- The Serendipitous Discovery of Sulfurized Linkages in DNA
- These Unique Linkages Are Now Thought to Influence Bacterial Fitness
- Phosphorothioated Bacterial DNA is Found in Human Microbiomes
Post-transcriptional sulfur modifications in transfer RNAs (tRNAs) are very diverse and are present in all domains of life. They are known to be important for proper tRNA function, and include 2-thiouridine (s2U) derivatives, 4-thiouridine (s4U), 2-thiocytidine (s2C), and 2-methylthioadenosine (ms2A). By contrast, sulfur-modified bases have not yet been discovered in messenger RNAs (mRNAs), and sulfur-modified linkages have not been discovered in RNA either, according to The Modomics RNA Modification Database. This also applies to DNA bases, but in 2005, Zhou et al. did discover sulfur-modified phosphorothioate linkages in Streptomyces lividans.
Sulfur-modified phosphorothioate linkages in DNA with Sp and Rp stereochemistry in DNA. Taken from commons.wikimedia.org and free to use.
This blog briefly recaps the serendipitous circumstances in which this stunning find was made, while also providing a synopsis of follow-on investigations on the biosynthesis of these puzzling linkages in DNA and their suggested functional significance.
Serendipity is defined as the occurrence and development of events by chance in a happy or beneficial way. It applies to the discovery of sulfur-modified phosphorothioate linkages in bacterial DNA, dubbed phosphorothiated (PT) DNA, in the beneficial sense, as this chance finding has expanded the molecular scope of epigenetics.
The discovery of PT-DNA in Streptomyces lividans (S. lividans, pictured here) reported in 2005 by Zhou et al. actually dates back to a 1988 report by these investigators, in which reparations of both plasmid and chromosome S. lividans DNA were partially degraded as smears rather than sharp bands in agarose gels during electrophoresis, as exemplified here. Although the origin of this phenomenon of DNA degradation, termed Dnd, was unknown at the time, the report stated that “[e>
xperiments to determine the nature of the Dnd modification in S. lividans and its sequence specificity are in progress”.
Microscopic image provided by the Centers for Disease Control and Prevention (CDC) of a slide culture of a Streptomyces sp. grown on tap water agar. Because S. lividans produces a large amount of secretory proteins, it has been widely used in the production and secretion of proteins of industrial interest. The Streptomyces genus are soil-dwelling filamentous bacteria, and their natural habitat has allowed them to produce most known natural antibiotics, as well as many other secondary metabolites and secreted enzymes that are important for economic and industrial use. Taken from commons.wikimedia.org and free to use.
This experimental determination proved to be quite difficult, as it took the next 17 years to obtain the compelling data that Zhou et al. published in 2005 in an article tilted A novel DNA modification by sulfur. The Dnd phenotype found in S. lividans DNA was associated with an 8,026-bp region comprised of five open-reading frames (ORFs) designated as dndA–E in a cluster named dnd locus. Analysis of these ORFs led to the prediction of five encoded proteins composed of anywhere from 126 amino acid residues (dndE) to 663 residues (dndC).
Protein-sequence homology searches suggested that the dnd cluster led to incorporation of sulfur or a sulfur‐containing substance into the genome of S. lividans, albeit in an unknown manner. Confirmation of sulfur incorporation involved bacterial propagation in minimal medium containing 35SO4 with later analysis of genomic DNA samples on an agarose gel, which demonstrated that 35S was associated with the high molecular weight DNA of wild‐type S. lividans, but not with a mutant lacking the dnd gene cluster.
|Main isotopes of sulfur (16S). Click here for details.|
abun-dance half-life (t1/2) mode pro-duct
32S 94.99% stable
33S 0.75% stable
34S 4.25% stable
35S trace 87 days β- 35Cl
36S 0.01% stable
Wang et al. published further studies in 2007 that involved 35S-labeling of genomic DNA with L-[35S>
-cysteine in bacteria harboring the dnd gene cluster, followed by digestions with nuclease P1 (Sp-selective) or snake venom phosphodiesterase (Rp-selective) and alkaline phosphatase (see Eckstein). The resultant mixtures were separated by reversed-phase HPLC with UV detection and scintillation counting, which led to identification of the phosphorothioate (PS)-containing dinucleotide 5’-d(GPSA)-3’ as having Rp stereochemistry (see above depiction) by comparison to authentic standards, which excluded the Sp diastereomer and the reverse 5’-d(APSG)-3’ compounds. This identification was confirmed using high-resolution mass spectrometric comparisons.
Importantly, Zhou et al. also found the Dnd phenotype in DNA of bacterial species of variable origin and diverse habitat, including marine organisms, suggesting that this sulfur modification is a widespread phenomenon. Exactly how widespread and sequence-specificity are discussed in a following section on Functional Significance, but it is biochemically important to first consider the biosynthetic pathways that have been deciphered for PT-DNA.
Biosynthetic Pathways for PT-DNA
Structures taken from wikipedia.org and commons.wikimedia.org and free to use.
According to Chen et al., the five proteins (DndA, DndB, DndC, DndD, and DndE) encoded by the dnd gene locus are necessary and sufficient for the process of DNA phosphorothioation in S. lividans. Among these five proteins, DndA has been found by use of L-[35S>
-cysteine to be a cysteine desulfurase, catalyzing the removal of sulfur from the substrate L-cysteine to give L-alanine and reconstituting a 4Fe-4S iron-sulfur cluster in DndC. It was therefore suggested that DndA-catalyzed sulfur mobilization is the first step of the DNA phosphorothioation procedure.
Generalized structure of a 4Fe-4S cluster comprised of four iron ions and four sulfide ions placed at the vertices in a cubane-like configuration, wherein the Fe centers are typically further coordinated by cysteinyl ligands. Taken from commons.wikimedia.org and free to use.
Pu et al. speculated that a free-radical was involved, and in 2019, the researchers published in vitro data on DndC-Cys-S-S•, a subsequently formed free-radical in which S• is derived from L-cysteine and reacts with anionic DNA P-O- to ultimately produce anionic DNA P-S- in the final PT-DNA linkage. Further mechanistic details remain to be elucidated, but readers interested in this working hypothesis can read the footnote at the end of this blog and consult Pu et al., an Open Access article.
Structure of pyridoxal 5′-phosphate (PLP) taken from commons.wikimedia.org and free to use.
Chen et al. add that cysteine desulfurases are involved in the syntheses of many kinds of sulfur-containing biomolecules, including thionucleosides in transfer RNAs (tRNAs). Typically, cysteine desulfurases contain a tightly bound pyridoxal 5′-phosphate (PLP) cofactor, which forms a Schiff base with a highly conserved lysine residue via condensation of the aldehyde group in PlP with the amino group in lysine.
To further elaborate on these details, Chen et al. determined the crystal structure of DndA from S. lividans. As the wild-type desulfurase did not afford suitable crystals, these were instead obtained with a C327S point-mutant, in which a serine replaced the catalytic cysteine Cys327. The structure of the homodimer shown here reveals the molecular mechanism that DndA employs to recognize the PLP cofactor. Moreover, Cys327, the catalytic cysteine of DndA, was found on a β strand, unlike other cysteine desulfurases.
Crystal structure of DndA homodimer from S. lividans. The structure is viewed perpendicular to the two-fold axis of the dimer. The two structural units are shown in magenta and green. Their bound PLP cofactors are presented as sticks, with carbon atoms yellow, nitrogen atoms blue, oxygen atoms red, and phosphorus atoms orange. Taken from Chen et al. and free to use. An open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
In a 2010 review, Deng and coworkers state that the predicted DndD protein of 663 amino acid residues has homology with the family of Structural Maintenance of Chromosomes (SMC) proteins associated with ATPase activity. These findings, and the fact that SpfD, a homolog of DndD in Pseudomonas fluorescens, showed specific ATPase activity, demonstrate that SpfD/DndD likely acts as an ATP-modulated DNA cross-linker, providing the energy required for stabilizing DNA secondary structures during the PT- modification process by hydrolyzing ATP.
Deng and coworkers further observed that DndE has 46% protein identity to phosphoribosylaminoimidazole carboxylase (NCAIR synthetase) from Anabaena variabilis, which is known to act at a condensing carboxylation step in purine biosynthesis. At that time, the biochemical function of DndE had not yet been demonstrated, but Cao et al. later went on to report a first-of-a-kind study on Dnd proteins, which included DndE.
According to Cao et al., recent genome mapping studies revealed two unusual features of PT modifications: short consensus sequences and partial modification of a specific genomic site in a population of bacteria. For example, only 12% of the GAAC/GTTC sites in E. coli B7A are converted to the GPSAAC/GPSTTC motif. To better understand the mechanism of target selection of PT modifications that underlies these features, Cao et al. characterized the DNA substrate recognition by DndC, D, and E in a cell-extract system from Salmonella, using a series of synthetic double-stranded oligonucleotides (ds-ODNs).
3D illustration of Salmonella bacteria for medical instruction.
The results revealed that the ds-ODNs underwent PT modification in vitro with the same modification pattern as in vivo, i.e., the GPSAAC/GPSTTC motif. However, they unexpectedly observed no significant effect on PT modification by sequences flanking the GAAC/GTTC motif, while PT also occurred in the GAAC/GTTC motif that could not be modified in vivo. Hemi-PT DNA also served as a substrate for PT-modifying enzymes, but not for single-stranded DNA. While further studies are needed to decipher these findings, these PT-modifying enzymes were discovered to function as a single large protein complex for the first time, with all of three subunits in tetrameric conformations.
Occurrence of PT-DNA
As was briefly mentioned in the introduction, PT-DNA is widespread in diverse microorganisms. According to Chen et al., an extensive investigation of bacterial genome sequences revealed that dnd gene cluster homologs are found in distantly related bacterial species from various geographic niches, including Escherichia coli, Pseudomonas, Salmonella, Bacillus, Oceanobacter, Shewanella, Hahella, Geobacter, Candidatus, Pelagibacter, and Nostoc, etc. These bacteria range from GC-rich Streptomyces to AT-rich Pelagibacter Ubique, from nonpathogenic Geobacter to pathogenic Salmonella, from soil-dwelling organisms to marine microbes, even including those from the environmental DNA (eDNA) of the Sargasso Sea.
In view of this widespread occurrence of PT modifications in microorganisms in and around Earth, the Zone was interested in whether PT modifications are found in humans or other mammals. According to the above-mentioned review in 2010 by Deng and coworkers, “a complete set of dnd homologs has not yet been found in mammals.” Although this review is dated, the Zone was unable to find any recent publications to disprove this.
Function of PT-DNA
Given the widespread occurrence of PT-DNA in microorganisms, the Zone was also interested in the known or proposed functions of these sulfur-modified linkages. After perusing the literature for an answer, it became apparent that PT-DNA does not have a known definitive role at this time. On the other hand, there are several intriguing suggestions, which will be outlined in the following subsections.
Streptomyces bacterial culture in an agar plate.
Nuclease Activity: The well-known protection of nucleic acids against degradation by some nucleases, which became a cornerstone for antisense and other forms of therapy, would seem to be an obvious possible role for PT modifications. However, Liu et al. reported the opposite, i.e. cleavage of PT-DNA by an endonuclease. Briefly, they observed that simultaneous expression of the dndA-E gene cluster from S. lividans and the putative Streptomyces coelicolor Type IV methyl-dependent restriction endonuclease ScoA3McrA leads to cell death in the same host.
A His-tagged derivative of ScoA3McrA cleaved sulfur-modified DNA, as well as Dcm-methylated DNA in vitro near the respective modification sites. Double-strand cleavage occurred 16–28 nucleotides away from the phosphorothioate links. DNase I footprinting demonstrated binding of ScoA3McrA to the Dcm methylation site, but no clear binding could be detected at the S-modified site under cleavage conditions. This is the first demonstration of an enzyme that specifically cleaves sulfur-modified DNA.
Human Gut Microbiome: In a 2020 publication, Sun et al. reported on the wide distribution of PT-containing microbes in the human gut microbiome for the first time. Based on these findings, it was suggested that “[t>
his work will guide future research on the transmission of PT among microbial communities in the micro-ecosystem of the human body. With the rapid growth of the human microbiome database, more PT-containing bacteria could be found to analyze the routine of PT transmission and discover the unknown motive force behind this phenomenon with effects on human health.”
Structure of 8-oxo-deoxyguanosine (8-oxo-dG) tautomer of 8-hydroxy-deoxyguanosine (8-OHdG). Taken from commons.wikimedia.org and free to use.
Anti-oxidant Property: Wu et al. reported that PT-DNA exhibits a mild anti-oxidant property, both in vivo and in vitro. It was found that 8-hydroxy-deoxyguanosine (8-OHdG or its tautomer 8-oxo-dG)—a sensitive marker of the DNA damage due to hydroxyl radical attack at the C8 of guanine—and reactive oxygen species (ROS) levels were significantly lower in dnd+ E. coli compared to a dnd− strain.
Furthermore, unlike traditional anti-oxidants, PT-DNA exhibited an unexpectedly high capacity for quenching hydroxyl radical. The phosphorothioate moiety donates an electron to either hydroxyl radical or guanine radical derived from hydroxyl radical, leading to a PS• radical. Then, a complex of PS• radical and OH− (i.e. the reductive product of hydroxyl radical) releases a highly reductive HS• radical, which scavenges more equivalents of oxidants in the way to poly-covalent sulfur compounds such as Sn, sulfite, and sulfate. These researchers concluded that “[t>
his plausible mechanism provides [a>
partial rationale as to why bacteria develop the resource-demanding PT modification on guanine-neighboring phosphates in [a>
Bacterial Fitness: In 2017, Kellner et al. reported that PT sulfur has redox and nucleophilic properties that suggest effects on “bacterial fitness in stressful environments.” They showed that PTs are dynamic and labile DNA modifications that cause genomic instability during oxidative stress. In experiments involving isotopic labeling coupled with mass spectrometry, they observed sulfur replacement in PTs at a rate of ∼2% h-1 in unstressed Escherichia coli and Salmonella enterica. Whereas PT levels were unaffected by exposure to hydrogen peroxide (H2O2) or hypochlorous acid (HOCl), PT turnover increased to 3.8-10% h-1 after HOCl treatment, consistent with the repair of HOCl-induced sulfur damage.
According to a review by Pullar et al., the powerful antimicrobial nature of HOCl has been well documented. It is the active ingredient in household bleach and the species responsible for the microbicidal properties of chlorinated water supplies. The production of HOCl by human and other mammalian neutrophils is an integral part of the ability of these cells to kill a wide range of pathogens.Taken from wikipedia.org and free to use.
PT-dependent sensitivity to HOCl extended to cytotoxicity and DNA strand breaks, which occurred at HOCl doses that were orders of magnitude lower than the corresponding doses of H2O2. The genotoxicity of HOCl in PT-containing bacteria suggests reduced fitness in competition with HOCl-producing organisms and during infections in humans.
Kellener et al. further note that, even without exogenous exposure to oxidants, PT modifications in the bacterial strains studied turn over relatively frequently, with ~2% of PTs replaced every hour, the time of one cell division in minimal media. There is no known PT removal enzyme, so PT turnover likely results from endogenous oxidants, metals, or alkylating agents modifying the PT sulfur, with this damage leading to desulfurization, phosphonate formation, or the DNA strand-break chemistry observed in vitro.
While these reactions of PT linkages and oxidizing agents may seem like esoteric biochemistry, the following recent report is the first connection between bacterial PT modifications and the quality of drinking water in supply systems.
Antibiotic Resistance to Disinfection: In a 2020 publication in Environmental Pollution, Khan et al. state that the mechanism driving the dissemination of antibiotic resistance genes (ARGs) in drinking water supply systems (DWSS) with multiple barriers remains poorly understood, despite several recent efforts. To address this, the researchers combined quantitative polymerase chain reaction (qPCR), metagenomics, and network analyses for the water treatment process and laboratory-scale experiments for chlorine treatment. Using model strains, they sought out to determine whether PT-DNA-producing dnd genes occurred in DWSS and facilitated the dissemination of mobilized colistin resistance-1 (mcr-1) and New Delhi metallo-β-lactamase-1 (blaNDM-1).
Microscopic image of Bdellovibrio. Taken from commons.wikimedia.org and free to use.
The results indicated that the relative abundance of dndB increased in the effluent, compared to the influent, in the water treatment plants. Presence of dndB copies had a positive correlation with the concentration of chloramine (ClNH2) disinfectant. Network analysis revealed Bdellovibrio (picture here) as a potential bacterial host for MCR genes, NDM genes, and dndB in the DWSS.
E. coli DH10B (Wild-type with the dndABCDE gene cluster and ΔdndB) model strains were used to investigate resistance to chlorine treatment at the concentration range of 0.5–3 mg/L. The resistance of the wild-type strain increased with increasing concentration of chlorine. PT-DNA protected MCR- and NDM-carrying bacteria from chloramine disinfection during the water treatment process. Kahn et al. therefore suggested that this higher relative abundance of ARGs in the effluent of the water treatment plants may be due to PT-DNA PT resistance to chloramine disinfection, which results in the enrichment of bacteria carrying MCR, NDM, and dndB.
Importantly, the researchers concluded that “[t>
his study provides a new understanding of the mechanism of ARG dissemination in DWSS, which will help improve the performance of drinking water treatment to control the risk associated with antibiotic-resistant bacteria.”
In a 2007 commentary in Nature Chemical Biology on the then recently reported existence of PT-DNA in bacteria, über-famous Prof. Fritz Eckstein (the innovator of phosphorothioate-modified nucleic acids) stated that “[t>
here is no doubt that this discovery has opened a new window and will stimulate research into novel aspects of DNA that have not been considered so far.” This prescient view has indeed proven to be the case for the reasons discussed in this blog and many more.
The Zone is especially intrigued by the occurrence of PT-DNA in human gut microbiomes and the likelihood that these PT-DNA-carrying bacteria can somehow influence human health, as suggested by Sun et al., mentioned above. It will be fascinating to learn what researchers find by looking further thought this “window.”
What do you think?
Your comments are welcomed, as usual.
Please feel free to share this blog with your colleagues or on social media.
In the above-mentioned 2007 commentary on PT-DNA by Eckstein, he observes that “incorporation of a sulfur atom into a phosphate presents an energetically uphill proposition. What comes to mind for the one required here is activation of the phosphate diester, for example by methylation, acylation, adenylation or phosphorylation of one of the oxygens, where this group is exchanged for sulfur by nucleophilic attack.” The Zone agrees with this mechanistic suggestion for P-O → P-S, especially in view of formal analogies to mechanisms for C-O → C-S in the biosynthesis of sulfur-containing tRNA nucleosides, such as 2-thiouridine (s2U), 4-thiouridine (s4U), and 2-thiocytidine (s2C), recently reviewed in detail by Čavužić and Liu.