; and quantification of constitutively expressed EGFP signals of control cell lines after treatment with the specified concentration of molnupiravir or favipiravir for 72 hours. Error bars represent standard deviations (n=3 biological replicates).

Figure 36 is a mutation spectrum of autonomously replicating RNA induced by molnupiravir or favipiravir. n = 6 (DMSOctrl) or n = 3 (other conditions) biological replicates. P values were calculated by two-tailed t-test (*p < 0.05, **p < 0.01, and ***p < 0.001).

FIG. 37 shows the design of the autonomously replicating RNA carrying EGFP and the corresponding host cells.

Figure 38 is the evolution experiment timeline of blue-shifted EGFP and the quantification of the corresponding flow analysis at the specified time point; and the mutation spectrum of EGFP mRNA after 2 days of molnupiravir treatment. The high baseline mutation frequency (such as DMSO conditions) is due to the cells being cultured for a long time before the experiment. Error bars represent standard deviation (n=3 biological replicates). P values are calculated by two-tailed t-test (*p<0.05, **p<0.01, and ***p<0.001).

Figure 39 is a nucleotide level mutation analysis of EGFP mRNA in the sorted cell population. Both RNA level and protein level mutations are marked. The details of the sorting are shown in the middle of Figure 38.

Figure 40 shows the percentage of the mutations in different cell populations. Error bars represent standard deviation (n=3 technical replicates).

FIG41 is a flow cytometric analysis of wild-type EGFP and two evolved mutants (Y66H, T203I), showing data from one of three biological replicates.

Figure 42 shows the transcription factors that respond to synthetic ligands using REPLACE. a, Design of self-replicating RNA carrying TetR-VP48 and host cells containing nsP4 and homologous TRE3G reporter genes. Note that this self-replicating RNA is similar to repRNA-v4, and the penultimate amino acid of nsP1 and nsP2 is valine. b, Experimental design and timeline of directed evolution of TetR-VP48. See Methods for details.

FIG. 43 is a nucleotide level mutation analysis of TetR mRNA in a sorted cell population (ie, 1% of cells with the highest mCherry signal).

Figure 44 shows the fluorescence intensity of the reporter gene activated by wild-type TetR-VP48 and evolved mutants at the indicated doxycycline (Dox) concentrations. Blue indicates the fold change of the reporter system signal under ligand and ligand-free conditions. Error bars indicate standard deviation (n=3 biological replicates).

Figure 45 shows the design of an autonomously replicating RNA carrying VP64-PadR, and the design of a host cell containing nsP4 and a reporter gene driven by a promoter containing a PadR binding site; and a timeline for the experimental design and directed evolution of VP64-PadR. See Methods for details.

FIG. 46 is a nucleotide level mutation analysis of PadR mRNA in the obtained cell population (without sorting).

Figure 47 is the fluorescence intensity of the reporter gene activated by wild-type VP64-PadR and evolved mutants at the specified sodium ferulate concentration. Blue represents the fold change of the reporter system fluorescence signal under ligand and ligand-free conditions. Error bars represent standard deviations (n=3-6 biological replicates).

Figure 48 is the fluorescence intensity of the reporter gene activated by wild-type VP64-PadR and the evolved mutants at the indicated sodium ferulate concentrations. The selected mutant is of particular interest because it exhibits an opposite ligand sensitivity compared to the wild type, which is also quantified in Figure 47. Error bars represent standard deviations (n=3 biological replicates).

Figure 49 is a timeline of the experimental design and directed evolution of VP64-PadR in the absence of a ligand (i.e., ligand-free evolution). In this experiment, the self-replicating RNA and host cells of Figure 45 were used. See Methods for details.

FIG. 50 is a nucleotide level mutation analysis of PadR mRNA obtained by flow cytometry in the ligand-free evolution experiment in FIG. 49 .

Figure 51 is the fluorescence intensity of the reporter gene activated by wild-type VP64-PadR and evolved mutants (from ligand-free evolution experiments) at the indicated sodium ferulate concentrations. The reporter gene expression fold change for each variant is indicated in blue. Error bars represent standard deviations (n=3 biological replicates). P values were calculated by two-tailed t-test (*p<0.05, **p<0.01, and ***p<0.001).

Figure 52 shows the ligand dose dependence and fitted dose response curves of wild-type VP64-PadR and VP64-PadR (D21G/D74V/M178T) mutants. The Hill coefficient and half inhibitory concentration (IC50) obtained by fitting for each material are indicated. Error bars represent standard deviations (n=3 biological replicates).

Figure 53 is the fluorescence intensity of the TRE3G reporter gene activated by wild-type dCasMINI-VPR and dCasMINI (S246R)-VPR. The fold change between the reporter levels of mutant and wild-type is indicated in blue. Error bars represent standard deviation (n = 5 biological replicates).

Figure 54 is the quantification of CD2 similar to h (using antibody staining). Ctrl indicates no sgRNA control. Error bars indicate standard deviation (n=4 biological replicates). P values were calculated by two-tailed t-test (*p<0.05, **p<0.01, and ***p<0.001).

Figure 55 is based on the base editing frequency of three endogenous sites in HEK293T cells after editing with wild-type dCasMINI or mutant dCasMINI (S246R). Error bars represent standard deviations (n = 4 biological replicates). P values were calculated by two-tailed t-test (*p<0.05 and **p<0.01). And the cutting frequency of four endogenous sites of CasMINI or mutant CasMINI (S246R) in HEK293T cells. Error bars represent standard deviations (n = 3 biological replicates). The sgRNA used is the sgRNA of literature ²⁰ .

FIG. 56 is an experimental design for testing whether RNA carrying MEK1 can adapt cells to challenge with therapeutic drugs.

Figure 57 is a summary of the mutations identified in MEK1 mRNA in cell populations adapted to three different MEK1 inhibitors. Each inhibitor had three biological replicates.

Figure 58 is a mutation analysis of MEK1 mRNA in cell populations adapted to three different MEK1 inhibitors. a, nucleotide level mutation analysis of MEK1 mRNA in cell populations collected before REPLACE experiment. b, nucleotide level mutation analysis of MEK1 mRNA in cell populations adapted to cobimetinib challenge. c, nucleotide level mutation analysis of MEK1 mRNA in cell populations adapted to trametinib challenge. d, nucleotide level mutation analysis of MEK1 mRNA in cell populations adapted to selumetinib challenge.

Figure 59 is a summary of mutations identified along the domains of MEK1. DD: docking domain; NES: nuclear export signal; NRR: negative regulatory region; KCD: kinase catalytic domain; AL: activation loop; DVD: multifunctional docking domain. The annotations on the domains were adapted from reference 60.

Figure 60 is the fluorescence intensity of SRE reporter gene activated by wild-type MEK1 and evolved mutants under the indicated conditions. Error bars represent standard deviation (n=3 biological replicates).

Figure 61 is an experimental design to test whether RNA carrying KRAS can adapt cells to a cell-intrinsic challenge (i.e., a dominant suppressive KRAS allele).

Figure 62 is a summary of mutations identified on KRAS(S17N) mRNA in a cell population adapted to dominant inhibition. For ease of comparison, the chart also includes the corresponding mutation information for experiments using wild-type KRAS. Three biological replicates were performed for each replicate RNA.

Figure 63 is additional features of the KRAS adaptive experiment. a, Nucleotide level mutation analysis of KRAS (wild type) mRNA in the cell population collected before the REPLACE experiment. b, Nucleotide level mutation analysis of KRAS (wild type) mRNA in the cell population collected on day 14. c, Nucleotide level mutation analysis of KRAS (S17N) mRNA in the cell population collected before the REPLACE experiment. d, Nucleotide level mutation analysis of KRAS (S17N) mRNA in the cell population that escaped dominant inhibition.

Figure 64 is a summary of mutations identified along the structural domains of KRAS. P-loop: phosphate binding loop; α3: α helix 3; HVR: hypervariable region. The annotations on the structural domains were adapted from 69.

Figure 65 is the fluorescence intensity of the SRE reporter gene activated by KRAS variants (top) and the corresponding effect on cell proliferation expressed by the number of cells relative to the baseline (bottom). Error bars represent standard deviations (n=5 or 6 biological replicates). For (d, h), p values were calculated by two-tailed t-tests (*p<0.05, **p<0.01, and ***p<0.001).

Figure 66 is the fluorescence intensity of the SRE reporter gene activated by KRAS variants (top) and the corresponding effect on cell proliferation relative to the number of cells at baseline (bottom). Error bars represent standard deviations (n=5-6 biological replicates). P values were calculated by two-tailed t-test (***p<0.001).

Figure 67 is the design of host cells carrying dCasMINI-VPR replicating RNA and nsP4, where the reporter gene is driven by the TRE3G promoter, and the U6 promoter-driven sgRNA targets the TetO site on the TRE3G promoter.

Figure 68 is the experimental design and directed evolution timeline of dCasMINI-VPR. See Methods for details.

FIG. 69 is a nucleotide level mutation analysis of dCasMINI mRNA in harvested cell populations (without sorting).

Figure 70 is the effect of different concentrations of nucleoside analogs on cell proliferation, quantified relative to the number of cells treated with DMSO. Error bars represent standard deviation (n=3 biological replicates).

DETAILED DESCRIPTION

The present invention can be implemented through the following embodiments, but the present invention is not limited thereto.

definition

As used herein, the term "RNA replicase" refers to RNA-dependent RNA polymerases that use RNA with a specific secondary structure as a template and replicate it.

As used herein, the term "autonomous replicating RNA" or "self-replicating RNA" refers to RNA that can replicate itself using its own RNA sequence as a template under the action of RNA replicase.

As used herein, the term "host cell" refers to a chassis cell capable of hosting autonomously replicating RNA.

As used herein, the term "replicase-limited host cell" refers to a host cell that expresses an RNA replicase at a moderate concentration and can maintain the long-term stable existence of self-replicating RNA.

As used herein, the term "mutagen" refers to a substance that can cause changes in the genetic material of a cell, causing gene mutations or chromosomal aberrations to exceed natural levels.

As used herein, the term "small molecule mutagen" refers to a mutagen of the chemical small molecule type.

As used herein, the term "nucleoside analog" refers to a structurally modified nucleoside (the portion of a nucleotide containing only a five-carbon sugar and a base).

As used herein, the term "positive-strand RNA virus" refers to a virus in which positive-strand RNA directly acts as mRNA, translates early proteins, and then, under the action of RNA replicase, uses the positive-strand RNA as a template to replicate to form a complementary strand (negative strand).

As used herein, the term "alphavirus" refers to a type of virus classified as Alphavirus in virus taxonomy.

As used herein, the term "Sindbis virus" refers to a positive-strand RNA virus called Sindbis virus under the genus Onychovirus.

As used herein, the term "repRNA-v1" refers to the replicative RNA obtained after the proline at position 726 of nsP2 of the Sindbis virus is replaced with leucine and the structural protein is replaced with EGFP-T2A-Puro.

As used herein, the term "repRNA-v2" refers to the replicative RNA obtained after glycine at position 539 of nsP1 of Sindbis virus is replaced by valine, proline at position 726 of nsP2 is replaced by leucine, glycine at position 806 of nsP2 is replaced by valine, and the structural protein is replaced by EGFP-T2A-Puro.

As used herein, the term "repRNA-v3" refers to the replicative RNA obtained after glycine at position 539 of nsP1 of Sindbis virus is replaced by alanine, proline at position 726 of nsP2 is replaced by leucine, glycine at position 806 of nsP2 is replaced by isoleucine, and the structural protein is replaced by EGFP-T2A-Puro.

As used herein, the term "repRNA-v4" refers to the replicative RNA obtained by replacing the glycine at position 539 of nsP1 of the Sindbis virus with alanine, replacing the proline at position 726 of nsP2 with leucine, replacing the glycine at position 806 of nsP2 with isoleucine, deleting nsP4 from the replicative RNA, and then using the EF1α promoter to drive the expression of nsP4 in the host cell, and replacing the structural proteins with EGFP-T2A-Puro.

As used herein, the term "evolution" refers to changes in inherited traits in a population from generation to generation at the cellular level.

As used herein, the term "directed evolution" refers to the process of artificially creating a large number of mutations in a laboratory, applying selection pressure in a specific direction according to a specific purpose, screening out protein molecules with desired characteristics, and realizing simulated evolution at the molecular level.

As used herein, the term "in vivo evolution" refers to the process of directed evolutionary mutation making occurring entirely within cells.

As used herein, the term "continuous in vivo evolution" refers to an experimental process in which cells obtained from the previous round of screening do not need to extract genetic material for in vitro mutagenesis, but can be directly used for the next round of mutagenesis after waiting for cell expansion.

As used herein, the term "vector" refers to a nucleic acid material that can transfer genetic material from one cell to another for the purpose of propagation, expression or separation.

As used herein, the term "prokaryotic vector" refers to a prokaryotic expression vector, which is a vector that can carry an inserted foreign nucleic acid sequence into prokaryotic cells for expression.

As used herein, the term "eukaryotic vector" refers to a eukaryotic expression vector, and in particular refers to a vector that can carry an inserted foreign nucleic acid sequence into eukaryotic mammalian cells for expression.

As used herein, the term "viral vector" refers to a virus-based vector that can carry inserted foreign nucleic acid sequences into mammalian cells for expression.

As used herein, the term "self-replicating vector" refers to a vector that can replicate autonomously in a mammalian host cell.

As used herein, the term "manufacturing biomacromolecules" refers to the process of engineering biomacromolecules such as proteins and nucleic acids using directed evolution to make them meet specific needs or purposes.

As used herein, the term "vaccine" refers to a biological product used for vaccination made from various pathogenic microorganisms or their partial proteins or nucleic acids and other biological macromolecules.

As used herein, the term "drug" refers to a substance that can prevent, treat and diagnose a disease.

As used herein, the term "mutant" refers to a nucleic acid molecule in which certain bases are mutated.

As used herein, the term "mutant library" refers to a population consisting of a large number (often in the billions) of nucleic acid molecules in which certain bases are mutated.

As used herein, the term "Darwinian adaptation" refers to a process similar to natural selection, in which cells utilize the REPLACE system to provide mutations when encountering changes in the external environment, and adapt to the new environment under the selection pressure provided by the environment.

As used herein, the term "orthogonal" means that the processes of introducing mutations into the host cell genome and the REPLACE system are independent of each other and do not affect each other.

As used herein, the term "fluorescent protein" refers to a class of proteins having a barrel-shaped structure and capable of emitting fluorescence.

As used herein, the term "transcription factor" refers to a DNA binding protein that can specifically interact with the cis-acting element of a gene and activate or inhibit the transcription of the gene.

As used herein, the term "Cas protein" refers to the abbreviation of CRISPR (clustered regularly interspaced short palindromic repeats) associated protein.

As used herein, the term "mini-Cas protein" refers to a Cas protein with a shorter protein sequence length than the classic Cas9, which is about a few hundred amino acids in length.

As used herein, the term "cancer-related key genes" refers to MEK1 and KRAS as used herein.

As used herein, the term "evolution platform" refers to a technology that can perform directed evolution of biological macromolecules.

As used herein, the term "DNA-based evolution platform" refers to a directed evolution technology that introduces mutations at the DNA level and constructs a mutant library.

As used herein, the term "RNA-based evolution platform" refers to a directed evolution technology that introduces mutations at the RNA level and constructs a mutant library.

As used herein, the term "synthetic biology" refers to an engineering discipline that specifically designs, modifies, or even resynthesizes organisms or cells to give them specific biological functions and to play a valuable role in fields such as medicine, agriculture, industry, and the environment.

As used herein, the term "engineering" refers to the process of changing the sequence of an existing biomacromolecule and optimizing its function to make it meet people's specific usage requirements.

It should be understood that other terms appearing in this article should be understood in accordance with the conventional meanings of the art as understood by those of ordinary skill in the art. In the event that a term defined in this article conflicts with a corresponding definition in other documents in the art, the term defined in this article shall prevail.

method

Plasmid construction

Plasmids used in this study were constructed by using Gibson assembly or restriction ligation. Primers used for cloning or sequencing were ordered from GENEWIZ. Amplification of DNA or cDNA fragments was performed using KOD One ^™ PCR Master Mix (TOYOBO, #KMM-101). Plasmids were expressed in E. coli strains. 5α (TSINGKE, TSC-C01), extracted with HiPure Plasmid DNA Mini Kit (Magen, #P100103), and finally confirmed by Sanger sequencing (RUIBO).

The relevant plasmid information is as follows:

Cell culture

Hamster fibroblast cell line BHK-21 (ATCC) and human cell line HEK293T (ATCC) were used in this study. BHK-21 cells were cultured in MEM/EBSS medium (HyClone, #SH3002401) supplemented with 10% FBS (Excell, #FSP500), 100 units/mL penicillin, 100 μg/mL streptomycin, and 1% MEM non-essential amino acid solution (Gibco, #11140050). HEK293T cells were cultured in DMEM medium (Gibco, #C11995500BT) containing 10% FBS (Excell, #FSP500), 100 units/mL penicillin, and 100 μg/mL streptomycin. All cell cultures were placed in a humidified incubator at 37°C with 5% _CO2 , and mycoplasma testing was performed daily.

Plasmid transfection and construction of stable cell clones

Plasmid transfection was performed by using Lipofectamine ^™ LTX reagent with PLUS ^™ reagent (Thermofisher #15338100) or PEI reagent (polyethylenimine, 1 μg/μL). PiggyBac (PB) transposon system or lentiviral packaging system was used for stable integration of plasmids. To construct a replicase-limited host cell line (i.e., a BHK-21 cell line stably expressing Sindbis viral RNA-dependent RNA polymerase (nsP4) on the BHK-21 genome), 2.5 μg of pEF1α-iRFP ^- NLS-P2A-ΔnsP3-nsP4 plasmid (or pEF1α-iRFP-NLS-TAG-ΔnsP3-nsP4 for cell lines with reduced nsP4 expression) and 0.5 μg of PiggyBac transposase plasmid were co-transfected into one well of a 6-well plate containing BHK-21 cells according to the instructions of the Lipofectamine ^{™ LTX reagent manual, with a cell confluency of ∼60%. The medium was changed 12 hours after transfection. After one week of culture, iRFP-positive BHK-21 cells were enriched by fluorescence-activated cell sorting (FACS) using a BD FACSAria™} III cell sorter. To report the activity of ligand-controlled synthetic transcription factors or Cas-based transcriptional regulatory proteins, the corresponding reporter plasmids were packaged into lentiviruses, and the resulting viral particles were used to transfect the replicase-limited BHK-21 cell line. Specifically, for TetR-VP48, the plasmid pTRE3G-puroR-P2A-NLS-mCherry-BGHployA-pEF1α/HTLV-mTagBFP2-NLS-WPRE was used; for PadR-VP64, the plasmid (O _PadR ) ₆ -phCMVmin-puroR-NLS-mCherry-hPEST-BGHployA-pEF1α/HTLV-mTagBFP2-NLS-WPRE was used; and for dCasMINI-VPR, the plasmid pTRE3G-puroR-P2A-NLS-mCherry-BGHployA-phU6-sgRNA_for_TetO-pEF1α/HTLV-tagBFP2-NLS-WPRE was used. For viral packaging, HEK293T cells were cultured to ∼70% cell confluence in one well of a 6-well plate and transfected with 3 μg of reporter plasmid, 2.66 μg of pCMVDR8.91 plasmid, and 0.34 μg of PMD2.G plasmid according to the instructions of the PEI reagent manual. The medium was replaced with 2 mL of fresh medium 8 hours after transfection. After 48 hours, the supernatant containing viral particles was centrifuged at 500 g for 3 minutes and filtered through a 0.45 μm filter. 500 μL of viral supernatant was added dropwise to one well of a 6-well plate in which the confluence of BHK-21 cells was ∼30%. The medium was replaced 12 hours after viral transduction. Four days later, BFP-positive BHK-21 cells were enriched by flow sorting. These polyclonal cells were cultured and used for subsequent experiments.

In vitro transcription and electroporation of autonomously replicating RNA

RNA was first transcribed in vitro and then electroporated into BHK-21 cells. To prepare the template DNA for in vitro transcription, 6 μg of plasmid DNA was linearized with 4 μL XbaI (NEB, #R0145) in an 80 μL reaction at 37°C for more than 12 hours. To terminate the enzyme digestion reaction, 4 μL 0.5M EDTA, 8 μL 3M sodium acetate, and 184 μL ethanol were added to the reaction. The mixture was thoroughly mixed and frozen at -80°C for at least 20 minutes, then centrifuged at 13,000 × g for 15 minutes at 4°C to precipitate the linearized DNA. After removing the supernatant, the usually invisible precipitate was carefully washed with 500 μL 75% ice-cold ethanol, and then centrifuged at 13,000 × g for 3 minutes at 4°C. The washing process was repeated twice. After completely removing the ethanol, the linearized DNA was resuspended in 10 μL of 10% (v/v) RNAse inhibitor solution (RNasin, Promega, #N2111) and immediately used as a template for in vitro transcription. The capping transcription reaction was performed using the prepared template DNA according to the user guide of the SP6 transcription kit (Thermo, #AM1340). For a 10 μL reaction, 5 μL of 2X NTP/CAP, 1 μL of 10X reaction buffer, 1 μL of GTP, 300 ng of linear template DNA, 1 μL of enzyme mix, and ~1 μL of RNA-free H ₂ O were used. After incubation at 37°C for 2 hours, the mixture was used immediately or stored at -80°C. If necessary, 1 μL of TURBO DNase was added to remove DNA, followed by incubation at 37°C for 15 minutes and purification by lithium choloride precipitation. For the electroporation step, the inventors modified the protocol from a previous literature ⁸¹ . Briefly, BHK-21 cells cultured in 100 mm dishes were harvested by trypsinization at 90% confluence and then washed twice with 10 mL of ice-cold DPBS. Then 4×10 ⁶ cells were suspended in 390 μL of ice-cold DPBS and mixed with 10 μL of in vitro transcribed RNA. The mixture was transferred to a frozen 2 mm gap sterile electroporation cuvette and electroporated using a Bio-rad GenePulser II with three exponential decay pulses at 3 s intervals. The settings of the device were as follows: field strength 2250 V/cm, capacitance 25 μF, and infinite resistance. After electroporation, the cells were allowed to recover at 0°C for 10 minutes, diluted in 10 mL of BHK-21 cell culture medium, and then transferred to a 100 mm tissue culture dish for normal culture.

Screening of mutation library of the penultimate amino acid of nsP1 and nsP2

To determine the most favorable combination of the penultimate amino acids of nsP1 and nsP2 for RNA host compatibility, the inventors constructed a two-site saturation mutation library based on 'NNK' (nsP1: G539X, nsP2: G806X; X represents one of the 20 protein-encoding amino acids). The construction of the mutation library involved two consecutive steps: First, an intermediate vector based on the Sindbis replicon plasmid (ML702|pSP6-nsP1-nsP3-nsP4-pSG-GFP-P2A-puroR) was constructed, nsP2 was deleted, and two restriction enzyme sites (NheI and EcoRI) were added. Second, nsP2 was PCR amplified with degenerate primers, and these fragments were connected to ML702 cut with NheI and EcoRI by Gibson assembly. The mutant library was then transcribed in vitro and electroporated into wild-type BHK-21 cells. After transfection, the cells recovered for 24 hours, and then 10 μg/mL of puromycin was added. The old medium was replaced with fresh medium containing puromycin every 2 days. After two weeks, cells were collected using TRIzol ^TM reagent (Thermo Fisher, #15596018) to extract RNA. About 3 μg of total RNA was reverse transcribed using HiScript III First Strand cDNA Synthesis Kit (Vazyme, #R312-02), and the reverse transcription primers used were replicon-specific primers. cDNA was PCR amplified using KOD One ^TM PCR Master Mix and target-specific primers. PCR products were sequenced using a fast NGS service (Tsingke Biotechnology) to quantify the ratio of each amino acid type of the penultimate amino acid of nsP1 and nsP2. Similar quantification was performed on the mutant library before screening. The enrichment multiple of each amino acid was quantified as the ratio after screening and before screening. Similar library screening experiments and quantification were performed on the replicase-limited BHK-21 stable cell line, in which nsP4 was replaced by NeoR.

Live cell imaging

During the culture of BHK-21 cells transfected with different versions of Sindbis viral RNA, 100 mm tissue culture dishes containing cells were removed from the incubator at designated time points in the relevant experiments for live cell imaging. Bright field and GFP fluorescence of cells were captured using LPlanFL PH2 4x/0.13 or LPlanFL PH2 10x/0.30 objective lens on FL automatic imaging system (Thermo Fisher Scientific). The obtained images were adjusted and cropped using ImageJ2 (https://imagej.net/software/imagej2/).

Streaming analysis

To analyze the expression of fluorescent proteins (i.e., mTagBFP2, EGFP, mCherry, and iRFP), single cell suspensions were prepared and flow cytometry was performed on a BD LSRFortessa ^TM cell analyzer. Data were analyzed by a MATLAB script (https://antebilab.github.io/easyflow/). Unless otherwise stated, positive cells were determined based on non-transfected control cells, and mean fluorescence intensity (MFI) was used to represent protein expression levels. To analyze RNA replication at early time points of cell entry, replicase levels (i.e., nsP4 expression levels) were approximated by iRFP levels. Cells with similar replicase levels were sorted, and the MFI of EGFP (expressed from autonomously replicating RNA) for each group of cells was calculated as a representative of RNA levels. For experiments analyzing the expression of CD2 using flow cytometry, the inventors used methods from previous ^literature20 . In brief, for each replicate, 2 μg of the corresponding plasmid (pEF1α-NLSdCasMINI-NLS-VPR-phU6-sgRNA_to_CD2-pEF1α/HTLV-mTagBFP2-NLS) was transfected into HEK293T cells using PEI reagent. Three days later, single cell suspensions were prepared, stained with FITC-CD2 antibody (Biolegend, #309206) in PBS plus 5% FBS at 4°C for 1 hour, and analyzed by flow cytometry. The data obtained were processed as described.

Immunofluorescence analysis of dsRNA

For immunostaining of viral double-stranded RNA (dsRNA), BHK-21 cells were cultured on 24-well glass plates 24 hours in advance. The cells were fixed with 4% paraformaldehyde for 20 minutes at room temperature, washed twice with PBS, and permeabilized with 100% ice-cold methanol for 30 minutes at -20°C. After washing twice with PBS, the cells were blocked with 1% IgG-free bovine serum albumin (BSA) for 30 minutes. The cells were then incubated with 0.5 μg/mL J2 antibody (Scicons #10010200) containing 1% BSA for 1 hour. Since the diluted J2 antibody may contain a small amount of denatured protein aggregates, which may cause high background in immunofluorescence detection, the diluted J2 antibody needs to be centrifuged at 13,000 × g for 3 minutes before use. After washing twice with 1% BSA, the cells were washed with 1/1000 diluted goat anti-mouse IgG H&L (Alexa Fluor ELISA Kit). 568) secondary antibody (Abcam) was incubated in the dark for 45 min and then washed twice with PBS in the dark. Then 5 μg/mL Hoechst 33342 was added to stain the nuclei at room temperature for 10 min in the dark and then washed twice with PBS in the dark. The cells were then imaged with a Plan-Apo 63x/1.4 oil objective (Leica) on a Dragon Fly200 high-speed confocal platform (ANDOR). Images were processed using ImageJ and in-house MATLAB scripts. In terms of image processing, the fluorescence Z-axis slice images were first merged, thresholded, and segmented according to the maximum value. For dsRNA immunofluorescence images, a gray value of 200 was used as the threshold, and the particle analysis plug-in in ImageJ was applied to the resulting binary images to quantify individual particles. For nucleus images, the threshold was calculated using the Huang ^method82 , and nuclear masks of individual cells were generated by the watershed algorithm after consecutive dilation and erosion operations. Particle analysis was applied to the nuclear mask images to quantify the size of the nucleus. The output of the particle analysis plugin was further processed in MATLAB to estimate the number of dsRNA per cell in each field of view. By comparing the images of the control and experimental groups, the true dsRNA signal was empirically defined as particles with at least 6 connected pixels. The signal intensity of a single dsRNA molecule was estimated by using the average intensity of the background-subtracted dsRNA signal of the smallest size. The background-subtracted true dsRNA signal in each field of view was then summed and divided by the estimated intensity of a single dsRNA and the number of cell nuclei to obtain the estimated number of dsRNA per cell in the field of view of interest.

RNA sequencing experiments and data analysis

In order to compare the transcriptomes between cells carrying different autonomously replicating RNAs, 300,000 BHK-21 cells carrying repRNA-v3 cells or repRNA-v4 cells (cultured for about 3 weeks after electroporation) were sorted using a BD FACSAria ^TM III cell sorter. Total RNA was extracted using TRIzol ^TM reagent (Thermo Fisher Scientific, #15596018), and the library was prepared using VAHTSUniversal V6 RNA-seq Library Prep Kit for MGI (Vazyme, #NRM604-2). The resulting library was sequenced by next generation sequencing on the DNBSEQ-T7 platform by Annouda Gene Technology. The paired-end 150bp sequences with low-quality sequences removed were aligned to the Mesocricetus auratus genome (GCF_017639785.1_BCM_Maur_2.0_genomic.fna) with the addition of the autonomously replicating RNA genome (nsP1, nsP2, nsP3, and GFP) using STAR2.7.8a. The genome annotation (GCF_017639785.1_BCM_Maur_2.0_genomic.gtf) was downloaded from NCBI and edited to include gene information on the autonomously replicating RNA genome. Transcript abundance was calculated by featureCounts and normalized and statistically compared using DESeq2. RPKM values were then calculated and biological pathway enrichment analysis was performed on the Reactome website (https://reactome.org/).

Mutation analysis of autonomously replicating RNA

To analyze the mutation rate and mutation spectrum of autonomously replicating RNA (repRNA-v4, unless otherwise stated) under different conditions, such as with or without replicase mutagens, and before or after evolution, the corresponding cells were collected and subjected to total RNA extraction and reverse transcription. The sequences of interest (e.g., nsP1-nsP3, EGFP, TetR-VP64, VP64-PadR, dCasMINI, MEK1, KRAS) were amplified from cDNA samples using KOD One ^TM PCR Master Mix and target-specific primers. The mutation rate and mutation spectrum were quantified by using NGS sequencing (Qingke Bio's Rapid NGS Service), and individual cDNA variants were obtained by Sanger sequencing of cDNA cloned into plasmids (RUIBIO). For the processing of NGS data, the cleaned reads were aligned to the corresponding reference sequences using BWA 0.7.17-r1188. The resulting Bam files were processed using in-house Bash and Python scripts to extract the required mutation information. For analysis of mutational spectra, the frequency of each mutation type (e.g., U to C) was calculated as the number of corresponding mutations divided by the total number of sequenced nucleotides. For analysis of mutations in evolved RNA, the percentage of nucleotides with mutations per nucleotide was expressed.

EGFP evolution experiment

The design of the system and the timeline of the experiment are shown in the top of Figures 37 and 38. The cells used in this experiment were cultured for about 2 weeks after electroporation and were initially used for mutation-related experiments in Figures 26, 28, 31, 32, and 36. The evolution experiment was performed in a 100 mm tissue culture dish. Cells at ~20% confluence were treated with 10 μM Molnupiravir (MCE, #HY-135853) for 2 days, and the culture medium was changed every day. The cells were then transferred to 5 100 mm culture dishes and allowed to grow for 3 days. Approximately 30 million cells were flow sorted to obtain blue-shifted EGFP mutants, of which 0.2% of the cells were sorted as positive. The sorted cells were cultured for about 7 days under the action of 10 μg/mL puromycin until enough cells were obtained for a second round of evolution, which was the same as the first round. Autonomously replicating RNA extracted from cells during the experiment (intermediate cell population or final cell population) was reverse transcribed and mutations were quantified by NGS sequencing. The cDNA library (containing the EGFP fragment) was cloned into a vector (pEF1α-EGFP mutant-P2A-Puro-BGH polyA-pEF1α/HTLV-iRFP-NLS), which was transformed into E. coli, and each mutant was identified by Sanger sequencing. To quantify the emission spectrum of EGFP variants, 600 ng of EGFP wild-type or mutant plasmids were transfected into HEK293T cells in 24-well plates using PEI reagent. After 12 hours of transfection, the medium was replaced with fresh medium, and the fluorescence intensity of the EGFP and BFP channels was measured by flow cytometry one day later.

TetR-VP48 (tTA) evolution experiments

The design of the system and the timeline of the experiment are shown in Figure 42. TetR-VP48 was cloned into a plasmid similar to repRNA-v4 (its three key amino acids are 'VLV' instead of 'ALI' in repRNA-v4), and the resulting self-replicating RNA was electroporated into BHK-21 cells expressing nsP4 (a stable cell line containing the ML700 plasmid). 24 hours after electroporation, the culture medium was replaced with fresh medium containing 5μg/mL puromycin and 100μg/mL neomycin. After 3 days of selection, a cell population (about 1 million cells) that stably carries the self-replicating RNA was obtained and used as the starting material for the directed evolution experiment in two 100 mm culture dishes. The steps of mutagenesis and selection are shown in Figure 42b. The cells in the two culture dishes were combined for FACS. The sorted cells were cultured for 3 days under the action of 80μg/mL puromycin and collected for RNA extraction, sequencing and mutation analysis. The cDNA library (containing the TetR gene) was cloned into a vector (pEF1α-TetR_mutant-VP48-BGH polyA-pEF1α/HTLV-iRFP-NLS), which was transformed into E. coli, and individual mutants were determined by Sanger sequencing. To quantify the ligand sensitivity of TetR variants, 400ng of TetR (wild type or mutant)-VP48 plasmid and 200ng of reporter plasmid (pTRE3G-Citline-NLS) were co-transfected into HEK293T cells in a well of a 24-well plate containing 1ng/μL doxycycline (Dox). After 12 hours of transfection, the medium was replaced with fresh medium containing Dox, and the fluorescence intensity was measured by flow cytometry one day later.

VP64-PadR evolution experiments

The design of the system and the timeline of the experiment are shown at the top of Figure 45. VP64-PadR ⁵⁶ (ENCU, gifted by Professor Ye Haifeng) was cloned into the repRNA-v4 plasmid, and the resulting self-replicating RNA was electroporated into BHK-21 cells expressing nsP4 (stable cell line containing ML700 plasmid) and carrying a stably integrated reporter plasmid ((O _PadR ) ₆ -phCMVmini-PuroR-NLS-mCherry-hPEST-BGHpolyA-pEF1α/HTLV-tag BFP2-NLS). 24 hours after electroporation, fresh culture medium containing 5 μg/mL puromycin and 10 μg/mL blasticidin was replaced. After 3 days of selection, a cell population (about 5 million cells) that stably carried the self-replicating RNA was obtained and used as the starting material for directed evolution experiments in 100 mm culture dishes. The experiment included two independent directed evolution activities, one to evolve ligand-resistant mutants and the other to evolve enhanced activation of target genes in the absence of ligands. The corresponding mutation and selection processes are summarized at the bottom of Figure 45 and Figure 49. On day 21, all cells in the first experiment and cells screened by mCherry in the second experiment were collected for RNA extraction, sequencing and mutation analysis. The cDNA library (containing the PadR gene) was connected to a vector (pEF1α-VP64-PadR_mutant-BGHpolyA-pEF1α/HTLViRFP-NLS), which was then transformed into Escherichia coli, and single mutants were determined by Sanger sequencing. To quantify the ligand sensitivity of PadR variants, 400 ng of VP64-PadR (wild type or mutant) plasmid and 200 ng of reporter plasmid ((O _PadR ) ₆ -phCMVmini-Puro-NLS-mCherry-hPEST-BGHpolyA-pEF1α/HTLV-mTagBFP2-NLS) were co-transfected into HEK293T cells in one well of a 24-well plate in a medium containing sodium ferulate (SF). 12 hours after transfection, the medium was replaced with fresh medium containing sodium ferulate, and the fluorescence intensity was measured by flow cytometry one day later.

CasMINI evolution experiment

The design of the system and the timeline of the experiment are shown in Figures 67 and 68. NLS-dCasMINI (v4) -NLS-VPR ²⁰ was cloned into the repRNA-v4 plasmid, and the resulting self-replicating RNA was electroporated into BHK-21 cells expressing nsP4 (stable cell line containing ML700 plasmid) and carrying a stably integrated reporter plasmid (pTRE3G-PuroR-P2A-NLS-mCherry-BGHpolyA-phU6-sgRNA_TetO-pEF1α/HTLV-tagBFP2-NLS). 24 hours after electroporation, the old culture medium was replaced with fresh culture medium containing 5 μg/mL puromycin and 10 μg/mL blasticidin. After 5 days of selection, a cell population (about 2 million cells) that stably carried self-replicating RNA was obtained and used as the starting material for directed evolution experiments in 100 mm culture dishes. On the 22nd day, cells were collected for RNA extraction, sequencing, and mutation analysis. The cDNA library (containing the dCasMIN gene) was connected to a vector (pEF1α-NLS-dCasMINI_mutant-NLS-VPR-BGH polyA-pEF1α/HTLV-iRFP-NLS), transformed into Escherichia coli, and mutants were identified by Sanger sequencing. The inventors only obtained one enriched mutation (S246R) in two replicates. The inventors characterized three functions of the mutant, including transcriptional activation function, base editing function and nuclease function. The information of sgRNA and the corresponding target gene information came from previous ^studies20 . As a transcriptional activator, the inventors characterized the ability of the wild type (dCasMINI (V4)) and mutant (S246R) to activate the reporter system used in the evolution process and the endogenous gene CD2. As a base editor or nuclease, the inventors tested two endogenous genes IFNγ and VEGFA. The plasmid information used in these experiments is detailed in Supplementary Table 1. In all tests, 2 μg of plasmid was transfected into HEK293T cells in one well of a 12-well plate. 12 hours after transfection, the medium was replaced with fresh medium, and 60 hours later, the cells were collected for flow cytometry analysis or sequencing.

MEK1 evolution experiments

The design of the system and the timeline of the experiment are shown in Figure 56. Human MEK1 cDNA was cloned into the repRNA-v4 plasmid, and the resulting self-replicating RNA was electroporated into BHK-21 cells expressing nsP4 (stable cell line containing the ML700 plasmid). 24 hours after electroporation (i.e., day 1), the culture medium was replaced with fresh culture medium containing 5μg/mL puromycin. After 4 days of selection (i.e., day 5), 2μM Molnupiravir was added to the culture medium. On day 7, therapeutic drugs (i.e., 2μM cobimetinib, 5μM trametinib, or 50μM selumetinib) were added to the culture medium. One week later (i.e., day 14), cells were collected for total RNA extraction, sequencing, and mutation analysis. The cDNA library (containing the MEK1 gene) was cloned into a vector (pEF1α-MEK1_mutant-BGHpolyA-pEF1α/HTLV-iRFP-NLS), which was transformed into E. coli, and individual mutants were identified by Sanger sequencing. To quantify the drug sensitivity of MEK1 variants, 400 ng of plasmid containing MEK1 (wild type or mutant) and 200 ng of SRE reporter plasmid (7×SREphCMVmini-NLS-mCherry-BGHpolyA-pEF1α/HTLV-mTagBF P2-NLS) were co-transfected into HEK293T cells in 24-well plates, and the culture medium contained 2 μM cobimetinib. After 12 hours of transfection, the culture medium was replaced with fresh culture medium containing cobimetinib, and the fluorescence intensity was measured by flow cytometry one day later.

KRAS(S17N) evolution experiment

The design of the system and the timeline of the experiment are shown in Figure 61. The cDNA of human wild-type KRAS or KRAS (S17N) was cloned into the repRNA-v4 plasmid, and the resulting self-replicating RNA was electroporated into BHK-21 cells expressing nsP4 (stable cell line containing the ML700 plasmid). 24 hours after electroporation (i.e., day 1), the culture medium was replaced with fresh culture medium containing 5μg/mL puromycin. After 4 days of selection (i.e., day 5), 2μM molnupiravir was added to the culture medium. After 10 days of culture, the cells were collected for total RNA extraction, sequencing, and mutation analysis. The cDNA library (containing the KRAS gene) was cloned into a vector (pEF1α-KRAS_mutant-BGHpolyA-pEF1α/HTLViRFP-NLS), which was transformed into Escherichia coli, and single mutants were identified by Sanger sequencing. To quantify the signaling activity of KRAS variants, 400 ng of plasmid containing KRAS (wild type or mutant) and 200 ng of SRE reporter plasmid (7×SRE-phCMVmini-NLS-mCherry-BGHpolyA-pEF1α/HTLV-mTagBF P2-NLS) were co-transfected into HEK293T cells in 24-well plates. After 12 hours of transfection, the medium was replaced with fresh medium, and the cell number and fluorescence intensity were measured by flow cytometry one day later.

Example

Example 1 - Engineering autonomously replicating RNA suitable for in vivo evolution

RNA replicons from alphaviruses (positive-strand RNA viruses) have been widely used not only to express heterologous proteins such as vaccine antigens in animal cells and ^{tissues37–39} but also in synthetic biology in ^mammals9,40 . However, they induce strong cytopathic ^effects35,41 , disrupting cellular physiological functions and preventing long-term propagation of replicons. Therefore, it is necessary to improve host cell compatibility by engineering alphavirus self-replicating RNA so that directed evolution can be performed in vivo using this RNA.

Based on work with replicons derived from Sindbis virus (an alphavirus), the inventors started with two versions of the replicon that had been shown to have low cytopathicity. ^{42, 43} The first version (repRNA-v1) contained a mutation in nonstructural protein 2 (nsP2), while the second version (repRNA-v2) contained this mutation and two mutations in the linker peptides of nsP1-nsP2 and nsP2-nsP3, respectively (i.e., the penultimate residues of nsP1 and nsP2; Figure 1). These replicons were transcribed in vitro and electroporated into BHK-21 (baby hamster kidney fibroblast) cells (see Methods for details). The inventors found that the number of cells surviving electroporation was significantly higher in the second version compared to the first version (Figure 2), indicating that perturbations in the processing of the polyprotein (i.e., P123 or P1234) significantly affected the cytopathicity. ^{42, 43} Therefore, the inventors screened these two linker residues (Figure 3, see Methods for details). The present inventors found that alanine was highly enriched at the penultimate residue of nsP1, while no amino acid was significantly enriched at nsP2 (Figures 3 and 4), suggesting that a single linker mutation of repRNA-v2 could improve cell compatibility. In addition, since the exact same mutation of the penultimate residue of nsP1 was shown to interfere with viral ^{replication42} , it is reasonable to infer that host cell compatibility may be improved by interfering with RNA replication.

Based on the above findings, the inventors attempted to further improve the compatibility of host cells by expressing the RNA replicase nsP4 in the host genome (rather than from the RNA itself). This will restrict the replication of RNA because the replication reaction will be subject to limited replicase and non-autonomous catalysis. The inventors then integrated a constitutively expressed nsP4 into the genome (see Methods for details) and re-constructed the library screen at the same residue position using RNA without nsP4 (a and b in Figure 5). Interestingly, the cells in this screen grew faster than the screen using repRNA-v2 (Figure 6). In addition, alanine was enriched again in the nsP1-nsP2 linker, and like the previous screen, no amino acid was significantly enriched in the other linker position (Figure 7).

Based on these screens, the inventors constructed two new versions of self-replicating RNA: repRNA-v3, which is based on repRNA-v2 and contains alanine and isoleucine as the penultimate residues of nsP1 and nsP2, respectively; and repRNA-v4, which contains the same mutations as repRNA-v3 but lacks nsP4 (and therefore requires replication in cells expressing nsP4). When constructing two versions of self-replicating RNA (repRNA-v3 and repRNA-v4), the inventors sought to compare the same self-replicating RNA with replicases expressed from different locations, either from the RNA itself (repRNA-v3) or from the host genome (repRNA -v4). Therefore, it was desired that both versions of RNA include the same penultimate residues of nsP1 and nsP2. Since alanine was enriched in both screens of nsP1-2 linker residues, the inventors selected alanine as the penultimate residue of both versions of nsP1. In contrast, no amino acid was significantly enriched in the nsP2-3 linker from either screen. Therefore, the inventors decided to select an amino acid (for the penultimate residue of nsP2) that might improve host compatibility when using a version with host-expressed replicase, as this version is the first choice for in vivo evolution experiments. Therefore, the inventors selected isoleucine because it was enriched in one of the repeated screens). The inventors compared the two new versions with repRNA-v2. The inventors found that the linker peptide mutation did improve cell compatibility (i.e., repRNA-v3 relative to repRNA-v2), and that limiting the level of replicase by expression from the host genome greatly enhanced host cell compatibility (i.e., repRNA-v4 relative to repRNA-v3) (Figures 8 and 9). These results suggest that when replicase levels are limiting, viral RNA replication appears to be accepted by host cells, resulting in ~10 ⁸ host cells carrying autonomously replicating RNA within a week (Figure 9). However, the above measurements were performed one week after electroporation, so the extent to which RNA is accepted by host cells at early and late stages (i.e., short-term and long-term stability) remains unknown.

Example 2 - Modified RNA has both short-term and long-term stability

The inventors first quantified the short-term stability of three autonomously replicating RNA versions (repRNA-v2 to v4) by measuring the EGFP expressed by the RNA at the early time point of entering the cell (Figures 10 and 11). For the version without limiting the replicase, the inventors found that the optimization of the linker peptide segment (repRNA-v3) improved the compatibility of the host cell compared with the unoptimized version (repRNA-v2), resulting in a higher maximum proportion of EGFP-positive cells and a higher proportion of EGFP-positive cells remaining at 72 hours (Figure 12). However, for these two RNAs, the proportion of EGFP-positive cells dropped to less than 10% within 3 days after electroporation, indicating that the host compatibility is generally low. In contrast, for the version with limited replicase (repRNA-v4), the proportion of EGFP-positive cells can be maintained as high as about 50% on the 3rd day, and a behavior pattern dependent on the replicase level is shown (Figure 13). Therefore, the replication mode of the limited replicase may achieve lower cell pathology to better maintain continuous RNA replication. Mechanistically, host transcriptome analysis showed that, compared with replicase-restricted replication, replication without replicase restriction (i.e., repRNA-v3) significantly upregulated genes involved in the interferon signaling pathway (Figure 14), which is a reflection of known virus-induced cellular ^{pathogenicity44} .

As the replicase level decreases, the short-term stability of RNA shows an increasing trend, because the EGFP-positive ratio shows a more stable maintenance state from the time when the EGFP level reaches a steady state (36 hours) to the end (72 hours) (Figure 13, left side of Figure 15). In order to test whether the reduction in replicase levels is causally related to the improvement in short-term stability, the inventors created a new host cell line (referred to as the "low replicase host") by inserting a stop codon in front of nsP4 (Figure 16). Compared with the previous host cell line (referred to as the "high replicase host"), the replicase level was further reduced. In this new host cell, the level of EGFP expressed from RNA (repRNA-v4) was lower than that of the previous host cell as expected (compare Figure 17 and Figure 13). Most importantly, in the new host cell, the short-term stability of the autonomously replicating RNA is indeed enhanced (Figure 15). These data collectively indicate that the reduction in replicase levels leads to a reduction in the RNA replication rate (i.e., a reduction in the EGFP level), which in turn leads to an enhanced short-term stability of the autonomously replicating RNA.

Based on these data, the inventors have constructed a simple RNA replication kinetic model, thereby proposing the hypothesis that the steady-state RNA level is proportional to the replicase level. By quantifying the steady-state RNA (EGFP) level under different replicase levels of two host cell lines, the inventors are pleased to find that for low replicase levels, the steady-state RNA level is linearly proportional to the replicase level (Figure 18). However, for high replicase levels (with high replicase hosts), it seems that the steady-state RNA level reaches a stable period after a linear stage (Figure 18), which may reflect that the cytopathological effects induced by RNA determine the carrying degree of host cells to RNA. Consistent with this, when the culture time is long, only cells with low replicase levels in high replicase hosts survive (Figure 19), which is significantly different from low replicase hosts (Figure 20).

In summary, these quantitative data help describe the process of RNA replication maintenance in the short term and illustrate the dynamics and function of the limited replicase mode of RNA replication. Kinetically, the limited replicase mode operates in a saturated kinetic state, and its steady-state level is regulated, which helps to improve host compatibility. Functionally, by enhancing host compatibility, this mode achieves the production of ~10 ⁸ cells (starting from 4×10 ⁶ cells) in one week (Figure 9). Importantly, RNA (repRNA-v4) can be stably propagated for more than a month (Figure 21, Figure 22), from which RNA is replicated uniformly in single cells, which is confirmed by immunostaining of replication intermediates-double-stranded RNA (dsRNA) (Figure 23, Figure 24, see Methods for details). By further quantifying the levels of viral genomes using RNA sequencing (see Methods for details), the inventors estimated that each cell could carry approximately 100 RNA molecules (Figure 25, the inventors used bulk poly-A RNA-seq to quantify the expression levels of autonomously replicating RNAs and endogenous genes. The inventors found that the expression level of repRNA-v4 (in high replicase hosts) was approximately 60% of the Gapdh gene, as measured by reads mapped to reads of nsP1-3. Since GapdhmRNA in mammalian cells is typically ¹⁰⁰⁰ molecules per cell3, the inventors inferred that the engineered autonomously replicating RNA, namely repRNA-v4, has at least 100 molecules per cell). Thus, up to more than 10 billion RNA molecules ( ¹⁰⁸ cells × 100 copies/cell) can then be diversified and selected within living mammalian cells.

Example 3 - Controllable RNA diversification by interfering with RNA replication accuracy

The inventors then explored the mutagenic capacity of this modified system, which can determine the evolution and adaptation efficiency of self-replicating RNA. The inventors first quantified the basic mutagenic rate of RNA and then tried to regulate this rate by intervening in the precision of the replicase.

In order to quantify the fidelity of replicase, the inventors used second generation sequencing to measure the mutation rate of RNA after maintaining 1, 3, 7, 14 and 21 days in cells (a and b of Figure 26). In order to establish a baseline, the inventors sequenced plasmid DNA and in vitro transcribed RNA, and found that sequencing and in vitro transcription can introduce a mutation rate of about 0.4 bases/kb in total together. Once entering the cell, the mutation on RNA accumulates at a speed of about 0.01 bases/kb per day, reaching about 0.22 bases/kb after 21 days (b of Figure 26, Figure 27), which contains four main mutation types (i.e. A to U, A to G, G to A and C to U) (Figure 28). Therefore, in the long-term RNA replication process, replicase does not introduce many errors, which is consistent with the situation that RNA is slowly replicated. The key is that such a low basic mutation rate provides an opportunity for a controllable mutagenesis rate.

To alter the fidelity of the replicase, the inventors measured the mutation rate of RNA in host cells carrying a mutant nsP4 known to reduce accuracy ⁴⁵ . The inventors observed only a slight but statistically significant increase in the mutation rate of the four mutation types, however this only very limitedly improved the diversification capacity of the RNA ( FIG. 29 ). The inventors next studied small molecule-mediated intervention of the RNA replicase, as some therapeutic nucleoside analogs fight RNA viruses by reducing the fidelity of the RNA replicase (e.g., molnupiravir used to treat COVID-19) ^46–50 . The inventors tested four different nucleoside analogs and compared their ability to eliminate the EGFP signal expressed by RNA, among which analogs with stronger ability to induce mutations resulted in faster accumulation of non-fluorescent EGFP mutants ( FIG. 30 , FIG. 31 ). The inventors found that among these four drugs, favipiravir and molnupiravir did not affect cell growth (Figure 70), and when added to the culture medium of the host cell containing the autonomously replicating RNA in a dose-dependent manner within 3 days, the EGFP signal (Figure 31, Figure 33) can be effectively reduced in a dose-dependent manner (Figure 32). By quantifying the RNA mutation rate during the entire time course of drug treatment, the inventors found that these two drugs can improve the mutagenic ability of RNA replicase by two orders of magnitude, achieving a mutation rate of up to 1 base/kb per day (Figure 34), which should be able to achieve rapid diversification based on RNA synthesis devices. Importantly, the enhanced mutagenic rate of the orthogonal RNA replication system is likely not to affect the genes expressed by the host genome, as can be seen from the stability of the iRFP signal (expressed together with nsP4) and the EGFP signal (Figure 35) in the control cell line.

The inventor further compared the mutation spectrum of RNA with nucleoside analogs or carrier control.The inventor found that molnupiravir, but not favipiravir, induces a relatively balanced mutation spectrum (Figure 36) by improving the mutation frequency of all four major conversion types on average.Therefore, the inventor chooses to use molnupiravir, which can induce a relatively extensive and balanced mutation spectrum, and allows the inventor to realize RNA diversification in a controlled manner.

Example 4 - Blue shift of EGFP signal based on REPLACE system

By linking the products of autonomously replicating RNA to phenotypes, in principle, these RNAs can mutate to produce diverse phenotypes that can be directed selected. Therefore, the inventors explored the effectiveness of orthogonal RNA replication systems through REPLACE. Through REPLACE, genes containing target protein gene sequences can be expressed through subgenomic promoters on RNAs, and these RNAs can be replicated, mutated, and directed selected.

The inventors attempted to test REPLACE by the color change of the EGFP signal, because EGFP itself is a material for verifying the directed evolution tool, because only one amino acid replacement is needed for GFP to become BFP ^51,52,19 . Specifically, the inventors placed the EGFP gene under the regulation of the subgenomic promoter of the autonomously replicating RNA (Figure 37), and then cultivated a cell group containing this RNA (see method for details). Subsequently, molnupiravir was added to the culture medium to accelerate mutagenesis, and this process lasted for two days (Figure 38). Thereafter, the flow analysis of GFP and BFP signals showed that a small part of cells deviated from the diagonal, which indicated that blue shift occurred in fluorescence (middle part of Figure 38). By cell sorting, the inventors directed and selected two cell groups showing different blue shift signals, and subjected them to a second round of mutagenesis and screening. After sequencing, the inventors observed that the vertical population had a highly enriched missense mutation (r.196U>C, i.e., the 196th nucleotide of the mRNA mutated from U to C), while another enriched missense mutation (r.608C>U) was present in the non-diagonal population (Figure 39, Figure 40). These two enriched mutations were associated with the conversion of signals from GFP to BFP or to Sapphire ^53,54 , respectively ⁵¹ (Figure 41). It is worth noting that although the two mutants that appeared blue-shifted signals only needed to mutate one residue each, the two residues that led to the two mutants were far apart, and the two mutants were easily detected after the first sorting, which proved the feasibility of the REPLACE system in evolving synthetic devices that require multiple mutations.

Example 5 - Using the REPLACE system to modify transcriptional regulatory proteins

The inventors next sought to use REPLACE to evolve synthetic transcriptional regulators, including ligand-controlled transcription factors (TFs) and Cas (CRISPR-associated) protein-based transcriptional regulators, which have been widely used in mammalian synthetic circuits7 ^.

The inventors first constructed synthetic TFs that could reduce or enhance ligand sensitivity. As a test of the REPLACE system, the inventors' ideal approach was to choose to reproduce the research results that other directed evolution methods have ^{achieved16,18} : evolving TetR to reduce its sensitivity to doxycycline. In short, the autonomously replicating RNA carrying TetR-VP48 (i.e., tTA) was continuously diversified for 7 days in the environment of relatively low doses of molnupiravir (2μM) (Figure 42a and b, see methods for details). At the same time, over a period of 2 weeks, the concentrations of doxycycline and puromycin in the culture medium were gradually increased, during which time TetR mutants that were able to bind to DNA (by resisting the increase in doxycycline concentration) and drive the expression of the puromycin resistance gene would be enriched. The inventors then sorted out cells expressing high levels of mCherry signals and extracted their RNA for reverse transcription and sequencing. Reassuringly, the inventors found three enriched missense mutations (>5%), of which the most enriched mutation, T103P, is directly located in the ligand binding pocket, and the other two mutations (R80C and E157G) are located near the pocket ⁵⁵ (Figure 43). Using a reporter system of HEK293T cells, the inventors found that TetR variants (determined from Sanger sequencing) showed increasing resistance to doxycycline (i.e., decreased sensitivity) as the number of mutated residues increased from one to three (Figure 44). In summary, these results indicate that REPLACE is able to construct new synthetic TFs through autonomously replicating RNA diversification and directed selection in mammalian cells.

The inventors next attempted to modify a newly developed ligand-controlled mammalian synthetic TF, VP64-PadR ⁵⁶ , whose DNA binding ability can be shut down by a drug approved by the US FDA (sodium ferulate). The purpose of the modification was to reduce its ligand sensitivity or to increase the dynamic range of its response. In order to reduce the sensitivity of the ligand, the inventors directed evolution of RNA containing VP64-PadR mutants by RNA diversification followed by selection with increasing concentrations of puromycin in the presence of the ligand (sodium ferulate) ( FIG. 45 , see methods for details). By reverse transcription and sequencing of RNA extracted from a population of surviving cells, the inventors identified eight enriched missense mutations ( FIG. 46 ), four of which (E28K, E92G, G153S, K166E) were located near the ligand binding pocket ⁵⁷ . Then, the inventors quantified the ligand sensitivity of the mutants obtained by first-generation sequencing through a reporter system, and found that mutants with single (E92G or G153S) or double (E28K/G153S) mutations near the binding pocket showed reduced sensitivity characteristics (Figure 47). Unexpectedly, mutations outside the binding pocket, by themselves or in concert with mutations near the binding pocket, can reduce (Figure 47) or even reverse (Figure 48) the sensitivity to sodium ferulate (T55I/Q78P, L72P/K166E, E28K/T55I/Q78P), which means that ligand binding has potential allosteric regulation and cooperative interaction between the two types of residues.

Next, the inventors attempted to expand the dynamic range of the response of VP64-PadR by constructing mutants with enhanced activation in the absence of a ligand. The inventors first performed RNA diversification and then selected RNA with enhanced target gene activation without the addition of sodium ferulate (Figure 49, see Methods for details). Finally, the inventors confirmed 6 enriched missense mutations, three of which (Y25C, E92K, D130G) were located near the binding pocket (Figure 50). It is worth noting that PadR variants containing one or more mutations in residues near the binding pocket (Y25C, Y25C/S174N, D130G/S174N, Y25C/E92K, D74V/R86K/D130G/S174N) can also show enhanced target activation in the absence of a ligand. However, compared with the wild type, the dynamic range of the reaction (quantified by the fold change of the target response with or without a ligand) is reduced (Figure 51). In contrast, for mutants carrying residues far from the binding pocket (D74V, D21G/D74V/M178T), the dynamic range of the response was at least comparable to that of the wild type. Importantly, one of the variants (D21G/D74V/M178T) exhibited a dramatically enhanced dynamic range (85-fold) compared to the wild type (38-fold) (Figure 51), with only a slight change in ligand affinity (Figure 52).

After establishing that REPLACE has the ability to transform ligand-controlled mammalian synthetic TF, the inventors further sought to evolve Cas-based transcriptional regulatory proteins. Miniaturized Cas12f protein, CasMINI ²⁰ , is about 500 residues in length and can be easily delivered into cells. It was previously transformed according to the principle of rational design ^20. The inventors hope to further improve the transcriptional regulatory ability of this protein through REPLACE. The inventors have diversified and screened the autonomously replicating RNA of CasMINI-v4 ²⁰ (dCasMINI-VPR) carrying endonuclease inactivation (Figures 67 and 68). Surprisingly, in two independent experiments, the inventors only confirmed one enriched mutation (S246R) (Figure 69). By quantifying the transcriptional activation ability through a reporter system, the inventors found that compared with the parent, the S246R variant had an activation ability of about 30% for a promoter containing TetO (used in the evolution experiment) (Figure 53), and its ability to activate an endogenous gene (CD2) using an sgRNA was increased by about 20% (Figure 54). The inventors further found that the evolved mutants did not lead to a significant increase in the related base editing ability and nuclease activity (Figure 55), which may be because the transcriptional regulatory activity may not be related to the activity of base editing or DNA cutting ²⁰ .

Together, these results demonstrate the power and versatility of REPLACE for constructing ligand-controlled or Cas-based synthetic transcriptional regulatory proteins in mammals, suggesting that this system has the potential to improve the functionality of other TFs in mammalian synthetic biology applications.

Example 6 - Self-replicating RNA carrying MEK1 achieves Darwinian adaptation

Through the orthogonality of the RNA replication system, the REPLACE system achieves a continuous mode of operation - RNA can replicate autonomously across generations, opening up a way to build cells in a Darwinian adaptive manner (i.e., through adaptive selection). Therefore, the inventors next tested whether mammalian cells can use REPLACE to evolve genes with important physiological significance, thereby autonomously adapting to new environments.

The inventors first explored whether cells carrying self-replicating RNA of the human MEK1 gene could adapt to an environment containing MEK1 allosteric inhibitors. Since it has been confirmed that in vivo or in vitro mutagenesis of MEK1 can produce inhibitor-resistant mutants of ^MEK158,19 , in principle, cells can use the evolvability of self-replicating RNA to survive in the presence of inhibitors. So the inventors used repRNA-v4 containing the wild-type MEK1 gene expressed by a subgenomic promoter to electrotransfect cells (day 0) (Figure 56, see methods for details). From the 5th day, molnupiravir was added to the culture to diversify the RNA. Then, starting from the 7th day, three independent MEK1 allosteric inhibitors were used to apply survival pressure to the cells, and the cells were collected for sequencing on the 14th day. Over a period of 7 days, RNA diversification and selection were carried out simultaneously in the inhibitor environment, achieving Darwinian adaptation (Figure 56).

Interestingly, the inventors found that surviving cells contained RNA highly enriched for MEK1 mutations (i.e., >5% in one or more biological replicates), and several mutations were enriched for all three inhibitors (Figures 57, 58), suggesting that Darwinian adaptation to allosteric inhibitors of MEK1 can occur through conserved mechanisms. Structurally, the mutations were located in different functional ^{domains59,58,60} (Figure 59): the inhibitor binding pocket (L115P, V211D, L215P), the negative regulatory domain (R47Q, Q56P, K57N), residues that may interfere with inhibitor function (H119Y), and the nuclear export signal (E39A) ⁶¹ . Many of these were previously discovered in mutagenesis screens for ^MEK158 . Sanger sequencing allowed the inventors to identify each MEK1 mutant, which carried an average of 1 to 3 mutations.

The inventors then used a fluorescent reporter system to measure MEK1 activity (i.e., an SRE reporter system that reflects MAPK/ERK signaling activity ^19,62 ) to confirm the resistance of different mutants to cobimetinib. The inventors found that in the presence of an inhibitor, all mutants except the H119Y mutation showed stronger reporter activity than the wild type (i.e., enhanced inhibitor resistance) ( FIG. 60 ). Puzzlingly, the basal signaling activity exhibited when two mutants were contained in the inhibitor binding pocket (L115P/V211D) was greatly reduced compared to single mutants (L115P and V211D), indicating that there is an interaction between the residues in the allosteric pocket ( FIG. 60 ). Interestingly, the combined mutations of the inhibitor binding pocket and the negative regulatory domain (Q56P/V211D and R47Q/K57N/V211D) dramatically enhanced the activity of MEK1 in the presence or absence of an inhibitor compared to mutants with only the binding pocket mutation ( FIG. 60 ). For one of the mutants (Q56P/V211D), the inhibitor even increased its activity (rather than inhibiting it) (Figure 60). These results highlight the ability of REPLACE to traverse the complex sequence-function landscape of MEK1 during cellular adaptation to MEK1 inhibition.

Example 7 - Escape from the dominant negative KRAS gene using the REPLACE system

Having established that REPLACE enables cells to adapt to external challenges (i.e., small molecule inhibitors), the inventors next set out to investigate whether and how cells equipped with autonomously replicating RNA could adapt to cell-intrinsic challenges. The inventors chose a proto-oncogene, KRAS, which is one of the most frequently mutated genes in ^cancer63 . In addition to small molecule-based cancer KRAS-targeted therapies, researchers have proposed and tested dominant-negative gene ^{therapies64,65} . Oncogenic KRAS activity (and cell proliferation) is inhibited by introducing a dominant-negative KRAS allele (S17N) ⁶⁶ . In principle, cells could become resistant to this treatment by reversing a single mutation (S17N), but there are other potential pathways to confer similar resistance.

Therefore, the inventors attempted to study whether and how cells carrying autonomously replicating RNA of the KRAS (S17N) allele overcome their (cell-intrinsic) inhibition of KRAS activity through Darwinian adaptation (Figure 61). As a control, the inventors also evolved cells carrying autonomously replicating RNA of wild-type KRAS (Figure 61). RNA was first continuously diversified for 9 days in a cell culture medium containing molnupiravir (Figure 61, see method for details), during which cells with increased KRAS activity and enhanced proliferation should have a greater survival advantage than cells with suppressed KRAS activity. By sequencing the RNA of the surviving population, the inventors found six enriched missense mutations on KRAS (S17N) that were not found on wild-type KRAS (Figure 62, Figure 63). Interestingly, the inventors found an enriched mutation that changed asparagine (asparagine) N17 to a hydroxy amino acid, threonine (N17T), rather than serine (serine, N17S) (Figure 62) that became in the wild type. Since either serine or threonine is essential for binding to guanine nucleotides—they can coordinate Mg2 ⁺ ions through their hydroxyl ^groups67 —the N17T mutation may be sufficient to reverse the effects of the dominant-negative phenotype of KRAS(S17N) ⁶⁸ . Several other mutations are located in important functional ^regions69 (Figure 64), including the switch I region (Y32C) responsible for effector ^engagement70 and a hydrophobic hub (L79F) mediating key allosteric ^{interactions71} . However, it remains difficult to determine which mutation is functionally responsible for escaping the dominant-negative KRAS.

Therefore, the inventors used the SRE reporter system to quantify the activity of each KRAS mutant (up to four mutations) and characterize their effects on cell proliferation. As a control, the inventors found that the signaling activity of wild-type KRAS was significantly enhanced compared to the dominant negative mutant (S17N), and the well-known oncogenic KRAS mutation, G12C, also resulted in higher activity (top of Figure 65). Consistent with the activity readings, both wild-type KRAS and KRAS (G12C) induced faster cell proliferation compared to KRAS (S17N), while the two variants themselves were indistinguishable in terms of cell proliferation (bottom of Figure 65). The inventors then tested confirmed mutants evolved from KRAS (S17N) and found that the N17T mutant reversed the dominant negative effect because it resulted in significantly higher activity levels than the wild type, as well as cell proliferation rates comparable to the wild type (Figure 65). Another mutation, L79F, further enhanced signaling activity, but did not enhance cell proliferation (Figure 65). These results suggest that T17 can better coordinate with Mg ²⁺ ions than S17 of wild-type KRAS, and the higher signaling activity may not lead to enhanced cell proliferation. For other mutants without N17T mutation, although they all enhanced KRAS signaling activity statistically, none of them caused a significant proliferation advantage (Figure 66), which means that these mutants may confer other types of survival advantages to cells during the period of escape from dominant negative inhibition.

Together, these results suggest that cells are able to escape the effects of dominant-negative KRAS through REPLACE-mediated Darwinian adaptation.

Example 8 - Discussion

Synthetic biology is often compared to disciplines such as electrical ^{engineering72} because both use engineering principles to construct pathways and systems that operate as designed. However, biological systems are inherently complex and naturally evolvable, which requires the theory and tools of synthetic biology to go beyond the scope of traditional disciplines to achieve the evolvability of constructed biological systems25 ^. To achieve this goal, the inventors developed REPLACE: RNA orthogonally replicated, continuously evolved, and stably propagated in mammalian cells. REPLACE not only enables the engineering of mammalian-compatible biomacromolecules directly in mammalian systems, but also provides synthetic evolvability to mammalian cells, allowing engineered cells to adapt to new environments using the evolvability of self-replicating RNA.

Compared to existing directed evolution methods, a key conceptual and technical advance of REPLACE lies in adapting and engineering the RNA virus replication system to in vivo directed evolution in mammalian cells. While viruses have previously assisted in the directed evolution of ^{biomacromolecules73,74,16,18} , REPLACE is fundamentally different because only a portion of the viral genome is fully propagated intracellularly and no live virus is produced. In contrast to RNA viruses (such as SARS-CoV-2) that accumulate widespread ^{mutations75,76} during replication, the Sindbis virus-derived RNA in the REPLACE system replicates in a replicase-limited mode and with saturating kinetics to ensure low levels of replication and base error accumulation, laying the foundation for chemically controlling the precision of the RNA replicase (and the resulting RNA mutagenesis). In contrast to natural RNA viruses, mutations in the autonomously replicating RNAs designed by the inventors arise primarily from perturbations to the RNA replicase rather than from steady-state (undisturbed) replication. This finding also suggests a potential adverse effect of widespread use of mutagenic nucleoside analogs to treat COVID- ^1949,77 – a double-edged sword that could significantly accelerate the emergence of new viral variants by inducing mutations and promoting rapid viral evolution.

The versatility of the REPLACE system enabled the inventors to use directed selection to engineer functional biomolecules, including synthetic transcriptional regulators, and to enable Darwinian adaptation of cells in response to challenges. These successful applications highlight three notable features of REPLACE. First, our RNA-based evolvable system is largely orthogonal to host cells and uniquely enables in vivo evolution of RNA machinery within host cells (as distinct from in vitro ^{evolution78,79} ). Each cell can accommodate approximately 100 copies, forming a large mutant library to ensure a wide variety of evolutionary experiments. Second, our system provides flexibility for both diversification and selection steps—diversification can not only confer a relatively broad mutation spectrum, but is also fully controllable; selection can be directed or fitness-based; and the two steps can even be performed simultaneously to achieve continuous evolution. Third, self-replicating RNA-based machinery provides a simple and powerful means to enable mammalian cells to accomplish adaptation, a task that is difficult, complex, and adaptive for traditional synthetic biology tools.

The inventors consider that REPLACE has a wide range of applications, including mammalian biomacromolecule engineering, synthetic biology, and developmental biology (e.g., for tracking cell lines). To further expand its scope of application, it can be improved in several aspects, including screening methods (e.g., incorporating negative selection), target modes (e.g., evolving functional RNA), and cell compatibility (e.g., further reducing cell pathology and expanding to other cell types).

Broadly speaking, the adaptive RNAs developed by the inventors are evolvable, which will promote new paradigms for the engineering and application of synthetic RNA devices in mammals. The current boom in RNA-based ^{therapeutics32,33,80} may promote the development of future evolvable and adaptable RNA-based therapeutic devices that can be delivered directly into cells in living animal tissues to address unforeseen challenges.

References

1. Hendel, S. J. & Shoulders, M. D. Directed evolution of inmammalian cells. Nat. Methods 18, 346–357 (2021).

2. Majors, B.S., Chiang, G.G. & Betenbaugh, M.J. Protein and Genome Evolution in Mammalian Cells for Biotechnology Applications. Mol. Biotechnol. 42, 216–223 (2009).

3. Morrison, M.S., Podracky, C.J. & Liu, D.R. The developing toolkit of continuous directed evolution. Nat. Chem. Biol. 16, 610–619 (2020).

4. Wang, Y. et al. Directed Evolution: Methodologies and Applications. Chem. Rev. 121, 12384–12444 (2021).

5. Molina, R.S. et al. In vivo hypermutation and continuousevolution. Nat. Rev. Methods Primer 2, 1–22 (2022).

6. Purnick, P.E.M. & Weiss, R. The second wave of synthetic biology: from modules to systems. Nat. Rev. Mol. Cell Biol. 10, 410–422 (2009).

7. Lienert, F., Lohmueller, J.J., Garg, A. & Silver, P.A. Syntheticbiology inmammalian cells: next generation research tools and therapeutics. Nat. Rev. Mol. Cell Biol. 15, 95–107 (2014).

8. Church, G.M., Elowitz, M.B., Smolke, C.D., Voigt, C.A. & Weiss, R. Realizing the potential of synthetic biology. Nat. Rev. Mol. Cell Biol. 15, 289–294 (2014).

9. Tan,

10. Gallup, O., Ming, H. & Ellis, T. Ten future challenges for synthetic biology. Eng. Biol. 5, 51–59 (2021).

11. Azam, M., Latek, R.R. & Daley, G.Q. Mechanisms of Autoinhibition and STI-571/Imatinib Resistance Revealed by Mutagenesis of BCR-ABL. Cell 112, 831–843 (2003).

12. Wang, L., Jackson, W.C., Steinbach, P.A. & Tsien, R.Y. Evolution of new nonantibody proteins via iterative somatic hypermutation. Proc. Natl. Acad. Sci. 101, 16745–16749 (2004).

13. Peck, S. H., Chen, I. & Liu, D. R. Directed Evolution of a Small-Molecule-Triggered Intein with Improved Splicing Properties in Mammalian Cells. Chem. Biol. 18, 619–630 (2011).

14.Hess, G.T. et al. Directed evolution using dCas9-targetedsomatichypermutation in mammalian cells. Nat. Methods 13, 1036–1042 (2016).

15.Ma,Y.et al.Targeted AID-mediated mutagenesis(TAM)enables efficientgenomic diversification in mammalian ce 846 lls.Nat.Methods 13,1029–1035(2016).

16. Berman, C.M. et al. An Adaptable Platform for Directed Evolution in Human Cells. J. Am. Chem. Soc. 140, 18093–18103 (2018).

17. Piatkevich, K.D. et al. A robotic multidimensional directedevolution approach applied to fluorescent voltage reporters. Nat. Chem. Biol. 14, 352–360 (2018).

18.English,J.G.et al.VEGAS as a Platform for Facile DirectedEvolutionin Mammalian Cells.Cell 178,748-761.e17(2019).

19. Chen, H. et al. Efficient, continuous mutagenesis in human cells using a pseudo-random DNA editor. Nat. Biotechnol. 38, 165–168 (2020).

20. Xu, X. et al. Engineered miniature CRISPR-Cas system formammaliangenome regulation and editing. Mol. Cell 81, 4333-4345.e4 (2021).

21. Weber, W. & Fussenegger, M. Emerging biomedical applications of synthetic biology. Nat. Rev. Genet. 13, 21–35 (2012).

22. Lim, W. A. & June, C. H. The Principles of Engineering Immune Cells to Treat Cancer. Cell 168, 724–740 (2017).

23. Sedlmayer, F., Aubel, D. & Fussenegger, M. Synthetic gene circuits for the detection, elimination and prevention of disease. Nat. Biomed. Eng. 2, 399–415 (2018).

24. Ruder, W.C., Lu, T. & Collins, J.J. Synthetic Biology Moving into the Clinic. Science 333, 1248–1252 (2011).

25. Castle, S. D., Grierson, C. S. & Gorochowski, T. E. Towards an engineering theory of evolution. Nat. Commun. 12, 3326 (2021).

26. Kheir Gouda, M., Manhart, M. & Balázsi, G. Evolutionary regain of lostgene circuit function. Proc. Natl. Acad. Sci. 116, 25162–25171 (2019).

27. Lezia, A., Csicsery, N. & Hasty, J. Design, mutate, screen: Multiplexed creation and arrayed screening of synchronized genetic clocks. Cell Syst. 13, 365-375.e5 (2022).

28. Cobb, R.E., Sun, N. & Zhao, H. Directed Evolution as a Powerful Synthetic Biology Tool. Methods San Diego Calif 60, 81–90 (2013).

29. Haseltine, E.L. & Arnold, F.H. Synthetic gene circuits: design with directed evolution. Annu. Rev. Biophys. Biomol. Struct. 36, 1–19 (2007).

30. Yokobayashi, Y., Weiss, R. & Arnold, F.H. Directed evolution of agenetic circuit. Proc. Natl. Acad. Sci. 99, 16587–16591 (2002).

31. Ravikumar, A., Arzumanyan, G.A., Obadi, M.K.A., Javanpour, A.A. & Liu, C.C. Scalable, Continuous Evolution of Genes at Mutation Ratesabove GenomicError Thresholds. Cell 175, 1946-1957.e13 (2018).

32. Sahin, U., Karikó, K. & Türeci, mRNA-based therapeutics—developing a new class of drugs. Nat. Rev. Drug Discov. 13, 759–780 (2014).

33. Paunovska, K., Loughrey, D. & Dahlman, J.E. Drug delivery systems for RNAtherapeutics. Nat. Rev. Genet. 23, 265–280 (2022).

34. Khan, K. H. Gene Expression in Mammalian Cells and its Applications. Adv. Pharm. Bull. 3, 257–263 (2013).

35. Strauss, J.H. & Strauss, E.G. The alphaviruses: gene expression, replication, and evolution. Microbiol. Rev. 58, 491–562 (1994).

36. Frolov, I. et al. Alphavirus-based expression vectors: strategies and applications. Proc. Natl. Acad. Sci. 93, 11371–11377 (1996).

37. Xiong, C. et al. Sindbis Virus: An Efficient, Broad Host 890 RangeVector for Gene Expression in Animal Cells. Science 243, 1188–1191 (1989).

38. Wolff, J.A. et al. Direct Gene Transfer into Mouse Muscle inVivo. Science 247, 1465–1468 (1990).

39. Zhou,

40. Wroblewska, L. et al. Mammalian synthetic circuits with RNAbindingproteins for RNA-only delivery. Nat. Biotechnol. 33, 839–841 (2015).

41. Jose, J., Taylor, A. B. & Kuhn, R. J. Spatial and Temporal Analysis of Alphavirus Replication and Assembly in Mammalian and Mosquito Cells. mBio 8, e02294-16 (2017).

42.Shirako,Y.&Strauss,J.H.Cleavage between nsP1 and nsP2initiates theprocessing pathway of Sindbis virus nonstructural polyproteinP123.Virology177,54–64(1990).

43.Frolov,I.et al.Selection of RNA Replicons Capable ofPersistentNoncytopathic Replication in Mammalian Cells.J.Virol.(1999)doi:10.1128/JVI.73.5.3854-3865.1999.

44. Guo, J.-T., Zhu, Q. & Seeger, C. Cytopathic and Noncytopathic InterferonResponses in Cells Expressing Hepatitis C Virus Subgenomic Replicons. J. Virol. 77, 10769–10779 (2003).

45. Rozen-Gagnon, K. et al. Alphavirus Mutator Variants PresentHost-Specific Defects and Attenuation in Mammalian and Insect Models. PLOS Pathog. 10, e1003877 (2014).

46. Urakova, N. et al. β-d-N4-Hydroxycytidine Is a PotentAnti-alphavirus Compound That Induces a High Level of Mutations in the ViralGenome. J. Virol. 92, e01965-17 (2018).

47. Delang, L., Abdelnabi, R. & Neyts, J. Favipiravir as a potential countermeasure against neglected and emerging RNA viruses. Antiviral Res. 153, 85–94 (2018).

48. Sheahan, T.P. et al. An orally bioavailable broad-spectrum antiviralinhibits SARS-CoV-2 in human airway epithelial cell cultures and multiple coronaviruses in mice. Sci. Transl. Med. 12, eabb5883 (2020).

49. Tian, L. et al. RNA-dependent RNA polymerase (RdRp) inhibitors: The current landscape and repurposing for the COVID-19 pandemic. Eur. J. Med. Chem. 213, 113201 (2021).

50. Kabinger, F. et al. Mechanism of molnupiravir-induced SARS-CoV-2 mutagenesis. Nat. Struct. Mol. Biol. 28, 740–746 (2021).

51. Heim, R., Prasher, D.C. & Tsien, R.Y. Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. 91, 12501–12504 (1994).

52. Sawano, A. & Miyawaki, A. Directed evolution of green fluorescent protein by a new versatile PCR strategy for site-directed and semi-randommutagenesis. Nucleic Acids Res. 28, e78 (2000).

53. Zapata-Hommer, O. & Griesbeck, O. Efficiently folding and circularlypermuted variants of the Sapphire mutant of GFP. BMCBiotechnol. 3, 5 (2003).

54. Stoltzfus, C.R. et al. Two-photon directed evolution of green fluorescent proteins. Sci. Rep. 5, 11968 (2015).

55. Ramos, J.L. et al. The TetR Family of Transcriptional Repressors. Microbiol. Mol. Biol. Rev. 69, 326–356 (2005).

56. Wang, Y. et al. A versatile genetic control system in mammalian cells and mice responsive to clinically licensed sodium ferulate. Sci. Adv. 6, eabb9484 (2020).

57. Park, S.C., Kwak, Y.M., Song, W.S., Hong, M. & Yoon, S. Structural basis of effector and operator recognition by the phenolicacid-responsivetranscriptional regulator Pad R. Nucleic Acids Res. 45, 13080–13093 (2017).

58. Emery, C.M. et al. MEK1 mutations confer resistance to MEKand B-RAFinhibition. Proc. Natl. Acad. Sci. 106, 20411–20416 (2009).

59. Roskoski, R. Classification of small molecule proteinkinaseinhibitors based upon the structures of their drug-enzymecomplexes. Pharmacol. Res. 103, 26–48 (2016).

60. Caunt, C. J., Sale, M. J., Smith, P. D. & Cook, S. J. MEK1 and MEK2 inhibitors and cancer therapy: the long and winding road. Nat. Rev. Cancer 15, 577–592 (2015).

61. Delaney, A.M., Printen, J.A., Chen, H., Fauman, E.B. & Dudley, D.T. Identification of a Novel Mitogen-Activated Protein KinaseKinaseActivation Domain Recognized by the Inhibitor PD184352. Mol. Cell. Biol. 22, 7593–7602 (2002 ).

62. Treisman, R. The serum response element. Trends Biochem. Sci. 17, 423–426 (1992).

63. Waters, A.M. & Der, C.J. KRAS: The Critical Driver and Therapeutic Target for Pancreatic Cancer. Cold Spring Harb. Perspect. Med. 8, a031435 (2018).

64. Adjei, A.A. Blocking Oncogenic Ras Signaling for Cancer Therapy. JNCIJ. Natl. Cancer Inst. 93, 1062–1074 (2001).

65. Weng, T.-Y. et al. DNA vaccine elicits an efficient antitumor response by targeting the mutant Kras in a transgenic mouse lung cancer model. GeneTher. 21, 888–896 (2014).

66.Stewart,S.&Guan,K.-L.The Dominant Negative Ras Mutant,N17Ras,CanInhibit Signaling Independently of Blocking Ras Activation*.J.Biol.Chem.275,8854–8862(2000).

67.Zhang,B.,Zhang,Y.,Wang,Z.&Zheng,Y.The Role of Mg2+Cofactor in theGuanine Nucleotide Exchange and GTP HydrolysisReactions of Rho Family GTP-binding Proteins*.J.Biol.Chem.275, 25299–25307(2000).

68. Feature -1_19.

69. Kwan, A.K., Piazza, G.A., Keeton, A.B. & Leite, C.A. The path to the clinic: a comprehensive review on direct KRASG12C inhibitors. J. Exp. Clin. Cancer Res. 41, 27 (2022).

70.Spoerner, M.et al.Conformational States of Human Rat Sarcoma(Ras)Protein Complexed with Its Natural Ligand GTP and Their Role for EffectorInteraction and GTP Hydrolysis*.J.Biol.Chem.285,39768–39778(2010).

71.Pantsar,T.et al.Assessment of mutation probabilities of KRASG12missense mutants and their long-timescale dynamics by atomisticmolecularsimulations and Markov state modeling.PLOS Comput.Biol.14,e1006458(2018).

72. Cheng, A. A. & Lu, T. K. Synthetic Biology: An EmergingEngineering Discipline. Annu. Rev. Biomed. Eng. 14, 155–178 (2012).

73. Esvelt, K. M., Carlson, J. C. & Liu, D. R. A System for the Continuous Directed Evolution of Biomolecules. Nature 472, 499–503 (2011).

74. Miller, S. M., Wang, T. & Liu, D. R. Phage-assisted continuous and non-continuous evolution. Nat. Protoc. 15, 4101–4127 (2020).

75. Lauring, A.S. & Andino, R. Quasispecies Theory and the Behavior of RNAViruses. PLoS Pathog. 6, e1001005 (2010).

76. Harvey, W.T. et al. SARS-CoV-2 variants, spike mutations and immunescape. Nat. Rev. Microbiol. 19, 409–424 (2021).

77. Robson, F. et al. Coronavirus RNA Proofreading:978 MolecularBasis andTherapeutic Targeting. Mol. Cell 79,710–727(2020).

78. Mills, D.R., Peterson, R.L. & Spiegelman, S. An extracellular Darwinian experiment with a self-duplicating nucleic acid molecule. Proc. Natl. Acad. Sci. 58, 217–224 (1967).

79. Ichihashi, N. et al. Darwinian evolution in a translation-coupledRNAreplication system within a cell-like compartment. Nat. Commun. 4, 2494 (2013).

80. Dykstra, P. B., Kaplan, M. & Smolke, C. D. Engineering synthetic RNA devices for cell control. Nat. Rev. Genet. 23, 215–228 (2022).

81. P., Lusa, S., Huylebroeck, D. & Garoff, H. In vitro mutagenesis of a full-length cDNA clone of Semliki Forest virus: the small6,000-molecular-weight membrane protein modulates virus release. J. Virol. 65, 4107–4113 (1991).

82.Huang,L.-K.&Wang,M.-J.J.Image thresholding by minimizing the measures of fuzziness.Pattern Recognition 28,41–51(1995).

Although the present application has illustrated and described exemplary embodiments for implementing the present invention, those skilled in the art will appreciate that the above embodiments should not be construed as limiting the present invention in any way, and that changes, substitutions and modifications may be made without departing from the spirit, principle and scope of the present invention, and that these modified embodiments are also within the scope of the present invention.

Claims (32)

1. An RNA replicase-assisted continuous in vivo evolution (REPLACE) system comprising autonomously replicating RNA and a host cell, wherein the autonomously replicating RNA is repRNA-v4, Wherein the autonomously replicating RNA lacks RNA replicase, and the host cell constitutively expresses the lacked RNA replicase. The host cell is a eukaryotic host cell. 2. The REPLACE system according to claim 1, wherein the host cell is an animal host cell. The REPLACE system according to claim 1 , wherein the host cell is a mammalian host cell. 4. The REPLACE system according to claim 1, wherein the host cell is selected from a hamster fibroblast cell line or is derived from a hamster fibroblast cell line. The REPLACE system according to claim 1 , wherein the host cell is selected from BHK-21 or is derived from BHK-21. The REPLACE system according to claim 1 , wherein the host cell is a replicase-limited eukaryotic host cell. The REPLACE system according to claim 1 , wherein the host cell is a replicase-limited animal host cell. The REPLACE system according to claim 1 , wherein the host cell is a replicase-limited mammalian host cell. 9. The REPLACE system of claim 1, wherein the host cell is selected from a replicase-limited hamster fibroblast cell line or a cell line derived from a replicase-limited hamster fibroblast cell line. 10. The REPLACE system according to claim 1, wherein the host cell is selected from the group consisting of replicase-limited BHK-21 or derived from replicase-limited BHK-21. The REPLACE system according to claim 1 , further comprising a mutagen that induces mutation. 12 . The REPLACE system according to claim 11 , wherein the mutagen is selected from any one of a small molecule mutagen and a nucleoside analog. 13. The REPLACE system according to claim 11, wherein the mutagen is selected from any one of favipiravir and molnupiravir. 14. An RNA replicase-assisted continuous in vivo evolution (REPLACE) method comprising: Provides autonomously replicating RNA; providing a host cell; Enables continuous evolution of autonomously replicating RNA within host cells; wherein the autonomously replicating RNA is repRNA-v4, wherein the autonomously replicating RNA lacks RNA replicase, and the host cell constitutively expresses the lacked RNA replicase, The host cell is a eukaryotic host cell. The REPLACE method according to claim 14 , wherein the host cell is an animal host cell. The REPLACE method according to claim 14 , wherein the host cell is a mammalian host cell. 17. The REPLACE method according to claim 14, wherein the host cell is selected from a hamster fibroblast cell line or is derived from a hamster fibroblast cell line. 18. The REPLACE method according to claim 14, wherein the host cell is selected from BHK-21 or is derived from BHK-21. The REPLACE method according to claim 14 , wherein the host cell is a replicase-limited eukaryotic host cell. 20. The REPLACE method according to claim 14, wherein the host cell is a replicase-limited animal host cell. 21. The REPLACE method according to claim 14, wherein the host cell is a replicase-limited mammalian host cell. 22. The REPLACE method of claim 14, wherein the host cell is selected from a replicase-limited hamster fibroblast cell line or a cell line derived from a replicase-limited hamster fibroblast cell line. 23. The REPLACE method according to claim 14, wherein the host cell is selected from the group consisting of replicase-limited BHK-21 or derived from replicase-limited BHK-21. 24. The REPLACE method of claim 14, further comprising providing a mutagenic agent that induces mutation. The REPLACE method according to claim 24 , wherein the mutagen is selected from any one of a small molecule mutagen and a nucleoside analog. 26. The REPLACE method according to claim 24, wherein the mutagen is selected from any one of favipiravir and molnupiravir. 27. Use of the REPLACE system of any one of claims 1 to 13 in the manufacture and optimization of biomacromolecules, vaccines, drugs, in the generation of mutant libraries, or in achieving Darwinian adaptation. 28. A vector for expressing the autonomously replicating RNA according to any one of claims 1 to 13. 29. A host cell comprising an expression vector for expressing the autonomously replicating RNA according to any one of claims 1 to 13. 30. A vaccine composition comprising RNA produced or producible by the REPLACE system of any one of claims 1 to 13. 31. A pharmaceutical composition comprising RNA produced or producible by the REPLACE system of any one of claims 1 to 13. 32. A delivery vector comprising RNA produced or produced by the REPLACE system of any one of claims 1 to 13.