Blue-white screening as a new readout for deoxyribozyme activity in bacterial cells

The 10–23 deoxyribozyme is considered as sequence-specific “molecular scissors” for RNA molecules. Extensive investigations have been reported for this deoxyribozyme in vitro and in eukaryotic host cells. However, few investigations are reported in the literature on the activity of this deoxyribozyme inside bacterial cells. The available reports focus on the cleavage of target mRNAs that encode for proteins which are responsible for the viability of bacterial colonies. Hence, the growth of bacterial cells was blocked and the main readouts in such studies were colony counts or optical density at 600 nm. In the current study, blue-white screening was utilized as a novel readout for analysis of the activity of the 10– 23 deoxyribozymes in viable bacterial cells. Two deoxyribozymes were designed to target the a-peptide fragment from b-galactosidase (lacZ) mRNA at two different positions, i.e. 50 untranslated region and translated region. Control experiments were performed utilizing DNA oligos that lacked the catalytic core. The 30–3 inverted thymidine modified deoxyribozymes were compared with unmodified ones to analyze the effect of such modification in prokaryotic cells. The activity of the designed deoxyribozymes caused a significant retardation in the formation of the blue-color in colonies with deoxyribozymes. Miller assay confirmed the blue-white screening results. This report showed a proof of concept for application of blue-white screening as a readout system for the activity of the 10–23 deoxyribozyme that is a model RNA-cleaving deoxyribozyme. The result of this report can promote future investigations on the activity of deoxyribozymes in prokaryotic cells.


Introduction
Protein enzymes were known as the sole biological catalysts until the 1980s. During the 1980s two natural RNA catalysts (ribozymes) were discovered for which a Nobel prize was shared between two scientists, Sidney Altman and Thomas Cech in 1989. 1 A few years later, reports on oligodeoxyribonucleotidemediated catalysis demonstrated catalytic ability of DNA molecules. 2 Since then, a growing number of deoxyribozymes (DNAzymes, DNA catalysts, Dzs) have been selected in vitro. 3 Early evolved deoxyribozymes were RNA-cleaving ones. 2b, 4 The catalytic cores of these deoxyribozymes, like many others, were anked by two binding arms that complementarily bind specic RNA substrates.
RNA-cleaving deoxyribozymes can be considered as sequence-specic "molecular scissors" for RNA molecules. The 10-23 deoxyribozyme which consists of a 15 nucleotide long catalytic core is a well-known example of an RNA-cleaving deoxyribozyme. It was shown that this deoxyribozyme can cleave biologically relevant RNA molecules in vitro in a site specic manner. 5 The activity of the 10-23 deoxyribozyme was also monitored in vivo. 6 The enzyme was shown to be capable to modulate gene expression at post-transcriptional level. 7 The in vivo targets of 10-23 deoxyribozyme were mostly viral RNA 8 or mRNA of eukaryotic cells. 9 Many genes from HIV-1, 10 hepatitis B and C, 11 inuenza A and B, 12 human papilloma virus, 13 severe acute respiratory syndrome coronavirus (SARS) 14 and Epstein-Barr virus have been the targets for 10-23 deoxyribozyme RNA cleavage. 15 Handful of genes responsible for cardiovascular, inammatory or central nervous system diseases were down-regulated with 10-23 deoxyribozyme. 6 Many cancer related genes such as multidrug resistance gene (MDR) have been targeted with this enzyme. 16 Currently, the 10-23 deoxyribozyme is a candidate therapeutic for several human diseases such as allergic bronchial asthma with T H 2 molecular signature. 17 The GATA-3 targeting 10-23 deoxyribozyme is now under clinical trial investigations by sterna biologicals and phase IIa for early-phase and late phase asthmatic responses aer allergen provocation has been completed. 18 Despite advances in application of 10-23 deoxyribozyme in eukaryotic and mammalian cells, limited numbers of reports are available on investigation of the 10-23 deoxyribozyme activity inside the bacterial cells. 6 A major challenge in application of deoxyribozymes in bacterial cells is dilution of the transfected oligonucleotide inside the fast dividing cells. Thus, all present reports used the 10-23 deoxyribozyme in an antibiotic-like manner. Such reports focus on targeting mRNAs of genes that were essential for viability and proliferation of the bacterial cells. The mRNAs for genes that are responsible for antibiotic resistance such as, mecR1 (ref. 23b) and bla1 (ref. 23a) related to the b-lactamase pathway are the most targets for application of 10-23 deoxyribozyme in bacterial cells. Besides, Tan et al. reported cleavage of sZ mRNA by 10-23 deoxyribozyme. 24 FtsZ gene is responsible for binary division of E. coli cells and its downregulation prevents bacterial colony formation. Therefore, in all the reported cases for prokaryotic application of 10-23 deoxyribozyme, inhibition of cell growth was the main result of the activity of the deoxyribozyme. Counting colonies on agar plate or analyzing OD 600 were the main readout systems for such studies.
Due to the aforementioned technical issues, the current applications of deoxyribozymes in prokaryotic cells are limited to 10-23 and cleavage of vital mRNAs. Although inhibition of bacterial growth has advantages and permits application for controlling bacterial populations, indeed it does not grant investigation of the activity of the deoxyribozyme inside of the prokaryotic cells. Cells where the deoxyribozyme are active die due to the blockade of the target genes and thus no means are available to understand the mechanism of action of the deoxyribozyme in viable bacterial cells. Fundamentally, it would be helpful to have a new readout system for investigation of the activity of various deoxyribozymes inside vital prokaryotic cells. In addition, with deeper understanding on the kinetics of deoxyribozymes inside prokaryotic cells, the scope is capable of broadening e.g. to utilize the prokaryotic-active deoxyribozymes as biological elements of biosensors.
The lactose (lac) operon consists of three genes for production of b-galactosidase (Lac Z), permease (Lac Y) and transacetylase (Lac A). The lac operon is widely used as a reporter of gene insertion in cloning at molecular biology labs by simple and visual blue-white screening. The enzyme b-galactosidase is a tetramer with identical four polypeptides and is responsible for formation of a blue color upon presence of a chromogenic substrate like 5-bromo-4-chloro-3-indolyl-b-D-galactopyranoside (X-gal). Deletion of a few amino acids from amino-terminus of monomers of the enzyme, causes the formation of a dimer which may also further breaks down to monomers. Administration of the "missing" amino acids in an oligopeptide, also called a-peptide results in re-structuring of the enzyme in active tetramer form. Molecular biologists hijacked this mechanism to analyze cloning efficiency. 26 To this endeavor, the host must be an E. coli strain with missing amino acids from amino-terminus of b-galactosidase. Upon transformation, the a-peptide gets provided into the cells. For a standard blue-white screening, the gene of interest is inserted in a multiple cloning site which is located inside the coding sequence of the a-peptide. Thus, upon gene insertion, the a-peptide is disrupted and a-complementation may not occur. In contrast, vectors with complete and active a-peptide will perform a-complementation. As a result, the colonies with vector that contains the gene of interest do not bear active tetramer b-galactosidase and cannot produce a blue color from X-gal, while colonies with vectors that lack the gene of interest will form a blue color. Accordingly, the color of the colonies will be a marker for of the activity of the b-galactosidase and consequently presence of the gene of interest.
This report provides a novel application for the well-known method of blue-white screening. The presented method doesn't utilize the screening as a readout of gene insertion but rather it pursues the catalytic activity of the 10-23 deoxyribozyme. Here, the fully active a-peptides were transformed into bacterial hosts. The 10-23 deoxyribozymes targeted the mRNA of the a-peptides. The post-transcriptional cleavage of the mRNA of the a-peptide was expected to diminish acomplementation and formation of the blue color. Hence, the proposed method is able to report the activity of the b-galactosidase in viable cells. Proof of principle results were shown for the activity of the 10-23 deoxyribozyme for site specic cleavage at 5 0 -UTR and translated regions.
In most of the few available reports on prokaryotic cells, the oligodeoxyribonucleotide sequences of 10-23 deoxyribozyme were transfected to the bacterial cell as naked single stranded DNA. As an example, in a case report, oligonucleotides bearing the sequence of 10-23 deoxyribozyme as monomers and dimers designed against b-lactamase gene and was investigated in TEM-1 producing E. coli. 19 Naked unmodied oligonucleotides are prone to degradation by nucleases. Various types of DNA modications are available to block nuclease degradation of oligonucleotides in vivo. Addition of a 3 0 inverted nucleotide at the 3 0 end of the oligo, 20 phosphorothioation of the oligo backbone, 21 modication at 2 0 position of the ribose sugar such as 2 0 -O-methyl modication 22 and application of locked nucleic acid (LNA) 13 on deoxyribozymes' binding arms are common examples. Among such a wide spectrum of DNA modications, only phosphorothioation of the 10-23 deoxyribozyme was investigated in prokaryotic cells. Phosphorothioated DNA oligos were applied to cleave b-lactamase in oxacillin and methicillin resistant S. aureus, WHO-2 strains. 23 In another strategy, vectors have been designed to transform the bacterial cells with deoxyribozymes 10-23. pssXGa vector 24 and pBlue-script-II KS (+) phagemid 25 were designed for E. coli. In such strategies, the deoxyribozymes are induced to be synthesized inside bacteria and thus there is less concern for nuclease degradation. In this report, modied and unmodied DNA oligos with 3 0 -3 0 inverted thymidine were compared for blockade of the activity of the bgalactosidase. In addition, the effect of the deoxyribozymes was dissected from antisense effect, by comparing the results of full deoxyribozymes and the ones without catalytic core.

Design of deoxyribozymes
The promoter of the lacZ-a peptide fragment was located in the reverse orientation from nucleotides 210-239 of the pGEM-T vector. Thus, the transcript template starts from position 209 in the reverse complement form down to position 2832, encompassing multiple cloning site and SP6 and m13 primer binding sites. The complete pGEM-T vector sequence with assigned fragments is depicted in ESI Appendix SI. † The mRNA transcript of the a-peptide fragment with pGEM-T vector was 378 nt long. The sequence of the transcript is provided in ESI Appendix SII. † The transcript coded for a fragment of lacZ apeptide from amino acid 67-115. The sequence of the a-peptide fragment from pGEM-T is presented in ESI Appendix SIII. † The start codon on the transcribed mRNA was at position 44-47.
The predicted secondary structure of the transcript was taken as a guide to choose the best cleavage sites. To support effective approach of the deoxyribozymes's binding arms to the target mRNA, positions at which no extensive stems were present have been selected as target sites. Extended binding arms would thermodynamically inhibit binding of the deoxyribozymes to the target. To address the effect of the position of the cleavage site, two target sites were chosen; one in the 5 0 -untranslated region (5 0 -UTR), upstream of the rst AUG and the second in the translated region. The 10-23 deoxyribozyme is known to be specic for cleavage at purine-pyrimidine junctions. The A 39 -C 40 junction (inside 5 0 UTR) and A 106 -C 107 junction (inside translated region) were selected for cleavage by the deoxyribozymes. The deoxyribozymes were applied in two formats; with 3 0 -OH and with 3 0 -3 0 inverted thymidine. The latter was used to investigate protection potentials of such modication in prokaryotic host cells. The effect of three control DNAs was analyzed as well; (1) to dissect the catalytic activity of the designed deoxyribozymes from antisense effect, DNA oligos without catalytic cores (WOCC) were analyzed in parallel. WOCC had three thymidines instead of the catalytic core.
(2) Additionally, a DNA oligonucleotide that do not target lacZ mRNA was co-transformed with pGEM-T vector as non-target DNA control (NTC). (3) Besides, in a negative control experiment, pGEM-T was transformed solely to E. coli competent cells. Fig. 1 and Table 2 depict the schematic structure of the target mRNA and sequences of the designed deoxyribozymes respectively.

Magnesium chloride optimization
Kinetic studies on 10-23 deoxyribozyme, similar to many others, conrmed strong dependence of the enzyme on bivalent metal ions e.g. Mg 2+ . 4 The activity of this deoxyribozyme is increased upon increment in Mg 2+ concentration in a hyperbolic manner. 4 However, cellular concentration of Mg 2+ is limited to submillimolar concentrations. Thereupon, the activity of deoxyribozymes gets hampered. Specic pumps, channels etc. guarantee a balanced concentration of Mg 2+ inside cells, and the Mg 2+ concentration of the culture media do not directly represent its cellular concentration. However, to assure that adequate amount of Mg 2+ is available, the bacterial growth at various concentrations of Mg 2+ was analyzed. The effect of Mg 2+ on the cell viability of E. coli colonies were investigated on agar plates. The highest concentration of Mg 2+ with no inhibition of cell growth was considered to be optimal. To this end, we investigated at which concentration of Mg 2+ the cell viabilities start to drop. Our result showed that Mg 2+ concentration above 40 mM reduces colony formation and thus, we decided to perform experiments with 35 mM Mg 2+ present in the culture media (Fig. 2).

Blue-white screening
Upon transformation and subsequent incubation on ampicillin containing agar medium, white colonies get formed in a time span of ca. 12 h. Aer appearance of white colonies, a solution containing both IPTG and X-gal (IPTG-X-gal) was sprayed over the agar plates. The solution was sprayed to ensure even  distribution of the solution. The color of the colonies started to change from white to blue. The complete color change took about 12 h and the colonies had shown a range of blue color intensities in the meantime. ESI Fig. SI and SII † show exemplary agar plates prior and aer IPTG-X-gal spray. The percentages of the blue color of samples with active deoxyribozymes were clearly less than control samples at early time points (Fig. SIII †). The photographs and analyzed data for the case of the 10-23 deoxyribozyme for untranslated region, A 39 , are shown in Fig. 3. The best effect was observed 1.5 h aer the spray. The difference faded at later time points in a way that at 12 h aer spray, almost no difference of the blue color was detectable in colonies. Bacterial colonies with different ranges of blue-color were grouped and plotted in Fig. 3B. Such presentation shows that the number of colonies with higher cyan percentage are reduced in MS and NMS in comparison to negative control.
In the case of the assays for untranslated region, the in negative control (NC), which contained only pGEM-T blue color started to appear rapidly and the number of colonies with high cyan percentage shied to about 60% of the population within 1.5 h. Such rapid color change was in favor of successful transcription of the mRNA of the a-peptide, fruitful translation of the mRNA and active complementation of the b-galactosidase. Formation of the blue color in samples with WOCC had no signicant change in comparison to the negative control and ca. 70% of the colony population had a cyan color between 14-27% aer 1.5 h. The color change in the other two samples, i.e. NMS and MS were stalled at early time points and the effect was more pronounced for MS. In a way that ca. 20% of the population of the colonies had minimal cyan color (0-6%) aer 1.5 h (Fig. 3A, second column). All the bacterial colonies could transform to complete blue color at longer time points, i.e. 12 h aer spray (Fig. 3A, last column). Fig. 4 depicts the average of the percentage of the cyan color in the colonies of agar plates at three-time points. For both cases of assays, with 10-23 deoxyribozymes for untranslated region, A 30 and translated region, A 106 , the most difference between the experimental setups was observed at 1.5 h aer spraying IPTG-X-gal. At this time point signicant reduction of the blue color was observed with modied sequences. Presence of the 3 0 -inverted thymidine at the 3 0 end of the deoxyribozymes was turned out to improve the catalytic performance of both deoxyribozymes that inside bacterial cells. The non-modied sequence could only reduce the cyan color in the translated region. In both cases, the WOCC DNA had no meaningful effect in comparison to the control experiments (Fig. 4).

Miller assay
The enzyme b-galactosidase is able to hydrolyze a wide range of articial b-D-galactopyranosides including the chromogenic substrate ortho-nitrophenyl-b-D-galactopyranoside (ONPG). The hydrolysis of the glycosidic bond results in formation of orthonitrophenol which is yellow. A method that measures the specic activities of the b-galactosidase has been presented by Miller in 1972. The assay reports the enzymatic activity as miller units.
The b-galactosidase activity of the transformed bacterial colonies were analyzed by Miller assay in this study. The yellow color was started to appear ca. 17 h aer transformation. Since then, the Miller assays were performed for negative control, non-target control DNA, WOCC, NMS and MS. As in blue-white screening, the signicant differences among the experimental setups were observed transiently. The most difference was at the 19 h time point. Consistent with blue-white screening results, at this time point, modied sequences of deoxyribozymes had signicant reduction effect in comparison to both negative control and non-target control in both target sites. Non-modied sequence of 10-23 was effective for reduction of the blue color in translated region, but no meaningful reduction was observed with non-modied sequence in the untranslated region, consistent with blue-white screening results. WOCC could not show signicant difference in either of the cases (Fig. 5).

Discussion
The RNA-cleaving 10-23 deoxyribozyme has been widely used for gene silencing in eukaryotic cells. 27 This deoxyribozyme was extensively applied for cleavage of mRNAs that are responsible for several diseases such as allergic reactions, inammation and cancer. 28 A few 10-23 deoxyribozymes have been gone through clinical trials and currently passed phase I and II. In addition, many viral mRNAs have been targeted for cleavage with 10-23 deoxyribozymes inside eukaryotic cells. 5 Despite such advances in application of 10-23 deoxyribozymes in eukaryotic cells, there are only few reports on the activity of this deoxyribozyme inside prokaryotic cells (Table 1). In addition, the available reports only target bacterial growth and survival. The b-lactamase pathway was the main target for 10-23 deoxyribozymes mediated cleavage in these reports.
To the best of our knowledge, the current report is the rst one for cleavage of an mRNA of a reporter gene with 10-23 deoxyribozymes in a viable prokaryotic cell. The target mRNA in this study was not intrinsically present in the host genome and was transformed simultaneously with the deoxyribozyme's oligonucleotide or its controls. An equivalent study for eukaryotic cells has been reported by Ackermann et al. for human embryonic kidney, HEK293 cell lines. 29 The target gene was provided with a eukaryotic expression vector. The transformation of the expression vector and deoxyribozyme's oligonucleotide was carried out simultaneously. The ratio of vector to oligonucleotide was 8500 : 1. Although the strength of the promoters and the cell division rates of the two studies are not directly comparable, this ratio was taken as a rule of thumb for this study.
In our study, we applied 0.02 pmol vector and 600 pmol deoxyribozyme for co-transformation into 50 ml of competent cells with total number of ca. 375 000 000 cells. The ratio of vector to oligonucleotide was ca. 32 000 : 1, and the ratio of deoxyribozyme to bacterial cells was 1 000 000 : 1. Despite excessive application of deoxyribozymes in this study, the transformed RNA-cleaving 10-23 deoxyribozyme was in a big challenge inside rapidly growing E. coli colonies. In other words, besides nuclease-mediated degradation of the transformed deoxyribozymes, 10-23 oligonucleotides were prone to dilute and vanish in bacterial descendants. Hence, we expected a transient knock-down effect which was conrmed by our results.
The blockade of formation of blue-color from X-gal was transient. In other words, the colonies with active enzyme were as blueish as control colonies at long time span (above 4 h from induction). The reason could be due to two parallel phenomena; (1) dilution of the deoxyribozymes upon bacterial division, (2) nuclease degradation. In this study, the deoxyribozymes had no antibiotic-like effect; therefore, the colonies were growing during the assays. This means, while the number of transformed DNA oligonucleotides was constant, the number of cells was increasing. Thus, it would be logical to deduce that within a colony there would be some bacterial cells that get a low concentration of the deoxyribozyme. At such low concentrations, the target mRNA (a-peptide) was not efficiently cleaved and cells scape from blockade of a-complementation. If only a few cells in a colony form the active b-galactosidase the presence of blue color will override the color of the colony i.e. the colony becomes blue, in spite of probable presence of bacterial cells that lack active b-galactosidase.
A thorough study must have been performed to nd the best time point with maximal difference with control samples. With 30 min intervals, we comprehensively investigated for the best time to visualize the effect. The optimal time for blue-white screening was between 1-2 h aer induction by spraying IPTG-X-gal solution. This time was indeed 11-12 hours aer transformation, due to the fact that it took about 10 h for formation of detectable colonies. The best time point for observing the difference in the Miller assay was 19 h, since it took about 17 h for initial observation of the yellow color from Miller assay.
Regarding the nuclease dependent degradation of oligonucleotides, various modications have been applied for 10-23 deoxyribozyme to increase the half-life of the DNA within eukaryotic cells. In contrast, the only modication in prokaryotic system was phosphorothioation. Here, we investigated the effect of the 3 0 -3 0 inverted thymidine. Our result demonstrated that MS (the 10-23 deoxyribozyme with modication) could knock-down b-galactosidase more efficiently in comparison to NMS (non-modied sequence). Formation of the blue color from X-gal and formation of yellow color from ONPG was meaningfully reduced in MS, although the effect was transient. NMS was effective only in translated region, while the MS could stall blue color formation in both target sites. Thus, we conclude that 3 0 -3 0 inverted thymidine is an efficient modication for application of deoxyribozymes in prokaryotic cells.
The DNA oligonucleotides at which the catalytic cores were replaced by three thymidines could not cause a decrease in the readout of any of the assays, expectedly. These oligonucleotides were modied with 3 0 -3 0 inverted thymidine and thus the absence of the effect was not the result of nuclease degradation. Hence, the blue-or yellow-color reductions were deduced to be the direct effect of 10-23 deoxyribozyme catalysis and cleavage of the mRNA of the a-peptide. In the absence of a-peptide, the four monomers of b-galactosidase could not form the active tetramer and the enzyme was blocked. Our result showed that similar reduction extent for translated and untranslated regions when modied deoxyribozymes were used. However, with non-modied sequences, the 10-23 deoxyribozyme could not efficiently cleave and block translation of a-peptide in untranslated region.
Selection of host and vector. In this report, pGEM-T was selected as a model vector for the a-complementation. The E. coli DH5a was chosen as the host strain since it has a deletion at amino-terminus (lacZDM15) and is suitable for blue-white screening. pGEM-T vector plasmid was purchased from Sina-Clone, Tehran, Iran. DH5a strain with ampicillin resistance was purchased from Persian type culture collection, Tehran, Iran.
Design of deoxyribozymes. The sequence of pGEM-T vector was retrieved from addgene web page. The transcript of apeptide fragment was recouped as the reverse complementary sequence aer the promoter, since the promoter was in reverse orientation. The sequence of the open reading frames was compared with the original sequence of the a-peptide with ExPASy and identity of the sequences were conrmed.
The secondary structure of the mRNA transcript of the apeptide was predicted with mfold from the DINAmelt web server. 30 The default parameters were used, and the secondary structure with the lowest free energy was exploited to identify the regions within the mRNA that are accessible to the deoxyribozyme. The deoxyribozymes with 9 nucleotides long binding arms at each site have been designed against target sequences at A 39 -C 40 and A 106 -C 107 . The le binding arms fully hybridize to the target DNA sequence except at the cleavage site's adenosines. The catalytic core of the 10-23 deoxyribozyme with the sequence of GGCTAGCTACAACGA were substituted the base pair at this adenosine. The adjacent cytosines at the cleavage sites and its 3 0 nucleotides were hybridized to the right binding arms. The designed binding arms were blasted to assure absence of non-target binding of the deoxyribozymes. The designed deoxyribozyme sequences were either modied (here mentioned as MS ¼ modied sequence) by adding a reversed 3 0thymidine or not modied (NMS). An oligonucleotide sequence from designed deoxyribozyme without catalytic core (WOCC) was considered as a control. The catalytic domain of 10-23 deoxyribozyme was substituted with three consecutive thymidines in WOCC oligonucleotides. Additionally, non-enzymatic, and non-target control DNA (NTC) which was a random oligonucleotide sequences was used to control the effect of transformation. Table 2 depicts the oligonucleotide sequences that have been used in this study.
Transformation of the chemically competent cells. For transformation, chemically competent cells were freshly prepared starting from 1.5 ml bacterial cells with OD 600 . The nal volume of competent cells was 100 ml from which 50 ml was used for transformation. Single stranded oligonucleotides were co-transformed with p-GEM plasmids for a-complementation simultaneously. Calcium chloride treated competent cells were incubated in presence of 18 ng i.e. 0.02 pmol pGEM-T vector and 600 pmol of oligonucleotides on ice for 30 min. Heat shock were performed by incubation at 42 C in a water bath for 90 s followed by 2 min incubation on ice. Subsequently, 400 ml of LB medium without antibiotic were added to the heat-shocked mixtures and incubated at 37 C for 1 h. Samples were centrifuged at 8000 rpm, room temperature and the pallets were resuspended in 50 ml of LB medium. The re-suspended bacteria were sub-cultured on agar plates containing ampicillin and optimized concentration of magnesium for blue-white screening assays and were sub-cultured in LB medium with ampicillin for miller assays.
Blue-white screening. Aer transformation, plating cells on agar and appearance of the rst colonies (12 h), 150 ml of a solution containing 1.2% IPTG (for induction of the expression of the b-galactosidase) and 0.5% X-gal (chromogenic lactose analogue) sprayed on the agar plates. The white colonies started to synthesize b-galactosidase and produce blue color. The agar plates were placed into a photo chamber with a yellow or black paper beneath. The yellow color was chosen since it is the most color to differentiate between white and blue color.
The chamber was equipped with white light LED strips, spanning all four vertical walls of the chamber. A camera was placed on top of the chamber and photographs were taken at xed intervals (every 30 min). Photos were analyzed with Adobe Photoshop CC2014 by reporting the percentage of the cyan in the Cyan-Magenta-Yellow-Black (CMYK) color system. In the CMYK system, each of the colors have a specic value between 0-100 that better denes the blue color of a gure in comparison to Red-Green-Blue (RGB) system. The cyan percentage of the white colonies in an un-induced plate was between 40-50% (Fig. SI †). Any increase in the cyan percentage was deduced to be upon activity of b-galactosidase; accordingly, the minimal cyan percentage of such plate was taken as the C background and was deducted from all cyan percentages of colonies of plates that got induced with IPTG-X-gal. The residual cyan percentages were normalized in a scale of 1-100. White color colonies have C N values below 20. Normalized C values, C N , were plotted by Graphpad Prism 7. All experiments were performed three times. The graphs show mean AE standard error of means (SEM).
Miller assay. Miller assay was performed according to standard protocols. 31 Briey, the transformed bacterial cells were transferred to 1.5 ml LB media with ampicillin and magnesium chloride. Next, 5 ml IPTG 20% was added to induce expression of b-galactosidase. At specic time points, 20 ml of the medium were mixed with 80 ml of the freshly made permeating buffer (0.1 M Na 2 HPO 4 , 20 mM KCl, 2 mM MgSO 4 , 0.0001% v/v mercaptoethanol). Then 2.5 ml chloroform and 2.5 ml 20% SDS were directly added to the mixture. Samples were vortexed-spinned and incubated for 20 min at 30 C. Consequently, 600 ml 1% ONPG was added. The ONPG substrate was supplemented with fresh mercaptoethanol (2.7 ml for each 1 ml of the solution) before to addition. Expressed b-galactosidases cleave ONPG to lactose and O-nitrophenol. The latter have a yellow color and absorbs at 420 nm. b-Galactosidase reaction was stopped by addition of 700 ml 10% Na 2 CO 3 , raising the pH of the solution to 9.0. Specic activity of the b-galactosidase was measured in Miller scale. Miller unit was dened as in eqn (1). Where, A 420 is absorption of yellow color produced in the test and A 550 is scattered from cellular debris that was corrected by the factor of 1.75 and A 600 determined cell density demonstrator.   Statistical analysis. All values of analyzed data are presented as mean AE standard deviation (SD) or standard error (SE) from three replications. The analysis method was one-way ANOVA followed by multiple comparison test (GraphPad Prism soware version 7). Differences among the groups were considered statistically signicant when the p value was less than 0.05 or indicated values in the graphs.

Conclusions
In summary, this report showed a proof of concept for application of blue-white screening as a readout system for activity of the model 10-23 deoxyribozyme. The result of this report can promote future investigations on activity of RNA-cleaving deoxyribozymes in vivo. The applied condition provided a transient inhibition effect on b-galactosidase activity. Stable knockdown effect is expected by employing methods that provide in situ production of single stranded deoxyribozymes in prokaryotic cells e.g. pBlue-scriptIIKS (+) 25 or MMLV-RT system. 7

Conflicts of interest
None of the authors declare conict of interest.