A disclosure gel for visual detection of live Bacillus anthracis spores

Abstract Aims To develop a gel formulation to trigger a visual signal for rapid disclosure of the location and extent of surface contamination with viable Bacillus anthracis spores. Methods and Results Methylumbelliferyl‐α‐d‐glucopyranoside was combined with hyaluronic acid to produce a gel that could be applied to a surface as a coating. It remained hydrated for a sufficient time for α‐glucosidase activity present in intact B. anthracis spores to cleave the substrate and release the fluorescent product, methylumbelliferone. The presence of B. anthracis spores could be disclosed at 5 × 104 CFU per reaction test well (0·32 cm2) both visually and using fluorescence detection equipment. Conclusions The disclosure gel provides a rapid, visual response to the presence of B. anthracis spores on a surface. Significance and Impact of the Study The disclosure gel demonstrates the first steps towards the development of a formulation that can provide nonspecialist users with a visual alert to the presence of B. anthracis spores on a surface. It is envisioned that such a formulation would be beneficial in scenarios where exposure to spore release is a risk, and could be used in the initial assessment of equipment to aid prioritization and localized execution of a decontamination strategy.


Introduction
Conventional detection and identification methods following a suspected release of a bacterial agent of concern, such as Bacillus anthracis, include sensitive and specific genetic and antibody-based technologies such as PCR and ELISA. Two problems encountered with such methods are surface sampling and differentiation between living and dead bacteria. Traditional plate count methods give an indication of viability but are time-consuming and do not avoid the issue of surface sampling, which may be of particular concern given the likelihood of nonuniform agent deposition on equipment surfaces. A simple approach to detect and locate living bacteria on a surface is the exploitation of active bacterial enzymes associated with a live cell to trigger disclosure of their presence.
One feature making B. anthracis favourable for hostile release is its exceptional ability to survive in the environment through the formation of metabolically inert spores. However, spores must remain able to interact with their surroundings to determine an optimal time to germinate and return to a vegetative state (Moir 2006). This process does not rely on a single signal, but rather a complement of redundant germination mechanisms ensures that they will not germinate until conditions are ideal for survival (Luu et al. 2011).
A number of enzymes have been identified in the outermost layer of the B. anthracis spore; the exosporium. These include alanine racemase (Steichen et al. 2003) and inosine-uridine-preferring nucleoside hydrolase (Redmond et al. 2004), which interact with the environment and may function to degrade the germinants L-alanine (Chesnokova et al. 2009) and inosine (Liang et al. 2008), respectively, to suppress germination in unfavourable conditions.
The natural course of colonization and infection of a mammal are also important stages in the lifecycle of the pathogen. To survive the onslaught of the immune system, the first-line of defence is again through enzymes within the exosporium. Superoxide dismutase (Baillie et al. 2005) and arginase (Weaver et al. 2007) are thought to play important roles in controlling intracellular survival in macrophages through protection from oxidative injury (Cybulski et al. 2009) and inhibition of nitric oxide synthesis within a macrophage (Raines et al. 2006). The activities of superoxide dismutase, inosine hydrolase and arginase have been recorded in intact endospores (Baillie et al. 2005;Liang et al. 2008) and semi-purified exosporium (Weaver et al. 2007).
This study focuses on exploiting the presence of a-glucosidase activity, which has been detected in intact Bacillus spores. The general role of a-glucosidases is to catalyse the hydrolysis of terminal, nonreducing, 1,4-linked D-glucose residues, with oligosaccharides being the favoured substrate. Glucosidase activity is poorly characterized in B. anthracis. However, it is known that a-D-glucosidase activity but not b-D-glucosidase activity is present in the organism (Sadler et al. 1984) and that in a close relative, Bacillus cereus, a-D-glucoside-hydrolysing enzymes are either cytoplasmic or extracellular, depending on the availability of glucose (Suzuki and Tanaka 1981). As many synthetic colorigenic and fluorogenic substrates are unable to penetrate the cytoplasmic membrane, only secreted or membranebound enzymes would allow detection of viable whole cells by this method, and thus these substrates are well-suited to enzyme-activated disclosure.
By combining a substrate to detect enzyme activity in viable spores with a highly hydrated polymer formulation which can easily be applied as a coating, a disclosure gel has been developed as a proof of principle concept to visually alert to the presence of B. anthracis endospores on a surface. The polysaccharide hyaluronic acid was selected as a carrier for the disclosure substrate due to its high moisture retention ability and viscoelasticity, combined with low immunogenicity and toxicity, which make it a potentially suitable gel for widespread use in disclosure.

Strains, media and growth conditions
Fourteen Bacillus strains and 11 other common environmental or skin-associated bacterial strains were used in this study (Table 1). Bacillus anthracis Sterne strain, which carries pXO1 but lacks pXO2, was used as a simulant for virulent B. anthracis as it can be handled in Advisory Committee on Dangerous Pathogens (ACDP) level 2 containment and has been shown to have comparable exosporium protein composition to virulent B. anthracis Ames (Redmond et al. 2004). Assay components were obtained from Sigma-Aldrich (Gillingham, Dorset, UK) and media components from Oxoid (Basingstoke, Berkshire, UK). Nutrient agar and nutrient broth were used for general growth of non-Bacillus strains. Bacillus strains were maintained on tryptone soy agar (TSA). The selective medium Brilliance Bacillus cereus agar was used to isolate bacteria of this type from soil.

Spore preparation
Spore stocks for each Bacillus species were prepared by growth on NBYS (Lecadet et al. 1980) agar plates at 28-30°C for 4-7 days. After sporulation was verified by phase contrast microscopy, with samples containing <1% vegetative cells, the bacteria were scraped off the plates and washed six times in sterile, distilled water with centrifugation at 15 000 g for 5 min. Spore preparations were quantified through serial dilution and plating on TSA. Stocks were adjusted to 10 8 CFU per ml with sterile, distilled water and stored at either À20°C with individual aliquots being thawed for each use (B. anthracis Sterne spores) or 4°C (other Bacillus spp.). Where dead spores were used for comparison, a spore stock at 1 9 10 8 CFU per ml was autoclaved for 20 min at 121°C before use. Sterility was confirmed by plating 100 ll of the autoclaved stock on TSA and incubating at 25°C for 3 days.

Preparation of vegetative bacterial suspensions
Vegetative bacterial cultures were prepared by inoculation of 20 ml nutrient broth and incubation at 30 or 37°C with shaking until an OD 600 of 0Á6-0Á8 was reached. The bacterial suspensions were then pelleted by centrifugation (3200 g for 5 min at 4°C) and washed three times in sterile phosphate buffered saline (PBS). Each pellet was finally resuspended in 2 ml sterile PBS and placed on ice. Vegetative bacterial suspensions were used immediately after washing. Each cell suspension was enumerated by serial dilution and plating on nutrient agar.
Water retention of the hyaluronic acid-based disclosure gel was measured by weight over time. Three individual 1 ml samples of the gel were pipetted onto glass slides and weighed. The slides were then incubated at 20°C, 85% relative humidity on an open tray. Weights were measured again after 3 and 6 h incubation.
To determine the effect of the formulation on germination of B. anthracis Sterne spores, a final concentration of 1 9 10 7 CFU per ml spores was added to disclosure gel and incubated for 3 h at 25°C. The percentage of germinating spores, as determined by a change to phase dark, was determined through visual inspection at 4009 magnification and compared to that of spores with germinants (100 mmol l À1 L-alanine and 10 mmol l À1 inosine) alone added.
Determination of a-glucosidase activity using the disclosure gel Five-microlitres of spore stock (at 1 9 10 8 CFU per ml) or washed vegetative bacterial suspension was added to 45 ll of the disclosure gel in a flat-bottomed black 96-well microplate (Nunc TM , Fisher Scientific, Loughborough, UK). Assay reactions were incubated at 25°C for 3 h unless otherwise stated. To stop the reaction and obtain optimal fluorescence, 50 ll of 0Á5 mol l À1 NaOH was added to each well. Fluorescence was measured at excitation/emission wavelengths of 365/455 nm, using an Infinite M200 Pro fluorescence microplate reader (Tecan, M€ annedorf, Switzerland). Wells without spores (blank) and with autoclaved (dead) spores were used as controls. Relative fluorescence was calculated by subtracting the fluorescence level of the blank (reagents only) wells from the test wells. Each plate included a calibration curve of 4-methylumbelliferone (4-MU) in DMSO, ranging from 0 to 50 pmol in a total volume of 100 ll, prepared in duplicate. The amount of 4-MU released in each test well was calculated using the slope of the calibration curve. Visual observation of the fluorescence output was carried out under a UV light source with a 485-655 nm emission filter.
Testing a-glucosidase activity in spores dried onto a surface Where spores dried onto the plate surface were to be tested, 5 ll spore stock was pipetted into individual wells and the plates were left uncovered in a microbiological cabinet overnight. Disclosure gel was added on top of the dried spores and incubated for 3 h at 25°C before the addition of 0Á5 mol l À1 NaOH and recording of fluorescence.
Testing the effects of soil interference on measurement of a-glucosidase activity Garden soil (lime-rich loamy soil) was suspended in sterile PBS to a concentration of 10 mg ml À1 . Tenfold serial dilutions of the soil suspension were made both with and without the addition of B. anthracis spore stock (1 9 10 8 CFU per ml final concentration) and 5 ll of each dilution was tested for fluorescence output after 3 h incubation with disclosure gel. Sterile PBS was used as a negative control sample.
Testing the effects of skin contact interference on measurement of a-glucosidase activity A pilot study was carried out to assess the possibility of interference with fluorescence output from the gel as a result of prior skin contact on a surface. This was crudely tested using polystyrene 6-well plates (Corning â , Sigma-Aldrich). Two individuals using unwashed hands touched the surface of each well within two 6-well plates multiple times for 5 min. To one plate that had been handled by each individual, B. anthracis spore stock at 1 9 10 9 CFU per ml was added in five drops, 2 ll each, around the surface of each well. The other plate handled by each individual had sterile water added in the same manner. Sterile, untouched plates were inoculated in the same way with spore stock or water for use as comparative controls. All plates had 990 ll disclosure gel added to the centre of each well, which spread to cover the surface. After 3 h incubation at 25°C and addition of 1 ml NaOH (0Á5 mol l À1 ) to each well, a 100 ll sample was transferred from each well into a black 96-well plate before fluorescence was measured. Statistical analysis was carried out using Student's t tests with the null hypothesis being that there would be no difference between assays carried out in the presence of soil or skin contact and those carried out in their absence. Differences with P values of <0Á05 were considered statistically significant.

Formulation development, water retention and germination support
A range of hyaluronic acid concentrations were tested for optimal rheological properties. To evaluate the effect of various inducers and ions on a-glucosidase activity from intact B. anthracis spores, a range of buffers and sugars were tested by inclusion in the formulation. Enzyme activity was also tested over a pH range of 5-7Á4 to determine the optimal pH conditions. Fluorescence output was reduced in the presence of PBS, acetate buffer and phosphate citrate buffer (all pH 7) compared with 150 mmol l À1 NaCl; and a pH of 7-7Á4 was found to be optimal for each buffer. Maltose, and to a lesser extent, sucrose, improved fluorescence output from spores, up to a final concentration of 100 mmol l À1 sugar (data not shown). Maltose was therefore included in the formulation as an inducer of a-glucosidase activity. Water retention by the disclosure gel after 3 h incubation was 86 AE 0Á2% and after 6 h was 69 AE 0Á3%. The rate of germination-initiation for B. anthracis spores in the disclosure gel was 98% but no outgrowth was observed.

Spore-associated a-glucosidase activity
Enzyme activity associated with intact, washed B. anthracis spores was measured by the detection of the fluorescent product 4-MU released during a 3 h incubation period in the presence of a-MUG. This method demonstrated visually detectable enzyme activity from intact spores against a background of dead (autoclaved) spores (Fig. 1). The level of fluorescent output continued to increase beyond the 3 h incubation used in the assay, so greater sensitivity could be achieved with longer incubation (data not shown). However, the given timeframe was chosen as a suitable window to achieve sufficient disclosure sensitivity whilst ensuring a reasonably rapid response could be obtained.

Effect of nutrient germinants and growth medium on a-glucosidase activity
The addition of freshly prepared nutrient germinants (100 mmol l À1 L-alanine and 10 mmol l À1 inosine) led to an increase in relative fluorescence (RFU) compared to the formulation without germinants for spores at a final concentration of 1 9 10 7 CFU per ml. From the eight experiments, the average increase was 62% with a standard deviation of AE6%. The increase in a-glucosidase activity may be due to new enzyme synthesis as a result of initiating germination, as seen previously (Chandrapati and Woodson 2003) or release of a-glucosidases from within the inner layers of the spore (Setlow et al. 2016).

Detection of low levels of B. anthracis spores
For spores both in suspension and dried onto a surface (0Á32 cm 2 area), 5 9 10 4 CFU could be consistently detected above background fluorescence both using fluorescence detection equipment (Table 2) and visually under UV light (Fig. 2). Visual observation of a fluorescent output from the disclosure gel could also be made where spores were present on a surface using a hand-held longwave UV lamp and only minimal shade from fluorescent laboratory lights. Spores that had been left to dry on a surface before the addition of the gel produced slightly lower average fluorescence output than spores in suspension, which would be expected due to lower accessibility to the substrate. Autoclaved spores could not be distinguished from background fluorescence.

Bacterial exclusivity of the disclosure formulation
The disclosure gel was tested against a panel of Bacillus species and other environmental or skin-associated bacterial species to determine its exclusivity. As expected, very little a-glucosidase activity was detected from the intact vegetative cells tested. The substrate a-MUG is unable to cross into the cytoplasm, which is where most known a-glucosidases are present. The absence of medium to support bacterial growth would also have reduced the incidence of any extracellular a-glucosidases being produced. Bacillus thuringiensis HD-1 Cry À had a similar level of a-glucosidase activity to B. anthracis in the disclosure gel but the other environmental B. thuringiensis strains tested did not (Table 3). Bacillus licheniformis spores also displayed notable a-MUG hydrolysis, albeit at a lower rate than both B. anthracis and B. thuringiensis HD-1 Cry À . Although a spore-associated a-glucosidase has not been described in the literature, B. licheniformis is known to produce both cytoplasmic and extracellular aglucosidases (Thirunavukkarasu and Priest 1984) so this activity is not unexpected. 1164 (AE303) 608 (AE311) 5 9 10 3 0Á17 (AE109) À706 (AE675) 5 9 10 5 Dead 0Á00 (AE58) À2032 (AE258) 5 × 10 5 5 × 10 4 CFU well -1 well -1 CFU 5 × 10 3 well -1 CFU

Effect of soil and skin contamination on a-glucosidase activity
Soil and skin contact were regarded as likely contaminants of surfaces suspected to have been exposed to bacterial agents of concern. They were, therefore, assessed for their potential interference on a-glucosidase activity from B. anthracis spores using the disclosure gel.
Garden soil suspension at a final concentration of 1 mg ml À1 in the gel showed considerable interference with fluorescence output, producing a significantly lower average RFU than the 'no soil' control (P = 0Á006), with wide variation (Fig. 3). No difference was observed between the other soil dilutions and the 'no soil' control. The soil suspension was confirmed as pH 7. The interference with fluorescence may therefore be due to the organic matter present within the soil, although further investigation would be necessary to confirm this.
As an indication of the levels of B. cereus or B. thuringiensis naturally occurring in the soil sample, the selective medium Brilliance Bacillus cereus agar was used to spread culture aliquots of the soil suspension at 10 mg ml À1 . An average of 140 blue colonies was observed per 1 ml of soil suspension, equating to fewer than one CFU per reaction well in the assay set-up. As expected from this result, no a-MUG hydrolysis was observed from soil suspension alone in the disclosure gel at any concentration (data not shown), suggesting that false-positive responses to disclosure gel application would not be caused by soil contamination with this soil type.
Skin contact on a surface prior to contamination with spores and application of disclosure gel resulted in a statistically significant reduction in fluorescence (P ≤ 0Á0001). The fluorescence output was reduced by almost half over the course of the 3 h incubation (Table 4). A small but not statistically significant (P = 0Á08) amount of fluorescence was observed on surfaces that had been handled but not exposed to B. anthracis spores. This is not sufficient to pose a risk of a visual false positive reaction from the gel.

Discussion
A polymer-based gel formulation has been developed to demonstrate the first steps towards visual determination of the presence and location of surface contamination with live B. anthracis spores. The results show that a-glucosidase activity present in intact B. anthracis spores can be used to cleave the synthetic substrate a-MUG to produce a fluorescent response which can be visualized under UV light within a few hours.
Initial evaluation of the gel has shown that it does not respond to a range of common environmental Bacillus or other bacterial species under the given conditions, affording a good degree of specificity for initial assessment of  equipment following a suspected release of biological agent. This would allow a more directed approach for sampling and deployment of highly specific and sensitive methods such as PCR. A number of strains tested are known to produce a-glucosidases; in other Bacillus species a-glucosidases have largely been reported in cytoplasmic or extracellular locations (Suzuki and Tanaka 1981;Thirunavukkarasu and Priest 1984;Castro et al. 1995) and Staphylococcus aureus (MalA) and Pseudomonas fluorescens produce cytoplasmic a-glucosidases (Guffanti and Corpe 1975). These and most other strains tested produced little or no fluorescence in response to the disclosure gel. Therefore, their presence on the surfaces is unlikely to interfere with the visual output of the formulation or produce false-positive signals unless present at very high inoculum levels, in the case of some Bacillus sp. spores, or in an active growth stage due to alternative nutrient sources; the lack of essential nutrients in the disclosure gel ensures it is not conducive to active metabolism and undesirable cell growth. Exceptions to this are some strains of B. thuringiensis and B. licheniformis, which had notable a-glucosidase activity in their respective intact spores, although at a lower level than B. anthracis. This cross-reactivity with some other Bacillus spp. could lead to false-positive disclosure, therefore further work to differentiate between Bacillus spp. would be beneficial to optimization. Preliminary assessments of the disclosure gel in the presence of substances that may interfere with either glucosidase activity or fluorescence output suggest that further stabilization of the gel may be beneficial for producing consistently reliable results in field environments. Soil showed a negative effect on fluorescence output only at the highest concentration tested (1 mg ml À1 ). However, the results showed huge variation between samples, which were not seen with lower concentrations of soil. The heterogeneity of the soil may be responsible for the variation seen in the fluorescence output, as it was untreated and not sieved before use. An interesting comparison would be made from testing a variety of soil types for their effects on the output from the disclosure gel. We would predict that soils with lower organic matter content would have a lesser interference effect on fluorescence.
Interestingly, B. anthracis spores applied to surfaces that had undergone considerable handling showed notably reduced fluorescence output from the disclosure gel. Secretions from sweat and sebaceous glands may be responsible for interference with fluorescence of the 4-MU released from hydrolysis of a-MUG within the gel. Further experiments will show whether this is the case or if the lowered fluorescence is a result of reduced a-glucosidase activity. The implications of the observed interference from both soil and skin contact could lead to a reduction in sensitivity of the disclosure gel in field situations. Studies into the locality, mechanism and purpose of observed a-glucosidase activity in intact spores have not been carried out and analyses of B. anthracis exosporium proteins have not identified any putative a-glucosidases (Steichen et al. 2003;Redmond et al. 2004). It may be that the aglucosidase is simply a cytoplasmic enzyme in the mother cell that becomes trapped in the outer layers of the spore before its release. However, Albert et al. (1998) described spore coat-associated a-glucosidase present in Geobacillus stearothermophilus spores, which was supported by Setlow et al. (2016). The authors predicted that this enzyme may have a role in early germination and outgrowth, which may also be the case for B. anthracis.
Future work will address increasing the sensitivity of the gel and assessing the potential benefits of adding buffering and stabilization components to the gel to improve its reliability in a field environment. Additionally, a number of factors would need to be assessed and addressed to move this approach towards a field-deployable solution. Conditions more closely mimicking a range of natural environments and scenarios would be needed to assess the performance of the gel in terms of maintenance of hydration, photo degradation of the fluorescent product 4-MU and possible effects of other contaminants such as residual growth medium. The effect of the gel on equipment and what the clean-up steps would be following disclosure are also important considerations.