Per NASA-HDBK-6022 (2010) samples are acquired using cotton swabs or 9′′ × 9′′ polyester wipes and processed via the associated NSA procedure. Both swab and wipe samples are premoistened with sterile DI water since they come in direct contact with flight hardware during sampling. As per NASA-HDBK-6022 (2010), swab samples are suspended in 10 mL 18.2 MΩ deionized water and wipes are suspended in 200 mL of planetary protection (PP) rinse solution for processing. PP rinse (buffered solution + Tween 80) was prepared in accordance with NASA-HDBK-6022 (2010). Samples then are processed via sonication for 2 min in an aqueous bath of 0.02% v/v Tween 80, at 19–27 kHz (Figures 1, 2). Samples then undergo “heat shock” at 80°C for 15 min to enable the detection of aerobic, mesophilic, and cultivable spores. Samples are then quickly cooled to 30–35°C. Portions (2 or 4 mL) of each sample are poured into sterile Petri dishes followed by a standard growth medium required by NASA-HDBK-6022 (2010), trypticase soy agar (TSA). The plates are incubated at 32°C for a period of 72 h, during which planetary protection engineers count colony forming units (CFUs) every 24 h.

Bacillus atrophaeus spores (∼1 μm) were obtained from Mesa Laboratories (#9372, Bozeman, MT, United States). A working spore stock solution was prepared containing 100 spores/mL. Following vortexing, 50 μL of the spore stock solution was transferred each into the three sample types: wipes, swabs, and control. Wipe samples refer to glass bottles containing only 200 mL of planetary protection (PP) rinse solution. PP rinse (buffered solution) was prepared in accordance with NASA-HDBK-6022 (2010). Swab samples refer to glass test tubes containing only 10 mL 18.2 MΩ deionized water. The wipe and swab sampling devices themselves (Polyester wipes, TX3211, Texwipe, Kernersville, NC, United States; Cotton Tipped Applicators, 806WC, Puritan, Guilford, ME, United States) were excluded during testing to minimize the potential variability of CFU recovery efficiency; the devices were previously validated by the Planetary Protection Office, thus we focused on comparing the methodologies only. The control samples refer to 50 μL of the working spore stock solution that was directly plated onto TSA plates.

Bacillus atrophaeus spores (∼1 μm) were obtained from Mesa Laboratories (#9372, Bozeman, MT, United States). B. atrophaeus has long been used as a model planetary protection organism as it grows well in the NSA incubation conditions (Aerobic, 32C, grows on TSA) and has a clear red pigmentation that can help distinguish it from potential contamination. Soil from APL’s campus was selected for its diverse, undefined, and abundant spore population instead of cleanroom samples that have a lower overall spore abundance and much higher variability in spore populations. Two types of solutions were made to directly compare NSA and MF: soil in DI water and B. atrophaeus in DI water. The soil solutions consisted of three trials (low, medium, and high concentrations, denoted as Soil concentration 1, 2, and 3, respectively) and the B. atrophaeus solutions consisted of one trial. Each trial consisted of one set of pour-plate samples, and two sets of membrane filtration samples for redundancy, denoted as “MF-APL 1” and “MF-APL 2,” each with 20 replicates. For the soil solutions, 1 g of dried soil was dried in open indoor air over the course of 1 week. It was then suspended in DI water and passed through a 20 μm filter to remove large solids. Spores were selected for by heat shocking the DI soil solution for 15 min at 80°C per NASA-HDBK-6022 (2010), thus assuming only spores remained in the soil samples after the heat shock. This starting solution was then serially diluted to a dilution factor of 1:10,000. For B. atrophaeus solutions, the B. atrophaeus spore stock was diluted to achieve approximately 1.25 spores/mL. Each sample consisted of 8 mL DI water to imitate a swab sample. A wipe-equivalent sample was not tested.

In accordance with the NASA Standard Assay, swab and wipe samples, as prepared in section “Preparation of Samples (Axenic Cultures; Jet Propulsion Laboratory)” or “Preparation of Samples (Soil and B. atrophaeus Solutions; Applied Physics Laboratory),” were sonicated, heat shocked, cooled, and processed via pour plating (NSA) (Figures 1, 2). For wipe samples, 4 mL was dispensed each into 12 Petri dishes and 2 mL was dispensed into 1 Petri dish, to total 50 mL of processed volume (PN 351029, Corning, Tewksbury, MA, United States) and filled with 1.5% TSA (Difco™ Tryptic Soy Agar PN 236920, BD Biosciences, Franklin Lakes, NJ, United States, or equivalent) using a peristaltic pump (Delta Scientific Medical Fill Master, or equivalent). For swab samples, 4 mL was dispended each into 2 Petri dishes and filled with TSA (alternatively 2 mL each into 4 Petri dishes). Plates were then incubated at 32°C and enumerated at 24, 48, and 72 h. Since only a fraction (50 mL for wipes and 8 mL for swabs) of the total sample volume (200 mL for wipes and 10 mL for swabs) is plated, the CFU enumerated at 72 h was multiplied by the pour fraction to extrapolate the final CFU. For wipe samples a pour fraction of 4× was applied to CFU results since 25% of the sample was processed (50/200 mL). For swab samples a pour fraction of 1.25× was applied to CFU results since 80% of the sample was processed (8/10 mL). A negative control consisting of a petri dish consisting of only poured TSA was included in each sample set to ensure no contamination occurred during the processing.

The membrane filtration equipment includes a 3-place laboratory manifold with hose barb cap, end cap, and manifold valves (PN 4889, Pall, Port Washington, NY, United States). The equipment also includes manifold standard adapters (PN 4891, Pall, Port Washington, NY, United States), 150 mL magnetic filter funnels (PN 424, Pall, Port Washington, NY, United States), Whatman 0.2 μm cellulose acetate membrane filters (PN 10404112, Cytiva, Marlborough, MA, United States), forceps, 1/4′′ inner diameter Tygon E-3603 laboratory tubing (Saint-Gobain, Malvern, PA, United States), and a 1 L glass filter flask.

The membrane filtration setup was assembled in a laminar flow environment or Class II, Type A2 biological cabinet and consisted of the following: The Pall laboratory manifold was first connected to a filter flask using Tygon tubing and a rubber stopper. The barbed end of the filter flask was attached to a vacuum source via Tygon tubing. The 150 mL magnetic filter funnels were placed into each of the three manifold openings via the manifold standard adapters, and subsequently, the 0.2 μm Whatman filters were placed inside the magnetic filter funnels using sterile forceps.

In accordance with the NASA Standard Assay, swab and wipe samples were sonicated, heat shocked, and cooled. To begin membrane filtration (MF), 5 mL of sterile 18.2 MΩ deionized water was filtered through each funnel to wet the filters. Next, the entire volume of each sample (10 mL for swabs or 200 mL for wipes) was poured into a filter funnel. For swabs, an additional 40 mL DI water was added to ensure a minimum of 50 mL is filtered through the funnel. The volumes were filtered via vacuum and the sample allowed to drain completely. A negative control consisting of only sterile 18.2 MΩ deionized water was included in each sample set to ensure no contamination occurred during the processing. The funnels and filters were rinsed of potentially inhibitory residues with an additional ∼50 mL of sterile water followed by filtration. Using sterile forceps, the membrane filters were removed from the funnels and placed into sterile Petri dishes prepared with TSA by slowly rolling the filter onto the solidified agar to avoid trapped air bubbles underneath the filter. Plates were then incubated at 32°C and enumerated at 24, 48, and 72 h. See Figures 3, 4 for process flow diagram. Since the entire sample volume is filtered for both wipe and swab samples (in the absence of sampling devices), the CFU enumerated at 72 h represents the final CFU (i.e., the pour fraction is 100%).

The membrane filtration equipment includes a Milliflex Plus Pump Single Head Kit (PN MXPPLUS01, MilliporeSigma, Burlington, MA, United States), Milliflex-100, 100 mL, 0.22 μm white gridded funnel and filter in one (PN MXGSWG124, MilliporeSigma, Burlington, MA, United States), Milliflex Cassette prefilled with Tryptic Soy Agar (PN MXSMCTS48, MilliporeSigma, Burlington, MA, United States), and forceps.

The membrane filtration setup was assembled in a laminar flow environment and consisted of the following: the sterilized pump head is seated onto the pump and wiped down with 70% isopropyl alcohol (IPA). The system is turned on and “manual” filtration mode is selected. Using sterile forceps, the mesh spacer from a funnel is moved to the pump head. Next the funnel is pushed down over the mesh spacer and pump head. The clear cover is removed from the funnel and temporarily set aside.

In accordance with the NASA Standard Assay, swab and wipe samples were sonicated, heat shocked, and cooled. To begin membrane filtration (MF), 8 mL of each sample was added to the funnel. For experiments conducted in this study, APL added 8 mL of sample instead of 10 mL. To ensure a minimum of 50 mL is filtered through the funnel, DI water was added. The “Start” button is pressed on the pump to initiate filtration. A negative control consisting of only sterile DI was included in each sample set to ensure no contamination occurred during the processing. The funnels and filters were rinsed of potentially inhibitory residues with an additional ∼50 mL of sterile water followed by filtration. The “Start” button was pressed again to initiate the automated drying process to remove excess moisture from the filter. Following filtration, the clear cover is returned to the funnel. The entire funnel assembly is removed from the pump head (mesh spacer is left behind) and fitted onto a TSA cassette by pushing down on the funnel assembly until a “click” is heard. The funnel is removed and the clear cover is moved to the cassette. Plates were then incubated at 32°C and enumerated at 24, 48, and 72 h. See Figures 5, 6 for process flow diagram. The entire sample volume (8 mL) was filtered (in the absence of sampling devices). Thus, the CFU enumerated at 72 h represents the final CFU (i.e., the pour fraction is 100%).

Colony forming units (CFU) recovery was compared between the two methods, NSA and MF, for wipe samples, using a known concentration of B. atrophaeus spores (Figure 7A and Table 1). Each method tested, and a control, consisted of thirty replicates. Non-parametric Mann–Whitney U analysis (significance level of 0.05) was conducted comparing the median CFU counts between methods. Mann–Whitney U test was utilized as data resulted in non-normal distribution indicating the need for a non-parametric test, as there is no assumption of normality or equal variances. The sample medians were statistically equivalent between each method, including the control (Table 2; z value for all method comparisons > critical value of −1.96 and <+1.96). The NSA method resulted in a considerably larger CFU recovery data range than the MF method or control, among which eight out of thirty replicates resulted in zero CFU after 72 h of incubation. These results are not surprising since, with the NSA method, only 50 out of 200 mL is plated; applying a pour fraction extrapolation factor (0.25) leads to greater variance, decreasing the accuracy of spore bioburden estimation. These results demonstrate that the NSA and MF methods are comparable in CFU recovery, but NSA has a higher variance. Moreover, MF allows for greater resolution of bioburden via processing of total volume compared with the NSA method.

CFU recovery was compared between the two methods, NSA and MF, for swab samples, using a known concentration of B. atrophaeus spores (Figure 7B and Table 1). Each method tested, and a control, consisted of thirty replicates. A non-parametric Mann Whitney U analysis (significance level of 0.05) was also used to compare the median CFU counts between methods, because data resulted in non-normal distribution. The sample medians were statistically equivalent between the NSA and MF methods (Table 2; z value of −1.693 > critical value of −1.96 and <+1.96). However, the control median was not statistically equivalent to either the NSA (Table 2; z value of −3.813 < critical value of −1.96) or MF (Table 2; z value of −3.013 < critical value of −1.96) methods. This recovery loss in the NSA and MF methods could be due to the lack of surfactant since swab samples are submerged in water and not PP rinse like the wipes; the recovery efficiency may be reduced due to spores adhering to the side of the glass during processing, as no surfactant is present to decrease surface tension. The NSA method resulted in a larger CFU recovery data range than the MF method, among which three out of thirty replicates resulted in zero CFU after 72 h of incubation. These results are not surprising since, with the NSA method, only 8 out of 10 mL is plated; applying a pour fraction extrapolation factor (0.8) leads to greater variance (albeit less so than wipe samples), decreasing the accuracy of spore bioburden estimation. These results demonstrate that the NSA and MF methods are comparable in CFU recovery. Moreover, MF allows for greater resolution of bioburden via processing of total volume compared with the NSA method.

CFU recovery was compared between the two methods, NSA and MF, for soil solutions using three different spore concentrations (low, medium, and high, denoted as soil concentration 1, 2, and 3, respectively). Results show that the MF method had higher mean CFUs than the NSA method (Figures 8A–C and Table 3). In two out of the three sample sets tested, the mean CFUs captured for the methods differed by less than one. T-Test was performed after evaluating the distribution of the differences between paired values, where the distribution of the differences should be approximately normally distributed for the T-Test to be implemented. For these sample sets, MF and NSA were found to be statistically equivalent (p-value > 0.05, Table 4). For one of the sample sets (concentration 1), the methods were found to not be statistically equivalent (p-value < 0.05) with MF collecting more CFUs. Altogether, the MF method is able to recover the same CFU number or more compared to the NSA method. MF and NSA perform similarly even at different concentrations. Moreover, the MF data generally had smaller standard deviation suggesting improved precision compared to NSA.

CFU recovery was compared between the two methods, NSA and MF, for B. atrophaeus solutions. There was less than 10% difference in the mean CFUs counted for each method (Figure 8D and Table 3). MF and NSA methods were shown to be statistically equivalent (p-value > 0.05, Table 4).

In the method comparison studies performed by JPL and APL, we excluded the wipe and swab sampling devices themselves during testing to minimize the potential variability of CFU recovery efficiency. Thus, for MF experiments, the pour fraction was 100%. To determine the pour fraction of MF with the sampling devices included, we measured the final volume poured out from each vessel (glass test tubes containing swabs and glass media bottles containing wipes) compared with the initial volume (10 mL for swabs and 200 mL for wipes). From 36 replicates, the average final volume was 9.55 mL for swabs (standard deviation = 0.07) and 185.11 mL for wipes (standard deviation = 1.94). The remaining volume was contained in the sampling devices (∼0.45 mL for swabs and ∼14.89 mL for wipes) and was unable to be extracted during processing. Subtracting the standard deviation values from the average final volumes, we calculated a 95% pour fraction for swabs and a 92% pour fraction for wipes. Thus, for implementation of MF as part of the NASA Standard Assay, we recommend using a 92% pour fraction overall for simplicity for both swabs and wipes.