The general synthetic procedure used was with a solution of KOH (Sigma-Aldrich, United States; 56 mg, 1.00 mmol) in EtOH (Sigma-Aldrich, United States; 10 ml) was cooled to 0°C, then morpholine (Sigma-Aldrich, United States; 87 μl, 1.00 mmol) and CS₂ (Sigma-Aldrich, United States; 151 μl, 2.50 mmol) were added to the solution successively. The resulting mixture was stirred at room temperature for 2 h, and then the solvent was reduced under vacuum. The first compound, Compound 1 (Dipotassium piperazine-1,4-dicarbodithiate, LRS01-057), was synthesized. ¹H NMR (400 MHz, DMSO-d₆) δ 4.34–4.22 (m, 4H), 3.51–3.39 (m, 4H). ¹³C NMR (100 MHz, DMSO-d₆) δ 215.02, 66.58, 50.01.

Following the general procedure, the title compound was synthesized in 57% as a white solid to Compound 2 (Sodium 4-(p-tolyl)piperazine-1-carbodithioate, LRS01-084) ¹H NMR (400 MHz, DMSO-d₆) δ 4.20 (s, 8H). ¹³C NMR (100 MHz, DMSO-d₆) δ 214.14, 49.51.

Following the general procedure using NaOH (Sigma-Aldrich, United States) as a base, the title compound was synthesized in 83% as a light yellow solid to synthesize Compound 3 (Sodium N-benzyl-N-methyldithiocarbamate, LRS01-075) ¹H NMR (400 MHz, DMSO-d₆) δ 6.98 (d, J = 8.1 Hz, 2H), 6.80 (d, J = 8.2 Hz, 2H), 4.41 (t, J = 5.1 Hz, 4H), 2.96 (t, J = 5.1 Hz, 4H), 2.16 (s, 3H). ¹³C NMR (100 MHz, DMSO-d₆) δ 214.73, 149.41, 129.78, 127.97, 116.15, 49.25, 49.09, 20.49.

Following the general procedure using NaOH as a base, the title compound was synthesized in 73% as a white solid for Compound 4 (Sodium N-allyl-N-methyldithiocarbamate, LRS01-077). ¹H NMR (400 MHz, DMSO-d₆) δ 7.28–7.19 (m, 4H), 7.19–7.12 (m, 1H), 5.43 (s, 2H), 3.24 (s, 3H). ¹³C NMR (100 MHz, DMSO-d₆) δ 215.70, 139.58, 128.42, 127.66, 126.75, 57.85, 40.99.

Following the general procedure using NaOH as a base, the title compound was synthesized in 56% as a white solid to Compound 5 (Sodium ((2S,3S)-1-ethoxy-3-methyl-1-oxopentan-2-yl) carbamodithioate, LRS01-072). ¹H NMR (400 MHz, DMSO-d₆) δ 5.83–5.70 (m, 1H), 5.08–5.02 (m, 1H), 5.02–4.97 (m, 1H), 4.73 (d, J = 5.8 Hz, 2H), 3.25 (s, 3H). ¹³C NMR (101 MHz, DMSO-d₆) δ 214.76, 135.06, 116.16, 57.57, 40.88.

Strains used here are S. pneumoniae TIGR4, D39, and Type 3, S. pyogenes M1, S. agalactiae ATCC 13813, and S. anginosus (Andrewes and Horder) Smith and Sherman emend. Whiley and Beighton 33,397™. For S. pneumoniae TIGR4, D39, and Type 3, S. agalactiae ATCC 13813, and S. anginosus M17 media (M17; BD Difco, United States) was prepared according to the manufacturer’s instructions. Briefly, 37.25 g of powder was suspended in 950 ml of Milli-Q grade water (≥18.0 MΩ cm⁻¹) and autoclaved at 121°C for 15 min before cooling to 50°C and adding 50 ml of a sterile 10% lactose (Beantown Chemical, United States) solution. For S. pyogenes, the culture medium was grown in Difco Todd Hewitt Broth (ThyB). For plating, Tryptic Soy Agar (TSA; Hardy Diagnostics, United States) was dissolved in Milli-Q water and autoclaved. After cooling autoclaved TSA, 5% defibrillated sheep’s blood (HemoStat Laboratories) of final volume and 20 μg/ml neomycin (Research Products international, United States) was added to the solution (for S. pneumoniae TIGR4 only). These plates (blood agar plates—BAP) were used for routine culture on solid media and “killing curve” serial dilution CFU counting. Bacteria from freshly streaked plates were placed into M17 and grown at 37°C in 5% CO₂, to an optical density (OD or OD₆₀₀) of 0.125 for growth curve assays and to an OD of ~0.300 for killing curve assays. To prepare working stocks, growing cultures are resuspended in fresh media +20% v/v glycerol and stored at –80°C. Aliquot viability and CFU counts were determined before use in experiments. Glycerol stock aliquots were diluted 1:5 into M17 with indicated copper and compound concentrations for assays.

Growth curve assay as described in Menghani et al. (2021). Briefly, copper stock solutions at 100 mM were prepared from CuSO₄ pentahydrate (VWR Life Sciences, United States) in Milli-Q water. Stock solutions of 100 mM Zn²⁺ were prepared from ZnSO₄ heptahydrate (VWR Life Sciences, United States). Stock solutions of 1 mM Compound compounds were prepared in Milli-Q water from solute derived and purified by the University of Arizona Center for Drug Discovery (ACDD). Sterile, individually wrapped clear 96 well polystyrene plates (Greiner Bio-One, USA) were arranged to test a range of concentrations combinations of Cu²⁺, Zn²⁺, and compounds. Frozen aliquots of bacteria were thawed and diluted five-fold into fresh M17 before adding 20 μl per well into a total well volume of 200 μl (1:50 total dilution). Assay plates were loaded into a Biotek Cytation5 (Biotek, Vermont, United States) pre-equilibrated to 37°C and 4% CO₂. Gas control settings were modified for an elevation of 720 m according to the manufacturer’s directions. The protocol-maintained temperature and CO₂, while measuring OD absorbance at 600 nm every 30 min for 16–20 h.

Killing curve assay as described in Menghani et al. (2021). Aliquots of bacteria were thawed and diluted 10-fold into assay conditions prepared in M17 media. Assay conditions included various concentrations of CuSO₄, Compound compounds, Zn²⁺, and Mn²⁺ (Sigma-Aldrich, United States). After exposure to the indicated conditions, bacteria were incubated at 37°C in 5% CO₂ for the stated time. Samples were serially diluted, plated on BAP, incubated overnight at 37°C in 5% CO₂, and counted to determine viable CFU. Colonies on each plate were counted and multiplied by the appropriate dilution factor based on which dilution it was to determine CFU/ml. For plates in which no colonies were visualized, this was deemed to be below the limit of detection (LoD) and is noted with a data point below the LoD line.

As performed previously (O’Brien et al., 2020) BL21 cells were transformed with DsRed2-pBAD and grown overnight in a shaker incubator at 250 rpm at 37°C on LB with ampicillin (Sigma-Aldrich, United States) 150 μg/ml. In the morning, the culture was centrifuged and resuspended on fresh media. Terrific broth (100 μg/ml amp) was inoculated to an optical density of 0.02. The culture was incubated at 37°C in a shaker incubator, and growth was monitored until 0.5–0.6 optical density, at which point the culture was induced with 0.1% arabinose (Tokoyo Chemical Industry), and the temperature was reduced to 16°C overnight. After 16–18 h, the culture was centrifuged, and the pellet was collected. The pellet was resuspended in buffer 50 mM TRIS (Sigma-Aldrich, United States) pH 7.4, 500 mM NaCl, 25 mM imidazole (Sigma-Aldrich, United States), 5% glycerol (RPI, United States), and lysed by sonication after adding protease cocktail of Leupeptin (Sigma-Aldrich, United States), Pepstatin A (Sigma-Aldrich, United States), phenylmethylsulfonyl fluoride or PMSF (Sigma-Aldrich, United States) and 0.2 mg/ml DNase (Sigma-Aldrich, United States). The lysate was centrifuged to remove debris, and the cell lysate was purified on an AKTA Pure via an affinity nickel column and buffer exchanged via a desalting column. The fractions were concentrated using a 10,000 mw CO (Millipore). The protein was left to mature overnight at room temperature. Concentration was calculated by sequenced and absorbance at 280 nm. Purity was assessed by 12% SDS PAGE. The protein was stored at 4°C and used within 2 weeks.

A relative affinity test was performed using the reversible fluorescence quenching of DsRed2 by copper (Eli and Chakrabartty, 2006; Sumner et al., 2006). The following was added to a 96 well plate (Greiner bio-one 655,096) in a total volume of 200 μl: 400 μM of freshly dissolved compound was added in 10 mM MOPS (Sigma-Aldrich, USA) pH 7.2 (chelated overnight with a Chelex 100 Biorad, United States and filtered sterilized through 0.2 μm membrane), 20 μM DsRed2, 100 μM copper (CuSO₄). In a Biotek Cytation5, the absorbance at 585 nm was monitored for 3 h to establish chelation of copper by the compounds and restore DsRed2 fluorescence.

To measure the internal metal concentration of the bacteria, strains were initially cultured on M17 + 5 mM lactose and frozen at −80°C in 20% glycerol. These glycerol stocks were used as the seed stock to inoculate 150 ml of M17 + 5 mM lactose. The bacterial culture was incubated at 37°C under 5% CO₂ until an OD of ~0.400 was reached. The culture was split into the indicated treatments and control. Incubation with treatments was performed at 37°C and 5% CO₂ for 30 min. Samples were quenched in a −3°C water bath to slow metabolism, followed by two washes of cold buffer (Tris 50 mM, NaCl 150 mM, EDTA 50 mM at pH 7.6), and centrifugation 3,500 ×g for 10 min at 4°C. Cold decanted samples were resuspended in 70% HNO_3, followed by overnight incubation at 65°C. After incubation, the samples were diluted to 2.5% HNO₃ using Milli-Q water at 18.1 MΩ. Bacterial plate counts were performed in TSA + 5% Sheep’s Blood through serial dilutions, as described above. Samples were analyzed for metal content using an iCAP PRO XDUO ICP-OES with a wavelength 324.8 nm copper. Standards were made using the iCap Series Multi-element test solution ICAP 6000 series Validator from Thermo scientific, and metal content of the washed samples was calculated using the Qtegra software. Experiments were performed in triplicate.

All mouse studies were conducted with prior approval and under the guidelines of the IACUC at the University of Arizona (IACUC protocol number 18-410, R35 GM128653) as previously performed (Menghani et al., 2021, 2022). All mice were maintained in a biosafety level 2 (BSL2) animal facility and monitored daily for signs of moribund. Eight-week-old female BALB/cJ mice (Jackson Laboratory) were anesthetized with 3% isoflurane and intranasally infected with an inoculum of 10⁷ CFU viable S. pneumoniae in 25 μl of Tris-buffered saline (TBS; 50 mM Tris, 150 mM NaCl, pH 7.4). Cohort controls were given 25 μl of TBS. At 4-, 8-, or 24-h post-infection, mice were treated with doses of intranasal Compound 3 or Compound 4 (approximately 1.6 mg/kg) in 25 μl of TBS. Mice were sacrificed by CO₂ asphyxiation and immediately dissected for lung and blood collection 48 h post-infection. Lung tissue was collected into 1.5 ml tubes containing 500 ml of phosphate-buffered saline (PBS; Gibco, United States) after a brief initial wash in 500 μl of PBS to remove any excess blood during dissection. The tissue was then homogenized and centrifuged for 30 s at 400 ×g. Blood samples (5 μl volume) were placed in a 45-μl volume PBS solution with heparin (10 UI/ml). Both lung and blood samples were then serially diluted at 1: 10 and plated on blood agar plates and incubated overnight at 37°C and 5% CO₂ for growth. The resulting bacterial colonies were counted for quantification and comparison.

J774A.1 macrophages (ATCC, United States) were maintained in a 37°C, 5% CO₂ incubator with Dulbecco’s Modified Eagle’s Medium (DMEM; Sigma-Aldrich, United States) containing fetal bovine serum (FBS [10%vol/vol]; Sigma-Aldrich, United States), glutamine (2 mM; Sigma-Aldrich, United States), penicillin (50 units/ml; Sigma-Aldrich, United States), streptomycin (50 mg/ml; Sigma-Aldrich, United States), and NaHCO₃ (0.015%). Cells were grown to 90% confluence in 12-well tissue culture plates (Greiner CELLSTAR®, Greiner Bio-One, United States) in 1 ml/well of growth medium. On the morning of the experiment, macrophages were washed twice with 1 ml PBS, and resuspended in 1 ml of “Serum-Free DMEM” growth medium without antibiotics, glutamine, NaHCO₃ or FBS but supplemented with 5 ng/ml IFN-γ (Bio Basic, United States), 400 ng/ml LPS (EMD Millipore, MilliporeSigma, United States) for “priming of macrophages” experiment. For “post hoc killing efficiency” experiment, macrophages were resuspended in 1 ml of Serum-Free DMEM containing 32 μM Compound 3, 5 ng/ml IFN-γ, and 400 ng/ml LPS.

Macrophages were incubated at 37°C, 5% CO₂ for 12 h. Glycerol stocks of TIGR4 S. pneumoniae kept at OD 0.3 are removed from-80°C storage, diluted into four 15 ml conical tubes of 5 ml total M17 + Lactose containing no additives (“Untreated”), 32 μM Compound 3, 250 μM CuSO₄, and 250 μM CuSO₄ + 32 μM Compound 3, respectively, for pre-treatment of bacteria experiment. Prior to incubation of bacteria in 37°C, 5% CO₂, an inoculum plate is made by serial diluting 100 μl from the no additives conical. After 15 min of incubation, bacteria are centrifuged at 4500 ×g for 10 min and resuspended in DMEM without antibiotics, glutamine, NaHCO₃ or FBS. Macrophages are removed from incubation, media is removed, washed with 1 ml PBS twice and then infected with 100 μl of S. pneumoniae solutions for both experiment types, corresponding to a multiplicity of infection (MOI) of 10 bacteria per macrophage. The 12-well tissue culture plates were centrifuged at 200 ×g for 2 min to facilitate co-culturing.

Wells were then washed twice with PBS at the given time-points (5, 15, and 30 min); each wash was followed by a 5-min incubation in a 37°C, 5% CO₂ incubator in DMEM containing gentamicin (50 μg/ml). Macrophages were lysed in 0.02% SDS by pipetting up and down in ddH₂O and serially diluted to determine the counts of viable intracellular bacteria. Data were normalized to the level of killing observed for the untreated TIGR4 bacteria for each assay.

Statistical significance was analyzed using Student t-test (two-tailed, unpaired), two-way ANOVA, or one-way ANOVA and Dunnett’s multiple comparisons test (Prism 9.20; GraphPad Software, United States). The p-values were as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

DMDC has known copper dependent toxicity against S. pneumoniae TIGR4, D39 and Type 3 (Menghani et al., 2021, 2022). Before investigating possible DMDC derivatives against the pneumococcus alone, it was tested if DMDC and copper was an effective broader spectrum treatment against known pathogenic streptococcal strains S. anginosus, S. pyogenes, and S. agalactiae as a “go/no go” checkpoint. DMDC with copper had significant bactericidal effect (at least a 90% decrease) after 4 h with S. pyogenes (Figure 1A) and 2 h with S. anginosus and S. agalactiae (Figures 1B,C). The effect of DMDC with copper against S. anginosus and S. agalactiae were the closest in comparison to S. pneumoniae, for which DMDC killed roughly 3 logs of bacteria at the two-hour time point, however, S. pyogenes was slower to die taking 4 h constantly to get to 1 log of killing. Therefore, with the potential for treatment of streptococcal species, and previously reported fungi and schistosomes (Menghani et al., 2021), derivatizing DMDC proceeded.

DMDC was derived into five compounds that retain putative copper ionophore properties (Table 1). The five derived compounds were named “Compound 1” through “Compound 5.” Each compound was initially screened by performing in vitro growth curves to determine if a growth defect was associated with either the compound alone or in combination with copper. These bacteria were started at lag phase growth. Next, successful compounds were tested in killing curves to determine if there was a bactericidal or bacteriostatic effect of copper, compound, or combination treatment. These curves were done on bacteria at log phase growth. While it is more beneficial to eliminate the bacteria than to just restrict growth, the growth curves offer an entry point to analyze compounds that use less resources that the killing curves and correlate with the killing curves. Compound 1, Compound 2, and Compound 5 had no effect, and thus, experiments with these compounds were discontinued (data not shown). Compound 3 and Compound 4 showed statistically significant growth defects for the combination treatments in comparison to an untreated control in TIGR4 S. pneumoniae (Figures 2A, C). Further, both Compound 3 and Compound 4 showed significant bactericidal activity in killing curves with 2 or more logs for each (99% killing; Figures 2B, D). From these data, it was concluded that Compound 3 and Compound 4 are viable copper-dependent bactericidal antibiotics that warrant further investigation.

The ability of the compounds to increase the concentration of copper inside TIGR4 was examined next. It was previously observed using DMDC plus copper that there was a significant increase in internal copper relative to copper or DMDC treatments alone (Menghani et al., 2022). Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) was used to measure the metal content within the bacteria and corroborate the mechanism seen previously. Compound 3 and Compound 4 showed significant increases in the combination treatment relative to the compound or copper alone increasing the bacteria associated copper by 100 x over the control (Figures 3A–D). Thus, Compound 3 and Compound 4 facilitate copper entry into the bacteria.

Determining copper affinity is notoriously difficult due to relatively high-affinity values. Thus, the ability to at least compare the binding rates of the compounds to DMDC using DsRed2, a protein that has its fluorescence quenched by copper, was pursued. The stoichiometry of DMDC to copper is 2:1 (Lei et al., 2015); this premise was used to provide the compounds in excess for the chelation assays. Using the compounds to compete for copper to reverse the quenching of copper to DsRed2, initial hierarchies in their ability to chelate copper and restore fluorescence were found (Figure 4). Interestingly, bactericidal compounds Compound 3 and Compound 4 exhibited more prolonged periods to quench DsRed2 as compared to parent compound DMDC. All had an initial dip implying that DsRed2 bound copper initially, but then the compounds took varying amount of time to equilibrate. Thus, while relative affinities and compound-mediated copper influx could play a role in the overall effectiveness of the antibiotic capacity of the compound, this process is not linear and likely highly dependent on the microbe-of-study’s phenotype and ability to flux metal in and out of the organism.

In a targeted small-molecule screen that identified DMDC as a potent copper-dependent antibiotic, combinations of many divalent metals with DMDC were investigated to determine if other metals can display toxicity (Menghani et al., 2021). To test the zinc-dependent toxicity with the derivatives, killing curves were performed using TIGR4 against combinations of zinc, Compound 3 or Compound 4, and combinations of compound + zinc. It was observed that the combination of 500 μM Zn²⁺ + 16 μM Compound 3 is bacteriostatic in that there was no significant growth between the untreated growth at 30 min compared to the treated grown at 120 min (Figure 5A). Further, by virtue of still seeing copper dependent toxicity and bacterial killing, it was observed that adding 500 μM Mn²⁺ does not rescue the copper-dependent toxicity seen with 250 μM Cu²⁺ + 32 μM Compound 3 (Figure 5B). These experiments were repeated for Compound 4 (Figures 5C,D). While 500 μM Zn²⁺ + 16 μM Compound 4 was found to be bacteriostatic, strikingly, the addition of 500 μM Mn²⁺ rescued the copper-dependent toxicity seen with 250 μM Cu²⁺ + 32 μM Compound 4 (Figures 5C,D). From these data, it can be concluded that Compound 3 and Compound 4 are dithiocarbamate compounds whose copper-dependent toxicity may be mediated by facilitating mismetallation events within the bacterium but potentially at different affinities as seen in Figure 4.

To directly test if the two compounds were promising antibiotics to progress along the drug discovery pipeline, they were tested in a murine pneumonia model previously used by our lab (Menghani et al., 2021). Mice were given a 100% lethal dose (LD₁₀₀) of S. pneumoniae TIGR4, then at 8 h later were given Compound 3 or 4. Then, the bacterial burden was measured after 48 h. A significant decrease in bacterial titers in the blood and lungs 48 h post-infection in 8-week-old BALB/c mice given 25 μl of 10 mM Compound 3 was observed (approximately 1.6 mg/kg; Figures 6A,B). Compound 3 was also given at 4 and 24 h and observed 2 log decrease at 4 h, and reduction at 24 h, but not a significant one. No significant decrease was noted for mice at 8 h given the same dosage of Compound 4 (Supplementary Figures 1A,B). Further, as previously observed with DMDC and copper (Menghani et al., 2022), the combination treatment of Compound 3 plus copper also made the pneumococcus more susceptible to activated macrophage mediated killing via a macrophage killing assay as compared to no treatment, copper, or Compound 3 alone (Figure 6C). These data show that Compound 3 is a promising antibiotic against S. pneumoniae TIGR4 with efficacy in vivo.

To determine if Compound 3 and Compound 4 are antibiotics that can be widely used against S. pneumoniae, killing curves were performed with these compounds against the D39 strain (serotype 2) and a serotype type 3 strain that is highly encapsulated (ATCC® 6,303™). These strains offer a small by representative sample of the more than 100 serotypes that exist for the pneumococcus. Compound 3 and Compound 4 were bactericidal in combination with copper against the D39 and the Type 3 strain, and more so (by several logs) than the effects of DMDC versus these strains as seen previously (Figures 7A–D) and to a greater degree than with TIGR4 at the 30, 60 and 120 min time point (Figure 2; Menghani et al., 2021). The intrabacterial concentration of copper with Compounds 3 and 4 against D39 and the type 3 strain was also measured. Because increased killing as compared to TIGR4 at the same concentrations (32 μM compound and 500 μM copper) was observed, lower respective concentrations (8 μM Compound and 125 μM copper) to make sure the bacteria were still viable at the time of the measurement were used. Compound 4 increased copper uptake in D39 and the type 3 strain to a similar degree as with TIGR4 (Supplementary Figures 2A–D). At these concentrations, only Compound 3 showed increased concentration of copper inside of the type 3 strain, and not D39 (Supplementary Figures 2A–D). However, not seeing the increase likely has more to do with the concentration used so viable cells were being compared.

Lastly, Compound 3 and Compound 4 were tested against S. anginosus, S. pyogenes, and S. agalactiae. Interestingly, there was variance between how each compound affected the survival of the Streptococcal species tested. Compound 3 only showed killing capacity vs. S. pyogenes killing roughly a log of bacteria in 4 h while the untreated grew roughly a log in this amount of time (Figures 8A, C, E). In contrast, Compound 4 demonstrated killing ability of all the tested strains killing roughly one log in each case (Figures 8B, D, F). Taken together, derivatizing DMDC is a viable strategy to targeting different strains of pneumococcus and other Streptococcal species, however, there is variance to the bacteria that warrants investigation to determine what makes a copper-based antimicrobial robust against one strain or species versus another.