ANTIMICROBIAL-RESISTANT ESCHERICHIA COLI FROM KARST W ATERS, SURFACES AND BAT GUANO IN SLOVENIAN CAVES PROTI PROTIMIKROBNIM UČINKOVINAM ODPORNA ESCHERICHIA COLI V KRAŠKIH VODAH, NA POVRŠINAH IN V GV ANU NETOPIRJEV V SLOVENSKIH JAMAH Sara SKOK 1 , Blaž KOGOVŠEK 1 , Rok TOMAZIN 2 , Samo ŠTURM 3 , Jerneja AMBROŽIČ AVGUŠTIN 4 & Janez MULEC 1,5* Abstract UDC 579:551.44(497.4) Sara Skok, Blaž Kogovšek, Rok T omazin, Samo Šturm, Jerneja Ambrožič Avguštin & Janez Mulec: Antimicrobial resistant Escherichia coli from karst waters, surfaces and bat guano in Slovenian caves Escherichia coli, one of the primary intestinal commensal bac- teria in humans and endothermic animals, is commonly con- sidered an indicator of faecal pollution. E. coli strains were iso- lated from karst rivers under different hydrological conditions, from footpaths in tourist caves and from bat guano. Isolates were tested for phenotypic resistance to ampicillin, chloram- phenicol, ciprofloxacin, nalidixic acid, tetracycline and trim- ethoprim. The highest percentage of antimicrobial resistant E. coli was found in karst waters, followed by those from surface swabs and from bat guano. Several isolates from rivers and swabs exhibited multidrug-resistant phenotype. Environmental conditions impact the populations of E. coli; a positive correla- tion between dissolved oxygen and E. coli counts, and a nega- tive correlation between conductivity and E. coli concentrations have been observed for karst rivers. Malenščica (Slovenia), a drinking water resource with an extensive catchment area, con- tained a relative high percentage of antimicrobial-resistant E. coli strains. None of the isolates from bat guano was resistant to ampicillin, chloramphenicol, and tetracycline. Future moni- toring of bats should consider a regular follow-up of indicative microbial disease indicators in fresh guano. Regular cleansing of tourist footpaths in caves and disinfection barriers at the cave Izvleček UDK 579:551.44(497.4) Sara Skok, Blaž Kogovšek, Rok Tomazin, Samo Šturm, Jerne- ja Ambrožič Avguštin & Janez Mulec: Proti protimikrobnim učinkovinam odporna Escherichia coli v kraških vodah, na površinah in v gvanu netopirjev v slovenskih jamah Bakterija Escherichia coli, eden najpomembnejših črevesnih ko- menzalnih mikroorganizmov pri ljudeh in drugih endotermnih živalih, velja za zanesljivega pokazatelja fekalnega onesnaženja. Sevi bakterije E. coli so bili izolirani ob različnih hidroloških raz- merah iz kraških rek, s turističnih pešpoti v jamah in iz gvana netopirjev. Izolati so bili testirani na izražanje fenotipske odpor- nosti proti ampicilinu, kloramfenikolu, ciprofloksacinu, nali- diksični kislini, tetraciklinu in trimetoprimu. Največji odstotek E. coli, odpornih proti protimikrobnim učinkovinam, je bil v kraških vodah, sledili so brisi jamskih površin in gvano netopir- jev. Pri večjem številu izolatov iz rek in brisov površin je bil izra- žen fenotip večkratne odpornosti proti testiranim učinkovinam. Okoljske razmere pomembno vplivajo na populacije E. coli; v kraških rekah sta bili opaženi pozitivna korelacija med raztopl- jenim kisikom in številom E. coli ter negativna korelacija med prevodnostjo in koncentracijo E. coli. Malenščica, vir pitne vode z velikim prispevnim območjem, je imela razmeroma velik de- lež sevov E. coli, odpornih proti protimikrobnim učinkovinam. Nobeden izmed izolatov iz gvana netopirjev ni bil odporen proti ampicilinu, kloramfenikolu ali tetraciklinu. Za prihodnji moni- toring netopirjev bi bilo smiselno razmisliti o rednem spreml- janju prisotnosti mikrobnih bolezenskih indikatorjev v svežem 1 Research Centre of the Slovenian Academy of Sciences and Arts, Karst Research Institute, Titov trg 2, SI-6230 Postojna, Slovenia, e-mails: blaz.kogovsek@zrc-sazu.si, janez.mulec@zrc-sazu.si, sara.skok@gmail.com 2 University of Ljubljana, Faculty of Medicine, Institute of Microbiology and Immunology, Zaloška 4, SI-1000 Ljubljana, Slovenia, e-mail: rok.tomazin@mf.uni-lj.si 3 Park Škocjanske jame, Slovenija, Škocjan 2, SI-6215 Divača, Slovenia, e-mail: samo.sturm@psj.gov.si 4 University of Ljubljana, Biotechnical Faculty, Department of Biology, Večna pot 111, SI-1000 Ljubljana, e-mail: jerneja. ambrozic@bf.uni-lj.si 5 UNESCO Chair on Karst Education, University of Nova Gorica, Glavni trg 8, SI-5271Vipava, Slovenia, e-mail: janez.mulec@zrc-sazu.si * corresponding author Received/Prejeto: 06.10.2020 DOI: 10.3986/ac.v49i2-3.9103 ACTA CARSOLOGICA 49/2-3, 265-279, POSTOJNA 2020 COBISS: 1.01 SARA SKOK, BLAŽ KOGOVŠEK, ROK TOMAZIN, SAMO ŠTURM, JERNEJA AMBROŽIČ AVGUŠTIN & JANEZ MULEC INTRODUCTION Local and global ecosystems are affected by various hu- man interventions, in terms of climate change, loss of biodiversity, and land transformation (Vitousek et al. 1997). Interventions are either short-lasting with a tran- sitory impact, or permanent, which commonly include physical changes of the habitat (Vitousek et al. 1997), and introduction of allochthonous material, including biota (Mösslacher et al. 2001; Vitousek et al. 1997). Conse- quently, the autochthonous communities can be affected and changed. Evaluation of human impact is an impor- tant step in following the dynamics and trends within the changing ecosystems to foresee the need for protection measures. Karst ecosystems are also subjected to such changes (Notenboom et al. 1994). For example, karst groundwater environments are highly vulnerable to ag- ricultural, urban or industrial impacts from both point- source and dispersed pollution (Kačaroğlu 1999). In many ways karst caves are an excellent model to observe and quantify different external impacts, for example on groundwater quality or cave-adapted biota in show caves. Some caves host bats – migratory animals and top preda- tors. Because of their sensitivity to changes in land use, agriculture and habitat fragmentation, bats are particu- larly good indicators of changes affecting wildlife and surrounding areas (Jones et al. 2009). Complete insight into ecosystem changes is ob- tained only by a simultaneous follow-up monitoring of physico-chemical parameters and the biota, including the microbiota. Escherichia coli, a Gram-negative bacte- rium from the Enterobacteriaceae family, is primarily a member of the gut microbiota in humans and other en- dothermic animals, and is therefore widely regarded as an indicator for faecal pollution, when it is isolated from other environments (Odonkor & Ampofo 2013; Jang et al. 2017). In the human gut, E. coli is regarded as a com- mensal microbe. However, due to the extreme plasticity of its genome, strains “equipped” with virulence factor genes can cause several intestinal and extraintestinal in- fections. Among them, most frequent are urinary tract infections and bloodstream infections in the European Union/European Economic Area, and they have both community and healthcare origins. These infections are usually treated with antimicrobials. The extensive use of antimicrobial agents in human and veterinary medicine and in agriculture, has led to a dramatic increase in the number of antimicrobial resistant bacteria (ARB) and antimicrobial resistance genes (ARG) encoded predomi- nantly on mobile genetic elements (MGE). Thus, the “cir- culation” of resistant, commensal and pathogenic E. coli between the clinical settlement and natural ecosystems enables the transfer of antimicrobial resistant (AMR) E. coli and ARG from human impacted areas (clinics, ani- mal farms, wastewater treatment plants, …) to pristine natural habitats, e.g., underground karst water, and their biota. From this point of view, it is important to detect and monitor the number of AMR-E. coli, their resistance profiles and putative MGEs, as well as possible entry points into such environments (Zhang et al. 2013). Recently it has been reported that the most impor- tant and/or prevalent detected resistance in E. coli strains from Europe is against aminopenicillins, fluoroquino- lones, third generation cephalosporins and aminogly- cosides. Currently, strains resistant against carbapenems remain rare in Europe (European Centre for Disease Prevention and Control 2018). In Slovenia in 2017, when the majority of samples for this study were collected, the highest proportion of resistant E. coli clinical isolates was detected in the case of aminopenicillins, where it reached 51.6 %, followed by fluoroquinolones and third genera- tion cephalosporins with proportions of 24.9 % and 12.5 % respectively. Combined resistance against third gen- eration cephalosporins, fluoroquinolones and aminogly- cosides was observed in 6.3 % of E. coli clinical isolates. Carbapenem-resistance was, however, not detected in the E. coli population surveyed in 2017 (European Centre for Disease Prevention and Control 2017, 2018). During the present study, flowing surface waters and groundwaters, solid surfaces impacted by orga- nized tours in show caves, and bat guano were screened for E. coli, and compared according to the phenotypic resistance of E. coli to selected antimicrobials, i.e., am- picillin (AMP), chloramphenicol (CHL), ciprofloxacin entrances reduce the concentration and transmission of E. coli significantly. A future, more detailed, study on characterization of additional E. coli isolates is needed to reveal their pathoge- neicity, mechanisms of antibiotic resistance, mobile genetic ele- ments, and gene transfer frequencies to other members of the karst microbiome. Key words: caves, water, bats, guano, swabs, E. coli, antimicro- bial resistance. gvanu. Redno čiščenje turističnih pešpoti v jamah in dezinfekcij- ska bariera ob vhodu pomembno zmanjšata številčnost in prenos E. coli. V prihodnosti bodo potrebne še nadaljnje in podrobnejše karakterizacije večjega števila izolatov E. coli za razumevanje nji- hove patogenosti, mehanizmov odpornosti proti antibiotikom, mobilnih genetskih elementov in frekvenc genskega prenosa na druge organizme v kraškem mikrobiomu. Ključne besede: jame, voda, netopirji, gvano, brisi, E. coli, od- pornost proti protimikrobnim učinkovinam. ACTA CARSOLOGICA 49/2-3 – 2020 266 ANTIMICROBIAL RESISTANT ESCHERICHIA COLI FROM KARST W ATERS, SURFACES AND BAT GUANO IN SLOVENIAN CAVES (CIP), nalidixic acid (NAL), tetracycline (TET) and trim- ethoprim (TMP). These antimicrobials were selected be- cause they were already reported from comparable envi- ronments (Reinthaler et al. 2003; Amaya et al. 2012). It is important to stress that highly karstifed zones can enable more intermicrobial contacts and potential gene transfer into the natural resistome via water flow, air streaming and migratory animals and humans. MATERIAL AND METHODS DESCRIPTION OF SITES AND SAMPLES W ATER SAMPLES Water samples were collected from five karst rivers (Črni potok, Malenščica, Pivka, Rak, Unica, Fig. 1). These riv- ers are part of the nearly 800 km 2 catchment area of the Malenščica and Unica springs. The Pivka River (and Črni potok to a minor extent) is the main allogenic recharge for the Postojnski jamski sistem (Postojna Cave System, 45°46ʹ57˝N, 14°12ʹ13˝E, 529 m a.s.l., total length 24 km), which comprises several caves (Postojnska jama, Otoška jama, Magdalena jama, Pivka jama, Črna jama), where the underground Pivka River is accessible and collects isotopically different waters from the aquifer. The un- derground Pivka River is impacted by human activities within the surrounding area, as is indicated by occasion- ally elevated concentrations of sulphates, chlorides, and organic and faecal pollutants (Mulec et al. 2019). In the underground it continues towards Planinska jama, where it converges with the Rak cave stream. Downstream from the underground confluence, the river emerges as the Unica (45°49ʹ13˝N, 14°14ʹ45˝E, 453 m a.s.l.). The Malenščica spring (45°49ʹ20˝N, 14°15ʹ19˝E) is exploited as a drinking water source for 21,000 inhabitants of the surrounding area (Petrič 2010). Sampling locations were selected upstream of the river sinks into the karst underground, along the under- ground water courses, and at the springs (see Fig. 1B for sampling sites). On-site measurements of temperature (T), electric conductivity (EC), pH and dissolved oxygen (DO) Fig. 1: Study sites: A – location of investigated caves (I – Županova jama, II – Postojnska jama and Planinska jama, III – Predjama, IV – Škocjanske jame); B – map of the area with water sampling sites (Map based on ARSO 2019; Cave Cadastre 2019). ACTA CARSOLOGICA 49/2-3 – 2020 267 SARA SKOK, BLAŽ KOGOVŠEK, ROK TOMAZIN, SAMO ŠTURM, JERNEJA AMBROŽIČ AVGUŠTIN & JANEZ MULEC were performed using a Multi-parameter portable meter (WTW Multiline Multi 3620 IDS). After sampling, sam- ples were transferred in a cool-box to the laboratory, where one millilitre of each specimen was placed directly onto microbiological media. The total ATP concentration of each sample served to estimate its total microbial biomass. Samples were collected under different hydrological conditions during 2017; from Črni potok in June, July, August and November, from Malenščica in September, from the Pivka in May, July, August, September, October and November, from the Rak in September, and from the Unica in September, October and November. SURFACE SW AB SAMPLES Swab samples from tourist footpaths were collected from four caves. In Postojnska jama, 780,000 tourists visited 5.3 km of the underground galleries in 2017. The surfaces of tourist footpaths are cleaned monthly with water jets. A disinfection barrier is installed at the entrance to the cave (Appendix 2). Predjama (Predjama Cave, 45°48ʹ56˝N, 14°7ʹ35˝E, 491 m a.s.l.) is explored in the total length of ap- proximately 17 km. 0.7 km is visited annually by approxi- mately 7,000 tourists. The first part of the cave serves as hibernaculum and swarming site for at least 13 bat species (Barbastella barbastellus, Eptesicus serotinus, Miniopterus schreibersii, Myotis capaccinii, M. daubentonii, M. oxygna- thus, M. myotis, M. nattereri, Nyctalus noctula, Pipistrellus pipistrellus, P. pygmaeus, Rhinolophus ferrumequinum, R. hipposideros). M. schreibersii presents almost ⅔ of indi- viduals, with approximately 1,200 to 1,500 hibernating bats (Presetnik et al. 2009). Bat droppings have accumulated in the form of heaps, and they are also scattered along the tourist footpath, from where the samples were collected. Škocjanske jame (Škocjan Caves, 45°39ʹ54˝N, 13°59ʹ38˝E, 317 m a.s.l., total length 6.2 km) are on the UNESCO W orld Heritage List and recognized as an underground karst wet- land under the Ramsar W etland Classification System. A 2.6 km section of the cave that is open to visitors throughout the year attracted 185,000 visitors in 2017. Nine bat species were documented in the cave (Barbastella barbastellus, Min- iopterus schreibersii, Myotis blythii, M. capaciinii, M. myotis, Pipistrellus pipistrellus, Plecotus macrobullaris, Rhinolophus euryale, R. ferrumequinum, R. hipposideros). The predomi- nant species, exceeding 8,000 individuals during the winter hibernation, is M. schreibersii (Presetnik et al. 2017). Con- sequently, parts of the tourist footpath are covered with bat droppings. Tourist footpaths are cleaned using water jets once per year in September. Županova jama (Županova Cave, 45°54ʹ47˝N, 14°38ʹ17˝E, 468 m a.s.l.) is a smaller tour- ist cave (0.71 km total length, with 0.61 km equipped for tourist access) visited by 5,000 tourists per year. On the tourist footpaths sampling sites for swabs were delimited with a template for bioburden control (5×4 cm, Copan Diagnostics Inc.) and sampled with a FLOQSwabsTM (Copan). Swabs were transferred into one millilitre of saline solution. In the laboratory, sam- ples were vortexed vigorously, and the liquid was col- lected in a separate tube. Additionally, each swab was centrifuged for 10 minutes at 4,000 RPM in its original tube to release any remaining liquid and associated mi- crobes. Subsequently the liquid was used for inoculation of microbiological plates and estimation of the total ATP . Swabs were collected from Postojnska jama, Pred- jama and Županova jama in November 2017 and in Škocjanske jame in September 2017. GUANO SAMPLES Guano scattered on the cave floor was sampled in Pred- jama in June 2018 and Škocjanske jame in April 2018. In addition, fresh guano was collected under roosting bat colonies in Škocjanske jame (the section named Schmid- lova dvorana and Rudolfova dvorana, 100 and 200 m from the cave entrance respectively), in August–Septem- ber 2018. Sterilized aluminium foil sheets were placed along the tourist footpath (which was not in use during guano collection) under bat colonies, and samples were collected after 7, 18 and 30 days of accumulation (Ap- pendix 1). Based on the results of field observations, the fresh guano probably belonged to a mixed population of several hundred individuals of Miniopterus schreibersii, Myotis myotis, Myotis blythii and Myotis capaciinii. In the laboratory, each guano sample was resus- pended in saline solution (1:10 w/v), serially diluted (up to 10 -3 ), and plated onto E. coli selection medium. MICROBIAL BIOMASS ESTIMATION, ISOLATION AND IDENTIFICATION OF E. COLI The ATP content in samples was estimated using AquaSnap Total testing instruments (Hygiena, USA) and expressed in RLU – Relative Light Units (where 1 RLU equates to 1 fmol of ATP) per millilitre (RLU /ml) or swabbed surface (RLU /20 cm 2 ). Ready-to-use microbiological media Compact Dry (Nissui Pharmaceutical, Japan) were used to esti- mate the concentration of heterotrophic aerobic bacte- ria (Compact Dry TC), coliforms and E. coli (Compact Dry EC) and enterococci (Compact Dry ETC). One set ACTA CARSOLOGICA 49/2-3 – 2020 268 ANTIMICROBIAL RESISTANT ESCHERICHIA COLI FROM KARST W ATERS, SURFACES AND BAT GUANO IN SLOVENIAN CAVES of TC plates was cultivated for 48 hours at 37 °C, and another set for 7 days at 20 °C. Numbers of bacteria were expressed as Colony-Forming Units (CFU) per millilitre for water samples or CFU per 20 cm 2 for surface swabs. Coloured colonies indicative for E. coli on Compact Dry EC plates were further purified on CCA agar (Coli- forms Chromogenic Agar, Conda Pronadisa, Spain) and subjected to identification by MALDI-TOF MS (Matrix- Assisted Laser Desorption/Ionization Time-Of-Flight Mass Spectrometry). Pure isolates from CCA plates were subsequently inoculated on 5.0 % defibrinated sheep- blood agar (BA) and incubated at 36±1 °C for 24 to 48 hours. BA was prepared at the Institute of Microbiology and Immunology, Faculty of Medicine, University of Lju- bljana, and contained 15 g agar (Sigma-Aldrich), brain- heart infusion broth (Becton Dickinson, Sparks, MD, USA), and 50 ml defibrinated reoxygenated sheep blood (Bio Gnost, Zagreb, Croatia) per 1000 ml. Bacterial iso- lates on the BA were subjected to identification using MALDI-TOF MS with an on-target formic acid extrac- tion technique. A 24 to 48 hours-old single colony was smeared onto the MALDI steel plate and overlain with 1 µl of 98 % formic acid. After drying, the sample was over- lain with 1 µl of photo-absrobent, α-cyano-4-hydroxy- cinnamic acid matrix solution in 50 % acetonitrile-2.5 % trifluoroacetic acid (Bruker Daltonik, Germany) and left to dry before subsequent analysis with a linear-mode microflex LT/SH MALDI-TOF MS system, Biotyper RTC software version 3.1 (Bruker Daltonik). The Bruker bacterial test standard (Bruker Daltonik) was used for calibration according to the manufacturer’s instructions. Quality of identification was assessed using the manufac- turer’s score value. A score of ≥2.000 indicated reliable species level identification, a score of 1.700 to 1.999 indi- cated reliable identification to the genus level, and a score of <1.700 was interpreted as unreliable identification. Identified E. coli isolates were designated (EC) and deposited at the Ex Culture Collection at the Department of Biology, Biotechnical Faculty, University of Ljubljana (Infrastructural Centre Mycosmo, MRIC UL, Slovenia). ANTIMICROBIAL SUSCEPTIBILITY TESTING Isolates from overnight incubation (minimum require- ment – belonging to genus Escherichia defined by MAL- DI-TOF MS Score value > 1.700), were tested for anti- biotic resistance against AMP (0.016–256 µg/ml), CHL (0.016–256 µg/ml), CIP (0.002–32 µg/ml), NAL (0.016– 256 µg/ml), TET (0.016–256 µg/ml) and TMP (0.002–32 µg/ml). Susceptibility testing was carried out using the gradient diffusion method on unsupplemented Mueller- Hinton agar (MHA, bioMérieux SA, France). The inocu- lum was prepared by suspending pure overnight cul- tures in saline solution, approximately corresponding to 1–2×10 8 CFU/ml. The inoculum was spread evenly over the entire surface of the MHA plates using a sterile cotton swab. E-test gradient strips (bioMérieux SA, France, for AMP, CHL, CIP, NAL, TET, and Bioanalyse, Turkey, for TMP) were applied on the agar surface within 15 minutes of inoculation of the plates. After 24 hours of ambient air incubation at 36±1 °C, MICs (Minimum Inhibitory Con- centrations) were determined visually according to the manufacturer’s instructions (bioMérieux, Bioanalyse). The results were interpreted according to the EUCAST (European Committee on Antimicrobial Susceptibility Testing) epidemiological cutoff values (ECOFF) for E. coli: AMP at 8 mg/l, CHL at 16 mg/l, CIP at 0.064 mg/l, NAL at 16 mg/l, TET at 8 mg/l and TMP at 2 mg/l (EU- CAST 2019). STATISTICAL ANALYSES Statistical analyses were performed using Daniel’s XL Toolbox, an open-source add-in for Microsoft Excel (Version 6.60, licensed under the Apache License, Ver- sion 2.0, Daniel Kraus, Würzburg, Germany). RESULTS IDENTIFICATION AND ENVIRONMENTAL CONDITIONS Samples that tested positive for the presence of E. coli on Compact Dry EC selection medium were further pro- cessed and analysed. Altogether 211 colonies were sub- jected to identification by MALDI-TOF MS. From these, 185 isolates were E. coli (highly probable species iden- tification – 100, secure genus identification – 77, prob- able genus identification – 8). The remaining isolates belonged to genera from the Enterobacteriaceae family (Enterobacter, Hafnia, Morganella, Proteus, Providencia, Serratia and Yersinia). Subsequently, 168 E. coli isolates were selected in the final dataset for antimicrobial sus- ceptibility testing. E. coli strains were retrieved from waters under dif- ferent environmental conditions (Tab. 1). Statistically significant positive correlations were observed between DO and concentrations of E. coli (r = 0.325, p = 0.008), ACTA CARSOLOGICA 49/2-3 – 2020 269 SARA SKOK, BLAŽ KOGOVŠEK, ROK TOMAZIN, SAMO ŠTURM, JERNEJA AMBROŽIČ AVGUŠTIN & JANEZ MULEC enterococci (r = 0.372, p = 0.002) and total heterotrophic bacteria grown at 20 °C (r = 0.342, p = 0.004). Negative correlations have been between EC and concentrations of E. coli (r = –0.446, p = 0.000), enterococci (r = –0.372, p = 0.001) and total heterotrophic bacteria grown at 20 °C (r = –0.265, p = 0.024). Whereas the ATP concentration values are indica- tive of abundant microbial biomass in all rivers, E. coli represented only a small proportion within the microbial communities. A comparatively high concentration of E. coli, relative to total microbial biomass, was detected in the Pivka River (sampling sites nos. 1 and 2, Fig. 1B) and even in water samples from the Malenščica spring (sampling site No. 11, Fig. 1B), a drinking-water source (Tab. 3). Surfaces of the tourist footpaths in caves varied in their microbiological load. ATP biomass estimator showed the lowest bioburden on tourist footpaths in Županova jama and relative uniform concentrations in Škocjanske jame and Predjama, both inhabitated by bats. Among the sites tested the cultivable microbial biomass was the lowest in Postojnska jama. This can be attributed to effective cleaning of the tourist footpaths one week be- fore the surfaces were sampled. The highest number of E. coli retrieved during the study was in Škocjanske jame, with more than 500 CFU/ 20 cm 2 (Tab. 2). In Predjama and Škocjanske jame, respectively 7 and 32 E. coli strains were isolated from non-fresh guano. Additionally, 34 strains were isolated from fresh guano sampled in Škocjanske jame. ANTIBIOTIC SUSCEPTIBILITY TESTING The MICs obtained were interpreted according to ECOFF, which divides the bacterial population into: • in vitro sensitive wild types (WT) – if they expressed MIC ≤ ECOFF they do not have phenotypically de- tectable acquired resistance mechanisms to a speci- fied antimicrobial agent, and thus represent a “nor- mal” population; • in vitro resistant/less susceptible non-wild types Tab. 2: Value ranges of bacterial indicators on swabbed surfaces. Cave Parameter Postojnska jama Predjama Škocjanske jame Županova jama ATP (RLU/ 20 cm 2 ) 3,804–27,036 2,784–17,952 2,776–22,240 1,480–24,304 TC-37 °C (CFU/ 20 cm 2 ) 5,067 11,400–41,019 9,800–170,667 17,400–35,600 Coliforms (CFU/ 20 cm 2 ) 324 92–3,480 548–1,532 812–3,400 E. coli (CFU/ 20 cm 2 ) 28 56–228 52–580 20–68 TC – heterotrophic aerobic bacteria Tab. 1: Ranges of environmetal parameters of the studied rivers. River Parameter Črni potok Malenščica Pivka Rak Unica T (°C) 3.8 8.8–10.2 4.9–15.1 9.0–11.6 6.9–11.1 EC (µS/cm) 103 382–393 302–394 358–425 332–452 pH 7.00 7.48–7.59 7.64–8.29 7.50–8.22 7.47–8.02 DO (mg/l) 9.28 9.10–10.06 6.22–13.92 8.23–11.68 10.47–12.90 DO (%) 75.5 83.6–94.3 66.4–117.0 80.4–108.7 100.2–111.0 ATP (RLU) 80 6–27 88–439 9–128 9–344 TC-20 °C (CFU/ml) 181 165–511 530–1850 249–1081 121–1300 TC-37 °C (CFU/ml) 6 32–160 30–1200 44–411 10–484 E. coli (CFU/ml) 5 1–9 1–54 1–23 1–24 Coliforms (CFU/ml) 6 9–40 2–308 18–205 3–228 ETC (CFU/ml) 2 0–4 0–84 0–22 0–36 TC – heterotrophic aerobic bacteria; ETC – enterococci ACTA CARSOLOGICA 49/2-3 – 2020 270 ANTIMICROBIAL RESISTANT ESCHERICHIA COLI FROM KARST W ATERS, SURFACES AND BAT GUANO IN SLOVENIAN CAVES (NWT) – in cases of MIC > ECOFF, when they have phenotypically detectable acquired resistance mechanisms. The division of isolates into WT and NWT does not consider clinical and pharmacological (pharmacokinetic and pharmacodynamic) factors and is therefore not re- lated to the patient (human, animal) and antimicrobial therapy regimen (EUCAST 2017). For most of the an- timicrobial drugs tested (AMP, CHL, TET, TMP), clini- cal breakpoints (CBS) have been set for interpreting the susceptibility-testing results for E. coli, and can therefore be classified as susceptible, susceptible-increased expo- sure, or resistant, according to obtained MICs. But be- cause these CBS values are set only for infections caused by E. coli in human patients, the results have been inter- preted only on the basis of their ECOFF, because this is not patient-species dependent. The highest percentages of NWT E. coli strains were found in samples collected from rivers (12.5 % for CIP, NAL and TMP , 9.7 % for TET, 8.3 % for AMP , 4.2 % for CHL), followed by those from contact surface isolates. E. coli isolates derived from guano samples expressed low percentages of in vitro phenotypic resistance. None of the isolates was resistant to AMP , CHL and TET. E. coli iso- lated from fresh guano expressed phenotypic resistance only to TMP (1.4 %, Fig. 2). Comparison of the river samples showed that iso- lates from Malenščica contained the highest percentage of NWT isolates (n = 12, 58.3 %), followed by those from the Unica (n = 12, 33.3 %), the Pivka (n = 28, 32.1 %) and the Rak (n = 18, 16.7 %). Neither of the two isolates from Črni potok expressed in vitro antimicrobial resistence. The highest percentage of NWT E. coli isolates from sur- face swab samples was detected among the Škocjanske jame samples (n = 18, 33.3 %), followed by those from Županova jama (n = 5, 20 %), and then those from Pred- jama (n = 15, 6.7 %). The single isolate from Postojnska jama expressed in vitro resistence to TMP . Several isolates from rivers and surfaces were mul- tidrug resistant, and some of them expressed no in vitro inhibition with MICs exceeding the highest tested value designated on E-test gradient strips. A strain from the ponor of the Pivka into Postojnska jama was resistant to all tested antimicrobials (Tab. 3; Tab. 4). Guano isolates were less resistant than those from both the water and surface isolates (Tab. 5). The highest percentage of E. coli in the microbial community (12.2–59.2 %) and the highest percentage of NWT E. coli (33.3 %) were detected in swabs from Škocjanske jame, followed by swabs from Županova jama (20.0 %) and those from Predjama (6.7 %). Altogether, ten surface swabs were sampled in Postojnska jama, but only one of them was positive for E. coli, despite a high total biomass in the sample (Tab. 4). Compared to the E. coli isolated from river water and swabs, the isolates from guano expressed the lowest in vitro resistance to the selected antimicrobials (n = 55, 9.0 %). The MIC ranges, MIC 50 , MIC 90 and mean MICs of six antibacterial drugs are presented in Tab. 6, and MIC distributions are shown in Fig. 3. The overall MIC rang- es for each of six antimicrobial drugs were 0.5–256 µg/ ml for AMP, 4–256 µg/ml for CHL, 0.008–32 µg/ml for CIP , 1–256 µg/ml for NAL, 0.75–256 µg/ml for TET and 0.125–32 µg/ml for TMP. NAL and AMP had the high- est mean MIC of 17.6 µg/ml and 15.9 µg/ml, respectively. The lowest mean MIC was for CIP with 0.221 µg/ml. Fig. 2: Percentage of NWT E. coli strains, isolated from karst water, surfaces and guano, resistant to AMP, CHL, CIP, NAL, TET and TMP. ACTA CARSOLOGICA 49/2-3 – 2020 271 SARA SKOK, BLAŽ KOGOVŠEK, ROK TOMAZIN, SAMO ŠTURM, JERNEJA AMBROŽIČ AVGUŠTIN & JANEZ MULEC Tab. 4: Antimicrobial resistance phenotype and percentage of E. coli isolates relative to cultivable bacteria in swabs. Strain Cave Antimicrobial resistance phenotype ATP (RLU) E. coli (%)* EC76 Škocjanske jame TET 2,776 59.2 EC77 Škocjanske jame TMP 3,068 18.3 EC78 Škocjanske jame AMP, TET, TMP 3,068 18.3 EC79 Škocjanske jame TET 3,068 18.3 EC85 Škocjanske jame AMP, CIP, NAL 4,656 12.2 EC86 Škocjanske jame TMP 3,144 14.6 EC110 Postojnska jama TMP 26,456 8.6 EC116 Predjama CHL 3,468 4.7 EC136 Županova jama AMP, CHL, NAL, TET 1,720 0.6 bold, no in vitro inhibition with MICs exceeding the highest value on E-test gradient strips: MIC ≥ 256 µg/ml for AMP , MIC ≥ 256 µg/ml for CHL, MIC ≥ 32 µg/ml for CIP , MIC ≥ 256 µg/ml for NAL, MIC ≥ 256 µg/ml for TET, MIC ≥ 32 µg/ml for TMP; relative to bacterial counts at 37 °C Tab. 3: Antimicrobial resistance phenotype and percentage of E. coli isolates according to total biomass and cultivable bacterial indicators from water samples (see Fig. 1B for sampling sites). Strain River / site no. Antimicrobial resistance phenotype ATP (RLU) / CFU/ ml # E. coli (%)* E. coli (%)** E. coli (%)*** EC40 Malenščica / 11 AMP 27 / 10 5 0.0090 2.4 7.7 EC42 Malenščica / 11 AMP 27 / 10 5 0.0090 2.4 7.7 EC45 Malenščica / 11 CIP , NAL 26 / 10 5 0.0050 1.0 3.1 EC55 Malenščica / 11 NAL, TMP 27 / 10 5 0.0090 2.4 7.7 EC66 Malenščica / 11 CIP , NAL, TMP 26 / 10 5 0.0050 1.0 3.1 EC67 Malenščica / 11 CIP , NAL 26 / 10 5 0.0050 1.0 3.1 EC68 Malenščica / 11 CIP , NAL 26 / 10 5 0.0050 1.0 3.1 EC61 Pivka / 1 TET, TMP 439 / 10 6 0.0035 4.0 6.1 EC126 Pivka / 1 CHL 207 / 10 6 0.0054 5.0 14.4 EC127 Pivka / 1 AMP, CHL, CIP , NAL, TET, TMP 207 / 10 6 0.0054 5.0 14.4 EC130 Pivka / 1 TET, TMP 207 / 10 6 0.0054 5.0 14.4 EC95 Pivka / 2 AMP, CIP , NAL, TET, TMP 135 / 10 6 0.0006 0.8 20.0 EC119 Pivka / 2 NAL 129 / 10 6 0.0043 3.6 11.3 EC120 Pivka / 5 NAL 206 / 10 6 0.0043 3.3 10.8 EC121 Pivka / 5 TMP 206 / 10 6 0.0043 3.3 10.8 EC44 Pivka / 7 AMP , TET, TMP 363 / 10 6 0.0005 0.7 1.0 EC38 Rak / 9 CIP 25 / 10 5 0.0020 0.5 1.9 EC39 Rak / 9 CIP 25 / 10 5 0.0020 0.5 1.9 EC65 Rak / 9 TMP 12 / 10 4 0.0100 0.3 0.8 EC27 Unica / 10 CIP 146 / 10 6 0.0022 2.4 4.8 EC87 Unica / 10 AMP, TET 344 / 10 6 0.0005 0.6 1.0 EC124 Unica / 10 CHL 174 / 10 6 0.0024 1.8 13.1 EC125 Unica / 10 TET 174 / 10 6 0.0024 1.8 13.1 bold, no in vitro inhibition with MICs exceeding the highest value on E-test gradient strips: MIC ≥ 256 µg/ml for AMP , MIC ≥ 256 µg/ml for NAL, MIC ≥ 256 µg/ml for TET, MIC ≥ 32 µg/ml for TMP; # , total biomass estimation (Hygiena, USA) *, relative to total microbial counts; **, relative to bacterial counts grown at 20 °C; ***, relative to bacterial counts at 37 °C ACTA CARSOLOGICA 49/2-3 – 2020 272 ANTIMICROBIAL RESISTANT ESCHERICHIA COLI FROM KARST W ATERS, SURFACES AND BAT GUANO IN SLOVENIAN CAVES Tab. 6: Susceptibility of E. coli from karst waters, surfaces and bat guano against six antimicrobials represented by MICs (n = 168). MIC (µg/ml) AMP CHL CIP NAL TET TMP Range 0.5–256 2–256 0.008–32 1.5–265 0.75–265 0.125–32 Mean 15.9 10.0 0.221 17.6 13.9 2.20 MIC 50 4 8 0.016 6 2 0.5 MIC 90 8 12 0.032 12 4 2 Tab. 5: Antimicrobial resistance phenotype of E. coli isolates in guanos. Strain Cave Relative guano age Antimicrobial resistance phenotype EC157 Škocjanske jame Non-fresh TMP EC163 Škocjanske jame Non-fresh NAL EC172 Škocjanske jame Non-fresh CIP EC201 Škocjanske jame Fresh TMP EC205 Škocjanske jame Fresh TMP Fig. 3: MIC distributions for six antimicrobials. Green columns represent wild type MIC ranges; red columns represent non-wild type MIC ranges. ACTA CARSOLOGICA 49/2-3 – 2020 273 SARA SKOK, BLAŽ KOGOVŠEK, ROK TOMAZIN, SAMO ŠTURM, JERNEJA AMBROŽIČ AVGUŠTIN & JANEZ MULEC DISCUSSION Inevitably the global use of large quantities of antimicro- bials to protect human and animal health, and to act as growth promotors, has led to the emergence and spread of ARB and ARG. In addition to compromising the treatment of infectious diseases, the natural resistome is constantly increasing in terms of the number and diver- sity of ARB and ARG. Pre-existing background levels of naturally occurring microbes, and those released due to discharge of human and animal waste, both contribute to the developing environmental resistome (Shao et al. 2018; Zhao et al. 2018). This study was designed to assess the presence and number of total and AMR-E. coli compared to the total microbial biomass in samples from different sites in the Slovenian karst ecosystem. Since resistant E. coli strains allow the introduction of ARG into the natural resistome, from where they can again be mobilized into different strains and even species, tracking and identifying ARB and ARG in groundwater resources and their environ- ment has become extremely important. Although the concentrations of antimicrobials and their metabolites in samples were not measured as part of the present study, their presence can be anticipated due to the the occur- rence of AMR strains within the environment (Huang et al. 2011). ANTIMICROBIALS IN KARST HABITATS The presence in the environment of NWT E. coli strains resistant to AMP , CHL, CIP , NAL, TET and TMP has al- ready been reported (Reinthaler et al. 2003; Amaya et al. 2012). Within the present study, resistance to at least one of the tested antimicrobials could be detected in 37 out of 168 E. coli isolates. Among these 23, 9 and 5 were from water, surface, and guano samples, respectively. β-lactams, e.g., AMP, are the first and the most widely used group of antimicrobials, mainly because of their efficiency compared to their low toxicity (Pang et al. 2016), and thus the presence of resistance genes in natural ecosystems is to be expected. They are consid- ered critically important by the World Health Organiza- tion (WHO) because of their wide use in treatment of infections caused by bacteria transmitted also from non- human sources or bacteria that acquire resistance genes from non-human sources (WHO 2017). Of 18 AMP-re- sistant strains, 6 were isolated from water, 3 from surfac- es and none from the guano samples. β-lactam-resistant strains have previously been described in karst drinking- water resources (Laroche et al. 2010). The prevalance of NAL resistance was rather high in all screened habitats (Fig. 2). Eight out of eleven quinolone-resistant isolates were retrieved from water samples (Tabs. 3–5). NAL, the first synthetic quinolone, has been in clinical use since 1967 and historically was used to treat urinary-tract infections caused by Gram negative bacteria (Emmerson & Jones 2003); nowadays it is used mostly as a screening test for (fluoro)quino- lone resistance. The second generation fluoroquino - lone, CIP, is one of the quinolones most prescribed in Europe (Van Doorslaer et al. 2011) and it is yet another critically important antimicrobial, especially in cases of invasive Campylobacter, Salmonella and multi-drug resistant Shigella infections. Quinolone-resistant infec- tions can also result from a non-human source of En- terobacteriales, including E. coli (WHO 2017). Limited biodegradability of CIP in conventional wastewater treatment plants (Robinson et al. 2005; Grenni et al. 2018) might be the reason for the relatively high preva- lence of associated resistance in karst waters (Fig. 2). Although resistance to both quinolones is due mainly to chromosomal mutations, horizontal transfer events of mutated chromosomal regions, mediated by phages, may play a significant role in AMR-resistance issues as recognized recently (Calero-Caceres et al. 2019; Bal- cazar 2020). Plasmid mediated quinolone resistance is also possible, and isolates like EC38, EC39 and EC27, with high MIC values for CIP and low MIC values for NAD, are commonly resistant to quinolones due to plasmid transfer (Hooper & Jacoby 2015). Tetracyclines, which were the first antimicrobials used as growth promotors in the 1940s, are used predom- inantly in veterinary medicine (Speer et al. 1992; Chopra & Roberts 2001). They are considered to be highly im- portant antimicrobials because the loss of their efficacy, due to the emergence of resistance, would have a signifi- cant impact on morbidity and mortality, especially for infections caused by Brucella, Rickettsia and Chlamydia (WHO 2017). Among the 11 TET-resistant isolates met in this study, 7 were from water samples and 4 from surface samples. TET resistance genes are among the most commonly described from water samples, includ- ing those from the karst water supply (Chen et al. 2017; Stange & Tiehm 2020). TMP-resistant strains were the second most com- mon AMR-E. coli encountered during the study. Among 13 isolates, 9 were from water samples, 4 from surfaces and, surprisingly, 3 from guano samples. TMP is an ef- fective and cheap antimicrobial, commonly used in combination with sulfametoxazole to treat urinary-tract infections (Sköld 2001) with a relatively long half-life (when the maximum concentration is reduced to half maximum concentration) of 20 to 100 days in the envi- ronment (Zuccato et al. 2006; Grenni et al. 2018). TMP is ACTA CARSOLOGICA 49/2-3 – 2020 274 ANTIMICROBIAL RESISTANT ESCHERICHIA COLI FROM KARST W ATERS, SURFACES AND BAT GUANO IN SLOVENIAN CAVES considered an especially important antimicrobial, mainly because TMP-resistant infections can also originate from non-human sources (WHO 2017). CHL is a broad-spectrum antibiotic. Its use in sys- temic therapy has been limited in many countries due to its toxicity, including apparent involvement in fatal oc- currences of granulocytopaenia, aplastic anaemia and thrombocytopaenia (Berendsen et al. 2010; WHO 2017), but is still used mostly in veterinary medicine (Grenni et al. 2018). Whereas CHL-resistant E. coli strains are therefore not expected to be detected frequently, during this study 3 strains were isolated from water and 2 from surfaces (Tabs. 3–4). Multiple drug-resistant strains were relatively com- mon in rivers and on contact surfaces, but were not pres- ent in isolates from either fresh or old bat guano (Tabs. 3–5). AMR-E. COLI IN KARST W ATER SYSTEMS AMR-E. coli strains were isolated from all the karst rivers except Črni potok, a small stream with a limited recharge area, that sinks into the Postojnski jamski sistem (Posto- jna Cave System). Of particular concern is the presence of resistant strains in samples from the Malenščica spring (Tab. 3), an important drinking-water source. Because the spring has an extensive recharge area it is difficult to pinpoint the exact faecal pollution point or points. Prob- able sources of AMR strains include agricultural runoff and wastewater effluents, although sewage treatment practices in the region have improved vastly in recent years. The prevalence of AMR-E. coli in the downstream underground flow is diminishing gradually, along with that of other microbial indicators (Tab. 3; Fig. 1). This can be attributed to a gradual die-off of bacteria along the un- derground course, but dilution with other waters within the karst aquifer cannot be excluded (Mulec et al. 2019). In addition, the successful dissemination of E. coli and other epidemiologically important microbes depends upon environmental factors. Favourable oxygen condi- tions support the viability of E. coli in water but, on the other hand, higher conductivity might have an adverse effect (Tab. 1). Low concentrations of antimicrobials can partly explain the higher prevalance of AMR-strains in waters. AMR-E. COLI ASSOCIATED WITH CONTACT SURFACES AND BATS IN CAVES It seems that cave conditions are not particularly adverse to microbial viability, due both to constant high humid- ity that prevents desiccation, and to the absence of UV irradiation. Tourists are probably the main source of AMR-strains on walking surfaces. They contribute sig- nificantly to the dissemination of resistant strains into and out of the cave and, in this respect, in spreading them to other ecosystems. A general low prevalence of E. coli in Postojnska jama can be attributed to regular cleaning of tourist footpaths and a disinfectant barrier at the tourist cave entrance. To a lesser extent the con- tamination of surfaces with ARB can also be attributed to water seepages. And, conversely, the use of water jets to clean surfaces in tourist caves might enhance the dispersion of ARB in the underground. The example of Postojnska jama demonstrates that those involved in tourist cave management should consider restrict- ing the dispersion of microbial biomass on shoes by in- stalling disinfectant barriers at cave entrances and exits (Appendix 2). The third source of ARB in the karst ecosystem derives from different animals, including bats. Inter- estingly, more AMR-E. coli strains were isolated from surface swabs than from bat guano in Škocjanske jame, even though both sampling sites were in the same part of the cave. As migratory animals, bats can travel reason- able distances and contribute significantly to the spread of minute biological material. For example, in its annual seasonal migrations between winter and summer shel- ters M. schreibersii can fly more than 100 kilometres; the longest recorded distance is 833 km (Dietz & Kiefer 2016). At least temporarily, the current presence of sta- ble bat populations in Škocjanske jame and Predjama indicates the relatively favourable states of both habi- tats (Presetnik et al. 2017). Bats, particularly weakened ones, come into contact with different surfaces in caves, including tourist footpaths. AMR-strains that colonize these surfaces can enter the bat population. The inevita- bly close contacts between individual bats in the colonies ensure rapid spread of pathogens through the popula- tion. Hence, monitoring of bats should include regular follow-up of indicative microbial disease indicators, in- cluding those in fresh guano. Their complete health sta- tus and numerical presence provide perfect indications of environmental disturbances and wildlife conditions (Jones et al. 2009). Whereas it is tempting to assume that bats are mi- nor contributors to the overall ARB and ARG load within the karst ecosystem, additional studies are needed to help find out whether and to what extent E. coli strains from guano are disseminated through the karst. ACTA CARSOLOGICA 49/2-3 – 2020 275 SARA SKOK, BLAŽ KOGOVŠEK, ROK TOMAZIN, SAMO ŠTURM, JERNEJA AMBROŽIČ AVGUŠTIN & JANEZ MULEC CONCLUSIONS The karst ecosystem is extremely vulnerable with respect to microbial contamination from different sources. Be- ing an important drinking water source and habitat for several rare and endangered species, it should be moni- tored for the presence of pathogenic and antimicrobial resistant bacteria, to detect and subsequently reduce or eliminate pollution sources. Shoe-sole disinfection should be introduced at the entrances and exits of all tourist caves. Applied along- side restriction of tourist movements in parts of the cave that have recently been flooded and/or have heavy bat guano accumulations can cut down on the dissemina- tion of bacteria among different habitats and potential hosts. Thus, the possibility of horizontal gene transfer of ARG between pathogenic and environmental bacteria is reduced. More intermicrobial contacts can be made at underground sites with active water flow and significant air streaming, as well as along routes shared by migratory animals and humans. Environmental conditions favouring the aquisition and maintainance of ARG, fitness costs and horizon- tal dispersal barriers, have not yet been studied in de- tail for E. coli. Further characterization of additional E. coli isolates from several sample sites is thus needed to help produce a broader picture on their pathogeneicity, mechanisms of antibiotic resistance, mobile genetic ele- ment content and transfer frequencies to other bacteria in karst habitats. ACKNOWLEDGEMENTS The authors acknowledge financial support from the Slovenian Research Agency (research core funding no. P6-0119 and project “Environmental effects and karst water sources: impacts, vulnerability and adaptation of land use”, no. J6-8266). Public Service Agency Škocjan Caves Park and Postojnska jama are thanked for pro- viding additional financial and logistical support. The project “Development of research infrastructure for the international competitiveness of the Slovenian RRI space – RI-SI-LifeWatch” co-financed by the Republic of Slo- venia, Ministry of Education, Science and Sport and the European Union from the European Regional Develop- ment Fund enabled some extra environmetal analyses. The authors also acknowledge David Lowe for providing language-editing assistance. REFERENCES Amaya, E., Reyes, D., Paniagua, M., Calderon, S., Rashid, M., Colque, P., Kuhn, I., Mollby, R., Weintraub, A. & C. Nord, 2012: Antibiotic resistance patterns of Escherichia coli isolates from different aquatic en- vironmental sources in León, Nicaragua.- Clinical Microbiology and Infection, 18, 9, E347-E354. htt- ps://doi.org/10.1111/j.1469-0691.2012.03930.x ARSO, 2019: Slovenian Environment Agency, Lidar data fishnet.- [Online] Available from: http://gis.arso. gov.si/ [Accessed 9th May 2019]. Balcázar, J.L., 2020: Implications of bacteriophages on the acquisition and spread of antibiotic resistance in the environment.- International Microbiology, 23, 475- 479. https://doi.org/10.1007/s10123-020-00121-5 Berendsen, B., Stolker, L., De Jong, J., Nielen, M., Tseren- dorj, E., Sodnomdarjaa, R., Cannavan, A. & C. Elliott, 2010: Evidence of natural occurrence of the banned antibiotic chloramphenicol in herbs and grass.- Ana- lytical and Bioanalytical Chemistry, 397, 1955-1963. https://doi.org/ 10.1007/s00216-010-3724-6 Calero-Cáceres, W., Ye, M. & J.L. Balcázar, 2019: Bacte- riophages as environmental reservoirs of antibiotic resistance.- Trends in Microbiology, 27, 7, 570-577. https://doi.org/10.1016/j.tim.2019.02.008 Cave Cadastre, 2019: Cave Register of the Karst Research Institute ZRC SAZU and Speleological Association of Slovenia. Postojna, Ljubljana. Chen, Z., Y u, D., He, S., Y e, H., Zhang, L., W en, Y ., Zhang, W ., Shu, L. & S. Chen, 2017: Prevalence of antibiot- ic-resistant Escherichia coli in drinking water sourc- es in Hangzhou City.- Frontiers in Microbiology, 8, 1133. https://doi.org/10.3389/fmicb.2017.01133 Chopra, I. & M. Roberts, 2001: Tetracycline antibiotics: Mode of action, applications, molecular biology, and ACTA CARSOLOGICA 49/2-3 – 2020 276 ANTIMICROBIAL RESISTANT ESCHERICHIA COLI FROM KARST W ATERS, SURFACES AND BAT GUANO IN SLOVENIAN CAVES epidemiology of bacterial resistance.- Microbiology and Molecular Biology Reviews, 65, 2, 232-260. htt- ps://doi.org/10.1128/MMBR.65.2.232-260.2001 Dietz, C. & A. Kiefer, 2016: Bats of Britain and Europe.- Bloomsbury Publishing, pp. 398, London. Emmerson, A. & A. Jones, 2003: The quinolones: dec- ades of development and use.- Journal of Antimi- crobial Chemotherapy, 51, S1, 13-20. https://doi. org/10.1093/jac/dkg208 European Committee on Antimicrobial Susceptibility Testing, 2017: Standard operating procedure: MIC distributions and the setting of epidemiological cut- off (ECOFF) values, EUCAST SOP: 10.0.- [Online] Available from: https://www.eucast.org/fileadmin/ src/media/PDFs/EUCAST_files/EUCAST_SOPs/ EUCAST_SOP_10.0_MIC_distributions_and_ epidemiological_cut-off_value__ECOFF__set- ting_20171117.pdf [Accessed 26th April 2019]. European Centre for Disease Prevention and Control, 2018: Surveillance of antimicrobial resistance in Europe: annual report of the European Antimicro- bial Resistance Surveillance Network (EARS-Net) 2017.- [Online] Available from: https://www.ecdc. europa.eu/sites/default/files/documents/AMR%20 2017_Cover%2BInner-web_v3.pdf [Accessed 26th April 2019]. European Centre for Disease Prevention and Control, 2019: Surveillance of antimicrobial resistance in Europe: annual report of the European Antimicro- bial Resistance Surveillance Network (EARS-Net) 2018.- [Online] Available from: https://www.ecdc. europa.eu/sites/default/files/documents/surveil- lance-antimicrobial-resistance-Europe-2018.pdf [Accessed 20th December 2019]. European Committee on Antimicrobial Susceptibility Testing, 2019: MIC and zone diameter distributions and ECOFFs.- [Online] Available from: http://www. eucast.org/mic_distributions_and_ecoffs/ [Ac- cessed 26th April 2019]. Grenni, P ., Ancona, V . & A. Barra-Caracciolo, 2018: Eco- logical effects of antibiotics on natural ecosystems: A review.- Microchemical Journal, 136, 25-39. htt- ps://doi.org/10.1016/j.microc.2017.02.006 Hooper, D.C. & G.A. Jacoby, 2015: Mechanisms of drug resistance: quinolone resistance.- Annals of the New York Academy of Sciences, 1354, 1, 12-31. https:// doi.org/10.1111/nyas.12830 Huang, C., Renew, J., Smeby, K., Pinkston, K. & D. Sedlak, 2011: Assessment of potential antibiotic contaminants in water and preliminary occur- rence analysis.- Journal of Contemporary Wa- ter Research and Education, 120, 4. https://doi. org/10.1142/9789812799555_0004 Jang, J., Hur, H.G., Sadowsky, M.J., Byappanahalli, M.N., Yan, T. & S. Ishii, 2017: Environmental Escherichia coli: ecology and public health implications-a re- view.- Journal of Applied Microbiology, 123, 3, 570- 581. https://doi.org/ 10.1111/jam.13468 Jones, G., Jacobs, D.S., Kunz, T.H., Willig, M.R. & P.A. Racey, 2009: Carpe noctem: the importance of bats as bioindicators.- Endangered Species Research, 8, 1-2, 93-115. https://doi.org/10.3354/esr00182 Kačaroğlu, F., 1999: Review of groundwater pollution and protection in karst areas.- Water Air and Soil Pollution, 113, 337-356. Laroche, E., Petit, F., Fournier, M. & B. Pawlak, 2010: Transport of antibiotic-resistant Escherichia coli in a public rural karst water supply.- Journal of Hy- drology, 392, 1-2, 12-21. https://doi.org/10.1016/j. jhydrol.2010.07.022 Mulec, J., Petrič, M., Kozelj, A., Brun, C., Batagelj, E., Hladnik, A. & L. Holko, 2019: A multiparameter analysis of environmental gradients related to hy- drological conditions in a binary karst system (underground course of the Pivka River, Slove- nia).- Acta Carsologica, 48, 3, 313-327. https://doi. org/10.3986/ac.v48i3.7145 Mösslacher, F., Griebler, C. & J. Notenboom, 2001: Bio- monitoring of groundwater systems: methods, applications and possible indicators among the groundwater biota.- In: Griebler, C. et al. (eds.) Groundwater ecology: A tool for management of wa- ter resources. Office for Official Publications of the European Communities, pp. 132-170, Luxembourg. Notenboom, J., Plénet, S. & M.J. Turquin, 1994: Ground- water contamination and its impact on groundwater animals and ecosystems.- In: Gibert, J. et al. (eds.) Groundwater ecology. Elsevier, pp. 477-504, London. Odonkor, S.T. & J.K. Ampofo, 2013: Escherichia coli as an indicator of bacteriological quality of water: an overview.- Microbiology Research, 4,1, e2-e2. htt- ps://doi.org/10.4081/mr.2013.e2 Pang, Y., Huang, J., Xi, J., Hu, H. & Y. Zhu, 2016: Effect of ultraviolet irradiation and chlorination on am- picillin-resistant Escherichia coli and its ampicillin resistance gene.- Frontiers of Environmental Sci- ence & Engineering, 10, 3, 522-530. https://doi.org/ 10.1007/s11783-015-0779-9 Petrič, M., 2010: Case study: Characterization, exploita- tion, and protection of the Malenščica karst spring, Slovenia.- In: Krešić, N. & Z. Stevanović (eds.) Groundwater Hydrology of Springs: Engineering, Theory, Management and Sustainability. Elsevier, Butterworth-Heinemann, pp. 428-441, Burling- ton. https://doi.org/10.1016/B978-1-85617-502- 9.00012-8 ACTA CARSOLOGICA 49/2-3 – 2020 277 SARA SKOK, BLAŽ KOGOVŠEK, ROK TOMAZIN, SAMO ŠTURM, JERNEJA AMBROŽIČ AVGUŠTIN & JANEZ MULEC Presetnik, P., Koselj, K., Zagmajster, M., Zupančič, N., Jazbec, K., Žibrat, U., Petrinjak, A. & A. Hudoklin, 2009: Atlas netopirjev (Chiroptera) Slovenije.- Cent- er za kartografijo favne in flore, pp. 152, Miklavž na Dravskem polju. Presetnik, P ., Šturm, S., Debevec, V . & T. Zorman, 2017: Netopirji v Parku Škocjanske jame.- Park Škocjanske jame, pp. 163, Škocjan. Reinthaler, F.F., Posch, J., Feierl, G., Wüst, G., Haas, D., Ruckenbauer, G., Mascher, F . & E. Marth, 2003: An- tibiotic resistance of E. coli in sewage and sludge.- Water Research, 37, 8, 1685-1690. https://doi.org/ 10.1016/S0043-1354(02)00569-9 Robinson, A.A., Belden, J.B. & M.J. Lydy, 2005: Toxic- ity of fluoroquinolone antibiotics to aquatic organ- isms.- Environmental Toxicology and Chemistry, 24, 2, 423-430. https://doi.org/10.1897/04-210R.1 Shao, S., Hu, Y ., Cheng, J. & Y . Chen, 2018: Research pro- gress on distribution, migration, transformation of antibiotics and antibiotic resistance genes (ARGs) in aquatic environment.- Critical Reviews in Bio- technology, 38, 8, 1195-1208. https://doi.org/10.10 80/07388551.2018.1471038 Sköld, O., 2001: Resistance to trimethoprim and sul- fonamides.- Veterinary Research, 32, 3-4, 261-273. https://doi.org/10.1051/vetres:2001123 Speer, B.S., Shoemaker, N.B. & A.A. Salyers, 1992: Bacte- rial resistance to tetracycline - mechanisms, trans- fer, and clinical significance.- Clinical Microbiol- ogy Reviews, 5, 4, 387-399. https://doi.org/10.1128/ cmr.5.4.387 Stange, C. & A. Tiehm, 2020: Occurrence of antibiotic resistance genes and microbial source tracking markers in the water of a karst spring in Germany.- Science of The Total Environment, 742, 10, 140529. https://doi.org/10.1016/j.scitotenv.2020.140529 Van Doorslaer, X., Demeestere, K., Heynderickx, P.M., Van Langenhove, H. & J. Dewulf, 2011: UV-A and UV-C induced photolytic and photocatalytic degra- dation of aqueous ciprofloxacin and moxifloxacin: Reaction kinetics and role of adsorption.- Applied Catalysis B: Environmental, 101, 3-4, 540-547. htt- ps://doi.org/10.1016/j.apcatb.2010.10.027 Vitousek, P.M., Mooney, H.A., Lubchenco, J. & J.M. Melillo, 1997: Human domination of Earth's eco- systems.- Science, 277, 5325, 494-499. https://doi. org/10.1126/science.277.5325.494 World Health Organization, 2017: Critically important antimicrobials for human medicine: ranking of anti- microbial agents for risk management of antimicro- bial resistance due to non-human use.- World Health Organization , 5 th revision. Licence: CC BY-NC-SA 3.0 IGO. Zhang, P ., Xu, P ., Xia, Z., Wang, J., Xiong, J. & Y . Li, 2013: Combined treatment with the antibiotics kanamy- cin and streptomycin promotes the conjugation of Escherichia coli.- FEMS Microbiology Letters, 348, 2, 149-156. https://doi.org/10.1111/1574-6968.12282 Zhao, Y., Ye, L. & X. Zhang, 2018: Emerging Pollutants– Part I: Occurrence, Fate and Transport.- Water En- vironment Research, 90, 10, 1301-1322. https://doi. org/10.2175/106143018X15289915807236 Zuccato, E., Castiglioni, S., Fanelli, R., Reitano, G., Bag- nati, R., Chiabrando, C., Pomati, F., Rossetti, C. & D. Calamari, 2006: Pharmaceuticals in the envi- ronment in Italy: causes, occurrence, effects and control.- Environmental Science and Pollution Research, 13, 1, 15-21. http://dx.doi.org/10.1065/ espr2006.01.004 ACTA CARSOLOGICA 49/2-3 – 2020 278 ANTIMICROBIAL RESISTANT ESCHERICHIA COLI FROM KARST W ATERS, SURFACES AND BAT GUANO IN SLOVENIAN CAVES Appendix 1: Collection of fresh guano on a sterile aluminium foil placed under colonies of roosting bats along the tourist footpath in Škocjanske jame, Rudolfova dvorana (Photo: S. Šturm). Appendix 2: Disinfectant barrier at the entrance to Postojnska jama (Photo: J. Mulec). APPENDIX ACTA CARSOLOGICA 49/2-3 – 2020 279