University of Birmingham Survival of the ovine footrot pathogen Dichelobacter nodosus in different soils

Dichelobacter nodosus ( D. nodosus ) is the causative agent of footrot in sheep; one of the most important health and welfare issues of sheep worldwide. For control programmes to be effective, it is essential that the transmission cycle of D. nodosus is understood and bacterial reservoirs in the environment are better de ﬁ ned. This study evaluated the survival of D. nodosus in different soils using soil microcosms. Cultivation independent and dependent methods were used to detect D. nodosus over 40 days from seeding in soil. A D. nodosus speci ﬁ c probe was used for quanti ﬁ cation by qPCR and viability was assessed by cell permeability to an intercalating dye, PMA, and by culture. Survival varied dramatically depending on soil type, matric potential (MP) and temperature. Our ﬁ ndings indicate that D. nodosus survival was higher at 5 (cid:1) C compared with 25 (cid:1) C in all soils and signi ﬁ cantly longer at both temperatures in clay soil ( > 44% clay) compared with other soil types. Survival under all conditions was longer than 30 days for both culture independent and dependent methods, this is substantially longer than previous studies and, if this is an infectious dose, longer than the current recommendation of resting a ﬁ eld for 14 days to prevent onward infection. © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).


Introduction
Ovine footrot is an infectious disease of sheep that is caused by an anaerobic bacterium Dichelobacter nodosus, which is the essential transmitting agent [23,44]. The disease is transmitted from sheep to sheep via pasture or bedding. It is characterized by interdigital inflammation, with or without separation of the keratinous hoof horn from the underlying dermis. Both conditions result in lameness and loss of body condition [41].
Footrot is the main cause of lameness in sheep in the UK [15,20] and accounts for serious economic losses in countries worldwide [35,40]. Recent reports indicate that footrot results in annual losses of between £24 and £80 million in the UK [29,41]. Its significant financial impact is due to a reduction in meat and wool production and the expenditure associated with treatment and prevention [13,35,40,41]. It is a painful condition and thus is an important and challenging welfare issue [11].
Footrot is seasonal in some areas of the world, particularly arid areas of Australia and India, where there are predictable periods of transmission [13]. In the UK, there is no seasonality for severe footrot, rather a series of mini-epidemics throughout the year [14] with increasing temperature and rainfall favouring spread of the disease [34]. Epidemics of interdigital dermatitis are also reported in spring [42]. Part of the strategy used in control and elimination programmes is empirical evidence that D. nodosus cannot survive for more than a few days off the feet of ruminants [1,4]. Survival away from the host is dependent on moist, mild conditions [12] and D. nodosus is reported to survive at an infectious dose for no longer than two weeks on pasture under optimal (warm and damp) conditions [4,43]. A recent study using qPCR demonstrated that D. nodosus can survive up to 14 days in soil microcosms at 5 C and for a further 24 days if powdered hoof horn was added to soil; survival was markedly reduced at 15 C however, the moisture content was not stated [7]. Nevertheless such work is important to move to an evidence-based approach to manage environmental contamination with D. nodosus.
To date, the environmental conditions and soil types where survival is greater, have not been elucidated and further work is warranted on D. nodosus survival outside the host, taking into account edaphic factors that result in dramatic differences in matric potential and water availability. Whilst a large number of studies have focused on Escherichia coli survival in soil, it is still uncertain how fluctuations in water availability affect survival [37]. Survival studies are uninformative without taking soil type into account if bulk soil is used in microcosms. For these reasons this study focused on the survival of D. nodosus in soil to investigate the impact of soil type, temperature and matric potential on longevity of viable D. nodosus in soil.

Material and methods
2.1. Cultivation of D. nodosus on solid and liquid media and its growth conditions A virulent strain of D. nodosus (VCS1703A) was cultured on 2e4% Trypticase Arginine Serine Hoof agar (TASH) and Trypticase Arginine Serine (TAS) (Becton, Sparks Maryland, USA; SigmaeAldrich, St. Louis, USA) and incubated at 37 C for 3e4 days in an anaerobic cabinet (Don Whitley Scientific Ltd., Shipley, UK) [33,36]. The cells were harvested in 2 ml phosphate buffer saline (PBS) solution to yield a final concentration of approximately 10 8 cells ml À1 .

Soil microcosms
Soil was grounded with mortar and pestle and sieved through 4 mm mesh to produce soil pores of more uniform size as reported previously [45]. Soil microcosms (50 ml Falcon tubes without aeration vents) containing 5 g of non-sterile Warwick soil (sandy loam) were set up in triplicate at À33 kPa, matric potential (MP defined as the suction required to extract water as soils dry out [3] calculated from previous wettingedrying curves of Warwick soil [9]; Fig. S1). The soil bulk density was maintained at 0.48 g/cm 3 . D. nodosus was inoculated to the soil microcosms to a final concentration of 10 8 cells g À1 soil at À33 kPa. Microcosms were incubated aerobically at 25 C in the dark and destructive sampling was performed at days 0, 1, 2, 3, 6, 10, 14, and 21. A sample of soil 0.5 g was taken for DNA extraction and 0.5 g for RNA extraction to which 1 ml of RNAlater Stabilization Solution (Life Technologies Ltd, Paisley, UK) was added and stored at À80 C.
Soil microcosms with different non-sterile soil types (20 g) including Warwick (sandy loam), Basilicata (sandy loam), Abanilla (sandy) and Stockton (clay) (soil characteristics Table 1; wettingedrying curves Fig. S1) were set up in triplicate, inoculated with D. nodosus culture (10 7 cells g À1 soil). The soil bulk density was maintained at 0.78 g/cm 3 . Microcosms were incubated at 25 C in the dark. Further, a second triplicate set of microcosms were incubated at 5 C and destructive sampling of both sets was performed at days 0, 3, 6, 9, 12, 20, 30, and 40. Uninoculated soil was used as control to determine D. nodosus background amplification.

Extraction of DNA from soil
DNA was extracted from 0.5 g soil samples using the FastDNA Spin kit for Soil following the manufacturer's instructions (MP Biomedicals, UK). DNA was then eluted into 70 ml of DNase Free water. Sterile soil was included as blank.

Extraction of RNA and synthesis of cDNA
RNA was extracted from Warwick soil samples collected at different time points using RNeasy Mini Kit (QIAGEN, Manchester, UK) and Griffith's method with some modifications. RNA integrity and concentration were measured using the Prokaryote Total RNA nano (Agilent Bioanalyser). DNase treatment of the total nucleic acid was performed using Turbo DNA free kit following the manufacturer's instructions. From DNA-free RNA, cDNA was synthesized using High capacity RNA-to-cDNA Kit (Applied Biosystems, Warrington, UK).

End point PCR and qPCR
PCR amplifications targeting D. nodosus 16S rRNA were performed using specific primers Cc and Ac [24] in an Eppendorf Mastercycler (Eppendorf, Hamburg, Germany) as described previously [28]. D. nodosus DNA from the strain, VCS1703A was included as a positive control and sterile water as a negative control.
The qPCR of the DNA and cDNA was performed in an Applied Biosystems, 7500 Fast real-time detection system (Applied Biosystems, Warrington, UK). The PCR targeted the rpoD gene (RNA polymerase sigma 70 factor, single copy number in D. nodosus genome) [6]. All PCR reactions were performed in triplicate as previously reported [28]. The rpoD copy number was estimated based on the standard curve obtained from analysis of ten-fold serial dilutions of DNA extracted from D. nodosus strain VCS1703A. The results were analysed using 7500 Fast System SDS software (Applied Biosystems, Warrington, UK).

Soil desiccation curves
To select the appropriate microcosm set up, drying out of soil was determined for up to 17 days using Warwick soil. Four different microcosm types were set up using 20 g soil in 50 ml falcon tube and included closed, open, partially closed and closed with aeration vents (two holes punched in lid and two in sides above soil level). Soil was prepared as previously reported [45]. Soil samples were moistened to approx. À33 kPa. The microcosms were incubated at 25 C in the dark and reduction in weight was recorded at each day until reductions in weights were constant.

Cultivation of D. nodosus from soil microcosms
Soil samples taken at different time points over 40 days from the inoculated microcosms were streaked (approximately 100 mg attached to the tip of the loop and soil suspension (1:10) in quarter strength Ringer's (100 ml)) on 2% TASH plates and incubated anaerobically at 37 C for 3e4 days. The colonies on 2% TASH plates were sub cultured on 4% TASH and again incubated anaerobically at 37 C for 3e4 days for isolating the pure colonies. The colonies obtained on the 4% TASH were confirmed as D. nodosus by Gram staining and a D. nodosus specific 16S rRNA gene colony PCR.

Live and dead cell microscopy and propidium monoazide treatment
The presence of live and dead cells in D. nodosus culture (cells were heat treated at 100 C for 15 min to obtain dead cells while untreated served as live cells) was analysed using 20 mM DAPI (4 0 , 6-Diamidino-2-Phenylindole, Dihydrochloride) and PI (Propidium Iodide) incorporation assay. D. nodosus cells were washed and resuspended in 0.85% NaCl, to which a mixture of DAPI and PI was added and incubated for 15 min in the dark. The cells were observed under fluorescent microscope to confirm the dead cells.
Anaerobically cultured D. nodosus cells on 2% TASH were  [30]. DNA was extracted from non-PMA treated live cells and PMA treated dead cells using DNeasy Blood and Tissue Kit (QIAGEN, Manchester, UK) following the manufacturer's instructions. Quantification of D. nodosus was determined by a qPCR assay targeting rpoD gene. Soil samples taken at different time points from Warwick soil microcosms were PMA treated as described above with some modifications. DNA was extracted from PMA treated samples using the FastDNA Spin kit for Soil (MP Biomedicals, UK) by following the manufacturer's instructions. A qPCR assay targeting rpoD was performed to determine the number of live cells in the given sample.

Turnover of D. nodosus DNA in soil
Soil microcosms containing 1 g Warwick soil were inoculated with D. nodosus in the form of live cells, dead but intact (UV treated), dead (Heat killed) and free DNA at a final concentration of 10 7 cells and rpoD genome equivalents in case of free DNA. The microcosms were moistened to À33 kPa and incubated at 25 C in the dark. Destructive sampling was performed in triplicate at days 0, 4, 7, 10, 14, 20, 25, 30, 40, 50 and 60. DNA was extracted from 0.5 g soil and D. nodosus was enumerated by a qPCR assay targeting rpoD gene.

Statistical analysis
The rpoD mean counts obtained from the qPCR were compared between test days for Warwick soil using two-sample T-Test in GraphPad Prism 5.0 to test for significant differences in the amount of D. nodosus detected at various time points. The exponential decay rates were calculated using the formula P (t) ¼ P 0 e Àrt , where P (t) ¼ Amount of D. nodosus rpoD genome equivalents at time t; P 0 ¼ Initial amount of D. nodosus rpoD genome equivalents at time t ¼ 0; r ¼ the decay rate; t ¼ time (number of periods). Models were built using MlwiN 2.1 [31] to investigate soil parameters including soil type, incubation time, temperature, pH, total organic carbon, moisture, sand, silt, clay in univariable and multivariable mixed effect models to determine whether all parameters together or individually predicted the survival of D. nodosus in soil.

Persistence of metabolically active cells in the soil microcosms
Survival studies indicated that D. nodosus had a biphasic decline in soil with a rate of 0.53 day À1 up to day 6 and 0.12 day À1 from 7 to 14 days (Fig. 1A). Metabolically active cells were detected using cDNA from rpoD RNA and declined in a biphasic manner as observed for DNA genome equivalents at 0.517 day À1 from day 1e6 and 0.021 day À1 from day 6e21 which was significantly lower than the initial rate of decline (p ¼ 0.0049); the decay rate was constant after day 6 (Fig. 1B).

Impact of aeration and survival of D. nodosus by soil type and temperature
For the first 10 days, moisture declined most rapidly in the open microcosms at 1.5% followed by ventilated, (1.4%) partially closed, (0.2%) and closed microcosms, (0.01%) (Fig. 2); after 10 days rates were constant. Therefore, the microcosm with aeration vents was chosen for further experiments.
We detected D. nodosus 16S rDNA by end point PCR in all soil types at each time point during the 40-day study at both 5 C and 25 C. D. nodosus colonies were observed in culture over 40 days and confirmed by staining and colony PCR with the exception of Warwick soil at 25 C on day 40 (Table S1).
At day 0, D. nodosus was present at 10 7 rpoD genome equivalents in all soil types and declined to below 10 5 rpoD genome equivalents over 40 days at 5 C and 25 C in Warwick, Basilicata and Abanilla soils (Fig. 3). There was a biphasic rate of decay in D. nodosus populations in the four soil types ( Table 2); Stockton (heavy clay) had a significantly slower rate of decay in both phases compared with other soil types. In the univariable mixed effects model, the decay in survival over time of genome equivalents of D. nodosus was a cubic function of day (days^1,^2 and^3) demonstrating a nonlinear decay over time. The %TOC, %moisture, %clay, %silt, were significantly associated with increased survival time, %sand was associated with a decreased survival time and pH was not associated with survival time when each of these variables were fitted in turn (Table S2). In the multivariable model, once survival over time was fitted, D. nodosus survived for significantly longer at 5 C compared with 25 C and in Stockton (heavy clay) soil compared with other soil types (Table 3). No other variables were significant because soil types were highly associated to %TOC, moisture, %clay, %silt (Table 1)

Direct evidence for survival of viable cells in Warwick soil microcosms
The heat-treated cells were dead as observed by PI staining (Fig. 4A). PMA treatment of dead cells resulted in a significant reduction in amplification (Fig. 4B) with background amplification of 10 2 cells, which was constant throughout the dilution series. These results confirm that PMA detected dead cells. At day 30, D. nodosus was present in Warwick soil at 10 5 rpoD genome equivalents as detected by qPCR but PMA treatment led to significant reduction (p ¼ 0.0002) to just above the detection limit of 10 2 rpoD genome equivalents. There was no significant difference in the number of live cells observed at 5 C and 25 C at day 30 (Fig. 4C).

Turnover of D. nodosus DNA in soil
DNA turnover in Warwick soil was fastest for free DNA (over 10 days) followed by UV and heat-treated cells. The live and intact UV treated cell survivals were not significantly different (Fig. 4D and Table 4). Rates of decline were biphasic for all with change in rate as follows: day 4 for heat-treated; day 7 for free DNA; day 10 for live cells and day 14 for UV treated cells (Fig. 4D).

Discussion
This is the first study to demonstrate that viable D. nodosus is detectable for 40 days off host in microcosms with aeration vents and that survival rate is affected by moisture, temperature and soil type. This is a step change in our understanding of survival of D. nodosus off host. A previous study of transmission of D. nodosus from floor to sheep indicated a 10-day off host survival of an infectious dose, defined by causing disease [43] and a microcosm   [7]. In the current study, populations of D. nodosus~10 6 cells were detected by qPCR in desiccated, cool, clay soil after 30 days (Fig. 3). Given that approximately 10 2 cells were viable at day 30 in Warwick (non-clay) soil where qPCR was 10 4 , it is likely that >10 4 cells were viable in clay soil at day 30 because of the overall higher load. In addition, the analysis of mRNA in the current study supported the observation of a significant viable D. nodosus population of 1.1 Â 10 4 at 21 days; this was in microcosms with constant moisture content. Although it is not possible to determine whether 10 4 is an infectious dose, it is a similar load to that detected on healthy and footrot affected feet by [44] and highlights that certain soils, in cool damp conditions, might be a source of an infectious dose of D. nodosus for more than 30 days. Non-sterile soil microcosms with four contrasting soil types were used that allowed soil to undergo drying to mimic the environmental conditions, survival in a range of MP's in soil can significantly impact on survival [32]. The wetting and drying curves of the four soils were determined so that the correct volume of water was added to standardize MP at À33 kPa average field moisture content for the UK. During trial desiccations studies the MP varied from À33 kPa to À10,000 kPA in heavy clay soil of Stockton and from À33 kPa to À37 kPa in Basilicata, a sandy loam (Fig. S2). This reflects the suction required to extract water as these soils dry out [3]. Previous work focused on D. nodosus survival in soil but did not consider changes in MP nor soil type and reported survival up to 14 days at 5 C and 7 days at 15 C respectively but water content was kept constant and MP was not determined in the soil used [7].
Studies on the survival of bacterial pathogens in soil have used viable cell counts [25,26,38] and molecular assays [18]. In the current study we used both, since D. nodosus does not form distinct colonies due to twitching motility: qPCR was used to enumerate cells and then RNA extraction and an in situ viability dye (PMA) were used provide a robust measure of viability. There was a significantly lower viable count than total count in the current study but still moderate loads of viable cells. Very few studies have used PMA on starved cells in soil therefore further work is needed to investigate cell integrity during survival.
The most significant reduction in D. nodosus DNA occurred in the first four days after inoculation with free DNA when water was present, in contrast, D. nodosus DNA persisted in soil for up to 60 days when the microcosm had dried out and prevented further degradation of DNA after the first week. Similar trends in turnover of free DNA and degradation were reported for Mycobacterium bovis DNA [45]. Their study also demonstrated that dead cells degraded quickly and that DNA did not persist at normal field moisture content for Warwick soil [45]. Thus it is likely that D. nodosus DNA detected in the current study is from living or, at worst, recently dead cells.
The biphasic decline in D. nodosus load in all soils related to water content indicates the critical drying out MP of each soil. In the case of Stockton this was À1033 kPa at 25 C but only À58 kPA in Abanilla. It is plausible that bacterial cells are so small that they do not respond in the same way to large changes in MP. Stockton soil showed the least decline in percentage water loss and survival mirrored changes in water content. qPCR provided clear evidence of different t d by soil type with slowest t d in heavy clay soil (Stockton). Edaphic factors are critically influential as clay soils can hold more water than sandy soils and dry out more slowly, however, in this experiment the MP increased dramatically in clay soil as drying out resulted in greatly increased suction. This did not impact on D. nodosus survival, in contrast to the impact of MP on plants [2], and therefore strongly suggests that bacteria can survive within clay lattices as they dry out but retain water with D. nodosus surviving in a clay soil with MP of À10,000 kPa. It is likely that the clay content improved survival of culturable cells in Stockton despite the high suction required to access water in a dry clay soil. Gram-negative bacteria studies in soil have demonstrated that clay content [25]; water availability [32]; organic carbon [39] and strain type [26] could all markedly impact survival, with a general trend of heavy, wetter clay soils with higher organic carbon aiding survival. Other features of clay include buffering capacity and better nutrient availability [5,10,16]. The decrease in particle size due to increased clay content increases the surface area of the soil and may result in protection from parasites, predators, toxins, UV and antibiotics [17,19,21]. The type of clay also influences survival and more work is needed on this. Selected studies have implicated the role of the natural community as a key driver in the survival of pathogens in soil where for certain soils the community had a negative effect on survival [27]. Following our analysis of the physicochemical properties affecting D. nodosus survival, work is needed to address the microbial community impacts but these are likely to be soil specific.

Conclusions
From these findings, we conclude that for the gram-negative pathogen D. nodosus, a temperature of 5 C compared with 25 C, and clay soil compared with other soil types, allowed survival of greater loads of D. nodosus over time than the previously reported studies. The results from field trials in Australia and from farmers' empirical experience in the UK suggest that sheep are not at risk from D. nodosus infection once fields have been rested for 14 days. The current study has indicated that D. nodosus can survive in soil with viable cells detected at day 40, and tolerate ultra-low soil MPs. Further work is warranted to determine how clay impacts pathogen survival in soil and whether surviving cells are infectious, and at what dose, and work investigating presence of cells from field samples would provide information on infection load in rested fields.

Conflicts of interest statement
None of the authors have any conflict of interest.