Intercepting biological messages: Antibacterial molecules targeting nucleic acids during interbacterial conflicts

Abstract Bacteria live in polymicrobial communities and constantly compete for resources. These organisms have evolved an array of antibacterial weapons to inhibit the growth or kill competitors. The arsenal comprises antibiotics, bacteriocins, and contact-dependent effectors that are either secreted in the medium or directly translocated into target cells. During bacterial antagonistic encounters, several cellular components important for life become a weak spot prone to an attack. Nucleic acids and the machinery responsible for their synthesis are well conserved across the tree of life. These molecules are part of the information flow in the central dogma of molecular biology and mediate long- and short-term storage for genetic information. The aim of this review is to summarize the diversity of antibacterial molecules that target nucleic acids during antagonistic interbacterial encounters and discuss their potential to promote the emergence antibiotic resistance.


Introduction
Bacteria live in dense polymicrobial communities constantly competing for resources and use either exploitative competition in which molecules like siderophores can be used to improve the acquisition of micronutrients; or interference competition in which cytotoxic molecules are used to inactive target cells (Granato et al., 2019).During evolution, bacteria have evolved a diverse array of weapons to inhibit the growth or kill competitors, which are broadly divided into contactindependent and contact-dependent antagonistic mechanisms (Peterson et al., 2020).These weapons specialized for biological conflicts evolved to target many cellular components essential for life, such as the genetic information flow through the central dogma, the cell wall, membranes, and key molecules like NAD + .As bacteria have been fighting these microscopic battles for millions of years using diverse antimicrobial molecules, it is not a surprise that studies on bacteria preserved in frozen glaciers identified the presence of antibiotic resistance genes that pre-dated human discovery of the first antibiotic (Mindlin and Petrova, 2017).In this review, we will examine molecules such as antibiotics, bacteriocins and effectors produced by bacteria and used during interbacterial conflicts to target DNA and several types of RNAs.We will end by highlighting the underappreciated but important role of these molecules in promoting antimicrobial resistance in natural environments.

Molecules of the central dogma
In molecular biology, the central dogma is an explanation of the flow of genetic information within a biological system.It refers to the information passing from DNA to RNA, and RNA to proteins (Crick, 1970;Morange, 2009).The machinery associated with their synthesis is among the most conserved, and (arguably) important molecules within a living cell.DNA and RNA are polymers of nucleotides, which are composed of a nitrogenous base, a pentose sugar, and a phosphate group (Rich, 1959;Minchin and Lodge, 2019).The bases are either purines (adenine or guanine), or pyrimidines (cytosine and thymine for DNA or uracil for RNA).The nucleotides are connected by phosphodiester bonds between the 5'-phosphate group and the 3'-hydroxyl group, while the bases adenine/thymine (or adenine/uracil for RNA) and guanine/cytosine establish hydrogen bonds (Rich, 1959;Minchin and Lodge, 2019).
DNA replication occurs in a semiconservative manner (Meselson and Stahl, 1958;Hanawalt, 2004).Helicases use energy of ATP hydrolysis to open the double-strand (Abdel-Monem et al., 1976;Oakley, 2019), DNA primases synthesizes RNA primers that will be used by DNA polymerases (Scherzinger et al., 1977;Oakley, 2019), while topoisomerases help in the unwinding process (Wang, 1971).Preservation of the integrity of the genomic information is fundamental for life and there are many DNA repair mechanisms that can either correct errors originated during replication or fix damages induced by external agents (Schärer, 2003).Damaged nucleotides can be repaired by base excision repair (BER) or nucleotide excision repair (NER) (Uphoff and Sherratt, 2017).BER recognizes abnormal bases in the nucleotides along the DNA molecule, such as uracil that spawn from cytosine deamination.BER includes the hydrolyzation of the abnormal base from the nucleotide, followed by the cleavage of the DNA by endonucleases (Uphoff and Sherratt, 2017).Meanwhile, NER removes an entire nucleotide that causes large distortions in the DNA double-helix, and includes the recognition of the lesion by the enzymes UvrA and UvrB, followed by incision at flanking sites of the distortion by UvrC endonuclease and displacement of the damaged strand by UvrD helicase (Uphoff and Sherratt, 2017).After excision from both BER or NER, DNA polymerase I and DNA ligase resynthesize DNA in the gap (Uphoff and Sherratt, 2017).A double-strand break (DSB) can be repaired by homologous recombination (HR) that preserves the previous genetic information or by nonhomologous end-joining (NHEJ), which can lead to the loss or alteration of the original information (Wyman et al., 2004;Shuman and Glickman, 2007).
The information stored in DNA is decoded into RNAs by RNA polymerases (RNAP) (Ebright, 2000).The transcribed RNA could be a transfer RNA (tRNA), a ribosomal RNA (rRNA) or a messenger RNA (mRNA).In bacteria, the 70S ribosome is composed by two subunits: the 30S subunit comprises the 16S rRNA and 21 proteins; while the 50S subunit contains the 23S rRNA, 5S rRNA and 33 proteins (Deutscher, 2009).In several cases, the final step in the expression of the information contained in genes is the synthesis of proteins (Rodnina, 2018), which begins with the association of the ribosome with an mRNA via interaction of the 30S subunit with the Shine-Dalgarno sequence in the mRNA (Shine and Dalgarno, 1974).The elongation follows as the codon in the mRNA is exposed to match the corresponding anti-codon of an aminoacyl-tRNA.The peptidyl transferase center of the ribosome establishes the peptide bond, which is mediated by a catalytic rRNA (Monro, 1967).Overall, fidelity and effectiveness of these steps are required for the maintenance of genetic information and its transfer into molecules that perform work inside living cells.

Bacterial antagonistic mechanisms
Bacteria inhabit complex environments where they interact and compete with other organisms, both prokaryotic and eukaryotic.Several systems specialized in biological conflict, both defensive and offensive, emerged during evolution to combat competitors, predators, and parasites (Figure 1).These systems participate in an arms race in which their genes have a high rate of evolution.Probably the most well-known antibacterial molecules are antibiotics, which are produced by a variety of organisms (Berdy, 2005).Antibiotics are bioactive secondary metabolites not synthesized by ribosomes (Berdy, 2005).They belong to different classes, usually based on their molecular strutures, and target several metabolic processes, including those related to the central dogma (Etebu and Arikekpar, 2016).These molecules are produced and secreted in the extracellular environment by ATP-binding cassette (ABC) transporters (Méndez and Salas, 2001).Producing-bacteria are protected from antibiotics by different mechanisms, including the synthesis of efflux pumps or specific enzymes that degrade/modify the antibiotic or its target (Darby et al., 2022) (Figure 1).
Bacteriocins are another type of biomolecule used in antagonistic encounters that are synthesized by ribosomes and can be divided into colicins and microcins (Cascales et al., 2007).Colicins are larger bacteriocins (>10 kDa) secreted by a diversity of bacteria, and Escherichia coli was the first and most extensively studied.Colicins have three domains: an N-terminal translocation domain, a central receptor-binding domain and a toxic C-terminal domain (Cascales et al., 2007).These proteins are released in the medium and are internalized by binding to specific outer membrane receptors.Colicin-producers encode immunity proteins that bind to the toxic domains to neutralize their effect (Cascales et al., 2007).The expression of these proteins is largely regulated by the SOS response to DNA damage (Walker, 1996;Cascales et al., 2007).Microcins consist of smaller polypeptides (<10 kDa) that require post-translational modification prior to secretion.Microcins target closely related species via binding to outer membrane receptors, and immunity is conferred either by a specific protein that interacts with the microcin or by efflux pumps (Duquesne et al., 2007) (Figure 1).
Many types of macromolecular complexes, named protein secretion systems, are key players in bacterial antagonist interactions (Klein et al., 2020).These include the T1SS, T4SS, T5SS and T6SS of Gram-negative bacteria and T7SS of Gram-positives (Figure 1) (Klein et al., 2020).The T1SS uses glycine-zipper proteins that form large aggregates in the producer outer membrane and kill target bacteria upon contact (García-Bayona et al., 2017).The bacteria killing T4SS apparatus is evolutionarily related to the conjugative machinery and relays on the coupling protein VirD4 for effector selection and translocation into competitors through an extracellular pilus (Souza et al., 2015).A subtype of T5SS mediating contact-dependent growth inhibition (CDI) is composed of two proteins, an outer membrane protein CdiB that anchors an exoprotein with a central receptor-binding and a C-terminal toxic domain (CdiA), which interacts with an outer membrane receptor at a target cell to deliver the toxic C-terminus (Aoki et al., 2005).The T6SS is a contractile nanomachine evolutionarily related to bacteriophage tails that fire an array of effectors inside target cells at each contraction event (Hood et al., 2010;Basler, 2015).The T7SS secretes effectors with an LXG N-terminal and C-terminal toxic domains and participates in bacterial competition in Grampositives (Cao et al., 2016).The vast array of macromolecules specialized in interbacterial conflicts reinforce their importance for bacterial fitness.
The protein complexes described above only mediate the secretion/translocation of the real key players in bacterial antagonism: the toxic molecules used to poison targets cells.In bacteria, there are two main types of toxic molecules: proteinaceous and small molecules (Ruhe et al., 2020).Proteinaceous antimicrobials contemplate ribosomesynthesized molecules, such as bacteriocins and effectors (Ruhe et al., 2020), while antibiotics are synthetized via the secondary metabolism (Walsh, 2016).Many effector proteins contain multiple domains, usually a conserved N-terminus that engage in protein export that varies according to the secretion system it is associated with (Ruhe et al., 2020); and a variable C-terminus that contains the toxic domains (Zhang et al., 2012;Ruhe et al., 2020).Effectors with this configuration are commonly known as polymorphic toxins (Zhang et al., 2012;Ruhe et al., 2020).Next, we will discuss these two main types of antibacterial molecules.

Proteinaceous antimicrobials targeting nucleic acids
A large variety of DNase and RNase domains have been predicted by in silico analysis of polymorphic toxins (Zhang et al., 2012).Most DNase effectors experimentally characterized to date belong to the His-Me finger superfamily (Pfam CL0263) or to the PD-(D/E)xK superfamily (CL0236).On the other hand, RNase effectors are more diverse and belong to the colicin D/E5 (CL0640), Ntox28 (PF15605), EndoU (CL0695) and PD-(D/E)xK (Table 1, Figure 2).

His-Me finger superfamily
The most representative superfamily of DNases is the His-Me finger, also known as HNH superfamily, named after the first characterized enzyme showing the conserved His-Asn-His residues (Wu et al., 2020).This superfamily is defined by the compact catalytic conserved ββα-fold, consisting of a β-hairpin followed by an α-helix in which a highly conserved histidine (H) is located at the end of the first β-strand and a metal-binding conserved residue in α-helix (Zn 2+ or Mg 2+ ) (Wu et al., 2020), thus the name His-Me finger.His-Me finger is thought to mediate nonspecific DNA cleavage, with the α-helix fitting into the DNA minor groove, which aligns the β-hairpin with the DNA phosphodiester backbone (Flick et al., 1998).For cleavage, the metal ion destabilizes the scissile phosphodiester and neutralize the negatively charged transition state (Maté and Kleanthous, 2004).The conserved H residue then activates a water molecule for a nucleophilic attack on the scissile phosphate to hydrolyze the bond (Yang et al., 2011).Even though the amino acid sequences of members of this superfamily are incredible variable, the compact ββα-fold and catalytic mechanism is well conserved (Jablonska et al., 2017;Wu et al., 2020).This fold is present in all kingdoms of life, and in bacteria the enzymes have variable functions spanning from genome maintenance to host defense and target offense (Wu et al., 2020).

PD-(D/E)xK superfamily
The PD-(D/E)xK superfamily is the second most abundant among bacteriocins and effectors with nuclease activity.Like the His-Me finger, proteins belonging to PD-(D/E)xK share small amino acid sequence similarity, but present conserved secondary structure signatures (Steczkiewicz et al., 2012).The conserved fold of this group comprise an α-helix followed by three antiparallel β-strands and a second α-helix followed by a final β-strand (αβββαβ) (Steczkiewicz et al., 2012).The catalytic residues are located in the second and third β-strand; the first α-helix has structural role and is related to the formation of the active site, while the second α-helix is involved in substrate binding (Wah et al., 1998).Conserved aspartic acid and glutamic acid (D/E) residues coordinate the metal ion (usually Mg 2+ ), while the conserved lysine (K) associates with a water molecule to hydrolyze the phosphodiester bond (Kelly et al., 2007).This superfamily includes enzymes related to DNA metabolism (Steczkiewicz et al., 2012).
The first PD-(D/E)xK effector was described in P. aeruginosa (TseT) and contains a Tox-REase-5 domain (PF15648) (Zhang et al., 2012;Burkinshaw et al., 2018).Homologs of TseT have been characterized in B. gladioli (TseTBg1 and TseTBg2) (Yadav et al., 2021), and degrade both DNA and RNA (Yadav et al., 2021).Interestingly, the DNase activity of TseTBg was affected by methylation.A DNA methylase (Dam BG ) is encoded next to the effector, and plasmids isolated from Dam BG -producing E. coli were not degraded by TseTBg1 or TseTBg2 (Yadav et al., 2021).In addition, point mutations in conserved aspartic acid (D) and lysine (K) of TseTBg1 and TseTBg2 abrogated DNase activity (Yadav et al., 2021).Another curiosity is that these effectors are encoded next to two cognate immunity proteins: one of them neutralizes the enzymatic activity in vitro while the second directly binds to the promoter region of the effector, acting as a transcriptional repressor (Yadav et al., 2021).
Enzymatic assays also showed the ability of additional PD-(D/E)xK superfamily members to degrade DNA in vitro.These include PoNe (Jana et al., 2019), RhsB (Pei et al., 2022), and IdrD (Sirias et al., 2020).Moreover, the T5SS effectors CdiA 2 -CT Ab30011 from A. baumannii (Roussin et al., 2019) and CdiA-CT E479 from B. pseudomallei (Nikolakakis et al., 2012) were experimentally shown to degrade nucleic acids, leading to cell growth arrest.The first induces target cell DNA damage, while the second is specific to tRNA Arg (Nikolakakis et al., 2012;Roussin et al., 2019).In summary, similar to the His-Me finger representatives, PD-(D/E)xK members can target both DNA and RNA molecules.

Colicin D/E5 superfamily
The first member of the Colicin D/E5 clan (CL0640) was isolated from E. coli and named colicin D (Timmis and Hedges, 1972).Later, a second member of this clan was identified in Shigella sonnei and called colicin E5 (Males and Stocker, 1982).This protein is homologous to colicin E3 in the receptor-binding and translocation domains but shows a distinct toxic domain (Yajima et al., 2006).Both colicin E5 and colicin D were shown to be ribonucleases that target tRNAs and cleave anticodon loops between the 34 and 35 nucleotides of queuine-containing tRNAs, and between the 38 and 39 nucleotides of tRNAs Arg , respectively (Ogawa et al., 1999;Tomita et al., 2000;Masaki and Ogawa, 2002).The catalytic domain found in these colicins were grouped with other metal-independent RNases as part of the BECR-fold (Barnase-EndoU-ColicinD/E5-RelE), which contain a similar structure composed of a α-helix and an anti-parallel β-sheet formed by four strands (Zhang et al., 2012).In colicin D, a large positively charged surface promotes tRNA binding and brings the anticodon loop close to a histidine residue located at the α-helix (His 611 ), which carries the catalytic function by acting as a general base (Yajima et al., 2004).Colicin E5 possesses a positively charged cleft that promotes RNA docking (Lin et al., 2005) and targets tRNA His , tRNA Tyr , tRNA Asn and tRNA Asp between their modified queuine nucleotide Q34 and U35 (Ogawa et al., 1999).The catalytic residues that participate in E5 enzymatic activity do not include a catalytic histidine that usually participate in RNA cleavage (Lin et al., 2005;Yajima et al., 2006), but instead residues R33 and K25 act as acid-base pairs (Inoue-Ito et al., 2012).
Besides colicin D and E5, other bacterial effectors have been described to belong to this clan (Table 1, Figure 2).Pyocin S4 from P. aeruginosa (Parret and De Mot, 2000) and klebicin D from K. pneumoniae (Chavan et al., 2005) have C-terminal domains that belong to the colicin D/E5 superfamily, and carocin S2 from P. carotovorum has ribonuclease activity in vitro (Chan et al., 2011).The CDI system has a variety of effectors that belong to this clan.The CdiA-CT EC869 and CdiA-CT EC3006 from E. coli are tRNases that have a different cleavage site located at the tRNA acceptor stem (Willett et al., 2015;Jones et al., 2017;Gucinski et al., 2019), the same is observed for CdiA-CT Kp342 from K. pneumoniae (Gucinski et al., 2019).CdiA-CT K96243 and CdiA-CT E478 from B. pseudomallei present the same activity as colicin E5 (Aoki et al., 2010;Nikolakakis et al., 2012).In summary, members of the colicin D/E5 superfamily target tRNA by cleaving at distinct sites.

EndoU superfamily
EndoU RNases comprise nucleases from eukaryotic and viral RNA-processing enzymes (Zhang et al., 2011) and polymorphic bacterial toxins (Zhang et al., 2012).As the letter "E" in the BECR fold, EndoU toxins are metal-independent ribonucleases that contain the typical four stranded β-sheet next to a α-helix structure (Zhang et al., 2012), and are predicted to have ribonuclease activity carried out by two histidine residues (Zhang et al., 2011;Michalska et al., 2018).This superfamily has been described to be related to Ribonuclease A (Mushegian et al., 2020).
Four EndoU antibacterial toxins were verified experimentally, and the results showed that this fold presents some diversity in its mode of action.The T7SS effector BC_0920 from Bacillus cereus has RNase activity (Holberger et al., 2012).MafB MGI-1NEM8013 , an outer membrane exported toxin from Neisseria meningitidis, is a nonspecific ribonuclease with a preference for urydilates (Jamet et al., 2015).CdiA-CT STECO31 , a T5SS secreted toxin from E. coli (Michalska et al., 2018), presents a specific cleavage site at the anticodon loop of tRNA Glu ; while CdiA-CT GN05224 from Klebsiella aerogenes shows tRNase activity in vivo (Michalska et al., 2018).
Even though bioinformatic analysis can broadly predict protein function, the precise mode of action of each nuclease within a superfamily requires empirical biochemical assays to accurately determine activity.

Other nuclease domains
Besides the nuclease groups mentioned above, other domains can be found in bacteriocins and effectors.Tde1 and Tde2 (type VI DNase effectors) from Agrobacterium tumefaciens have a Ntox15 domain (Zhang et al., 2012;Bondage et al., 2016), which is a polymorphic toxic domain characterized by an all α-helical fold and conserved HxxD catalytic residues (Zhang et al., 2012).Both effectors display DNase activity (Bondage et al., 2016).Several WapA proteins from B. subtilis display tRNAse activity, such as WapA-CT 168 , WapA-CT natto and WapA-CT T-UB-10 ; however, the toxic domains remain undetermined (Koskiniemi et al., 2013).In addition, Wap-CT PY79 was hypothesized to display tRNAse activity based on sequence similarity (Stempler et al., 2017).
A recently discovered effector with no detectable domain and DNase activity is Tce1 (T6SS contact-independent antibacterial effector 1) from Yersinia pseudotuberculosis (Song et al., 2021).Tce1 is a Ca 2+ -and Mg 2+ -dependent enzyme that displays an interesting mechanism of target-cell delivery, which can be either dependent or independent of contact (via the outer membrane receptors BtuB and OmpF) (Song et al., 2021).
Also recently, new polymorfic toxin C-teminal domains (PTs) were described (Nachmias et al., 2022).The toxic domains of PT1 and PT7 were shown to be non-specific DNases that did not show sequence or structural similarity to any known nuclease (Nachmias et al., 2022).PT1 is likely secreted by the T6SS, while PT7 is probably secreted via the T7SS (Nachmias et al., 2022).
Other toxins with undetectable domains but with experimentally characterized nuclease activities comprise carocin S1 and S3 from P. carotovorum (Chuang et al., 2007;Wang et al., 2020), pyocin S3 from P. aeruginosa (Duport et al., 1995), and the T6SS effector Hcp-ET3 from E. coli (Ma et al., 2017).The characterization of these and other new toxic domains is an interesting source of information to the discovery of novel enzymatic activities.

Deaminases
Deaminases are enzymes that induce the deamination of nucleotides and are related to salvage pathways of purines and pyrimidines (Nygaard, 1993).Several deaminase domains have been predicted in polymorphic toxins (Iyer et al., 2011;Zhang et al., 2012).The first characterized T6SS deaminase effector was DddA (dsDNA deaminase toxin A) from Burkholderia cenocepacia (Mok et al., 2020).DddA promotes deamination of cytosine and its conversion to uracil in dsDNA, leading to a DNA mismatch during replication that needs to be repaired by the base excision repair (BER) pathway (Uphoff and Sherratt, 2017;de Moraes et al., 2021).An example of deaminases targeting ssDNA is the T6SS effector SsdA (ssDNA deaminase toxin A) from Pseudomonas syringae, which deaminases cytosine into uracil (de Moraes et al., 2021).Sublethal doses of DddA are related to an increase in the frequency of mutations, with a preference for C/G to A/T substitutions (Mok et al., 2020;de Moraes et al., 2021).The action of these mutagenic effectors can promote antibiotic resistance in natural settings (de Moraes et al., 2021).

ADP-ribosyltransferases
ADP-ribosyltranferases (ARTs) are enzymes able of transferring an ADP-ribose from the cofactor β-nicotinamide adenine dinucleotide (NAD + ) into certain targets, which could be either amino acids or nucleotides (Mikolčević et al., 2021).In bacteria, many ARTs are virulence factors involved in pathogenesis that modify specific host cell proteins to manipulate cellular functions (Yoshida and Tsuge, 2021).These ARTs can be classified into two families: diphtheria toxin (DTX) with the conserved residues H-Y-E, and cholera toxin (CTX) with the conserved residues R-S-E (Mikolčević et al., 2021).
Among the weapons used in interbacterial antagonism, Tre23 (type VI secretion ADP-ribosyltranferase effector 23) from Photorhabdus laumondii is an ART from the H-Y-E clade that transfers ADP-ribose to 23S rRNA (Jurėnas et al., 2021).This modification occurs at the 23S rRNA GTPase-associated site of the ribosome, which is necessary for elongation during translation, thus stopping protein synthesis (Jurėnas et al., 2021) (Figure 2).Another RNA modifying toxin is RhsP2 from P. aeruginosa (Bullen et al., 2022).Interestingly, this enzyme displays the conserved residues Y-E and E from the two DTX and CTX ART families (Bullen et al., 2022).RhsP2 ADP-ribosylates a series of non-coding RNAs in target cells, including 4.5S rRNA, 6S rRNA, tRNAs, hindering multiple essential pathways (Bullen et al., 2022).Thus, ART toxins provide another layer of antagonistic strategies that bacteria use to interfere with molecules of the central dogma.

Antibacterial small molecules targeting nucleic acids
Bacteria produce several classes of antibiotics that target nucleic acids, such as aminoglycosides, tetracyclines and macrolides (Table 1, Figure 2).The structural diversity of these molecules provides distinct opportunities for inhibition of the information flow thought the central dogma.Some antibiotics can induce DNA cleavage, inhibit DNA gyrases/ topoisomerases or RNA polymerases, or bind to ribosomal RNAs to interfere with protein synthesis.
Among the antibiotics that induce DNA cleavage there are bleomycins, calicheamicin and daunorubicin.The bleomycin group comprises bleomycins, phleomycins, tallysomycin and zorbamycins (Hecht, 2000).Bleomycins are glycopeptides first isolated from Streptomyces verticillus (Umezawa et al., 1966) that promote oxidative cleavage of double-strand DNA in a sequence-specific manner (Takeshita et al., 1978;Kross et al., 1982).These antibiotics rely on the presence of molecular oxygen and a redox active metal like Fe 2+ or Cu + (Burger et al., 1981;Hecht, 2000).Bleomycins are composed of four functional domains: metal-binding, DNA-binding, linker region connecting the two previous domains, and a disaccharide moiety that promotes cell selectivity (Boger and Cai, 1999).The metal-binding domain is responsible for the specificity of DNA sequence (Sugiyama et al., 1986), which consists mainly of GT dinucleotides but can also be GC and AT (Kross et al., 1982).Phleomycins, tallysomycins and zorbamycins have slightly different sequence specificity but cleave DNA in a similar mechanism (Kross et al., 1982).Calicheamicin belongs to the enediynes group of antibiotics and was first isolated from Micromonospora echinospora ssp.calichensis (Zein et al., 1988).It promotes double-strand DNA cleavage in a sequence-specific manner, preferentially at AGGA, TCCT and ACCT (Zein et al., 1988).The mechanism of cleavage requires the removal of hydrogen atoms (abstraction) from the DNA backbone (Lee et al., 1991).Daunomycin from Streptomyces peucetius can intercalate and form complexes with DNA, leading to chromosome fragmentation (Marco et al., 1975).
Some antibiotics promote DNA degradation by arresting topoisomerases.Type II topoisomerases function by promoting metal-dependent DNA double-strand breaks, followed by ATP-dependent translocation of DNA segments and rejoining the separated DNA ends (Gentry and Osheroff, 2013).The DNA gyrase and topoisomerase IV (topo IV) are type II topoisomerases found in bacteria and are composed of two domains: GyrA and GyrB, and ParC and ParE, respectively (Levine et al., 1998).The GyrA or ParC domains interact with DNA, while GyrB or ParE bind and hydrolyze the ATP necessary for enzymatic function (Levine et al., 1998).Some groups of antibiotics bind to the ATP-binding site of GyrB and ParE to inhibit the activity of the topoisomerase complex, thus generating DNA breaks and the collapse of the replication fork (Anderson et al., 2000;Maxwell and Lawson, 2003).These antibiotics comprise coumarins and cyclothialidines from Streptomyces spp.(Goetschi et al., 1993;Oblak et al., 2007), kibdelomycin from Kibdelosporangium sp.(Phillips et al., 2011), and amycolamicin from Amycolatopsis sp.(Sawa et al., 2012).
Transcription is another seductive target for antibacterial natural products.Rifamycin from Amycolatopsis rifamycinica (Sensi, 1959) is a macrolide antibiotic that blocks transcription by binding to the β-subunit of the RNA polymerase, thus stopping DNA-dependent RNA synthesis via transcript elongation arrest (Campbell et al., 2001;Floss and Yu, 2005).
The ribosome is the center of protein synthesis.It is a large ribonucleoprotein complex composed of two subunits (30S and 50S) forming the 70S bacterial ribosome.The 30S subunit contain the 16S rRNA, while the 50S subunit contain the 23S rRNA and 5S rRNA (Deutscher, 2009).These nanomachines are one of the favorite targets when it comes to bacterial growth inhibition by antibiotics.Most of these antibacterial molecules inhibit ribosome activity by binding directly to the rRNAs and arresting translation by acting as allosteric inhibitors.Here we focused only on antibiotics produced by bacteria that interfere with protein synthesis by binding to rRNAs.
The 50S subunit is also widely affected by antibiotics.Erythromycin, lincomyicin, blasticidin, viomycin and capreomycin target the 23S rRNA.Antibiotics from the macrolide class are produced by diverse Actinomycetes (Dinos, 2017) and can bind to the 23S rRNA at the nascent peptide exit tunnel (Schlünzen et al., 2001;Vázquez-Laslop and Mankin, 2018).Lincomycin from Streptomyces lincolnensis (Mason et al., 1962) binds to the peptidyl transferase cavity at the ribosomal A site (Douthwaite, 1992).Blasticidin S from Streptomyces griseochromogenes (Takeuchi et al., 1958) and sparsomycin from Streptomyces sparsogenes (Owen et al., 1962) bind to the 23S rRNA at the ribosomal P site (Johnston et al., 2002;Hansen et al., 2003).Tuberactinomycins, such as capreomycin from Streptomyces capreolus (Herr Jr and Redstone, 1966) and viomycin from Streptomyces puniceus (Finlay et al., 1951), can interact with both 30S and 50S ribosomal subunits by binding to 16S rRNA at helix 44 and to 23S rRNA at helix 69 (Johansen et al., 2006).In summary, antibiotics collectively work in several steps to prevent the information flow through the central dogma.

Contribution to the development of antibiotic resistance
During the evolutionary arms race in which bacteria developed several weapons to inactivate or kill competitors, immunity mechanisms to prevent self-intoxication and protect sister-cells evolved concomitantly.For proteinaceous antibacterial molecules like effectors and bacteriocins, the expression of a specific immunity protein is usually the most common mechanism of defense (Zhang et al., 2012;Ruhe et al., 2020).For small molecules like antibiotics, there are several mechanisms that could render a cell resistant: (1) target modification by specific enzymes; (2) target bypass via mutations in the targets that lead to reduced affinity; (3) degrading or modifying proteins that act on the molecules; (4) reduced intake via altered membrane permeability; (5) efflux pumps that export the molecules (Darby et al., 2022).
During interbacterial competitions, effectors and bacteriocins that target the DNA contribute to the emergence of antibiotic resistance by increasing the rate of mutagenesis in cells that receive a sublethal dose.The deaminase T6SS effector DddA has been shown to increase the rate of C/G to T/A mutation, leading to emergence of rifamycin resistance by introducing point mutations in the rpoB gene, which encodes the β-subunit of RNA polymerase (de Moraes et al., 2021).In addition, cleavage of the 16S rRNA by colicin E3 promotes faster tRNA-mRNA translocation in ribosomes, thus making it less sensitive to inhibition by the antibiotic viomycin (Lancaster et al., 2008).
In general, DNA damage induced by bacteriocins or effectors activate the SOS response, which can induce the activation of the translesion DNA repair pathway and promote mutations (Patel et al., 2010).The mutagenesis can also be responsible for altering gene expression or characteristics of membrane channels important for antibiotic internalization (Livermore, 1990).Mutations in the promoter region of OmpF (outer membrane protein F) leads to its downregulation, thus conferring β-lactam resistance in E. coli (Delcour, 2009).Similarly, point mutations in OmpF in Enterobacter aerogenes reduce outer membrane permeability and promote resistance to β-lactam antibiotics, which act by inhibiting peptidoglycan synthesis (Dé et al., 2001).
In addition to contributing to an increase in the mutation rate of target cells, antibacterial molecules (e.g., lipases and peptidoglycan hydrolases) can promote the lysis of target cells and the release of extracellular DNA, which could be uptaken by the attacker bacterium and incorporated into its genome, thus stimulating horizontal gene transfer and the spread of genes encoding antibiotic resistance.Examples of this include the T6SSs of Vibrio cholerae and Acinetobacter baylyi (Borgeaud et al., 2015;Cooper et al., 2017;Ringel et al., 2017).Curiously, V. cholerae have its T6SS gene cluster under the control of competence regulators (Borgeaud et al., 2015), demonstrating the relationship between the bacterial competition and horizontal gene transfer events.

Perspectives
Nucleases are possibly the most ancient biological weapons and likely used in periods prior to the development of individual cells surrounded by membranes.Their activities are among the chemical armaments used in biological conflicts across all organizational levels.For example, endonuclease domains of the His-Me superfamily are found in nucleic acid-degrading snake toxins, bacterial polymorphic toxins, bacterial restriction-modification systems conferring antiviral immunity, and eukaryotic apoptosis systems (Zhang et al., 2012;Trummal et al., 2014;Jablonska et al., 2017).There is still a wide array of predicted nucleic acids-targeting enzymes that require further empiral characterization.While it is possible to extropolate the possible activities of predicted groups based on similarities to known enzymes, such as Ntox18, Ntox19, Ntox22 and Ntox30 that are expected to be metal-independent RNases (Zhang et al., 2012), there are Ntox groups for which the nature of catalysis could not be predicted (Zhang et al., 2012).
The large number of antibacterial molecules targeting the central dogma and the number of resistance mechanisms promoting immunity to these molecules, call our attention to the fact that antibiotic resistance is an ancient and naturally occurring phenomenon widespread in the environment.It is important to note that these molecules attacking the central dogma act as part of a miscellaneous arsenal of toxins that damage other cellular components and their combined effect dictates the aftermath of antagonistic interactions.Experimental data confirmed that antibiotic resistance can arise solely by competitive interactions between bacteria without previous antibiotic exposure (Koch et al., 2014).Bacteria joined an arms race millions of years prior to the discovery of antibiotics and studying the mechanisms and outcomes of antagonistic interaction might help us anticipate the emergence of antibiotic resistance in different settings.

Figure 1 -
Figure 1 -Antagonistic strategies used by bacteria to counteract competitors.(A) Contact-independent antagonism.Colicins, microcins and antibiotics (red hexagon) reach targets by binding to OMRs (outer membrane receptors) prior to internalization.Autointoxication is prevented by immunity proteins, degrading/modifying proteins or efflux pumps (blue circles).Outer membrane vesicles (OMVs) deliver toxins to competing bacteria by membrane fusion.(B) Contact-dependent antagonism.T5SS presents CdiB anchored in cell membrane and CdiA extended.Receptor-binding domain (RBD) of CdiA interacts with OMR of targets to translocate CdiA-CT (red) into competitors.T6SS is anchored in the cell membrane and upon contraction propelled into target cell to deliver toxins (red hexagon).T7SS effectors (red hexagons) secreted into target cells upon contact.Outer membrane exchange (OME) events can transfer toxic proteins (red hexagons) that reach targets.Nanotubes are membrane extensions that connect two bacteria to transport toxins (red hexagons).Cognate immunity proteins produced by attacking bacteria are represented by blue circles.Created with BioRender.com.

Figure 2 -
Figure 2 -Antibiotics, bacteriocins, and effectors targeting nucleic acids.Schematic representation of the information flow through the molecules of the central dogma (DNA, RNA and protein).Antibiotics, bacteriocins, and contact-dependent effectors targeting nucleic acids either by binding and inhibition or by enzymatic cleavage are indicated.Molecules were grouped according to their protein domains: His-Me finger (green), PD-(D/E)xK (blue), Colicin D/E5 (light grey), E3-rRNAse (dark grey), antibiotics (orange), others (light red).The complete list of molecules is described inTable 1. Created with BioRender.com.

Table 1 -
Antibacterial molecules that target nucleic acids.