In silico two-dimensional substrate-Rz complexes were predicted using the Bimolecular Structure Prediction by RNAstructure 6.4 (Mathews Lab, University of Rochester, NY, United States) (Bellaousov et al., 2013). Three-dimensional structures were modeled using the RNAfold Web Server (Gruber et al., 2008) and RNAComposer (Popenda et al., 2012) and visualized with the Visual Molecular Dynamics program (VMD) (Humphrey et al., 1996).

The RNA oligonucleotides, including the acpP mRNA fragment as a substrate (labeled mRNA _acpP ) and the ribozymes, long-armed (Rz_long) and short-armed (Rz_short), were purchased from Future Synthesis (Poland, Poznań). Plasmids based on pUC57-Kan (cloning vector with a kanamycin resistance marker; 2,579 bp; ori pMB1) containing the Rz sequences, designed using SnapGene software (www.snapgene.com), were purchased from BioCat (Germany, Heidelberg). Plasmid DNA and RNA oligonucleotides were dissolved in autoclaved Milli-Q (MQ) water. Sample concentrations were measured using a DeNovix DS-11 spectrophotometer.

The E. coli strains used were BL21(DE3), which encodes T7 RNA polymerase (Jeong et al., 2009), and TOP10, which has high transformation efficiency (Yang and Yang, 2012; Liu et al., 2018). The following media and reagents were used for bacterial cultures: Lysogeny Broth—LB (VWR), Lysogeny Broth with Agar—LA (VWR), Mueller Hinton Broth—MHB (Difco), Davis Minimal Medium (Sigma) with a final concentration of 0.4% D-(+)-glucose (Sigma), kanamycin—Kan (Gibco, Fluorochem), magnesium chloride—MgCl₂ (Sigma), calcium chloride—CaCl₂ (Sigma), isopropyl β-d-1-thiogalactopyranoside—IPTG (BioShop), glycerol (Avantor), and ultra-pure MQ water—Integral 10. Growth media, 0.1 M CaCl₂, and 50% glycerol, were autoclaved. Stock solutions of D-(+)-20% glucose, 50 mg/mL Kan, 1 M MgCl₂, and 1 M IPTG were sterilized using sterile PES syringe filters with a 0.22 μm pore size (GenoPlast Biotech).

In vitro formation of Rz and mRNA substrate complexes was monitored by isothermal titration calorimetry (ITC) using a Nano ITC (TA Instruments). RNAs were dissolved in 50 mM Tris–HCl pH 7.5 (Tris(hydroxymethyl)aminomethane/Tris—Merck, Hydrochloric acid/HCl—Honeywell), 10 mM MgCl₂ (Sigma), and incubated separately at 37 °C for 10 min outside the calorimeter. The final RNA concentrations and ratios were: substrate: Rz_short (32 μM:7 μM) and substrate: Rz_long (42 μM:6 μM). Then, an excess of mRNA substrate dilution of 70 μL was placed in the titration syringe and the ribozyme dilution of 300 μL was placed in the calorimeter sample cell. The reference cell was filled with MQ water. The system was equilibrated for 200 s to establish a stable baseline before the first injection. Next, 17 injections of 3 μL of the mRNA substrate were titrated every 500 s into the sample cell containing Rz and stirred at 350 rpm. Data were analyzed using NanoAnalyze software (TA Instruments).

In vitro cleavage reactions were performed in the buffer (50 mM Tris–HCl pH 7.5, 10 mM MgCl₂). The mRNA substrate and ribozymes were first denatured separately for 1 min at 90 °C. Next, the samples were cooled to 37 °C to allow refolding of tertiary structures. The reactions were then initiated by mixing the RNAs and incubating at 37 °C. Samples of the substrate and Rz complexes were collected over time and the reaction was stopped by adding 50 mM EDTA (ThermoFisher Scientific). Next, the samples were stabilized with formamide (FORMAzol, Molecular Research Center) in the 1:1 RNA: FORMAzol ratio. Control samples, containing either the substrate or Rz alone, were loaded with 6x Loading Dye (ThermoFisher Scientific). Before loading, the RNA samples were denatured for 3 min at 55 °C (short-armed Rz) and 55 °C or 70 °C (long-armed Rz). Substrate cleavage was analyzed by separating the samples on 15, 18% or 20% denaturing polyacrylamide gels with 7 M or 8 M urea, 10% APS, and TEMED (40% Acrylamide-Bis Solution—Bio-Rad, urea—Sigma, Ammonium Persulfate/APS—Sigma, N, N, N′, N′-Tetramethylethylenediamine/TEMED—Sigma) in 1xTBE buffer (Tris—Merck, Boric Acid—Sigma, Disodium ethylenediaminetetraacetate dihydrate/Na₂EDTA—Sigma) using a vertical gel electrophoresis system (ASG-250 Gel System, C. B. S. Scientific). After the gel polymerized, the wells were rinsed and a pre-run electrophoresis without samples was performed in 1xTBE running buffer for 30 min at 400 V. The wells were then rinsed again, samples were loaded, and the main run was performed for 3 h 30–45 min at 400 V. After electrophoresis, gels were stained with SYBR Gold (Invitrogen, ThermoFisher Scientific) dissolved in 1xTBE. Gels were imaged using a Molecular Imager Gel Doc XR + system (Bio-Rad) and the band intensities were measured using the Image Lab Software (Bio-Rad).

Competent cells of E. coli TOP10 were prepared to store plasmids as reserve stocks in the bacterial bank using the E. coli Transformer Kit (A&A Biotechnology). Competent cells of E. coli BL21(DE3) were prepared as follows. The strain was plated on LA and incubated overnight at 37 °C. A single bacterial colony was inoculated into 10 mL of LB medium and incubated overnight at 37 °C with shaking at 600 rpm (Thermomixer comfort, Eppendorf). 1 mL of the overnight culture was added to 100 mL of fresh LB and incubated at 37 °C with shaking at 165 rpm (Incubator Shaker Series, Innova 44) until the culture reached optical density at 600 nm (OD₆₀₀) between 0.2 and 0.3. The culture was divided into two 50 mL sterile falcon tubes and incubated on ice for 20 min. The cells were then centrifuged for 15 min at 4 °C at 5,000 × g (Centrifuge 5810 R, Eppendorf). Supernatants were removed and each pellet was resuspended in 10 mL of ice-cold 0.1 M CaCl₂, incubated on ice for 40 min, and again centrifuged for 15 min at 4 °C at 5,000 × g (Centrifuge 5810 R, Eppendorf). The supernatants were removed and each pellet was resuspended in 0.6 mL of ice-cold 0.1 M CaCl₂. Next, 0.6 mL of 50% ice-cold glycerol was added. The cells were mixed, aliquoted into 100 μL portions, and stored at −80 °C.

E. coli TOP10 competent cells were transformed with plasmids according to the E. coli Transformer Kit (A&A Biotechnology). Plasmid DNA containing mutated inactive ribozymes was transformed into high-efficiency NEB 5-alpha competent E. coli cells, provided with the Q5^® Site-Directed Mutagenesis Kit, following the manufacturer’s protocol. E. coli BL21(DE3) competent cells were transformed using the heat-shock method. Briefly, aliquots of E. coli BL21(DE3) competent cells and plasmids were thawed on ice for 15 min. Fifty nanogram of plasmid DNA was then added to each 100 μL aliquot of competent cells and stirred gently. A no-DNA control sample with sterile water was always transformed to monitor for contamination. The mixtures were incubated on ice for 30 min. Next, the competent cells were heat-shocked at 42 °C for 2 min and placed on ice for an additional 2 min. After that, 1 mL of LB medium (pre-heated to 37 °C) was added to each mixture and incubated at 37 °C for 1 h with gentle shaking at 300 rpm (Thermomixer comfort, Eppendorf). Subsequently, the cells were centrifuged for 1 min at 5000 rpm (Centrifuge 5415 R, Eppendorf). One milliliter of LB was removed and the bacterial pellet was resuspended in the remaining volume (100 μL) of the medium. Next, the cells were plated on LA plates containing Kan and then incubated overnight at 37 °C.

A single bacterial colony was inoculated into a 1 mL of medium (MHB, Davis). Overnight cultures in MHB were supplemented with Kan (final concentration of 50 μg/μL) and were prepared either with or without Mg²⁺, while cultures in Davis medium contained neither Kan nor Mg²⁺. The bacteria were incubated overnight at 37 °C with shaking. The next day, experiments were performed using either glass falcon tubes or 96-well microplates.

For the experiments with falcon tubes, each overnight culture was diluted 100x in 2 mL of the same fresh medium. MHB cultures were supplemented with or without Mg²⁺ (final concentrations: 2.5, 5, or 10 mM). During incubation at 37 °C with shaking at 600 rpm, OD at 600 nm was measured every 30 min or 60 min using a spectrophotometer SP-830 Plus Metertech. After 2.5 h of growth for MHB and 4 h for Davis medium, ribozyme expression was induced by adding IPTG to a final concentration of 1 mM. Each reading was preceded by a blank measurement with clear unsupplemented medium, so the raw data could be used directly in the graphs.

For experiments in microplates, overnight cultures were diluted to the identical final dilution of 100x in a fresh medium to a final volume of 100 μL per well. For each biological replicate, each culture condition had 6 technical replicates, grown with or without 1 mM IPTG. IPTG was added at the beginning of the experiment, prior to adding bacteria. Two independent biological experiments were performed in MHB and two or three in Davis medium. Twenty-four wells containing clear unsupplemented medium were used for background subtraction and as a sterility control. Each microplate was sealed with a transparent sterile film (Microplate Sealing Films, EXCEL Scientific). OD at 600 nm was measured using a microplate reader (Biotek Synergy H1) every 30 min using Gen5 software. The sterility control readings were subtracted from the OD of the bacterial culture. Data from technical replicates for every culture were averaged and analyzed using a two-way ANOVA test. Graphs were prepared with GraphPad Prism 8 software. A p-value < 0.05 was considered statistically significant.

The 20-h growth kinetics of bacterial cultures containing different Rz plasmids, the control plasmid without Rz, and a no-plasmid control, were monitored in MHB medium incubated with tetracycline and with 1 mM IPTG using the microplate method described above. Tetracycline concentrations of 16, 8, 4, 2, 1, or 0.5 μM were prepared following the broth microdilution protocol of the Clinical and Laboratory Standards Institute method M07-A10. Experiments were performed in two biological replicates, each with 2–5 technical replicates.

Previous studies (Dryselius et al., 2003; Good et al., 2001), including ours (Castillo et al., 2018; Równicki et al., 2019), showed that blocking the GAGUAUGAG region of the acpP mRNA, which encompasses the start codon (underlined), with a complementary peptide nucleic acid oligomer, leads to the inhibition of ACP expression and subsequent E. coli growth. Within this acpP mRNA sequence, the nucleotide triplet motif (5′-NUX-3′) at which Rz can cleave is present. Therefore, we selected the GUA sequence in the acpP mRNA as the most likely position for efficient cleavage by Rz (Kore, 1998). We designed two Rz variants differing in the length of their hybridization arms and targeted to the mRNA _acpP substrate cleavage site: Rz_short with 7- and 9-nucleotide-long arms, and Rz_long with two 19-nucleotide arms (Table 1; Figure 2A). For each variant, the Rz catalytic core was either kept in its original form or was extended by one adenine based on in silico predictions suggesting improved binding affinity of the mRNA substrate with the ribozyme. As such an extended catalytic core had not been previously tested, we decided to explore both possibilities.

Alignment of the mRNA _acpP sequences of the E. coli BL21(DE3) strain used in this study with those of selected pathogenic E. coli strains revealed 100% similarity across the target region (Supplementary Table S1). This result underscores the potential for using the mRNA encoding the ACP protein as a potent ribozyme target in pathogenic strains as well.

In silico methods were used to generate the tertiary structure models of the complexes of Rz_long and Rz_short with the mRNA substrate (Figures 2B,C). These models illustrate the Rz arms hybridizing with the mRNA substrate at predicted positions flanking the start codon.

Next, we used ITC to characterize binding between ribozymes and the mRNA _acpP substrate. Titration of the substrate into the ribozymes yielded sigmoidal binding isotherms, indicating an exothermic and spontaneous binding process (Figure 3). The mRNA substrate and Rz_short form a complex with a 16-base pair (bp) interaction region, while the Rz_long complex involves a 38 bp interaction, as predicted in silico (Figure 2A). For both complexes, the data fit to a 1:1 binding model, with association constants (K_a) in the range of 10⁷–10⁸ M⁻¹, confirming the formation of stable complexes. The calculated dissociation constants (K_d) indicate that the complex with Rz_long is more tightly bound than the one with Rz_short (Table 2). Thermodynamic parameters obtained in this study are comparable to previously reported values obtained for similar 15 bp (Mikulecky and Feig, 2004) and 23 bp Rz-substrate complexes (Reymond et al., 2009).

Next, we tested the in vitro cleavage of the mRNA substrate with Rz ribozymes. Analysis of the polyacrylamide gels revealed cleavage products for both the short- and long-armed Rz variants. However, under all tested conditions, Rz_short was consistently faster and more efficient at cleaving the mRNA substrate than Rz_long (Figure 4; Supplementary Figure S1). Attempts to optimize the Rz_long reaction conditions, e.g., by varying incubation temperature or substrate:ribozyme ratio, did not improve the cleavage efficiency to the level exhibited by Rz_short (Supplementary Figure S1). The higher in vitro cleavage efficiency observed for Rz_short compared to Rz_long is likely due to the tighter binding between Rz_long and the substrate (Table 2), which could hinder the dissociation of cleavage products from Rz_long, thereby reducing the rate of enzyme turnover, and making the cleavage of Rz_long less detectable (Supplementary Figure S1).

Based on the higher activity of the Rz_short variants, evidenced by more intense product bands, we further characterized their cleavage kinetics. As anticipated, we observed two distinct cleavage products, the appearance of which corresponded with a time-dependent decrease in the full-length substrate (Figures 4A,B). The gels showing cleavage at different molar ratios of the substrate and Rz_short are shown in Supplementary Figure S2. Analysis of band intensities at different ribozyme to substrate ratios indicated that substrate cleavage was most efficient when the substrate and Rz_short were present in equal molar quantities (Supplementary Figure S3). An excess of Rz_short did not improve cleavage efficiency, corroborating the 1:1 stoichiometry determined by ITC.

A comparison of the in vitro activity of Rz_{short_3A}, which contains the three-adenosine core, with the activity of the Rz_{short_4A} variant, which contains an extra adenine, showed greater product formation and faster disappearance of the substrate band for Rz_{short_3A} (Figure 4). Specifically, Rz_{short_3A} achieved complete substrate cleavage and elimination within 45–60 min, while Rz_short__4A cleaved approximately 75% of the substrate in the same time frame (Figure 4C). Nevertheless, Rz_{short_4A} also cleaved the RNA substrate completely after 180 min at a 1:1 ratio (Supplementary Figure S3).

To express Rz in bacteria for in vivo testing, we designed four cassettes encoding different Rz variants within the pUC57-Kan plasmid (Supplementary Table S2). Each cassette contains a T7 promoter and a transcription terminator for the phage T7 RNA polymerase, as well as additional strong terminators to prevent the synthesis of the ribozyme by other bacterial polymerases (Supplementary Figure S4). These flanking terminators, ilvBN and ECK120016882 (indicated as “terminator_for” and “terminator_rev*” in Supplementary Table S2) are naturally occurring (Chen et al., 2013). The main component of the cassette is the Rz, with two different arm lengths (Rz_short or Rz_long) specifically tailored to hybridize with the mRNA _acpP (Table 1; Figure 2). Additionally, the Rz variants are either standalone or integrated within a tRNA scaffold to protect them from enzymatic degradation [Rz_short(tRNA) and Rz_long(tRNA)]. An extra linker was added to the 5′-end of Rz to connect it to tRNA (Huang et al., 2019) (Supplementary Figure S5), as linkers at the 5′-end of tRNA were found to increase the effectiveness of mRNA cleavage by Rz (Medina and Joshi, 1999). The original linker from Huang et al. (2019) included an EcoRI site, so we modified its sequence to remove any restriction sites within the linker.

To optimize Rz activity in vivo and to be able to incorporate additional controls, we also engineered extra elements within the Rz cassette. One such element is an auto-cleaving hammerhead ribozyme, incorporated as an additional cleavage tool because of its high cleavage efficiency at 3′ termini compared to other ribozymes (Chowrira et al., 1994) (Supplementary Figure S4_1). This auto-cleaving segment cleaves at the GUC site and was positioned after the Rz or the 3’ tRNA terminus to shorten the transcript. This way, following cis-cleavage of the auto-cleaving segment within the cassette, only 3 extra nucleotides remain (Supplementary Figure S4_2), instead of the 48 nucleotides of the T7 terminator (Supplementary Figure S4_1). Without this self-cleaving element, Rz or Rz(tRNA) would possess an excessively long 3′-terminal tail that could interfere with forming the complex between Rz and the mRNA _acpP substrate. Another element, designed as a contingency if the acpP transcript cleavage were to fail, is a site for inserting an additional gene. We included restriction sites (EcoRI/SpeI) for cloning the red fluorescent protein (RFP) gene, which would enable conversion of Rz into a control ribozyme that would target and cleave mRNA _rfp . This arrangement would allow monitoring of cleavage of an mRNA encoded on the same plasmid by measuring RFP fluorescence (Równicki et al., 2017). The designed cassettes also contain other restriction sites (e.g., SpeI, AgeI) to facilitate fragment replacements through restriction enzyme digestion and ligation. Moreover, all sequences can be easily modified using polymerase chain reaction (PCR). Finally, to distinguish growth inhibition caused by mRNA steric blocking versus cleavage, we prepared two catalytically inactive ribozyme variants (Supplementary Figure S6).

We introduced the plasmid constructs containing the Rz variants into E. coli BL21(DE3). The presence of the T7 RNA polymerase gene, under the control of the lacI repressor and lac UV5 promoter, along with the T7 promoter sequence on the plasmid vector, ensures that upon successful transformation, Rz expression occurs intracellularly and can be controlled by IPTG (Supplementary Figure S7).

When culturing bacteria containing a plasmid that encodes antibiotic resistance, it is standard practice to supplement the medium with the corresponding antibiotic. However, to avoid potential interference with the ribozyme-catalyzed reaction, we chose to conduct the experiments without the antibiotic. Aminoglycoside antibiotics, including Kan, bind various RNAs, and could potentially bind Rz and inhibit its transesterification reaction (Stage et al., 1995). To determine whether Kan affected our results, we performed experiments with and without Kan, and compared bacterial growth (Supplementary Figure S8). The results showed similar OD₆₀₀ levels for the control strain and those transformed with Rz-encoding plasmid constructs, regardless of the presence of Kan or IPTG induction. Based on these results, we conducted subsequent experiments without Kan supplementation.

Furthermore, in vitro studies have shown that Mg²⁺ ions are important for forming an active complex between Rz and its RNA substrate (Orita et al., 1995), with Rz catalytic activity increasing with Mg²⁺ concentration (Inoue et al., 2004). Consequently, we investigated whether in vivo Rz activity is affected by supplemental Mg²⁺. We supplemented the medium with 2.5, 5, and 10 mM Mg²⁺. The results showed that higher Mg²⁺ concentrations reduced growth-inhibitory effects arising from Rz_long, even with IPTG induction (Supplementary Figure S9). Thus, contrary to in vitro findings (Inoue et al., 2004; Orita et al., 1995), our data showed that excess Mg²⁺ adversely affects in vivo Rz performance. While Mg²⁺ influences in vitro cleavage, in vivo these ions appear to support bacterial growth rather than enhance cleavage. Based on these observations, we conducted subsequent experiments without supplemental Kan or Mg²⁺.

We also monitored the effect of IPTG on the growth of bacterial cultures. We found that in MHB medium, the OD₆₀₀ of Growth Control and Plasmid Control (containing the pUC57-Kan vector) was consistently lower (by approximately 0.1) in the presence of IPTG compared to cultures grown without IPTG (Supplementary Figures S10, S11). Therefore, the lower OD of bacteria containing Rz could not be conclusively attributed to Rz activity because it could also result from the metabolic burden imposed by IPTG addition. To address this issue, we changed the culture medium from MHB to Davis minimal medium. In Davis medium, the OD values for Growth Control with and without IPTG, were similar, as were OD values for Plasmid Control with and without IPTG (Supplementary Figures S10–S12). We therefore concluded that the reduced OD of bacteria containing Rz_long, observed after IPTG addition, was likely due to plasmid expression and ribozyme activity.

We also monitored bacterial growth in MHB and Davis media using a microplate reader, as opposed to the above experiments performed in falcon tubes. As expected, the highest OD₆₀₀ was observed for the Growth Control (Supplementary Figure S13). In both media without IPTG, the growth curves for E. coli containing plasmids with Rz_long and Rz_short showed OD₆₀₀ values 0.1 to 0.2 lower than the Growth Control, demonstrating statistically significant growth inhibition. Importantly, after IPTG induction, both Rz_long and Rz_long(tRNA) visibly inhibited bacterial growth in both MHB and Davis media (Supplementary Figure S13).

Interestingly, in experiments using MHB medium in both falcon tubes (Supplementary Figures S8–S11) and microplates (Supplementary Figure S13), we consistently observed a delayed onset of the logarithmic growth phase in bacteria containing Rz_long grown without IPTG. This led us to hypothesize that a component within the medium was causing T7 promoter leakage and expression of Rz_long. The MHB medium contains starch (1.5 mg/mL = 7.14%), a saccharide similar to other saccharides present in plant-based media (Krefft et al., 2022), and structurally similar to IPTG, which may induce the expression of the lac promoter in E. coli BL21(DE3). This likely explains the unintended expression of Rz_long in MHB without IPTG. In contrast, glucose, which is present in Davis minimal medium, does not cause such leakage (Krefft et al., 2022).

The OD of the E. coli culture with the pUC57-Kan vector (Plasmid Control) was higher than that of bacteria containing a plasmid with the ribozyme cassettes, indicating that the observed growth inhibitions can be attributed to ribozyme expression. The 20- to 24-h kinetics confirmed that inhibition of bacterial growth was most pronounced in Davis medium (Supplementary Figure S14; Supplementary Table S3). We also noted a lower OD for bacterial cultures with the Plasmid Control compared to the Growth Control after 24 h (Supplementary Figure S15), likely due to the burden imposed by the plasmid (Diaz Ricci and Hernández, 2000; Ow et al., 2006). However, comparing the OD measured for the Rz_long and Rz_long(tRNA) cultures to the OD of the Plasmid Control in Davis medium with IPTG (Supplementary Figure S15) shows that the reduction in bacterial growth is due to the activity of the expressed ribozymes and not merely the presence of the plasmid vector itself.

Since the kinetic experiments confirmed that Davis medium provided optimal conditions for observing bacterial growth inhibition in cultures with Rz, we used this medium to test Rz activity in cells, both without and with IPTG induction. We compared modified ribozymes containing an extra adenine in the core (4A) to ribozymes with the original three-adenine (3A) catalytic core (Figure 5). Although statistically significant differences in OD₆₀₀ were observed without IPTG induction (Figure 5A), as expected, adding IPTG significantly accentuated the differences between the Rz variants (Figure 5B).

Figure 5B shows that Rz_short(tRNA)_{_4A} and Rz_short(tRNA)_{_3A} (green lines) were both inactive, exhibiting similar growth patterns to the Growth and Plasmid controls. Their standalone variants, Rz_{short_3A} and Rz_{short_4A}, were moderately active (red lines) with the 3A variant (light red line) demonstrating stronger inhibitory activity than its 4A counterpart, particularly between 9 and 18 h; however, after 20 h, their activity was similar. The Rz_long(tRNA) 3A and 4A variants (pink lines) demonstrated stronger and similar inhibitory effects, with slight differences only between 10 and 15 h, and the final OD after 24 h decreased by approximately 50% compared to the Growth Control (Supplementary Table S4).

Importantly, both Rz_long variants (blue lines, Figure 5B) effectively suppressed bacterial growth, exhibiting the largest inhibitory effects. The Rz_{long_3A} construct reduced final OD₆₀₀ values by approximately 70% compared to the Growth Control, representing the most potent inhibitory effect among all tested constructs. The differential effects of tRNA fusion on Rz_short versus Rz_long variants suggest that structural context influences activity, with the tRNA fusion abolishing activity in Rz_short constructs while preserving it in Rz_long constructs.

Since the kinetic experiments shown in Figure 5 indicated that Rz_long and Rz_long(tRNA) were the most active constructs, we redesigned these ribozymes into inactive variants (Supplementary Figure S6) to determine whether the observed inhibition arises from mRNA blocking or cleavage (Supplementary Figure S16). Interestingly, the growth inhibition of bacterial cells containing the inactive ribozymes (Supplementary Figure S16) was greater than the inhibition caused by their corresponding active versions (Figure 5).

Notably, all active ribozyme variants not only reduced final OD values but also delayed the onset of the exponential growth phase, with Rz_long variants showing delays of approximately 4–6 h compared to controls. For Rz_long, the variant with four adenines (4A) inhibited bacterial growth until ~12 h while the variant with three adenines (3A) did so for ~7 h (Figure 5B). Although their inhibitory effects were similar at 18.5 h, thereafter Rz_{long_3A} showed greater growth inhibition. At 24 h, bacterial growth relative to the Growth Control was 57% for Rz_{long_4A} and 33% for Rz_{long_3A} (Supplementary Table S4), suggesting that the original catalytic core design maintains optimal activity in vivo. This statistically significant ~24% difference in final OD between the cultures containing Rz_{long_4A} and Rz_{long_3A} suggests the involvement of mRNA cleavage, as the 3A variant is more catalytically active in vitro. If growth inhibition were the result of only translational blocking, the inhibitory activities for Rz_{long_3A} and Rz_{long_4A} should be the same, as they share identical binding arms.

To investigate whether ribozyme activity in cells could enhance antibiotic efficacy, we performed a preliminary study using tetracycline. We chose tetracycline due to its relatively high MIC of 16 μM (Wojciechowska et al., 2020) against the E. coli BL21(DE3) strain. This experiment was carried out in MHB because it is a commonly used medium for MIC testing. Analysis of the OD of bacterial cultures incubated with tetracycline showed that the presence of Rz variants reduced the MIC of tetracycline to 8 μM for Rz_short, Rz_short(tRNA), Rz_long(tRNA) (both the 4A and 3A variants), and Rz_{long_3A}, and to 4 μM for the Rz_{long_4A} variant (Figure 6; Supplementary Figure S17). Consequently, bacteria expressing Rz variants were more susceptible to tetracycline. This synergistic effect represents a twofold reduction in MIC for most ribozyme variants and a fourfold reduction for Rz_{long_4A}. The Plasmid Control showed no change in tetracycline MIC compared to the Growth Control, confirming that the enhanced tetracycline susceptibility results from ribozyme activity rather than plasmid burden.

The growth kinetics of cultures incubated with tetracycline and Rz 4A ribozyme variants are shown in Supplementary Figure S18. At increasing tetracycline concentrations, Rz_long(tRNA) most strongly suppressed the logarithmic growth phase compared to the other Rz 4A variants. However, the standalone Rz_long variant caused the greatest reduction in the tetracycline MIC. The growth kinetics of cultures incubated with tetracycline and the Rz 3A variants (Supplementary Figure S19) showed a similar trend to the 4A ribozymes, but with slightly less growth inhibition.