Re-evaluating the antibacterial properties of DMARD and pro-drug sulphasalazine against autoimmune bacterial triggers after eighty years
Ian Edwin Cock, Michael Wellesley Whitehouse
TL;DR
This paper investigates how sulphasalazine and its breakdown products affect bacteria linked to autoimmune diseases, finding that one breakdown product is particularly effective against these bacteria.
Contribution
The study re-evaluates the antibacterial properties of sulphasalazine and its metabolites against bacterial triggers of autoimmune diseases.
Findings
Sulphasalazine's breakdown product sulfapyridine showed potent antibacterial activity against Proteus spp. and Klebsiella pneumoniae.
Combining 5-aminosalicylate with sulfapyridine enhanced antibacterial effects, showing synergy and additive effects against certain bacteria.
Sulphasalazine's pro-drug properties are highlighted as effective against bacterial triggers of inflammatory diseases.
Abstract
Sulphasalazine (SSZ) has been used to treat a range of inflammatory conditions since the 1940s. It functions as a pro-drug that, upon azoreduction by selected gastrointestinal bacteria (including the bacterial triggers of some inflammatory diseases), releases an antioxidant protective molecule, 5-aminosalicylate (5-AS), and the antibacterial molecule sulfapyridine (SP). SSZ, 5-AS and SP were evaluated for growth inhibitory activity against some bacterial triggers of rheumatoid arthritis (Proteus spp.), ankylosing spondylitis (Klebsiella pnumoniae), multiple sclerosis (Acinetobacter baylyi and Pseudomonas aeruginosa) and rheumatic fever (Streptococcus pyogenes). These bacteria have previously been reported to have azoreductase activity and therefore they may locally convert the SSZ pro-drug into 5-AS and SP. The potency of all compounds, as well as a combination of 5-AS and SP, were…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3- —Griffith University
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsInflammatory mediators and NSAID effects · Organoselenium and organotellurium chemistry · Rheumatoid Arthritis Research and Therapies
Introduction
Sulphasalazine, SSZ, (aka salazosulfapyridine, Salazopyrin^®^, Azasulfidine^®^) is a disease-modifying, anti-rheumatic drug (DMARD), first introduced in the 1940s by Pharmacia, Uppsala, Sweden following pioneering clinical studies in the late 1930s by Svartz (1942a 1942b, 1948, 1988, Supplementary Data). SSZ controls many of the clinical features of spondylarthropathy (axial or peripheral articular involvement), psoriasis, rheumatoid arthritis, uveitis and some other inflammatory conditions. This non-biological (pro-) drug has withstood the test of time (Haagsma 2005; Rains et al. 1995). It is still prescribed more than 80 years after its clinical introduction, sometimes in combination with methotrexate. Such a long use as an antirheumatic drug is rare amongst the so-called DMARDS’s. SSZ is selectively bioactivated in anaerobic tissues by reductive cleavage of the diazo bond (Fig. 1). Additionally, it is site-directed i.e. a “magic bullet”, as described by Paul Ehrlich (Svartz 1942), delivering both an antioxidant (5-aminosalicylate) to minimise tissue damage, and an antibiotic (sulfapyridine) to control bacterial pathogens that trigger chronic inflammation. Indeed, it has been stated that “sulphasalazine is effective in a proportion of patients, is relatively safe and has a rapid onset of action. These factors may make sulphasalazine the useful for treating early rheumatoid arthritis.” (Barx1992).
Fig. 1. The metabolism of SSZ by colonic microflora
An important feature of SSZ’s (and its metabolites) clinical action is its relatively selective effect on anaerobic infections that can sustain tissue damage in chronic infections. Bacterial triggers have been identified for multiple autoimmune inflammatory diseases, including Proteus spp. (rheumatoid arthritis), Klebsiella pneumoniae (ankylosing spondylitis), Acinetobacter spp. and Pseudomonas aeruginosa (multiple sclerosis), and Streptococcus pyogenes (rheumatic fever). It has been postulated that inhibiting the growth of these pathogens may be useful in reducing the incidence and severity of the prodromal symptoms and tissue degeneration in people genetically susceptible to these diseases, although this remains to be tested in in vivo test models (Cock and Cheesman 2019). Also, SSZ has shown antibacterial activity against P. mirabilis, with an MIC of 625 µg/mL (Ansaru et al. 2022). The same study also reported noteworthy inhibitory activity against several other gastrointestinal bacteria (MICs 312.5–1250 µg/mL). However, SSZ’s effects against other bacterial triggers of autoimmune diseases remain unestablished. Several previous studies that have reported antibacterial effects for SSZ have examined the faecal microflora of patients undergoing SSZ treatment (Kanerud 1994; Neumann et al. 1987). Whilst these studies indicate that SSZ treatment reduces the faecal load of bacterial triggers of autoimmune diseases, they fail to take into account additional factors that may influence the gastrointestinal microbiome. Additionally, MIC values were not reported in those studies, making comparisons between (and within) studies impossible.
Antibacterial activity has also been reported for 5-aminosalicylate (5-AS) and sulfapyridine (SP) (Das Mahapatra et al. 2024), although their activity against the bacterial triggers of autoimmune diseases has been relatively neglected. The earlier studies examined bacterial susceptibility using agar diffusion assays but MIC values were not reported. Further studies are required to quantify this activity, and to screen against an extended panel of bacteria, particularly against those that trigger autoimmune diseases. Surprisingly, given that SSZ is converted 5-AS and SP in the gastrointestinal tract, we were unable to find studies that evaluate combinations of these compounds for antibiotic activity enhancement. For the first time, this study quantifies the antibacterial activity of a combination of 5-AS and SP against bacterial triggers of four autoimmune diseases and compares the potency to that of the parent molecule SSZ. Additionally, the class of interactions between the combination components are reported for the first time using ƩFIC and isobologram analysis.
Methods
Test compounds
Sulphasalazine (≥ 97% purity), 5-aminosalicylate (95% purity) and sulfapyridine (≥ 99%) were purchased from Sigma-Aldrich, Australia. All test compounds were individually dissolved in deionised water, containing 1% dimethyl sulfoxide (DMSO) (Sigma-Aldrich, Australia) to give 10 mg/mL stock concentrations. The stock solutions were stored at 4 °C in the dark until use. For bioactivity testing, the stock solutions were diluted to the required concentrations in deionised water containing 1% DMSO.
Bacterial cultures
The bacteria selected for screening are known triggers of autoimmune inflammatory diseases in genetically susceptible individuals (Cock and Cheesman 2019). Reference strains of Proteus mirabilis (ATCC 21721), Proteus vulgaris (ATCC 20719), Klebsiella pneumoniae (ATCC 31488), Acinetobacter baylyi (ATCC 33304) and Pseudomonas aeruginosa (ATCC 39324) were purchased from the American Type Culture Collection, USA. A clinical isolate strain of Streptococcus pyogenes was provided from the School of Environment and Science teaching laboratory, Griffith University, Australia. This bacterial stain has previously been reported to be resistant to multiple antibiotics, including β-lactam antibiotics (Cheesman et al. 2021). All bacteria were cultured in nutrient broth (Oxoid Ltd., Australia). Streak nutrient agar (Oxoid Ltd., Australia) plates (5 mm depth) were tested in parallel to ensure the purity of all bacterial cultures. All bacterial species were cultured at 37 °C for 24 h and were passaged and maintained in nutrient broth at 4 °C until used. For all bacterial assays, the bacterial cultures were adjusted to a cell count of approximately 10^8^ cells/mL (0.5 McFarland turbidity).
Evaluation of antibacterial activity
Susceptibility of the growth of each bacterium to sulfasalazine and its component moieties (sulfapyridine and 5-aminosalicylate) was assessed using a modified disc diffusion assay (Tiwana et al. 2020). A volume of 10 µL of each test was individually loaded into a paper disc ands applied to the surface of an agar plate that had been spread with 100 µL of the relevant bacterial culture and incubated for 24 h at 37 °C under aerobic conditions. Following the incubation period, the zones of inhibition (ZOIs) were measured to the nearest whole millimetre using a ruler and are expressed as means ± standard error of the mean (SEM). Standard discs containing penicillin-G (10 µg/disc), chloramphenicol (10 µg/disc), erythromycin (10 µg/disc) and tetracycline (10 µg/disc) were purchased from Oxoid, Australia and used as antibiotic positive controls. Filter paper discs infused with 10 µL of distilled water (containing 1% DMSO) were included on each plate as a negative control. All test and control were performed in triplicate in three independent experiments (n = 9) and are expressed as means ± SEM.
Determination of minimum inhibitory concentration (MIC) for bacterial growth
The MICs of sulfasalazine and its component moieties (sulphapyridine and 5-aminosalicylate) were evaluated by standard methods (Tiwana et al. 2020). Growth control wells (bacteria without test substances) and sterility controls (media without bacteria) were also included in duplicate for each plates. All plates were incubated at 37 °C for 24 h. p-Iodonitrotetrazolium violet (INT; Sigma-Aldrich, Australia) was dissolved in sterile deionised water to prepare a 0.2 mg/mL stock INT solution. A volume of 40 µL of the INT stock solution was added into all wells of a sterile 96-well plate and incubated for a further 6 h at 37 °C. MICs were visually determined as the lowest dose at which colour development was inhibited. All tests and controls were tested in duplicate in two separate experiments (n = 4), and are presented as mean values. The classification of all bacteria as susceptible or resistant to antibiotics was based on CSLI guidelines.
Combinational studies
Fractional inhibitory concentration (FIC) assessment
The combinational effects between SP and 5-AS were determined by calculating the sum of fractional inhibitory concentrations (ΣFIC) for each combination (Tiwana et al. 2020). FIC values for the individual components (a and b) were calculated using the following equations, where (a) represents the activity of the test sample and (b) represents that of the conventional antibiotic:
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{F}\mathrm{I}\mathrm{C}\left(\mathrm{a}\right)=\left(\frac{\mathrm{M}\mathrm{I}\mathrm{C}\left[\mathrm{a}\:\mathrm{i}\mathrm{n}\:\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{b}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\:\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h}\:\mathrm{b}\right]}{\mathrm{M}\mathrm{I}\mathrm{C}\left[\mathrm{a}\:\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{p}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{l}\mathrm{y}\right]}\right)$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{F}\mathrm{I}\mathrm{C}\left(\mathrm{b}\right)=\left(\frac{\mathrm{M}\mathrm{I}\mathrm{C}\left[\mathrm{b}\:\mathrm{i}\mathrm{n}\:\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{b}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\:\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h}\:\mathrm{a}\right]}{\mathrm{M}\mathrm{I}\mathrm{C}\left[\mathrm{b}\:\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{p}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{l}\mathrm{y}\right]}\right)$$\end{document}The ΣFIC was subsequently calculated using the formula ΣFIC = FIC(a) + FIC(b) and used to define the type of drug interaction. The effects were defined as synergistic (ΣFIC ≤ 0.5), additive (ΣFIC > 0.5-1.0), indifferent (ΣFIC > 1.0–4.0) or antagonistic (ΣFIC > 4.0) (Cheesman and Cock 2017).
Varied ratio isobologram studies
Where a synergistic interaction was identified, it was further evaluated using different ratios of the two components to determine optimal ratios for potential therapeutic use. Nine combinations containing 10 to 90% SP and 5-AS respectively were evaluated. All ratios were tested in duplicate in two independent experiments (n = 4). Mean data points for four replicates of each ratio were plotted on an isobologram and analysed to determine the optimal combination ratios to ascertain synergy. Data points on or below the 0.5:0.5 line on the isobologram indicated synergy; data points above the 0.5:0.5 line, up to and including the 1.0:1.0 line indicated an additive interaction; data points above the 1.0:1.0 line indicated non-interactive combinations.
Results
Growth inhibition of Proteus spp. bacterial triggers for rheumatoid arthritis
Proteus mirabilis growth was susceptible to sulfasalazine (SSZ) and its component moieties (sulfapyridine (SP) and 5-aminosalicylate (5-AS) (Fig. 2a). The pro-drug produced a ZOI of 8.5 ± 0.5 mm ZOIs, which indicates moderate growth inhibitory activity. Interestingly, both of the component moieties had substantially better P. mirabilis growth inhibitory activity, as determined by the larger ZOIs. SP was a particularly good inhibitor of P. mirabilis growth, with a ZOI of 11.2 ± 0.4 mm. Notably, this inhibition was comparable to the ZOIs produced against the chloramphenicol control (14.3 ± 0.6 mm). Interestingly, this P. mirabilis strain is a multi-antibiotic-resistant strain (Nel et al. 2020) that was not inhibited by any of the other conventional antibiotics screened here. Similar susceptibilities were seen for P. vulgaris (Fig. 2b), although this bacterium was substantially more susceptible to these compounds, compared to P. mirabilis. Indeed, a ZOI of 14.5 ± 0.5 mm was measured for SP against this bacterium. This is particularly noteworthy and was substantially superior to the growth inhibition produced by chloramphenicol (8.2 ± 0.4 mm) or tetracycline (6.7 ± 0.3 mm).
Fig. 2. Antibacterial activity measured as mm zones of inhibition (ZOI) of sulfasalazine (SSZ) and its metabolites sulfapyridine (SP) and 5-aminosalicylate (5-AS) against a P. mirabilis, b P. vulgaris, c K. pneumonia, d A. baylyi, e P. aeruginosa, f S. pyogenes. Pen = penicillin-G (10 µg/disc); Chl = chloramphenicol (10 µg/disc); Ery = erythromycin (10 µg/disc); Tet = tetracycline (10 µg/disc); NC = negative control (1% DMSO). Results are expressed as mean ZOIs from three independent experiments, each with internal triplicates (n = 9) ± SEM. * indicates results that are significantly different to the negative control (P < 0.01)
Growth inhibition of K. pneumoniae, a bacterial trigger for ankylosing spondylitis
SSZ, SP and 5-AS also inhibited the growth of K. pneumoniae (Fig. 2c) albeit, with similar efficacy to the inhibition of the Proteus spp. (as judged by the ZOI size). In particular, the ZOIs measured for SP (11.3 ± 0.6 mm) compare favourably to the inhibition by the chloramphenicol antibiotic controls (13.8 ± 0.4 mm) and was substantially better than erythromycin (6.8 ± 0.4 mm). Quite notably, this bacterial strain was completely resistant to all other conventional antibiotics tested.
Growth inhibition of A. baylyi and P. aeruginosa bacterial triggers for multiple sclerosis
The SSZ pro-drug and the SP and 5-AS component moieties also inhibited A. baylyi growth, (Fig. 2d). The SP moiety (ZOI = 11.3 ± 0.6 mm) was also substantially more potent than the SSZ pro-drug (7.8 ± 0.4 mm) or the 5-AS metabolite (8.3 ± 0.3 mm). SSZ and 5-AS had substantially lower efficacy (ZOI of 7.8 ± 0.4 mm and 8.3 ± 0.3 mm respectively), indicating moderate to low inhibitory activity. In contrast, the chloramphenicol control was a potent A. baylyi growth inhibitor (ZOI = 16.3 ± 0.6 mm). This bacterium was resistant to all other antibiotics tested.
Interestingly, P. aeruginosa was also relatively resistant to most of the conventional antibiotics tested (Fig. 2e). This strain was strongly inhibited by chloramphenicol (ZOI = 9.7 ± 0.3 mm) in contrast to tetracycline (ZOI = 6.8 ± 0.4 mm). It was completely resistant to all other antibiotics at the doses tested. By contrast, SP was a good inhibitor of P. aeruginosa growth (ZOI = 12.3 ± 0.3 mm). The 5-AS moiety was also a good growth inhibitor (9.6 ± 0.3 mm), although the parent SSZ pro-drug showed substantially lower activity (ZOI = 7.4 ± 0.4 mm). These findings therefore indicate that SP was the most effective inhibitor of the growth of both these bacterial triggers for multiple sclerosis.
Inhibition of S. pyogenes, a bacterial trigger for rheumatic fever
SSZ, SP and 5-AS were all strong inhibitors of S. pyogenes growth. SP was again the most potent antibacterial agent (Fig. 2f). Indeed, a ZOI of 16.3 ± 0.3 mm was measured for SP against this bacterium. This inhibition compared favourably to that of the conventional antibiotics. Only chloramphenicol (ZOI = 12.6 ± 0.4 mm) inhibited this S. pyogenes strain, whilst all other conventional antibiotics were completely ineffective, indicating that SP in particular, but also SSZ and 5-AS, inhibit the growth of the bacterial trigger of rheumatic fever and therefore may be useful in preventing and treating rheumatic fever in genetically susceptible people (although this remains to be tested in in vivo test models), as well as treating other diseases caused by this bacterium.
Quantification of minimum inhibitory concentration (MIC)
The relative antimicrobial potency was further evaluated by determining the MIC values using liquid dilution MIC assays (Table 1). Consistent with the antibacterial screening assays, SP was the best inhibitor of the growth of all of the bacterial triggers of the selected autoimmune diseases, with MICs ranging from 78 µg/mL (against P. mirabilis) to 625 µg/mL (against A. baylyi).
Table 1MIC values (µg/mL) of the prodrug (SSZ), its component moieties (SP, 5-AS) and some other antibiotics against some bacterial triggers of autoimmune diseaseSSZP. mirabilisP. vulgarisK. pneumoniaeA. baylyiP. aeruginosaS. pyogenes6251250625250012502500SP781561566253123125-AS62525002500250050005000Antibiotic ControlsPen----2.5-Chl1.252.51.92.51.251.25Ery--1.9--2.5Tet-1.251.9-1.9-NC------SSZ, sulfasalazine; SP, sulfapyridine; 5-AS, 5-aminosalicylate; -, no inhibitory activity was detected
Assessing fractional inhibitory concentration (FIC) against Proteusspp., a bacterial trigger for rheumatoid arthritis
A range of interactions were noted for combinations of SP and 5-AS when tested against these bacterial triggers of autoimmune inflammatory diseases (Table 2). This combination was synergistic against a single bacterium (P. vulgaris). This is particularly interesting as the mathematical model used to determine the class of interaction defines synergy as at least a four-fold increase in potency compared to the sum of the activities of the individual components when used as monotherapies. Thus, the SP + 5-AS combination may be particularly useful for the prevention and treatment of RA. The majority of the combinations were either non-interactive (33%) or additive (50%). Additive interactions, whilst not as large as synergistic combinations, may be still useful for treating bacterial infections. Therefore, the SP + 5-AS combination might also be valuable against P. mirabilis, K. pneumoniae and S. pyogenes. Thus, as well as having benefits against prodromal rheumatoid arthritis, this combination would also be useful in the prevention and treatment of ankylosing spondylitis and rheumatic fever. The non-interactive interactions, whilst not providing extra benefits compared to either treatment alone, also do not decrease each other’s activity. Therefore, there would not be decreases in efficacy when combining the compounds (or when they naturally dissociate in the GI tract, to the individual compounds). Interestingly, no antagonistic interactions were noted.
Table 2. Calculated FIC values for the extract and antibiotic combinations against some bacterial triggers of autoimmune inflammatory diseasesΣSingle Component Therapies MICs (µg/mL)Combinations (SP + 5-AS) SSZ
SP
5-AS
SP FIC
5-AS FIC
ΣFIC
Class of interaction
P. mirabilis 625 78 6250.230.46 0.69
Additive
P. vulgaris 1250 156 25000.150.32 0.47
Synergy
K. pneumoniae 625 156 25000.380.5 0.88
Additive
A. baylyi 250062525000.871.252.12Non-interactive P. aeruginosa 1250 312 50000.751.82.55Non-interactive S. pyogenes 2500 312 50000.420.48 0.9
Additive SSZ = sulfasalazine; SP = sulfapyridine; 5-AS = 5-aminosalicylate; FIC = fractional inhibitory concentrations; ΣFIC = sum of fractional inhibitory concentractions; MIC = minimum inhibitory concentration. Noteworthy MIC values (< 500 µg/mL) are indicated in bold text. Synergistic combinations are indicated in bold; additive combinational effects are indicated in italics
Varied ratio combination studies (isobolograms)
The SP + 5-AS combination produced synergistic activity against a single bacterium (P. vulgaris). Therefore, the combination was further evaluated across a range of ratios against this bacterium to evaluate the ideal ratios to use therapeutically (Fig. 3). Interestingly, the FIC(SP): FIC(5-AS) values aligned with the x (SP) axis, indicating that SP was the main growth inhibitory component of the mixture, and that the 5-AS component potentiated its activity. Of further note, all ratios containing ≤ 60% SP were synergistic. Therefore, any of these ratios would be particularly useful against P. vulgaris infections.
Fig. 3. Isobologram for varying ratio combinations of sulfapyridine (SP) and 5-aminpsalicylate (5-AS) against P. vulgaris, tested at various ratios. Results represent mean MIC values of four replicates. Ratio = % extract: % antibiotic. Ratios lying on or underneath the 0.5:0.5 line are considered to be synergistic (Σ FIC ≤ 0.5). Points between 0.5:0.5 and 1.0:1.0 are deemed to be additive (Σ FIC > 0.5-≤1.0)
Discussion
These studies highlight the potential of SSZ as a pro-drug (a drug precursor) that can be activated in situ, specifically where needed rather than in a distal organ such as the kidney, lung or liver. So SSZ is a useful therapeutic option for ulcerative colitis, in an environment carrying the largest population of drug-susceptible pathogens. SSZ is rather less effective for treating inflammatory bowel disease (IBD) located mainly in the small intestine. There is increasing evidence that bio-transformed products in the colon can communicate with the brain and assist in regulating immune-aggressive activities elsewhere throughout the body (Yemula and Sheikh 2024). Notably, 70% of the body’s immune system is associated with brain-gut connections (Mayer 2011).
SSZ is also used to treat extra-enteric inflammation such as rheumatoid arthritis (RA), urinary tract infections (UTIs), where Proteus species and Klebsiella species are debilitating (local) pathogens. Nanna Svartz’s efforts to introduce SSZ was in the context of RA following speculations that this disease might be a consequence of unknown infections (Svartz 1942, 1958). Those studies demonstrate Professor Svartz’s foresight that a molecular approach designing new, synthetic medications for chronic disease was a helpful addition to the relevant patterns of drug discovery mainly based on phytochemicals (such as artemisians for malaria, taxins for cancer, and salicin for fevers and pain).
Notably, SSZ and its metabolites inhibited the growth of all of the bacterial triggers of autoimmune diseases screened in this study. Proteus mirabilis was particularly susceptible to SSZ, with a MIC value of 625 µg/mL calculated in our study. Notably, this MIC is identical to that of a recent study (Ansaru et al. 2022). However, whilst that study reported noteworthy anti-P. mirabilis activity for SSZ, it did not screen the metabolites of the pro-drug for activity. In contrast, our study also screened sulfapyridine and 5-aminpsalicylate for the ability to inhibit P. mirabilis and found that sulfapyridine was a particularly potent inhibitor of the growth of this bacterium (MIC = 78 µg/mL). Interestingly, P mirabilis (as well as other Proteus species) have been reported to have azoreductase activity shown in vitro by dye decolourisation assays (Table 3; Mohanty and Kumar 2021; Olukanni 2011; Roxon et al. 1967). Taken together, these findings support the description of SSZ as a ‘magic bullet’ drug for the treatment of rheumatoid arthritis. SSZ functions as a pro-drug, which is ‘activated’ by the pathogens that it targets. Following oral administration, SSZ travels through the gastrointestinal tract until it encounters bacteria (including Proteus spp.) with azoreductase activity. The pro-drug is then metabolised to release sulfapyridine, which also inhibits the very bacteria responsible for its release, as well as 5-aminpsalicylate, which has anti-inflammatory effects. Thus, the bacterium activates/releases the drug that induces its own inhibition, at the site of the bacterial infection, thereby fulfilling Ehrlich’s definitions of a ‘magic bullet’ drug.
Table 3. Selected bacteria with reported azoreductase activityBacterial speciesGrowth conditionsCommentsOxidoreductase enzyme (where known)References Acinetobacter baumannii AerobicReduction of reactive blue 224 (RB-224) dyeAzoreductaseUllah et al. 2024 Bacillus badius AerobicRequires NADH and NADPH cofactorsAzoreductaseMisal et al. 2011 Bacillus cereus tested in anaerobic conditions97% Reactive red 120 dye decolourisation ion 72 hAzoreductase; NADH cofactorModi et al. 2010; Pricelius et al. 2007 Bacillus flexus AerobicIndanthrene Blue RS dyeAzoreductaseMohanty and Kumar 2021; Enterococcus faecalis Anaerobic> 95% decolourisation (multiple dyes) within 12 h.Azoreductase; NADH, NADPH, FMN cofactorsHandayani et al. 2007; Dhanve et al. 2009; Chen et al. 2008Escherichia coli JK109Anaerobic82% decolourisation of Direct Red 71 in 12 h.Azoreductase; NADPH cofactorJin et al. 2009; Phugare et al., 2011Klebsiella oxytoca GS-4-08AerobicSpectrometric measure of change in [NADH]Azoreductase activity; NADH cofactorHua and Yu 2019 Klebsiella pmneumoniae AerobicReduced methyl orangeAzoreductase; NADH cofactorDixit and Garg 2021Pigmentiphaga kullae K24Not definedDecolourised Orange I dyeAzoreductase; NADPH cofactorChen et al., 2009 Proteus mirabilis AerobicSeveral, including Indanthrene Blue RS dyeAzoreductaseMohanty and Kumar 2021; Olukanni 2011 Proteus vulgaris AerobicDecolorised tetrazine dyeAzoreductaseRoxon et al. 1967 Pseudomonas aeruginosa Aerobic94% decolourisation (multiple dyes) after 7 days.AzoreductaseMohanty and Kumar 2021; Crescente et al. 2016Shewanella oneidensis MR-1MicroaerophillicDecolourises methyl redAzoreductase; NADH cofactorYang et al., 2013 Staphylococcus aureusa AerobicDecolourised methyl redAzoreductase; NADPH cofactorChen et al. 2005 Streptococcus faecalis AerobicMultiple dyesAzoreductaseChung et at. 1993 Streptococcus mutans AerobicIdentification of azoreductase genesAzoreductaseKrieger et al. 2022 Streptococcus pyogenes AerobicDecolorisation of Brown 706 dyeAzoreductaseKhan et al., 2021 Xenophilius azovorans Not definedDecolourised Orange IIAzoreductase; NADH or NADPH cofactorBurger and Stolz 2010
Azoreductase activity has also been reported for the bacterial triggers of ankylosing spondylitis (Klebsiella pneumoniae), multiple sclerosis (Acinetobacter spp. and Pseudomonas aeruginosa) and rheumatic fever (Streptococcus pyogenes) (Table 3). Therefore, each of these bacteria may also metabolise SSZ to release sulfapyridine and 5-aminosalicylate. Notably, sulfapyridine was substantially more potent at inhibiting the growth of these bacteria than SSZ, with MIC values approximately 25% of the MIC calculated for the parent pro-drug. Indeed, strong growth inhibition was noted for sulfapyridine against K. pneumoniae,* A*,* baylyi*,* P. aeruginosa* and S. pyogenes, with MIC values of 156, 625, 312 and 312 µg/mL respectively.
Of perhaps greater interest, the metabolites displayed substantially greater antibacterial activity when tested in combination than either drug tested separately. Indeed, synergy was recorded for the combination of sulfapyridine and 5-aminpsalicylate against P. vulgaris. Notably, the mathematical model used in our study to evaluate the class of combination interaction defines synergy as an increase in activity of ≥ 4 fold of that of the individual components (Cheesman et al. 2017). Thus, azoreduction of SSZ not only releases the active component sulfapyridine but is also releases a synergiser (5-aminpsalicylate). Taken together, this results in anti-P. vulgaris activity that is approximately 32-fold more potent than that of the pro-drug SSZ.
Potentiation was also noted when sulfapyridine and 5-aminpsalicylate were tested in combination against P. mirabilis,* K. pneumoniae* and S. pyogenes. Whilst these were classified as additive effects rather than synergy, these results also highlight the additional benefits of having the two SSZ metabolites localized together. Indeed, the ƩFIC mathematical model used herein defines additive effects as a 2 to 4-fold increase compared to the individual components. Thus, the growth inhibitory activity of the SSZ metabolites in combination would be between 8-fold (assuming a 2-fold increase due to the additive combination) and 16-fold (assuming a < 4-fold increase due to the additive combination), compared to that of the pro-drug SSZ against those bacteria.
The antibacterial screening studies reported herein only tested against planktonic bacterial cultures. Notably, many of the bacterial triggers of autoimmune diseases are capable of forming biofilms as a protective mechanism. Biofilms generally decrease the antibiotic efficacy of antibacterial therapies, rendering them less effective (Srivastava et al. 2024). Therefore, future studies are required to evaluated the effects of SSZ, SP and 5-AS against biofilms of these bacteria. Additionally, the effects of SSZ and its metabolites need to be tested against beneficial species of the gastrointestinal microbiota. This is particularly relevant for other bacterial species with azoreductase activity, as the conversion of SSZ to its metabolites by those species may result in dysbiosis, leading to unexpected consequences. Further research is required.
Whilst our study did not examine the toxicity of SSZ or its metabolites, the safety data for these molecules has been extensively reported previously (Ye et al. 2024; Lehr et al. 1940). However, future studies also need to evaluate the toxicity of a combination of sulfapyridine and 5-aminpsalicylate, and to use that data to calculate safety indexes. Of further note, sulfapyridine has low solubility in aqueous solutions (< 1 mg/mL at 22 °C), which may impact the usefulness of this drug combination. The solubility of this molecule approximately doubles at 37 °C, which may partially mitigate this issue (Delgado et al. 2013). Additionally, the solubility of sulfapyridine increases substantially as the pH decreases. Indeed, the solubility in pH 5.5 increases to 120 mg/mL. Thus, sulfapyridine would have good solubility in the stomach, although this would decrease as it passes through the duodenum (pH ~ 6.1) to the disrtal ileum (pH ~ 7.5). Future studies are required to examine the bioavailability of solubilised sulfapyridine as it moves through the gastrointestinal tract.
Conclusions
SSZ has been used clinically since the 1940s to treat the clinical symptoms of a variety of inflammatory conditions, including spondylarthropathy, psoriasis, rheumatoid arthritis, uveitis etc. It is selectively activated by specific gastrointestinal bacteria (including bacterial triggers of some inflammatory diseases) via reductive cleavage of the diazo bond. This releases 5-AS, which mitigates tissue damage, and SP, which inhibits the growth of bacterial pathogens that trigger chronic inflammation. Notably, the antibacterial properties of the SSZ metabolite SP in vitro is substantially greater than for the parent pro-drug. Furthermore, combining sulfapyridine and 5-AS (as occurs when the diazo bond in SSZ is reduced), greatly increases the potency of the treatment of the combination compared to that of either metabolite (or the pro-drug) when tested separately in the assay model used in our study. Whilst this study highlights the potential of SSZ and its metabolites for inhibiting the bacterial triggers of rheumatoid arthritis (Proteus spp.), ankylosing spondylitis (Klebsiella pneumoniae), multiple sclerosis (Acinitobacter spp. and Pseudomonas aeruginosa), it is noteworthy that our study screened against a single strain for each bacterial species. Future studies should test these drugs against an expanded bacterial panel, including antibiotic resistant strains. Notable, whilst antibacterial activity was reported in our study for SSZ and its metabolites, the assays used did not discriminate between bacteriostatic and bactericidal activity and further studies are needed. Additionally, SSZ and is metabolites (alone and in combination) should also be screened against bacterial triggers of other inflammatory diseases. The bioavailability and safety of these drugs also requires further evaluation. Additionally, as the ‘magic bullet’ hypothesis for SSZ is reliant on the localised conversion of SSZ to SP and 5-AS, further testing is required to determine whether these bacterial species efficiently reduce SSZ to its metabolites in vivo. Furthermore, whilst we have demonstrated that 5-AS potentiates the antibacterial activity of SP, the potentiating mechanism was not evaluated and future studies are required to determine how this is achieved. As modulation of intracellular antibiotic concentration is a major method of antibiotic potention, efflux pump assays and electron microscopy studies to examine the bacterial membrane structure may be particularly useful.