Hypothiocyanite and host-microbe interactions

Summary

The pseudohypohalous acid hypothiocyanite / hypothiocyanous acid (OSCN⁻ / HOSCN) has been known to play an antimicrobial role in mammalian immunity for decades. It is a potent oxidant that kills bacteria but is non-toxic to human cells. Produced from thiocyanate (SCN⁻) and hydrogen peroxide (H₂O₂) in a variety of body sites by peroxidase enzymes, HOSCN has been explored as an agent of food preservation, pathogen killing, and even improved toothpaste. However, despite the well-recognized antibacterial role HOSCN plays in host-pathogen interactions, little is known about how bacteria sense and respond to this oxidant. In this work, we will summarize what is known and unknown about HOSCN in innate immunity and recent advances in understanding the responses that both pathogenic and nonpathogenic bacteria mount against this antimicrobial agent, highlighting studies done with three model organisms, Escherichia coli, Streptococcus spp., and Pseudomonas aeruginosa.

Introduction to Hypothiocyanite

The most potent antimicrobial oxidants produced by the innate immune system are the hypohalous and pseudohypohalous acids produced when heme peroxidases catalyze the reaction between hydrogen peroxide and a halide or pseudohalide ion (Cl⁻, Br⁻, SCN⁻) (Arnhold & Malle, 2022, Day, 2019). The most reactive and best studied of these oxidants is hypochlorous acid (HOCl), the active ingredient of household bleach and a significant antimicrobial effector produced by neutrophils (Ulfig & Leichert, 2021). This review focuses on the role of the abundant, but less well understood pseudohypohalous acid hypothiocyanous acid (HOSCN). The pKa of HOSCN is 5.3, meaning that it exists in chemical equilibrium with the OSCN⁻ hypothiocyanite ion and is largely present as hypothiocyanite at at physiologically-relevant pHs (Ashby, 2012). In this review, we will generally refer to this mixture as HOSCN for the sake of brevity.

We will first briefly summarize the chemistry of HOSCN, what is known about HOSCN production by the immune system, and when and where host-associated bacteria might be expected to encounter HOSCN, and then review what is known about how bacteria respond to and defend themselves against this oxidant, highlighting substantial recent progress in this area and outstanding questions that remain to be addressed by future research.

Of the six types of mammalian heme peroxidase enzymes, three are established to catalyze the formation of substantial quantities of antimicrobial (pseudo)hypohalous acids from (pseudo)halides and H₂O₂: lactoperoxidase (LPO), myeloperoxidase (MPO) and eosinophil peroxidase (EPO)(Paumann-Page et al., 2017, Arnhold & Malle, 2022) (Figure 1). The two-electron oxidation of thiocyanate (SCN⁻) catalyzed by the peroxidase enzymes is what forms HOSCN. Only MPO can synthesize HOCl (Paumann-Page et al., 2017, Arnhold & Malle, 2022). HOSCN is considered a pseudohypohalous acid because of its chemical similarity to the hypohalous acids (HOCl, HOBr, HOI), while having a non-halide precursor, SCN⁻(Ashby, 2012).

Figure 1.

Substrate specificity of the heme peroxidases LPO (lactoperoxidase, blue sphere), EPO (eosinophil peroxidase, orange sphere), and MPO (myeloperoxidase, red sphere). SCN⁻ is the preferred substrate for all three peroxidases, and only MPO can generate HOCl. LPO is capable of oxidizing Br⁻, but at much slower rates than MPO or EPO. Made with Biorender.

LPO is considered the main producer of HOSCN and is secreted in the saliva, tears, breastmilk, and lacrima of mammals (Arnhold & Malle, 2022, Ashby, 2012). Both MPO, released during the degranulation of leukocytes (Winterbourn et al., 2016), and EPO, produced by eosinophils (Ramirez et al., 2018), are also capable of oxidizing SCN⁻ to form HOSCN (Arnhold & Malle, 2022). Notably, SCN⁻ outcompetes the halide substrates (I⁻, Cl⁻, Br⁻) for the active site of peroxidases (Arnhold & Malle, 2022), meaning that even peroxidases capable of producing the more reactive hypohalous acids will produce large amounts of HOSCN if SCN⁻ is present. At average serum Cl⁻ and SCN⁻ concentrations, for example, MPO produces equal amounts of HOCl and HOSCN (van Dalen et al., 1997). Non-enzymatic oxidation of SCN⁻ by HOCl or reactive chloramines also produces HOSCN (Ashby, 2012, Xulu & Ashby, 2010).

Like the hypohalous acids (HOCl, HOBr, and HOI), HOSCN is a strong oxidant capable of killing or inhibiting bacterial cells. Unlike HOCl, which non-specifically oxidizes a wide range of cellular components (Ulfig & Leichert, 2021, Pattison et al., 2012, Skaff et al., 2009), HOSCN mainly damages bacterial cells via oxidation of thiols, including cysteine, cysteine-containing molecules (like glutathione), or other low-molecular weight thiols (like bacillithiol), via an unstable sulfenyl thiocyanate (R-S-SCN) intermediate, forming sulfenic acids (R-SOH) or either inter- or intra-molecular disulfide bonds (R₁-S-S-R₂) (Barrett & Hawkins, 2012, Pattison et al., 2012, Winterbourn et al., 2016, Fuentes-Lemus et al., 2021, Aune & Thomas, 1978, Oram & Reiter, 1966, Skaff et al., 2009). HOSCN is a more reactive thiol oxidant than OSCN⁻ (Ashby, 2012), meaning that antimicrobial activity of HOSCN is higher at lower pH.

Oxidative damage caused by HOSCN inhibits glycolysis and other central metabolic pathways, transporters, respiration, and can lead to protein aggregation, albeit less dramatically than HOCl (Groitl et al., 2017, Ulfig & Leichert, 2021, Thomas & Aune, 1978). Also unlike HOCl, the oxidative damage caused by HOSCN is largely reversible, since HOSCN cannot oxidize thiols to sulfinic (R-SO₂H) or sulfonic (R-SO₃H) acid (Ashby, 2012, Thomas & Aune, 1978). While it has not been demonstrated directly, we suspect that HOSCN also damages other biologically important and oxidation-sensitive thiols, such as those in iron-sulfur clusters and mononuclear iron enzymes (Imlay, 2014, Pattison et al., 2012). We refer the reader interested in more detail on heme peroxidases and the chemistry of HOSCN to the following excellent comprehensive reviews (Arnhold & Malle, 2022, Ashby, 2012, Ulfig & Leichert, 2021, Day, 2019).

The molecular mechanisms by which host-associated bacteria respond to and defend themselves against reactive oxygen and reactive chlorine compounds have been the subject of extensive study (Sen & Imlay, 2021, Ulfig & Leichert, 2021). By contrast, until recently, the mechanisms by which bacteria survive stress caused by HOSCN remained almost entirely unknown. Before reviewing recent work which has made significant progress in this area, we will first consider where the precursors of HOSCN come from, how much HOSCN is likely to be present in different locations in the body, and what evidence exists for a role for HOSCN in modulating host-microbe interactions.

Hypothiocyanite in the Human Body

There are few direct measurements of in vivo HOSCN concentrations in any biological system. Older methods depend on the oxidation of thiol-containing reagents (e.g. 2-nitro-5-thiobenzoate; TNB), and require careful controls to eliminate interference by other oxidants (Below et al., 2018, Pruitt et al., 1983, Tenovuo et al., 1982). These methods were used, for example, to measure HOSCN in human saliva (up to 60 μM)(Pruitt et al., 1983, Tenovuo et al., 1982), but have not been used widely for other fluids. An HOSCN-specific ion chromatography technique has been recently described for the determination of HOSCN concentrations in saliva (maximum value measured = 3.9 mg L⁻¹, or approximately 50 μM)(Below et al., 2018), but this technique has not yet been widely used and it is unclear whether it will be applicable to other biological fluids.

Many fluorescent probes exist for the in vivo detection of reactive oxygen, nitrogen, and chlorine species (ROS, RNS, RCS) (Chen et al., 2016, Wu et al., 2019), but none of these, to our knowledge, have been tested to determine if they react with HOSCN. The presence of HOCl in biological samples can be inferred from the detection of HOCl-specific biomarkers (e.g. 3-chlorotyrosine or glutathione sulfonamide)(Kettle et al., 2014, Harwood et al., 2008), but no specific biomarkers produced by HOSCN oxidation of biological molecules are known. A recently described fluorescent biosensor based on the HOCl-sensing transcription factor NemR is reported to respond to HOSCN (Kostyuk et al., 2022), but it also responds to HOCl, so similarly lacks specificity.

Indirect inference of the presence of HOSCN therefore relies on the fact that enzymatic HOSCN synthesis requires three components which are easier to measure in situ: a heme peroxidase, H₂O₂, and SCN⁻ (Arnhold & Malle, 2022, Ashby, 2012). We can thereby infer where in the body and under what circumstances bacteria are likely to be exposed to HOSCN based on the abundance of these three components (Figure 2).

Figure 2.

Sources of hypothiocyanite (HOSCN) in the human body. Thiocyanate (SCN⁻) is derived from food via bacterial metabolism of glucosinolates and cyanogenic glycosides, and accumulates in fluids (e.g. saliva, airway secretions, or blood plasma). Dietary cyanide (CN⁻) is converted to SCN⁻ in the liver. Heme peroxidases LPO (lactoperoxidase, blue spheres), EPO (eosinophil peroxidase, orange spheres), and MPO (myeloperoxidase, red spheres) catalyze the production of HOSCN from SCN⁻ and H₂O₂. SCN⁻ also reacts non-enzymatically with MPO-produced hypochlorous acid (HOCl) to produce HOSCN. The concentrations of HOSCN in the airway, plasma, and intestine are unknown. Made with Biorender.

Heme Peroxidases

Three main mammalian heme peroxidase enzymes catalyze the formation of antimicrobial (pseudo)hypohalous acids from (pseudo)halides and H₂O₂: lactoperoxidase (LPO), myeloperoxidase (MPO), and eosinophil peroxidase (EPO)(Arnhold & Malle, 2022).

LPO, which is constitutively secreted in the saliva, lacrima, and breastmilk of mammals, works as a maintenance enzyme, setting it apart from MPO and EPO, which are released during inflammation (Day, 2019, Courtois, 2021). LPO regulates the oral microbiome (Courtois, 2021) and keeps the breastmilk free of pathogens during the first few weeks after delivery (Gothefors & Marklund, 1975). Mice secrete LPO in their intestines, but humans do not (Rigoni et al., 2017), and murine neutrophils contain less MPO than human ones (Rausch & Moore, 1975, Nauseef, 2022). This adds a layer of complexity to the use and interpretation of animal models in studies of the role of HOSCN in host interactions, as mice may have different concentrations of HOSCN than humans.

MPO is contained in the granules of neutrophils alongside a host of other bactericidal proteins (Winterbourn et al., 2016). Neutrophils are the most abundant and quickest responding immune cell. They quickly phagocytize invading pathogens and expose them to a wide range of toxic antimicrobials. Upon phagocytosis, degranulation occurs in the phagosome, releasing MPO, which utilizes Cl⁻ that has been imported by CFTR and CIC-3 (Painter et al., 2006, Moreland et al., 2006), and NADPH oxidase-derived H₂O₂ (Winterbourn et al., 2006). Therefore, within the phagosome, MPO primarily produces HOCl. However, MPO can be released into the extracellular space during inflammation, during neutrophil cell death or oxidative burst. If SCN⁻ is available, as noted above, it is the preferred substrate for MPO, leading to the production of HOSCN (Arnhold & Malle, 2022).

EPO is released extracellularly by eosinophils. Eosinophils are migrant immune cells found in blood and tissue in the thymus, adipose tissue, uterus, and lungs (Marichal et al., 2017), where eosinophils are also expanded at sites of allergic inflammation, and specifically in the bronchial area during inflammation (Nakagome & Nagata, 2018). EPO is released extracellularly because eosinophils mostly play a role in the killing of parasites and cancer cells, which are typically too large to be phagocytosed (Ramirez et al., 2018). The substrates of EPO are Br⁻, NO₂⁻, and SCN⁻, with SCN⁻ being the preferred substrate (Arnhold & Malle, 2022). Recently, Gurtner et al. discovered that there are specialized eosinophils contained to the gastrointestinal tract, which expand during inflammatory bowel disease, suggesting that EPO may contribute to HOSCN exposure of intestinal microbes during colitis (Gurtner et al., 2022).

Hydrogen Peroxide

The concentrations of H₂O₂ and SCN⁻ affect how much HOSCN peroxidases are able to produce. As stated above, H₂O₂ is released in the phagosome of neutrophils as a by-product of NADPH oxidase (NOX), and therefore will also be released during cell death and oxidative burst of neutrophils, meaning that the concentration of H₂O₂ is much higher at sites of inflammation (Winterbourn et al., 2016, Decoursey & Ligeti, 2005). NOX2 is the main producer of H₂O₂ used by MPO. Dual oxidase enzymes (DUOX) also produce H₂O₂, including that which LPO uses to generate SCN⁻ in mucosal airways (Sarr et al., 2018). Additionally, H₂O₂ is produced by some bacteria, including commensal oral streptococci and the pathobiont Streptococcus pneumoniae (Mraheil et al., 2021, Okahashi et al., 2022).

Thiocyanate

Thiocyanate (SCN⁻) is mostly acquired from the diet. Many food plants contain either glucosinolates, which contain isothiocyanate groups (Wittstock et al., 2017), or cyanogenic glycosides, which release cyanide when metabolized by intestinal bacteria (Cressey & Reeve, 2019). Glucosinolates are found in vegetables of the order Brassicales (cabbages, turnips, mustard, horseradish, etc.)(Clarke, 2010), while cyanogenic glycosides are abundant in cassava, stone fruits, and almonds, among other food plants (Jones, 1998, Nyirenda, 2020). Cyanide is detoxified into SCN⁻ by the mitochondrial enzyme rhodanese, largely in the liver (Buonvino et al., 2022). The opportunistic pathogen Pseudomonas aeruginosa, like some other bacteria, produces significant amounts of cyanide (Letoffe et al., 2022), which may increase SCN⁻ concentrations in P. aeruginosa-infected individuals. Glucosinolates are also metabolized by intestinal bacteria, releasing SCN⁻, which is then taken up and distributed through the body (Bouranis et al., 2021). The pathway by which the abundant intestinal anaerobe Bacteroides thetaiotaomicron metabolizes glucosinolates has recently been described (Liou et al., 2020).

SCN⁻ concentrations have been measured in many human body fluids (summarized in (Chandler & Day, 2012, San Gabriel et al., 2020)). Blood plasma levels in non-smokers range from 5 to 50 μM (Lorentzen et al., 2011, Madiyal et al., 2018), with substantially higher concentrations found in tears (150 μM)(van Haeringen et al., 1979), gastric fluid (250 – 300 μM)(Das et al., 1995), airway fluids (from 30 – 650 μM in the lung and up to 1.2 mM in the nose) (Lorentzen et al., 2011, Wijkstrom-Frei et al., 2003), and in saliva (0.5 – 3 mM) (van Haeringen et al., 1979, Madiyal et al., 2018). Vegans may have higher average SCN⁻ levels than non-vegans (Leung et al., 2011) and, because of the cyanide in tobacco, SCN⁻ concentrations are much higher in smokers (Scherer, 2006, Madiyal et al., 2018).

Given these facts, HOSCN is likely to exist at substantial concentrations in the oral cavity and the respiratory tract and is probably present under any inflammatory circumstances when activated neutrophils are releasing MPO and H₂O₂ into the extracellular space, although the exact physiological concentrations and relative contributions of HOSCN to antimicrobial effects remain unclear in most cases (Ashby, 2012, Chandler & Day, 2012, Ulfig & Leichert, 2021).

Hypothiocyanite and Host-Microbe Interactions

Mammalian cells are resistant to damage by HOSCN (Barrett & Hawkins, 2012, Chandler et al., 2013), in contrast to HOCl, which causes extensive damage to both bacteria and host cells (Hawkins, 2020, Ulfig & Leichert, 2021). Mammalian selenocysteine-containing thioredoxin reductase efficiently reduces HOSCN to SCN⁻, while bacterial thioredoxin reductase has no such activity and is inhibited by HOSCN (Chandler et al., 2013). This has led to the hypothesis that SCN⁻ plays an antioxidant role in innate immunity, since it reacts efficiently with tissue-damaging HOCl and HOCl-derived chloramines to generate HOSCN (Ashby et al., 2004, Xulu & Ashby, 2010), protecting the host while retaining antimicrobial activity (Chandler & Day, 2012).

The role of HOSCN in host-microbe interactions during a disease state has only been studied in a few sites in the body. For example, it is known that the oral cavity has a high concentration of LPO, and therefore a steady secretion of HOSCN. LPO has long been understood to play a role in the prevention of dental caries (Courtois, 2021). In mice, decreased expression of LPO in the gut is linked to increased microbial dysbiosis and colitis (Lin et al., 2022).

The role of HOSCN in lung disease has been mostly studied in the context of cystic fibrosis (CF). CF is caused by mutations of the cystic fibrosis transmembrane regulator (CFTR), which exports Cl⁻ and SCN⁻ across the cellular membrane and impacts the amount of substrate available to LPO and MPO in the airway (Xu et al., 2009). Low SCN⁻ concentrations in saliva are a biomarker for CFTR activity, which correlates with CF disease severity (Malkovskiy et al., 2019), and nebulized SCN⁻ has anti-inflammatory and anti-microbial effects in a mouse model of the CF lung (Chandler et al., 2015).

Bacterial Responses to HOSCN

Bacterial responses to HOSCN have only been investigated in three types of model organisms: Escherichia coli, Streptococcus spp. and Pseudomonas aeruginosa (Figure 3). These experiments have not used a consistent method of treating bacterial cells with HOSCN, which can make comparisons among them challenging. The two most common methods are a bolus addition of pre-synthesized HOSCN (25 – 800 μM), which allows the exact concentration of HOSCN to be controlled at time of addition, and treatment by adding LPO, SCN⁻, and H₂O₂ (or an H₂O₂-generating enzyme) simultaneously (referred to here as the LPO-SCN⁻ system), which may more closely mimic the gradual and ongoing exposure of HOSCN bacteria would face in a host environment, but makes it more difficult to control the concentration and determine that any effects seen are exclusively from stress caused by HOSCN, as opposed to H₂O₂ toxicity or other activities of LPO (Arnhold & Malle, 2022, Sen & Imlay, 2021). LPO-SCN⁻, in fact, protects some bacteria against H₂O₂ toxicity (Adamson & Carlsson, 1982, Ashby et al., 2009), further complicating interpretation of these experiments. There are also substantial differences in treatment conditions among these experiments, including the growth media or buffers in which the bacteria were exposed to stress, cell density, and bacterial growth phase, all of which can impact the effect of strong oxidants (e.g. HOCl or H₂O₂) on bacteria (Ashby et al., 2020, Verspecht et al., 2021), and might also be relevant in experiments using HOSCN. HOSCN is much less reactive than HOCl (Pattison et al., 2012), which may reduce the impact of cell density in HOSCN stress assays. Experiments with S. pneumoniae suggest that bacteria react only slowly with HOSCN (Shearer et al., 2022a, Shearer et al., 2022b), suggesting that even bolus addition of HOSCN may result in a relatively slow flux of oxidant into the cells.

Figure 3.

Known responses of E. coli, Streptococcus spp., and P. aeruginosa to HOSCN. All listed genes are upregulated in response to HOSCN exposure. Bolded genes have been experimentally demonstrated to aid in bacterial survival of HOSCN stress. Made with Biorender.

Escherichia coli

The LPO-SCN⁻ system was first tested on E. coli in 1976, when Reiter et al. observed bactericidal effects of milk on different Gram-negative bacteria (Reiter et al., 1976), followed by work from Aune and Thomas correlating oxidation of E. coli thiols by the LPO-SCN⁻ system with inhibition of respiration (Aune & Thomas, 1978, Thomas & Aune, 1978). Garcia-Graells et al. later observed that the LPO-SCN⁻ system was bacteriostatic to E. coli in milk at room temperature, but when combined with high hydrostatic pressure, could kill the cells when inoculated at a low cell density (Garcia-Graells et al., 2000). This group also isolated a pressure-resistant strain of E. coli that was able to withstand the combination of the LPO-SCN⁻ system and high hydrostatic pressure (Garcia-Graells et al., 2003), but the genetic basis of that resistance was not identified.

In 2004, Sermon et al. investigated transcriptional responses to overnight treatment with the LPO-SCN⁻ system in rich media using a promoter fusion strategy (Sermon et al., 2005a). Some known stress response genes were upregulated, including recA, dnaK, and sodA, suggesting induction of DNA damage, protein unfolding, and superoxide stress responses, respectively (Imlay, 2008, Maslowska et al., 2019, Roncarati & Scarlato, 2017). The LPO-SCN⁻ system also induced expression of several other genes, including corA and mgtA, which encode metal ion transporters, cysJ, which encodes the alpha subunit of sulfite reductase, tcyP, which encodes a cysteine transporter, and fruB, which encodes a fructose transport protein. The contributions of these genes to HOSCN response remain mostly unclear. Deletions of corA, cysJ, or tcyP sensitized E. coli to LPO-SCN⁻ in HEPES buffer, but deletion of mgtA did not. The phenotype of the ΔcorA mutant was dependent on Ni²⁺ transport, and addition of Ni²⁺ greatly enhanced the toxicity of LPO-SCN⁻ (Sermon et al., 2005b). The mechanism by which it does so has not yet been established, nor has the impact of Ni²⁺ on HOSCN toxicity in any other organism been examined. Further work from the same group revealed that mutations that reduce the permeability of the outer membrane by disrupting porins increase tolerance to LPO in rich media (De Spiegeleer et al., 2005), presumably by reducing the ability of HOSCN to cross the outer membrane, especially under near-neutral pH conditions, where most of it will be present as negatively-charged and membrane-impermeable OSCN⁻.

More recently, we found that E. coli encodes a potent HOSCN reductase enzyme, called RclA (Meredith et al., 2022). RclA is a flavin-dependent oxidoreductase that reduces HOSCN to SCN⁻ in vitro with very high efficiency, and E. coli mutants lacking rclA are sensitive to bolus addition of HOSCN in minimal glucose medium. RclA homologs are present in many other bacterial species (Derke et al., 2020), suggesting that this is a common defensive strategy. RclA homologs from S. pneumoniae, B. thetaiotaomicron, Limosilactobacillus reuteri, and Staphylococcus aureus are all capable of protecting the E. coli rclA mutant against bolus addition of HOSCN (Meredith et al., 2022). In E. coli, rclA transcription is induced by the redox-sensitive transcription factor RclR following exposure to HOSCN (Meredith et al., 2022) and is transcribed in an operon with two downstream genes of unknown function called rclB and rclC (Parker et al., 2013). Mutants lacking either rclB or rclC are sensitive to bolus addition of HOSCN in minimal glucose medium (Meredith et al., 2022). The rclBC genes are generally conserved only in enterobacteria (Derke et al., 2020). RclR also responds strongly to reactive chlorine compounds (Parker et al., 2013), and since the HOCl-responsive E. coli transcription factor NemR also appears to detect HOSCN (Kostyuk et al., 2022), we suspect that many cysteine-dependent HOCl-sensitive regulators in bacteria (da Cruz Nizer et al., 2020) may also respond to HOSCN, but this remains to be determined. A homolog of RclC has recently been characterized in uropathogenic E. coli (UPEC) and has been implicated in aiding with membrane stability when cells are exposed to HOCl (Sultana et al. 2022), but the mechanism by which it does so remains unknown.

Streptococcus spp.

The relationship between oral Streptococcus spp. and HOSCN was first examined over 50 years ago, when Oram and Reiter showed that catabolic enzymes of some oral streptococci were inhibited by the LPO-SCN⁻ system in milk (Oram & Reiter, 1966), results confirmed 20 years later in cell-free extracts (Carlsson et al., 1983). Inhibition by LPO-SCN⁻ in saliva occurred at pH 5, but not at pH 7 (Lumikari et al., 1991), suggesting that toxicity in these Gram-positive bacteria was due to protonated HOSCN, not OSCN⁻, presumably due to the ability of uncharged HOSCN to cross the cell membrane (Ashby, 2012). Carlsson et al. reported that crude lysates of the commensal oral streptococci S. sanguinis and S. mitis contained NAD(P)H-dependent HOSCN reductase activity (Carlsson et al., 1983). These species were able to resist inactivation of metabolic enzymes by exposure to LPO-SCN⁻ in minimal salts, while the cavity-causing pathogen S. mutans had no such activity and was not able to recover from stress treatment. In continuous culture in artificial saliva medium, addition of LPO-SCN⁻ allowed S. sanguinis to outcompete S. mutans (van der Hoeven & Camp, 1993), consistent with roles for LPO and HOSCN in shaping the composition of the oral microbiome, a hypothesis that has become widely established in the literature (Courtois, 2021), although a detailed understanding of the mechanisms involved remains to be established (Ashby et al., 2009). The genomes of many commensal oral streptococci, including S. sanguinis and S. mitis, encode homologs of RclA, while the genome of S. mutans does not (Meredith et al., 2022, Derke et al., 2020), but it has not yet been rigorously shown what factors contribute to HOSCN response and resistance in oral streptococci.

S. pneumoniae, which can be a commensal inhabitant of the nasopharynx and is a major cause of bacterial pneumonia and other infections (Collaborators, 2023, Morimura et al., 2021), is exposed to an environment containing HOSCN when it colonizes the human lung. S. pneumoniae, like some oral streptococci, produces abundant H₂O₂ (Carlsson et al., 1983). While the LPO-SCN⁻ system is capable of killing S. pneumoniae (Gingerich et al., 2020), this species is more HOSCN-resistant than other tested lung pathogens (Shearer et al., 2022a). HOSCN resistance in this bacterium is partially due to the ability of S. pneumoniae to import glutathione and glutathionylate proteins after HOSCN treatment, and mutants lacking either the glutathione transporter GshT or glutathione reductase Gor are significantly more sensitive to both bolus addition of HOSCN in minimal salts solution and to the LPO-SCN⁻ system in rich media (Shearer et al., 2022c). S. pneumoniae also possesses a homolog of RclA, called Har. Unlike E. coli RclA, Har appears to be constitutively expressed. When Har was knocked out in combination with disruptions in glutathione import and recycling, S. pneumoniae growth was completely inhibited by the LPO-SCN⁻ system in rich media (Shearer et al., 2022b).

Pseudomonas aeruginosa

The cystic fibrosis transmembrane conductance regulator (CFTR) that is defective in cystic fibrosis patients is an SCN⁻ transporter, and some reports suggest that patients have lower SCN⁻ concentrations in their lungs than healthy controls (Chandler & Day, 2012, Malkovskiy et al., 2019, Xu et al., 2009). The resulting changes in HOSCN levels in the cystic fibrosis lung may play a role in the inflammation and sensitivity to bacterial infection in that environment (Bojanowski et al., 2021, Chandler & Day, 2012). Therefore, the ability of P. aeruginosa, a major pathogen of the CF lung (Greenwald & Wolfgang, 2022), to respond to and survive stress caused by HOSCN is important to understand. P. aeruginosa is substantially more sensitive to HOSCN than S. pneumoniae (Shearer et al., 2022a), and is more sensitive at pH 6.8 than at pH 7.4 (Chandler et al., 2013), consistent with protonation of HOSCN being an important determinant of its toxicity for bacteria.

Farrant et al. have explored the genetic responses of P. aeruginosa PA14 to both HOCl and HOSCN (Farrant et al., 2020). They identified a homolog of RclR which upregulates the expression of rclX, encoding a predicted peroxiredoxin, in response to both oxidants (Farrant et al., 2020). Mutants lacking either rclR or rclX were sensitive to bolus addition of HOSCN in rich medium and in both planktonic and biofilm states in artificial sputum medium, but the physiological substrate of RclX has not been identified. This is consistent with a conserved role for RclR homologs in HOSCN response, even in organisms like P. aeruginosa which lack RclA, RclB, or RclC homologs (Derke et al., 2020). In separate work, Groitl et al. found that when P. aeruginosa PA14 was exposed to bolus addition of HOSCN in minimal glucose medium, extensive protein aggregation occurred, inducing the heat shock response as well as a response consistent with membrane damage (Groitl et al., 2017). They also observed that when ppk, encoding polyphosphate (polyP) kinase (Bowlin & Gray, 2021), was deleted, susceptibility to protein aggregation and growth inhibition by HOSCN was increased, consistent with the role of polyP in protecting bacteria against protein aggregation caused by HOCl (Gray et al., 2014). Whether polyP is an important contributor to HOSCN resistance in other bacteria is unknown.

P. aeruginosa PA14 is the only bacterium for which whole-genome transcriptomic response to HOSCN has been reported, by both Farrant et al. and Groitl et al. (Farrant et al., 2020, Groitl et al., 2017). Both groups identified substantial overlaps between HOSCN and HOCl responses (da Cruz Nizer et al., 2020), including upregulation of protein-stabilizing chaperones, sulfur-containing amino acid metabolism, and the RclR and MexT regulons. There were, however, some differences in their results, including in the activation of other HOCl-sensitive regulators by HOSCN (e.g. NemR). These differences are probably due to variations in treatment conditions, further illustrating the challenge of extrapolating stress response networks from a small set of experiments. More transcriptomic experiments with different species under different HOSCN treatment regimens will be needed to systematically characterize how bacteria regulate gene expression in response to HOSCN and how this may or may not overlap with responses to other antimicrobial oxidants.

Outlook and Future Directions

The study of bacterial HOSCN stress response is in its infancy. The results summarized above suggest overlaps with both ROS and RCS responses, particularly in the areas of redox homeostasis (e.g. glutathione metabolism), cysteine metabolism, and prevention of protein aggregation (Figure 3), but considerable work needs to be done to more fully characterize this response. It will be particularly important to expand studies of HOSCN resistance beyond the few species that have been studied so far, since many pathogens and commensal bacteria live in HOSCN-containing environments (Figure 2). Most intriguing, perhaps, is the question of what specific defenses exist against HOSCN. Contrary to the current paradigm (Chandler et al., 2013), it is clear that many bacteria contain highly active HOSCN reductases (RclA / Har)(Meredith et al., 2022, Shearer et al., 2022b), but what other defenses exist, and how do they work? Are there transcription factors that respond specifically to HOSCN? RclBC in E. coli and RclX in P. aeruginosa are notable examples of proteins that defend against HOSCN by unknown mechanisms (Meredith et al., 2022, Farrant et al., 2020), but we think it likely that other such defenses exist. Systematic experiments to comprehensively characterize HOSCN defense in diverse bacteria are clearly needed.

The role bacterial HOSCN resistance genes may play in host colonization and pathogenesis have yet to be determined in vivo in any model system. Recent progress in S. pneumoniae is encouraging for testing how HOSCN resistance contributes to pathogenesis in this important lung pathogen (Gingerich et al., 2020, Shearer et al., 2022a, Shearer et al., 2022b, Shearer et al., 2022c). Future work will also need to be done to explore the role of HOSCN resistance in vivo in other pathogens, especially those that infect the lungs and oral cavity (e.g. P. aeruginosa, S. aureus, S. mutans, etc.). One intriguing unexplored area is the role of bacterial HOSCN resistance in the gut. The fact that E. coli, B. thetaiotaomicron, and many other intestinal bacteria contain conserved HOSCN resistance genes (Derke et al., 2020, Meredith et al., 2022) suggests that they encounter HOSCN in that environment, but we are not aware of any experiments directly addressing this question. Results in mice suggest a possible link between LPO and colitis (Lin et al., 2022), but how this might translate to humans or be affected by diet or other factors remains unknown. The question of the physiological role of HOSCN in host-microbe interactions is an important one, and recent progress in establishing the molecular mechanisms of HOSCN resistance will finally make it possible to address this question directly.