Abstract
Regulated antisense RNA (asRNA) expression has been employed successfully in Gram-positive bacteria for genome-wide essential gene identification and drug target determination. However, there have been no published reports describing the application of asRNA gene silencing for comprehensive analyses of essential genes in Gram-negative bacteria. In this study, we report the first genome-wide identification of asRNA constructs for essential genes in Escherichia coli. We screened 250,000 library transformants for conditional growth-inhibitory recombinant clones from two shot-gun genomic libraries of E. coli using a paired-termini expression vector (pHN678). After sequencing plasmid inserts of 675 confirmed inducer-sensitive cell clones, we identified 152 separate asRNA constructs of which 134 inserts came from essential genes while 18 originated from non-essential genes (but share operons with essential genes). Among the 79 individual essential genes silenced by these asRNA constructs, 61 genes (77%) engage in processes related to protein synthesis. The cell-based assays of an asRNA clone targeting fusA (encoding elongation factor G) showed that the induced cells were sensitized 12 fold to fusidic acid, a known specific inhibitor. Our results demonstrate the utility of the paired-termini expression vector and feasibility of large-scale gene silencing in E. coli using regulated asRNA expression.
Keywords: Antibiotic, antisense RNA, Escherichia coli, essential gene, operon
Introduction
During the past few decades, bacterial pathogens have become increasingly resistant to antibiotics, limiting treatment options for infections caused by drug-resistant bacterial pathogens (Boucher et al., 2009). As we face growing antibiotic resistance, the development of novel antibiotics continues to stagnate. Therefore, there is an urgent need for the discovery of new antibacterial agents to target drug-resistant bacteria, especially Gram-negative pathogens (Boucher et al., 2009).
Regulated asRNA expression has been used effectively to study gene functions in different bacterial systems, including Streptococcus mutans (Wang & Kuramitsu, 2005), S. aureus (Ji et al., 2001, Forsyth et al., 2002), and E. coli (Nakashima & Tamura, 2009). By blocking the expression of its target gene, an asRNA increases the sensitivity of bacteria only to specific inhibitors for a protein encoded by that target gene (Forsyth et al., 2002, Young et al., 2006). This differential sensitivity screening assay has been used to validate mechanisms of action for known antibiotics (Forsyth et al., 2002, Ji et al., 2004) and to discover novel antibacterial inhibitors (Young et al., 2006, Wang et al., 2007). Furthermore, hundreds of S. aureus asRNA strains have been configured into a TargetArray which was employed to study mechanisms of action of antibacterial inhibitors (Donald et al., 2009, Xu et al., 2010). Thus, regulated asRNA expression has a great potential for antibiotic drug discovery.
However, the regulated asRNA approach has seen limited success in Gram-negative bacteria, including E. coli. There have been no published reports describing the adoption of the regulated asRNA approach for comprehensive genome-wide essential gene determination and/or silencing in Gram-negative bacteria. It has been recognized that asRNA-mediated down-regulation of gene expression in E. coli is inefficient for reasons not yet clearly understood (Wagner & Flardh, 2002). Attempts to improve the efficiency were rather frustrating initially (Engdahl et al., 2001). Several years ago, a series of expression vectors were designed such that expressed asRNA molecules have paired termini to enhance their stability and hence gene knock-down efficiency in E. coli (Nakashima et al., 2006). In this report, we present a first genome-wide attempt to obtain cell-growth inhibitory E. coli asRNA constructs through phenotypic screening two shot-gun genomic libraries based on a paired-termini expression vector, pHN678 (Nakashima et al., 2006). Our results will stimulate further studies of gene functions, coordinated gene expression on operons and interactions of cellular processes via regulated asRNA in E. coli. Furthermore, the collection of the E. coli asRNA clones generated using this approach will be a valuable tool in the antibiotic drug discovery, especially for therapeutics targeting Gram-negative bacterial pathogens.
Materials and methods
Construction of random genomic libraries
Genomic DNA was extracted from E. coli MG1655 cells (American Type Culture Collection, Manassas, VA) using the Wizard Genomic DNA Purification Kit (Promega, Madison, WI) followed by partial digest with Sau3AI or CviKI-1 (NEB, Ipswich, MA). The resulting DNA fragments (200–800 bp) were purified from agarose gels using the Zymoclean Gel DNA Recovery Kit (Zymo Research, Orange, CA). Plasmid vector was digested with BamHI (if Sau3AI was used to digest the genomic DNA) or SnaBI (if CviKI-1 was used), dephosphorylated using Antarctic Phosphatase (NEB) and then ligated with the inserts using the T4 DNA ligase (Life Technologies, Carlsbad, CA). Ligation mixtures were transformed into E. coli DH5α competent cells (Life Technologies) and plated onto LB agar plates plus 34 μg mL−1 chloramphenicol. Cloning efficiency of the pHN678 library was determined by colony PCR using the following primers: 5′-CGACATCATAACGGTTCTGGCAAAT-3′ (forward) and 5′-GACCGCTTCTGCGTTCTGATTT-3′ (reverse) (Eurofins MWG Operon, Huntsville, AL). The primers were designed so that in the absence of an insert, a 290 bp band should be detected by PCR. Any band larger than this size would indicate the presence of a cloned DNA.
Inducible phenotypic screening for inducer-sensitive clones
Colonies from the random genomic libraries were individually picked with sterile tooth picks, inoculated into wells of 96-well microplates (Corning #3370, Fisher, Pittsburgh, PA) containing LB broth plus chloramphenicol, and grown overnight at 37 °C for 16 hours. Each 96-well microplate was then replica plated onto two sets of Nunc’s Omni Trays (Rochester, NY) using a 96-pin replicator (V&P Scientific, San Diego, CA). Both trays contained LB agar plus chloramphenicol, with one of them supplemented with IPTG (inducing plate). A positive cell clone (PT18, targeting rplF and rpsH genes) was included in each microplate as a positive control. Inducer-sensitive clones were identified via growth defects (lethal or defective growth) present only on the inducing plates. The inducer-sensitivity of these clones was confirmed again prior to plasmid insert sequencing. Each inducer sensitive clone was given a clone number beginning with a prefix PT since the paired-termini vector pHN678 was used. The clone names of Library C clones are affixed with a letter “C” to differentiate from those from the Sau3AI digested library.
Insert DNA sequencing and bioinformatics analysis
Plasmids were isolated from confirmed inducer-sensitive clones and sequenced at Eton Bioscience Inc. (San Diego, CA) to determine the DNA sequences of the inserts and their orientations. The DNA sequences were then compared to the annotated genomic sequence of E. coli MG1655 (Genbank accession number NC_000913) to determine the origin of DNA inserts and their orientation using NCBI BLAST. The essentiality of the corresponding target gene was determined based the Profiling of E. coli Chromosome (PEC) database (http://www.shigen.nig.ac.jp). The operon structure for relevant genes targeted by asRNA was obtained from the RegulonDB (http://regulondb.ccg.unam.mx/) to determine if other essential genes are present in the targeted operon.
Selective asRNA cell sensitization
To quantitatively measure the IPTG-induced growth inhibition in E. coli asRNA cell clones (e.g., fusA cell clone, PT44), seven-point IPTG dose response curves were obtained as described previously (Xu et al., 2006). To determine the initial inducer conditions appropriate for sensitizing asRNA cell clones, IPTG concentrations causing between 70% to 80% cell growth inhibition for asRNA clones were determined. One asRNA clone (PT44) targeting fusA gene (which encodes elongation factor G) was studied in more detail to demonstrate selective cell sensitization. Specifically, an exponential growth culture of PT44 was inoculated into fresh LB broth plus chloramphenicol and appropriate IPTG concentrations (and no IPTG control). The inoculum was combined in a microplate with seven-point serial dilutions of fusidic acid, a known inhibitor of elongation factor G, and cell growth in each well of the microplate was monitored as described previously (Xu et al., 2006). Fusidic acid dose response curves of cell growth under various inducer treatments were generated by using the 8 h time-point optical density data and graphed using Prism software (GraphPad, San Diego, CA). Once the optimal IPTG concentration was obtained, additional cell-based assays were performed against a panel of known antibiotics of different chemical classes to obtain fold sensitization values [the ratio between the IC50 value (the concentration at which cell growth inhibited 50% compared to control) under non-induced condition and that of induced condition].
Results
Construction of random genomic libraries using shot-gun cloning
Based on the result of a time-course Sau3AI digestion (Fig. S1A, Supporting Information), the optimal partial digestion time was 4 hours to generate DNA fragments of appropriate size for library construction. Ligation mixtures were transformed into E. coli DH5α competent cells. Insert cloning efficiency analysis (Fig. S1B, Supporting Information) indicated that the cloning efficiency for this library was 92 %. To increase the randomness of the genomic DNA generated, an alternative genomic library was constructed using a blunt-end producing restriction endonuclease (CviKI-1). The cloning efficiency of the CviKI-1 based library (termed Library C) was 90%.
Screening for inducer-sensitive clones
To screen for inducible growth-inhibitory recombinant clones, transformants were grown overnight with chloramphenicol in the presence or absence of inducing IPTG. An example of screening plates and sensitive clones is shown (Fig. S1C, Supporting Information). A total of 1,500 confirmed IPTG-sensitive clones were obtained from screening 250,000 individual transformants. Only 675 of the 1,500 confirmed clones were sequenced (see below). An example of inducer-dependent inhibition of growth of asRNA clone PT113 is shown in Fig. S1D (Supporting Information).
AS constructs targeting E. coli essential genes
Plasmid DNAs from a total of 675 confirmed inducer sensitive clones were sequenced. It was determined that enough clones were analyzed because more analysis leads to identification of duplicates, suggesting that the phenotypic screening process under the condition scheme is approaching saturation. Among the sequenced clones, 134 separate clones contained insert DNA sequences derived from and in antisense orientation to known essential genes based on PEC database (Table 1). For most of the essential genes targeted by asRNAs, multiple gene-silencing asRNA constructs were discovered, with rplF gene (encoding 50S ribosomal subunit protein L6) being “hit” the most (17 times) (Table 1). Because many essential genes engaging in a cellular process are usually clustered in an operon, many essential operons are targeted by a multitude of asRNAs, especially the operons for ribosomal protein genes. For example, rplN operon which contains 11 essential genes was “hit” by 17 unique asRNAs (Fig. 1A, with 2 asRNAs not shown due to space limit). On an individual gene level, four unique asRNAs were found to target fusA gene (Fig. 1B) while another four target rpoC gene (Fig. 1D).
Table 1.
E. coli genes targeted by asRNA constructs, other essential genes present on the same operons and insert DNA genome coordinates.
| Gene | Product encoded a | Other essential genes within operon b | asRNA clone insert coordinates c, clones and clone types |
|---|---|---|---|
| accC | acetyl-CoA carboxylase, biotin carboxylase subunit | accB | (3404969-3405089): PT312 |
| acpP | acyl carrier protein (ACP) | fabD, fabG | (1150883-1150973): PT1192C, PT1193C |
| bamA | outer membrane protein assembly factor, forms pores; required for OM biogenesis; in BamABCD OM protein complex | lpxD, fabZ, lpxA, lpxB | (2242734-2242870): PT164 |
| fusA | protein chain elongation factor EF-G, GTP-binding | rpsL, rpsG | (3470872-3471062): PT3, PT43, PT499 |
| (3469579-3469688): PT4, PT144, PT212, PT220, PT221, PT226 | |||
| (3470419-3470651): PT44, PT313 | |||
| (3470656-3470867): PT257 | |||
| glnS | glutamyl-tRNA synthetase | none | (705881-706317): PT200 |
| (705276-705473): PT418 | |||
| gltX | glutamyl-tRNA synthetase | none | (2518066-2518186): PT311, PT319, PT320, PT375, PT475, PT497 |
| infB | fused protein chain initiation factor 2, IF2:membrane protein/conserved protein | nusA | (3311410-3311552): PT129 |
| ispB | octaprenyl diphosphate synthase | (3332571-3332694): PT157 | |
| leuS | leucyl-tRNA synthetase | nadD, holA, lptE | (673825-674065): PT7, PT46 |
| parE | DNA topoisomerase IV, subunit B | none | (3172272-3172325): PT1589C |
| pheS | phenylalanine tRNA synthetase, alpha subunit | thrS, infC, rplT, pheM, pheT | (1796927-1797014): PT1157C, PT1173C, PT1561C |
| pheT | phenylalanine tRNA synthetase, beta subunit | thrS, infC, rplT, pheS, pheM | (1794440-1794587): PT79, PT335, PT368 |
| (1794205-1794396): PT153 | |||
| (1794205-1794357): PT412 | |||
| (1794809-1794914): PT1108C, PT1261C | |||
| (1794112-1794204): PT422 | |||
| (1795187-1795315): PT1608C | |||
| prfA | peptide chain release factor RF-1 | hemA, prmC | (1264626-1264716): PT69, PT123, PT147, PT316 |
| (1264548-1264716): PT381 | |||
| rplB | 50S ribosomal subunit protein L2 | rpsJ, rplC, rplD, rplW, rpsS, rplV, rpsC, rplP, rpmC, rpsQ | (3449247-3449437): PT1316C |
| rplC | 50S ribosomal subunit protein L3 | rpsJ, rplD, rplW, rplB, rpsS, rplV, rpsC, rplP, | (3450082-3450342): PT117 |
| (3450200-3450427): PT1363C | |||
| rplD | 50S ribosomal subunit protein L4 | rpsJ, rplC, rplW, rplB, rpsS, rplV, rpsC, rplP, rpmC, rpsQ | (3450082-3450342): PT117 |
| (3450200-3450427): PT1363C | |||
| (3449574-3449772): PT1578C | |||
| (3449913-3450111): PT1603C | |||
| rplE | 50S ribosomal subunit protein L5 | rplN, rplX, rpsN, rpsE, rpsH, rplF, rplR, rpmD, rplO, secY | (3445153-3445259): PT121 |
| (3444964-3445113): PT1254C | |||
| (3444903-3444963): PT1297C | |||
| rplF | 50S ribosomal subunit protein L6 | rplN, rplX, rplE, rpsE, rpsH, rpsN, rplR, rpmD, rplO, secY | (3444089-3444227): PT18, PT35, PT263, PT265, PT266, PT267, PT268, PT269, PT272, PT275, PT276, PT277,, PT355, PT362, PT365 |
| (3443744-3444084): PT45 | |||
| (3443474-3443735): PT1116C | |||
| rplL | 50S ribosomal subunit protein L7/L12 | rplJ, rpoB, rpoC | (4178602-4178650): PT304 |
| rplM | 50S ribosomal subunit protein L13 | rpsI | (3376226-3376437): PT1306C, PT1386C |
| (3376658-3376776): PT1330C, PT1533C | |||
| (3376438-3376588): PT1518C, PT1587C | |||
| rplN | 50S ribosomal subunit protein L14 | rplF, rplX, rplE, rpsE, rpsH, rpsN, rplR, rpmD, rplO, secY | (3446053-3446165): PT55, PT334 |
| rplO | 50S ribosomal subunit protein L30 | rplN, rplX, rplE, rpsN, rpsH, rplF, rplR, rpsE, rplD, secY | (3442251-3442600): PT189, PT293 |
| (3442191-3442269): PT1228C | |||
| (3442537-3442667): PT1280C | |||
| rplQ | 50S ribosomal subunit protein L17 | rpsK, rpsM, rpsD, rpoA | (3437674-3437796): PT91 |
| (3437801-3438054): PT235, PT241 | |||
| (3437565-3437690): PT1367C | |||
| (3437978-3438088): PT1518C, PT1587C | |||
| (3437691-3437966): PT1610C | |||
| rplR | 50S ribosomal subunit protein L18 | rplN, rplX, rplE, rpsN, rpsH, rplF, rpsE, rpmD, | (3443474-3443735): PT1116C |
| (3443279-3443402): PT1385C | |||
| rplV | 50S ribosomal subunit protein L22 | rpsJ, rplC, rplW, rplB, rpsS, rplD, rpsC, rplP, rpmC, rpsQ | (3447932-3448050): PT141, PT142 |
| (3448126-3448303): PT1272C, PT1247C | |||
| rplW | 50S ribosomal subunit protein L23 | rpsJ, rplB, rplC, rplD, rpsS, rplV, rpsC, rplP, rpmC, rpsQ | (3449247-3449437): PT1316C |
| (3449574-3449772): PT1578C | |||
| rplX | 50S ribosomal subunit protein L24 | rplN, rplE, rpsN, rpsH, rplF, rplR, rpsE, rpmD, rplO, secY | (3445539-3445602): PT1602C |
| rpmD | 50S ribosomal subunit protein L30 | rplN, rplX, rplE, rpsN, rpsH, rplF, rplR, rpsE, rplO, secY | (3442605-3442849): PT80, PT479 |
| (3442251-3442600): PT189, PT293 | |||
| (3442537-3442667): PT1280C | |||
| rpoA | RNA polymerase, alpha subunit | rpsM, rpsK, rpsD, rplQ | (3438419-3438661): PT300, PT328 |
| (3438059-3438210): PT326 | |||
| (3438300-3438414): PT379 | |||
| (3437978-3438088): PT1518C, PT1587C | |||
| rpoC | RNA polymerase, beta prime subunit | rplJ, rplL, rpoB | (4186643-4186713): PT11, PT50 |
| (4184366-4184694): PT306 | |||
| (4185255-4185547): PT1130C | |||
| (4184355-4184497): PT1509C | |||
| rpsA | 30S ribosomal subunit protein S1 | none | (961514-961694): PT15 |
| (961506-961694): PT239 | |||
| (962090-962383): PT1408C | |||
| (962523-962628): PT1566C | |||
| rpsB | 30S ribosomal subunit protein S2 | tsf | (190451-190554): PT17, PT172 |
| rpsC | 30S ribosomal subunit protein S3 | rpsJ, rplC, rplD, rplW, rplB, rpsS, rplV, rplP, rpmC, rpsQ | (3447274-3447439): PT1124C |
| (3447272-3447441): PT1259C | |||
| rpsE | 30S ribosomal subunit protein S5 | rplN, rplX, rplE, rpsN, rpsH, rplF, rplR, rpmD, rplO, secY | (3443076-3443203): PT36 |
| (3442605-3442849): PT80, PT479 | |||
| rpsH | 30S ribosomal subunit protein S8 | rplN, rplX, rplE, rpsN, rpsE, rplF, rplR, rpmD, rplO, secY | (3444556-3444785): PT12 |
| (3444089-3444227): PT18, PT35, PT263, PT265, PT266, PT267, PT268, PT269, PT272, PT275, PT276, PT277, PT355, PT362, PT365 | |||
| rpsI | 30S ribosomal subunit protein S9 | rplM | (3376226-3376437): PT1306C, PT1386C |
| rpsK | 30S ribosomal subunit protein S11 | rpsM, rpsD, rpoA, rplQ | (3440016-3440145): PT74 |
| (3439846-3439981): PT1308C | |||
| rpsL | 30S ribosomal subunit protein S12 | rpsG, fusA | (3472339-3472689): PT93 |
| rpsM | 30S ribosomal subunit protein S13 | rpsK, rpsD, rpoA, rplQ | (3440016-3440145): PT74 |
| (3440150-3440288): PT148 | |||
| rpsN | 30S ribosomal subunit protein S14 | rplN, rplX, rplE, rpsE, rpsH, rplF, rplR, rpmD, rplO, secY | (3444556-3444785): PT12 |
| (3444903-3444963): PT1297C | |||
| (3444668-3444857): PT1364C | |||
| (3444766-3444902): PT1534C, PT1549C | |||
| rpsP | 30S ribosomal subunit protein S16 | trmD, rplS | (2743908-2743982): PT84, PT133 |
| (2744137-2744339): PT1293C | |||
| rpsR | 30S ribosomal subunit protein S18 | none | (4423294-4423920): PT113 |
| (4424028-4424446): PT1275C | |||
| rpsS | 30S ribosomal subunit protein S19 | rpsJ, rplC, rplD, rplW, rplB, rplV, rpsC, rplP, rpmC, rpsQ | (3448126-3448303): PT1160C, PT1272C, PT1247C |
| (3448304-3448550): PT1411C, PT1502C | |||
| mutLd | methyl-directed mismatch repair protein | yjeE | (4396833-4396880): PT1540C, PT1542C |
| priBd | primosomal protein N | rpsR | (4423724-4423857): PT1303C, PT1505C |
| rimMd | 16S rRNA processing protein | rpsP, trmD, rplS | (2743908-2743982): PT84, PT133 |
| rplAd | 50S ribosomal subunit protein L1 | rplJ, rplL, rpoB, rpoC | (4176746-4176904): PT1351C |
| rplId | 50S ribosomal subunit protein L9 | rpsR | (4424183-4424555): PT70 |
| (4424028-4424446): PT1275C | |||
| (4424321-4424446): PT1349C | |||
| (4424306-4424446): PT1546C | |||
| (4424462-4424572): PT1579C | |||
| rplKd | 50S ribosomal subunit protein L11 | rplJ, rplL, rpoB, rpoC | (4176746-4176904): PT1351C |
| (4176522-4176719): PT1520C, PT1551C, PT1580C | |||
| rpsOd | 30S ribosomal subunit protein S15 | nusA, infB | (3309412-3309551): PT1499C |
| rpsUd | 30S ribosomal subunit protein S21 | dnaG, rpoD | (3208846-3209066): PT1156C |
| tufAd | protein chain elongation factor EF-Tu (duplicate of tufB) | rpsL, rpsG, fusA | (3469077-3469201): PT32, PT599 |
| (3468846-3469072): PT323 (1 mismatch at 3469061) | |||
| tufB | protein chain elongation factor EF-Tu (duplicate of tufA) | thrU, glyT | (4174117-4174241): PT32, PT599 |
| (4174246-4174472): PT323 |
From PEC database.
From RegulonDB website.
Based on Genbank accession number NC_000913.
Denotes non-essential genes which share operons with other essential gene(s).
Figure 1.
Representative E. coli operons targeted and silenced by asRNAs. Solid dark arrows are essential genes while gray arrows are non-essential genes based on PEC database. All open arrows are asRNAs, each capable of inhibiting the growth of E. coli cells upon induction by 1 mM IPTG. A. Operon rplN contains 11 essential genes and one non-essential gene. Every essential gene is targeted by at least one asRNA construct. Two additional asRNAs are not shown due to space limit. B. Operon rpsL contains three essential genes plus one non-essential gene (tufA). Four and one asRNA constructs were discovered to silence fusA and rpsL genes, respectively. Immediately below the rpsL operon, an unlinked operon containing three non-essential genes (tufB, thrT and tyrU) and two essential tRNA genes (glyT and thrU) was shown to illustrate common nucleotide regions of tufA and tufB genes potentially targeted by two independent asRNA constructs (PT32 and PT323). PT323 was apparently derived from tufB gene since there is one nucleotide mismatch located at genome position 3469061 of tufA gene. C. Operon rpsF showing that five independent asRNA constructs targeting non-essential genes, illustrating their capacity to independently silence the essential gene (rpsR) within the same operon. D. In operon rplK, five asRNA constructs target essential genes while two independent asRNA constructs target non-essential genes. Operon structures were based on information from RegulonDB website.
While plasmids of the majority of the remaining confirmed inducer sensitive clones were found to contain insert sequences corresponding to non-essential genes (data not shown), 18 separate asRNA constructs were identified to have derived their inserts from non-essential genes which share operons with known essential genes (Fig. 1B, 1C and 1D; Table 1). For example, five asRNA constructs, originated from non-essential genes (priB and rplI) of the rpsF operon, can individually inhibit cell growth upon induction (Fig. 1C). Similarly, each of the two asRNA constructs, originated from the non-essential rplK and/or rplA genes, inhibits cell growth when induced by IPTG (Fig. 1D). These results strongly suggest that induced asRNA silences gene expression at the operon level. Interestingly, an asRNA construct (PT32) targets both tufA and tufB genes because the asRNA complements mRNA regions of both genes 100% (Fig. 1B). Additionally, another asRNA construct (PT323) derived its sequences from the tufB gene located on an unlinked operon which also has two essential genes (glyT and thrU) (Fig. 1B). While PT32 could derive its sequences from either tufA or tufB due to identical sequences in the relevant regions, PT323 shares 100% homology with tufB gene but has one mismatch nucleotide with tufA sequence (data not shown). Combining essential genes directly targeted with those indirectly targeted (via operon effects) by asRNA constructs, a total of 79 essential genes can be silenced or knocked down by 152 separate asRNA constructs (Table 1 and Table S1, Supporting Information). It is noteworthy that among these 79 essential genes, 61 (77%) are involved in processes related to protein synthesis (Table S1, Supporting Information).
Selective cell sensitization
To determine if an asRNA clone can be sensitized towards inhibitors targeting the essential protein involved, the fusA asRNA clone (PT44) was studied in cell-based assays. The strategy is to inhibit the function of an essential protein to a threshold level (but without totally shutting down the cells’ growth) whereby the cell is sensitive to any additional assault on the same essential protein. First, a seven-point IPTG dose response curve for PT44 was generated (Fig. 2A) with the IC50 value of 28 μM. Subsequently, the inducer concentrations for sensitizing PT44 clone against fusidic acid (which targets EF-G) were further optimized (Fig. 2B). Our results indicated that at 45 μM IPTG, the asRNA clone exhibits 12 fold increase (IC50 at 0 μM divided by IC50 at 45 μM) in sensitivity to the specific inhibitor (Fig. 2B). The optimized cell-based assay was performed against serial dilutions of 9 other antibiotics (Fig. 2C). Results showed that the fusA asRNA clone was the most sensitive to fusidic acid (12 fold), followed by erythromycin (5 fold) and tetracycline (4 fold), both are well known antibiotics targeting protein synthesis (Fig. 2C).
Figure 2.
Inducer optimization and selective sensitization of fusA targeted PT44 clone. A. Inducer dose response curve for PT44 clone. X axis shows the log values of inducer concentrations while Y axis shows the percent inhibition values (as compared to no-induction cell control) of cells treated under various inducer concentrations shown. B. PT44 cells become more sensitive to fusidic acid as the amount of IPTG increases. X axis shows the log values of inhibitor concentrations. Y axis is the same as in panel A. C. Induction of asRNA to fusA mRNA (in PT44 clone) selectively sensitizes the cells to fusidic acid which specifically targets EF-G encoded by fusA gene. Also included in the assay were 9 other antibiotics targeting essential processes such as cell wall synthesis (cefotaxime), DNA replication (novobiocin), RNA transcription (rifampicin), fatty acid synthesis (triclosan), protein synthesis (streptomycin, tetracycline, kanamycin and erythromycin) and folate synthesis (trimethoprim). For each condition, four replicate data points were obtained per experiment and each experiment was performed three times (n=3) with SD error bars shown.
Discussion
It was recognized that conditional silencing by introduced asRNAs in Gram-negative bacteria is less efficient than in Gram-positive bacteria (Wagner & Flardh, 2002). Specifically, while global essential genes in S. aureus (Ji et al., 2001, Forsyth et al., 2002) and S. mutans (Wang & Kuramitsu, 2005) have been identified by regulated asRNAs, the adoption of such approach in Gram-negative bacteria has not been reported (Good & Stach, 2011). Although the reasons for such discrepancy are not well defined, one possible explanation lies in the reduced stability of plasmid-borne artificial asRNAs in E. coli probably due to the presence of RNase E in this bacterium (Xu et al., 2010), but not in S. aureus. For this reason, Nakashima and colleagues (Nakashima et al., 2006) designed a series of E. coli plasmid vectors which produce RNA molecules with paired-termini to increase the asRNA stability and conditional gene silencing. Targeted antisense fragment cloning using such paired-termini vectors has produced asRNA constructs which have shown to knock-down or silence the expression of a number of essential genes in E. coli (Nakashima et al., 2006). In this communication, we report a genome-wide application of regulated asRNA expression in E. coli using the vector pHN678. Here, we demonstrated that employing this paired-termini vector indeed identified a large number of asRNA constructs targeting E. coli essential genes and, to a lesser extent, some non-essential genes which share operons with essential genes.
While asRNA constructs targeting essential genes of a number of cellular processes in E. coli were identified (Table 1 and Table S1), particularly striking was the observation that the asRNAs predominately silence the expression of essential genes (77% of total genes) involved in protein synthesis processes (tRNAs, tRNA synthetases, transcription, ribosomal proteins and translation factors) (Table S1). We speculate that this bias may have been caused by high basal level (leaky) promoter (Ptrc) activity from the vector in the absence of IPTG (Nakashima & Tamura, 2009) during the library transformation process. It was possible that for those asRNA constructs derived from essential genes which normally are expressed at low levels, even the background level of asRNAs could render the E. coli clones unable to grow into colonies after transformation. In contrast, asRNA clones in which highly expressed genes are being targeted would be able to grow into colonies and selected during the subsequent phenotypic (+IPTG) screens. This hypothesis is supported by data from DNA array based E. coli gene expression profiling (Tao et al., 1999). For example, 53 of the 79 essential genes (67%) targeted by asRNA constructs (Table S1) are within the top 10% highly expressed genes among the 4290 ORFs examined when E. coli cells were grown exponentially in LB broth plus glucose (Tao et al., 1999). To increase the diversity of asRNA clones identified, possible technical improvements include replacing Ptrc with a more stringent promoter element on the cloning vector or employing a number of plasmid vectors each containing a promoter with different range of activities (Xu et al., 2010; Nakashima et al., 2006).
The recovery of 18 asRNA constructs derived from 10 non-essential genes which share operons with essential genes provides strong support for a hypothesis that expressed asRNAs silence gene function in E. coli at the operon level. The mechanism of asRNA inhibition in S. aureus was examined previously by Young and coworkers (Young et al., 2006) who demonstrated that asRNAs exert their inhibition by eliciting degradation of mRNAs upstream (5′) of the regions where the asRNAs bind, which lends support to our hypothesis. If the hypothesis is confirmed, an asRNA construct or synthetic oligonucleotide could inhibit as many as 11 essential genes simultaneously on the rplN operon (Fig. 2A), rendering it difficult for multiple resistant mutations to occur in multiple genes. If such multi-gene mechanism of gene silencing turns out to be prevalent among bacteria, it will facilitate design and development of antisense-based antimicrobial therapeutics which are “polypharmaceutical” (Good & Stach, 2011) or “multi-“targeting” (Silver, 2007): antibiotics (e.g., most of the successful antibiotics in clinical use) target or interact with two or more bacterial target proteins.
In this study, two genomic libraries were constructed successfully and screened for inducible growth-inhibitory asRNA clones. The asRNA constructs discovered could knock down or silence the expression of 79 E. coli essential genes. While the genes being targeted are not yet comprehensive, likely due to a leaky Ptrc promoter of pHN678, this communication represents a first published report to successfully apply regulated asRNA technology to discover E. coli asRNA clones at the genome level. Such conditional asRNA clones will not only stimulate studies of global functions of genes and operons in E. coli but also facilitate discovery and development of novel antimicrobial agents to combat multi-drug resistant pathogens.
Supplementary Material
Acknowledgments
Funding for this project has been provided by NIH grant SC3GM083686 (to H. H. Xu). We greatly appreciate National Institute of Advanced Industrial Science and Technology (AIST) of Japan for kindly sharing plasmid vector pHN678. Technical assistance from Jonathan Chen, Dolly Foti, Hawra Karim, Marcela Arenas, Edie Bucar, Isba Silva, Michael Boateng-Antwi, Miriam Gonzales, Virginia Tan, Alfonso Brito and Marlyn Rios was greatly appreciated. J. Tain was supported by a BioSecurity Scholarship from a Department of Homeland Security grant (2009-ST-062-000018 to H. H. Xu). H. Crummer was supported by a Bridge to the Future program funded by NIH grant 5R25GM049001. All authors have no conflict of interest to declare.
References
- Boucher HW, Talbot GH, Bradley JS, et al. Bad bugs, no drugs: no ESKAPE! An update from the Infectious Diseases Society of America. Clin Infect Dis. 2009;48:1–12. doi: 10.1086/595011. [DOI] [PubMed] [Google Scholar]
- Donald RG, Skwish S, Forsyth RA, et al. A Staphylococcus aureus fitness test platform for mechanism-based profiling of antibacterial compounds. Chem Biol. 2009;16:826–836. doi: 10.1016/j.chembiol.2009.07.004. [DOI] [PubMed] [Google Scholar]
- Engdahl HM, Lindell M, Wagner EG. Introduction of an RNA stability element at the 5′-end of an antisense RNA cassette increases the inhibition of target RNA translation. Antisense Nucleic Acid Drug Dev. 2001;11:29–40. doi: 10.1089/108729001750072100. [DOI] [PubMed] [Google Scholar]
- Forsyth RA, Haselbeck RJ, Ohlsen KL, et al. A genome-wide strategy for the identification of essential genes in Staphylococcus aureus. Mol Microbiol. 2002;43:1387–1400. doi: 10.1046/j.1365-2958.2002.02832.x. [DOI] [PubMed] [Google Scholar]
- Good L, Stach JE. Synthetic RNA silencing in bacteria - antimicrobial discovery and resistance breaking. Front Microbiol. 2011;2:185. doi: 10.3389/fmicb.2011.00185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ji Y, Yin D, Fox B, Holmes DJ, Payne D, Rosenberg M. Validation of antibacterial mechanism of action using regulated antisense RNA expression in Staphylococcus aureus. FEMS Microbiol Lett. 2004;231:177–184. doi: 10.1016/S0378-1097(03)00931-5. [DOI] [PubMed] [Google Scholar]
- Ji Y, Zhang B, Van SF, et al. Identification of critical staphylococcal genes using conditional phenotypes generated by antisense RNA. Science. 2001;293:2266–2269. doi: 10.1126/science.1063566. [DOI] [PubMed] [Google Scholar]
- Nakashima N, Tamura T. Conditional gene silencing of multiple genes with antisense RNAs and generation of a mutator strain of Escherichia coli. Nucleic Acids Res. 2009;37:e103. doi: 10.1093/nar/gkp498. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nakashima N, Tamura T, Good L. Paired termini stabilize antisense RNAs and enhance conditional gene silencing in Escherichia coli. Nucleic Acids Res. 2006;34:e138. doi: 10.1093/nar/gkl697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Silver LL. Multi-targeting by monotherapeutic antibacterials. Nat Rev Drug Discov. 2007;6:41–55. doi: 10.1038/nrd2202. [DOI] [PubMed] [Google Scholar]
- Tao H, Bausch C, Richmond C, Blattner FR, Conway T. Functional genomics: expression analysis of Escherichia coli growing on minimal and rich media. J Bacteriol. 1999;181:6425–6440. doi: 10.1128/jb.181.20.6425-6440.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wagner EG, Flardh K. Antisense RNAs everywhere? Trends Genet. 2002;18:223–226. doi: 10.1016/s0168-9525(02)02658-6. [DOI] [PubMed] [Google Scholar]
- Wang B, Kuramitsu HK. Inducible antisense RNA expression in the characterization of gene functions in Streptococcus mutans. Infect Immun. 2005;73:3568–3576. doi: 10.1128/IAI.73.6.3568-3576.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang J, Kodali S, Lee SH, et al. Discovery of platencin, a dual FabF and FabH inhibitor with in vivo antibiotic properties. Proc Natl Acad Sci U S A. 2007;104:7612–7616. doi: 10.1073/pnas.0700746104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xu HH, Real L, Bailey MW. An array of Escherichia coli clones over-expressing essential proteins: a new strategy of identifying cellular targets of potent antibacterial compounds. Biochem Biophys Res Commun. 2006;349:1250–1257. doi: 10.1016/j.bbrc.2006.08.166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xu HH, Trawick JD, Haselbeck RJ, et al. Staphylococcus aureus TargetArray: comprehensive differential essential gene expression as a mechanistic tool to profile antibacterials. Antimicrob Agents Chemother. 2010;54:3659–3670. doi: 10.1128/AAC.00308-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Young K, Jayasuriya H, Ondeyka JG, et al. Discovery of FabH/FabF inhibitors from natural products. Antimicrob Agents Chemother. 2006;50:519–526. doi: 10.1128/AAC.50.2.519-526.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.