Multimodal biofilm control strategies for spacecraft water systems: Evaluating coatings, nutrient removal, and biocides for improved sustainability

The organisms used in the experiments are direct and frequent isolates from the ISS water system (Zea et al., 2020), and are a subset of the organisms selected by Velez Justiniano et al. for their previous monoculture work (2021). Three bacteria and one fungus were used in co-culture for all experiments conducted and include the bacteria Burkholderia contaminans, Methylobacterium organophilum, and Ralstonia insidiosa, and a filamentous fungus, Coniochaeta mutabilis (formerly Lecythophora mutabilis), see Supplemental Table 1 for internal and external strain identifiers. Genomes for the strains of B. contaminans and M. organophilum used here have been published (Velez Justiniano et al., 2023). The bacteria were provided by Boeing, and the fungus was provided by Marshall Space Flight Center (MSFC). Stocks of the organisms were stored at −80 C in a 50/50 mixture of glycerol and cell culture in their corresponding nutrient broth. Streak plates of each organism were prepared on R2A agar (BD Difco) to ensure culture purity. To prepare the inoculum, individual colonies were used to inoculate flasks of sterile full-strength tryptic soy broth (TSB; BD Difco) for bacteria or Sabouraud dextrose broth (SDB; BD Difco) for fungus. Inoculum flasks consisted of 25 mL of sterile nutrient broth in a slip-capped 125 mL baffled flask and were placed in a shaker-incubator at 30 C and 225 rpm. M. organophilum and C. mutabilis were incubated for approx. 48 h while B. contaminans and R. insidiosa were incubated for approx. 24 h. Once dense cultures were achieved, they were washed via three several steps of centrifugation (8,000 x g) and resuspension in fresh sterile 0.4% NaCl solution (pH 5.5). After excess nutrient broth was washed away, the optical density of each culture was measured at 600 nm. The density was adjusted so addition of 100 μL of each culture to the Center for Disease Control (CDC) reactor (BioSurface Technologies, Bozeman, MT) (ASTM, 2021) resulted in initial densities of 10⁴ CFU/mL for each of the four organisms.

Inconel® 718 (a nickel-chromium alloy) and Teflon^TM (polytetrafluoroethylene) cylindrical coupons (12.7 mm diameter, 3.8 mm thickness) were procured (BioSurface Technologies; Bozeman, MT). Prior to coating or use in the reactor, coupons were washed by soaking and sonication in soapy water for approx. 5 min followed by a rinse step and additional sonication of approx. 5 min in Milli-Q water (resistivity 18.2 MΩ·cm). The coating tested was provided by The Sherwin-Williams Company (Cleveland, OH) and is a commercially available product known as Sher-Loxane® 800, product numbers B80W501 (extra white) and B80V500 (standard temperature hardener). The coating was applied with a paintbrush such that a homogenous thickness of coating was present on the flat surface of the coupon (Figure 1). After at least 48 h of drying at room temperature, the coupon was flipped over so the other flat surface could be coated with the freshly mixed coating. The round edges of the coupons were not coated, but the painted coating dripped over the side in some areas. No additional wash step was completed after coating. Contact angles were measured for uncoated and coated surfaces with a VCA 2500 XE system and surface topography was measured with a Bruker Dektak stylus profilometer (Chi et al., 2015) (Supplemental Table 2).

Figure 1.

Coupons for sampling biofilm accumulation in the CDC reactor.

CDC biofilm reactors (BioSurface Technologies) were assembled and sterilized empty in an autoclave at 121 C and 21 PSI for at least 30 minutes. Uncoated coupons were loaded before the reactor was autoclaved (8 rods loaded with 3 coupons per rod). For experiments that utilized coated coupons, the reactor was autoclaved without coupons. After autoclaving, coated coupons were loaded and sterilized using the same method outlined previously (Mettler et al., 2022). In summary, the reactor was placed in a Labconco class II biosafety cabinet, coupon holder rods were removed, coated coupons were loaded, and then rods were submerged in 99% isopropanol for at least 30 s before air drying and replacement into the reactor.

The medium used in the experiments was developed by researchers at the Center for Biofilm Engineering (CBE) in collaboration with MSFC (Table 1). The medium was designed to mimic the ISS wastewater tank nutrient levels and chemical composition (50% influent from humidity condensate and 50% influent from urine distillate) during a biofouling event. An early version of the recipe was published by Velez-Justiniano et al. (2021). Slight modifications were made to the formulation, including the addition of minor/trace nutrients (MTN), removal of KCl, and reduction of KI concentrations (Sandvik et al., 2022). The new version of the medium was termed microbial ersatz MTN. To prepare the medium, Milli-Q water was autoclaved for 3 h in 20 L glass carboys. After autoclaving, filter sterilized stock solutions were added to the carboy. A 10 mL aliquot of medium was removed from the carboy with a sterile serological pipette for measuring pH. The pH was adjusted on the aliquot to 5.55 ± 0.05 with hydrochloric acid (1 M) or sodium hydroxide (1 M) and scaled to adjust the medium in the carboy. Sterile HCl or NaOH was added to the carboy and the pH was checked again. The iterations tested in the experiments here include a full-strength version of microbial ersatz MTN and a no phosphorus version which excluded the phosphate stock. The no phosphorus medium was prepared in carboys that had not held full strength medium, or if they had, they were acid washed with 10% HCl and rinsed thoroughly with Milli-Q water.

Silver (I) fluoride (99+%, Thermo Scientific) was used in experiments featuring biocide dosing. The biocide was prepared at a concentration of 216.66 mg/L so that when it was added to the reactor, it was present at a concentration of 2 ppm. The silver (I) fluoride was added to Milli-Q water and filter sterilized into a sterile container with tubing connected to the reactor.

The CDC reactor experiments were conducted to evaluate control strategies alone and in combination with one or two other control strategies. Three biological replicates of each experiment were completed. Figure 2 presents a schematic to visualize the iterations of each experiment. The format of Figure 2 is carried through the data and microscopy figures that present the results of the experiments.

Figure 2.

Biofilm control methods used in each of the eight iterations of experiments. The empty section in the upper left represents the base condition of unmitigated biofilm growth. This layout is used throughout the results to present data for each experiment.

Sterile reactors loaded with coupons were filled to the effluent port with fresh, sterile medium (approx. 325 mL). For inoculation, 100 μL of each washed culture was added to the reactor and the stir bar was set to 125 ±10 rpm. The full consortium was used for each experiment and computational fluid dynamics models indicate the stir rate resulted in shear stresses of 0.365 +/− 0.074 Pa across the surfaces of all 24 coupons (Johnson et al., 2021). Experiments took place at room temperature and with ambient oxygen concentrations. About 5–10 min after inoculation, a sample of bulk fluid was taken from the reactors to validate the cell concentrations at the start of the experiments via viable plate counts.

After 24 h batch operation, continuous flow was initiated. Fresh, sterile medium was pumped (Masterflex L/S, Cole Parmer) into the reactor at 2 mL/min, creating a residence time of about 162 min. The medium used for continuous flow was the same as that used for the batch phase operation. For reactors receiving biocide, the first dose corresponded with the initiation of continuous flow, and silver fluoride was dosed every 24 h at 3 mL/min for 1 min.

The first sampling point took place at the completion of 24 h batch conditions, just prior to the start of continuous flow. Sampling consisted of removing a reactor rod and gently dipping it into 30 mL of sterile 1M phosphate buffered saline (PBS). After rinsing, the coupons were removed and put into sterile 50 mL Falcon tubes containing 10 mL sterile PBS. The biofilm was removed from the coupon surface and disaggregated via a vortex mixing and sonication series of 30 s each for a total of 2 min 30 s (vortex, sonicate, vortex, sonicate, vortex). For Days 1, 3, and 5, three coupons were sampled for viable plate counts of attached biofilm. For Day 7, two coupons were sampled for plate counts and one was sampled for microscopy. Planktonic samples were taken with a sterile serological pipette from the open port on the CDC reactor where the rod was removed. The rod was then replaced with a sterile dummy rod to maintain similar fluid dynamics in the reactor. Bulk reactor fluid was also sampled 6 h after the biocide dose in experiments that received silver fluoride to monitor effects of the silver fluoride dose on planktonic populations. Planktonic samples were vortex mixed for 30 s to disaggregate potential biofilm aggregates. Disaggregated biofilm and planktonic samples were plated on selective agar for enumeration (see Table 2 for the agar used). Agar plates were allowed to dry at room temperature before placing them into a 30 C incubator. Enumeration took place after the formation of visible colonies, typically after 2 days for general counts, B. contaminans, and R. insidiosa, and 4–5 days for C. mutabilis and M. organophilum. The recipes for selective agar were made specifically for the strains used in this study. The antibiotics and their respective concentrations were determined experimentally to select for only the desired organism within the defined consortium.

During Day 7 sampling, coupons were stained for confocal analysis. After gently rinsing in PBS, coupons were placed into sterile 24-well plates. The well was then flooded with freshly prepared LIVE/DEAD^TM BacLight^TM stain and incubated in the dark for 15 min at room temperature. The stain was removed from the well and the coupons were washed three times with filter-sterilized Milli-Q water. Next, the well was flooded with a 50/50 solution of sterile 1X PBS and Calcofluor White and incubated in the dark for 10 min at room temperature to stain fungal cells. After the incubation period, the stain was removed, and the coupons were washed with filter-sterilized Milli-Q water three times. The wells were then flooded with a solution of 4% paraformaldehyde and 2.5% glutaraldehyde. The 24-well plate was covered with a foiled lid and wrapped several times with parafilm. The wrapped plate was then placed into a closed plastic bag and stored at 4 C until confocal microscopic imaging at excitation wavelengths of 405 nm, 483 nm, and 568 nm on a Leica DM6 Multiphoton Upright Confocal Microscope. At least four random positions on each coupon were imaged from each biological replicate. The published images are a single field of view representative of all biological replicates.

The effects of Sher-Loxane® 800, phosphorus removal, and silver fluoride dosing were evaluated on a species-by-species basis. To investigate the effects on the entire biofilm and planktonic community, the viable plate counts on the non-selective R2A plates were used. Mixed effects ANOVA models and statistical comparisons were generated using MiniTab (version 21.1).

The viable biofilm densities are presented in Figure 3. The average of the three biological replicates is presented in the graphs and consisted of three technical replicates each for Days 1, 3, and 5 (n=9 coupons total), and two technical replicates each for Day 7 (n=6 coupons total).

Figure 3.

Viable biofilm density for each of the eight iterations of biofilm control combinations. Continuous flow was initiated on Day 1 after the sample was taken. The grey dashed line represents the limit of detection, y=1.69. Error bars are standard deviation.

The experiments featuring unmitigated biofilm growth (no coating, full strength medium, and no biocide; Figure 3A) fostered the highest viable biofilm density of all the experiments. R. insidiosa was present at a viable density of approx. 8.4 log₁₀CFU/cm² on Teflon coupons and 8.1 log₁₀CFU/cm² on Inconel coupons by Day 7 and made up the majority of the biofilm. B. contaminans was present at the next highest density of 6.8 log₁₀CFU/cm² on Teflon and 6.0 log₁₀CFU/cm² on Inconel. C. mutabilis and M. organophilum did not contribute very much to the viable cell density of the biofilm and were only present at densities ranging from 3–4 log₁₀CFU/cm² on Inconel and Teflon. All species in the biofilm appear to reach steady state by three days after inoculation.

In experiments featuring the no phosphorous medium as the only method of biofilm control (Figure 3B), the general trend is a reduction in viable biofilm density by 0.5–1 log₁₀CFU/cm². The experiments with silver fluoride dosing as the sole method of biofilm control (Figure 3C) featured a reduction in viable biofilm density from Day 1 to Day 3 (biocide was dosed on Day 1 immediately after sampling). R. insidiosa still made up the majority of the biofilm, with a final viable density of about 5.2 log₁₀CFU/cm². B. contaminans, C. mutabilis, and M. organophilum were all present in the Day 7 biofilm near the limit of detection, ranging from 1.3–2.3 log₁₀CFU/cm².

The experiments on uncoated coupons with nutrient limitation and biocide as the biofilm control methods (Figure 3D) featured a reduction of about 0.5 log₁₀CFU/cm² in R. insidiosa viable density compared to experiments with just silver dosing (Figure 3C). B. contaminans density remained about the same compared to experiments with just silver dosing, while C. mutabilis and M. organophilum density dropped to below the limit of detection by Day 7.

The coated coupons effectively delayed viable biofilm accumulation by several orders of magnitude for at least three days after inoculation (Figure 3E vs 3A). This delay presented an excellent opportunity for additional control methods to further suppress biofilm accumulation. The final viable biofilm density was similar to the unmitigated case, reaching approx. 8.0 log₁₀CFU/cm² for B. contaminans and 7.5 log₁₀CFU/cm² for R. insidiosa (a total cell density of approx. 8.5 log₁₀CFU/cm²). An important outcome in the reactors with coated coupons was a change in the composition of the biofilm. The biofilms on uncoated coupons were always dominated by R. insidiosa while the biofilms on coated coupons were mainly dominated by B. contaminans in reactors not dosed with biocide. The densities of C. mutabilis and M. organophilum in this experiment (Figure 3E) were also several orders of magnitude lower than in the unmitigated case (Figure 3A).

Mirroring the comparison between Figure 3A and 3B, the inclusion of nutrient limitation as a biofilm mitigation method on coated coupons (Figure 3F) reduced viable biofilm accumulation by approx. one order of magnitude compared to biofilms with the coating as the only method of biofilm control (Figure 3E).

The greatest reduction in biofilm accumulation occurred when silver fluoride dosing was incorporated as a biofilm control method with the coating (Figure 3G) and when all three strategies were implemented together (Figure 3H). In these experiments, the biofilm density was significantly reduced on Day 1 compared to the unmitigated case (Figure 3A) and the density was generally highest on Day 1. Throughout the experiments, the biofilm density remained low, aside from R. insidiosa on Day 5 in the experiment with coating and silver fluoride (Fig 3G). By Day 7, the experiments with silver fluoride and Sher-Loxane® 800 showed a viable biofilm density of 3.2 log₁₀CFU/cm², which is five orders of magnitude lower than the unmitigated case. However, the experiment featuring all three methods of biofilm control still outperformed the coating and silver dose approach, resulting in a biofilm density of about 1.5 log₁₀CFU/cm².

Mixed effects statistical analysis was fit to the log-transformed CFU data to investigate the resulting viable biofilm density in each treatment method more definitively. The species were evaluated individually, and the biofilm was evaluated as a whole, using counts from the R2A agar plates. The ANOVA models used the experiment replicate as a random factor, and medium, biocide, coating, material, and day all as fixed factors. The p-values associated with the mixed effects F-tests for most individual factors are presented in Table 3, see Supplementary Table 3 for all remaining p-values.

Additionally, follow-up Tukey pairwise comparisons were performed for the three-way interactions of medium, biocide, and coating. These specific comparisons were performed as they were the biofilm control methods investigated. The Tukey pairwise comparisons for biofilm densities are presented in Supplementary Table 4. The comparisons indicate that for R. insidiosa and the whole biofilm (R2A), the addition of each control method contributes significantly to reduction of viable cells. Further, the incorporation of additional biofilm control methods may not significantly reduce viable biofilm accumulation for the other organisms in the biofilm; specifically, the coating for B. contaminans and removal of phosphorus from the medium for C. mutabilis and M. organophilum.

In addition to biofilm measurements, bulk fluid samples from the reactors were taken to monitor the reactor planktonic populations. The averages of the three biological replicates are presented in Figure 4. The planktonic populations generally mirror the trends observed in the viable biofilm accumulation (Figure 3) in all reactors aside from those treated with silver fluoride. In the reactors receiving biocide (3 mL in 60s every 24 h), the planktonic population was significantly reduced and there was a cyclical suppression and rebound of viable planktonic (detached biofilm) cells. Mixed effects statistical analysis was fit to the log-transformed CFU data to more definitively investigate the resulting viable planktonic cell density in each treatment method. The models used the experiment replicate as a random factor, and medium, biocide, coating, and day were all fixed factors. The material type was not included as a factor in planktonic analysis because both Teflon and Inconel were present in the same reactor. The p-values associated with the F-tests for all individual factors are presented in Table 4, see Supplementary Table 5 for all remaining p-values.

Figure 4.

Viable planktonic cell density for each of the eight iterations of biofilm control combinations. Continuous flow began on Day 1 after the sample was taken. The grey dashed line represents the limit of detection, y=1.3. Error bars represent standard deviation.

Tukey pairwise comparisons were also performed for the three-way interaction of media, biocide, and coating presence in the reactor. The comparisons are present in Supplementary Table 6. Results indicate that the biofilm control methods had minimal effect on the planktonic densities of C. mutabilis and M. organophilum in the reactor. This result is also likely due to the low presence of the organisms in the reactor without any biofilm control method incorporated, meaning there was not much reduction in planktonic density that could occur with the biofilm control methods. The Tukey comparisons indicate that the incorporation of biofilm control methods had a greater impact on the planktonic densities of B. contaminans, R. insidiosa, and the whole community (R2A counts).

Confocal microscopy images were generated for biofilms that had accumulated on the coupons after seven days in the reactors. Figures 5 and 6 present 3D projections of these images for Inconel and Teflon surfaces, respectively. The greatest reduction in biofilm accumulation can be seen when comparing panels C and D to G and H in Figures 5 and 6. Panel E in each figure presents biofilms on Sher-Loxane 800® coating grown in full strength medium without biocide dosing. The biofilms appear much denser than those grown on uncoated coupons (panel A), however, some of the fluorescence in the Sher-Loxane 800® images is due to the coating itself. The coating became highly fluorescent in the wavelengths used to image the biofilm (excitation wavelengths of 405 nm (blue), 483 nm (green), and 568 nm (red)) and appears in different false colors due to image processing to improve visibility of the cells on the surface. The microbial cells present on the coating have more intense fluorescence than the coating, so the cells remain distinguishable from the surface.

Figure 5.

Confocal images of biofilms accumulated on Inconel CDC coupons seven days after inoculation. Viable bacterial cells are in green, non-viable bacterial cells are in red, and fungal cells are in blue. The fluorescence of the coating can be seen in the three images on the lower right. The scale bar for each image is 30 μm.

Figure 6.

Confocal images of biofilms accumulated on Teflon CDC coupons seven days after inoculation. Viable bacterial cells are in green, non-viable bacterial cells are in red, and fungal cells are in blue. The fluorescence of the coating can be seen in the three images on the lower right. The scale bar for each image is 30 μm.

Side views of the tallest portion of the biofilm presented in Figures 5 and 6 are provided for biofilms grown in full strength medium without biocide dosing to provide additional context regarding biofilm structure that may be lost in the 3D projections (Figure 7). It can be seen that the biofilms on the uncoated coupons (panels A and C) had a much larger maximum thickness than those on the Sher-Loxane 800® coated coupons (panels B and D).

Figure 7.

Side views of tallest segments of biofilms on A) uncoated Inconel, B) coated Inconel, C) uncoated Teflon, and D) coated Teflon grown in full strength medium without biocide dosing. Scale bar in each image is 30 μm.

Phosphorus was identified as a major limiting nutrient during the development of the microbial ersatz MTN medium (Sandvik et al., 2022) and has been reported as a limiting nutrient in the ISS water processing assembly (WPA) (Diaz et al., 2021). Phosphorus is pervasive in biomolecular processes, lending to its capacity as a limiting nutrient (Elser, 2012). In spacecraft, phosphorus may be removed using nutrient filters, resins, or biological methods such as plant growth—all of which would need external validation for incorporation into the water system. As can be seen in Figure 3, the removal of phosphorus generally tended to reduce viable biofilm accumulation by a half a logarithm to a full order of magnitude with the planktonic densities mirroring that trend (Figure 4).

Silver has long been used to control microorganisms, whether in textiles (Radetić, 2013), material coatings (Knetsch and Koole, 2011; Schierholz et al., 1998), or as a biocide (Sim et al., 2018). It has broad antimicrobial effects including interaction with membrane proteins, disruptions of electron transfer and respiration, interactions with DNA and proteins, and generation of reactive oxygen species (Li et al., 2018). Many experiments investigating silver compatibility in spacecraft systems have been conducted (Adam, 2009; Beringer et al., 2014; Callahan et al., 2007; Petala et al., 2016; Petala et al., 2017; Roberts et al., 2007; Wallace et al., 2016; Wallace et al., 2017), and NASA has expressed interest in moving away from iodine treatment towards silver treatment for potable water (Li et al., 2018). The use of silver fluoride at 2 ppm was determined based on unpublished results from the CBE investigating a wide range of biocides for use in spacecraft water systems. The daily dosing of silver fluoride to the reactors effectively suppressed additional biofilm accumulation during the continuous flow phase. In the ISS water system, reducing planktonic cell density is a key goal in addition to reducing biofilm density because planktonic cells (and/or detached biofilm clumps) can cause clogging issues downstream in the water recovery process. Also, planktonic cells can serve to inoculate surfaces within the wastewater tank (represented by the CDC reactor) or further downstream, leading to additional biofilm formation.

Combining the coating, a daily silver fluoride dosing, and phosphorus removal resulted in the lowest viable biofilm accumulation, ultimately leading to a biofilm total cell density of 1.74 log₁₀CFU/cm² (Inconel) and 1.77 log₁₀CFU/cm² (Teflon) at seven days, which is just above the limit of detection of 1.69 log₁₀CFU/cm². None of the control methods alone came anywhere near this level of biofilm reduction. Testing all three biofilm control methods alone and in combination with each other was important to show that under the conditions tested, good biofilm control will almost certainly not be achieved by one method alone.

In spacecraft water systems, the goals of biofilm control extend beyond logarithmic reduction of viable microbial cells. The accumulation of biofilm and biofilm detachment (even dead cells) contributes to the potential clogging of valves and pipes. As such, it is important to evaluate the biofilm thickness and morphology in addition to the viable cells within. To investigate the biomass and overall clogging potential of the biofilms in various states of control, confocal imaging provides important insight.

Figures 5 and 6 feature representative confocal images of the biofilms after seven days in CDC reactors for each of the eight iterations of experiments conducted. The uncoated Inconel coupons (Figure 5) seemed to foster thinner, near monolayer, biofilms with fewer large aggregates than the uncoated Teflon coupons (Figure 6), which aligns with the viable biofilm density data (Figure 3). However, when the coating was utilized, there were minimal differences in biofilm morphology (or density) between Inconel and Teflon, which is also supported by the CFU data (Figure 3). Figure 7 shows the differences in thickness for all the biofilms grown in full strength microbial ersatz MTN and without biocide dosing. This figure was generated by finding the x-slice of the confocal images in Figures 5 and 6 with the tallest segment of biofilm. Despite the general monolayer biofilm morphology on uncoated Inconel (Figure 7A), the clump of biofilm in the image results in a biofilm with a larger maximum thickness than the biofilms on Sher-Loxane® 800 coated coupons. These tall clumps of biofilm are concerning for their potential for sloughing off and clogging pipes. The clogging potential is even more drastic with the uncoated Teflon biofilm (Figure 7C).

A key morphological difference between biofilms on the uncoated and coated coupons was the presence and absence of fungal hyphae, respectively. On the uncoated coupons, especially on Teflon, fungi were present in the hyphal (filamentous) morphology. The hyphae add significant height to the biofilm, sometimes upwards of 200 μm as can be seen in Figure 7C and add additional surface area for bacterial colonization. Both viable and non-viable bacteria (green and red, respectively) appear to have colonized the fungal hyphae (blue) in Figure 7C. The increased height further enhances clogging potential in small diameter pipes and valves like those found in spacecraft water systems. The additional surface area provided by the hyphae for bacterial colonization means that any treatment present at the substratum (such as a coating) would not be effective against bacteria colonizing the hyphae. For this reason, it is especially encouraging that this coating fostered minimal fungal cell attachment, and that the fungal cells present were in the conidial morphology (Figures 5 and 6, lower right corner).

Some methods used in this study are similar to those reported by Velez Justiniano et al. (2021), such as a subset of the organisms investigated and a modified version of the medium used in their experiments (Sandvik et al., 2022; Velez Justiniano et al., 2021). There has been extensive collaboration between the authors, additional researchers at the CBE, and researchers at MSFC. The results that have been reported by Velez Justiniano et al. (2021) examine relevant organisms (including similar organisms to those used here) as single-species cultures in the planktonic phase. As such, this is the first time to our knowledge, that results of experiments featuring an ISS relevant selection of organisms grown as biofilms in consortium using a relevant medium have been published.

Each of the approaches for biofilm control evaluated in this research have been explored individually by other researchers for spacecraft applications. Four peer-reviewed studies on coatings for use in spacecraft water systems have been published (Demir et al., 2022; Flores et al., 2023; Li Sip et al., 2023; Mettler et al., 2022). All four studies featured Pseudomonas aeruginosa, an organism that has not been isolated from the ISS wastewater tank, potable water dispenser, or condensate. The media used in the tests varied between researchers, but a simulated wastewater was not employed by any of them. One group used Luria Broth with potassium nitrate as an artificial urine medium (Flores et al., 2023), while the others used TSB (Li Sip et al., 2023; Mettler et al., 2022), PDB (Mettler et al., 2022), or PBS (Demir et al., 2022). All the studies reported promising results, but tests only lasted 72 h or less.

One published study evaluated numerical models of nutrient limitation in the WPA. The researchers evaluated the effects of carbon and phosphorus limitation to understand biofilm accumulation in the WPA during typical operation and potential dormancy conditions (Diaz et al., 2021). The models also investigated the effects oxygen availability would have on open and closed systems. Though the study discusses bacterial metabolism, no specific organisms were listed as the basis of the model and fungal microorganisms were not included, which as reported in the results here, contribute greatly to biomass and biofilm thickness. While the study was an interesting mathematical evaluation, no biological tests were completed to support the reported results.

Several studies examining biocide use for biofilm reduction in spacecraft water distribution systems have been published. A long-term study (3 years) found that iodine dosing showed promising reduction of biofilm as long as concentrations were maintained between 1–3 mg/L (Schultz et al., 1992). The researchers used stainless steel, a relevant material (Callahan et al., 2007), and used non-sterile Milli-Q water as the medium and inoculum. They frequently isolated organisms in the resulting biofilm included Pseudomonas cepacia (now known as Burkholderia cepacia), Pseudomonas pickettii, Acinetobacter sp., Bacillus sphaericus, and a Methylobaterium sp., which is very similar to the defined consortium used in the research presented here. The same group published an additional study evaluating iodine and ozone for biofilm kill and removal in spacecraft water systems (Schultz et al., 1994). Their results showed that 25 mg/L and 250 mg/L iodine. and 2.3 mg/L ozone effectively kill attached biofilm but do not remove biofilm. The biocides were evaluated as a treatment method on established biofilm rather than as a maintenance method with regular dosing.

One paper describes a combination of nutrient limitation and biocide dosing for evaluation of dormancy conditions in spacecraft water systems (Beitle et al., 2024). The consortium used in the experiments was very similar to the one used here, but with the addition of Cupriavidus metallidurans, another frequent isolate of the ISS wastewater tank (Zea et al., 2020). The biocide used was silver fluoride and the target nutrient for removal was phosphorus. The experiments also featured a “synthetic wastewater,” but the specific composition is not listed. This study is an exciting synthesis of two methods of biofilm control, but unfortunately it only describes the experimental setup and to the best of our knowledge, the results have not been published. When results are published, they may corroborate results from this study that show multiple approaches of biofilm control are necessary for proper suppression of biomass and biofilm accumulation.

Approaches used in this research could also be used to evaluate biofilm control in Earth-based water systems. It is likely that using organisms relevant to drinking water distribution systems, industrial water systems, or wastewater systems and the use of a medium relevant to the target system could result in similar biofilm reduction when multiple control strategies are combined. Space research can sometimes help inform or inspire applications on Earth, as may be the case here.