Occurrence of Enterococcus species with virulence markers in an urban flow-influenced tropical recreational beach

Marine Pollution Bulletin 82 (2014) 26–38

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Occurrence of Enterococcus species with virulence markers in an urban flow-influenced tropical recreational beach Asmat Ahmad a,⇑, Ayokunle Christopher Dada a,b,⇑, Gires Usup c, Lee Yook Heng d a

School of Biosciences and Biotechnology, Faculty of Science & Technology, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Malaysia Institute of Ecology and Environmental Studies, Obafemi Awlowo University, Ile-Ife, Nigeria c School of Environmental & Natural Resource Sciences, Faculty of Science & Technology, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Malaysia d School of Chemical Sciences and Food Technology, Faculty of Science & Technology, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Malaysia b

a r t i c l e

i n f o

Keywords: Antibiotic resistance Beach water Enterococcus Sequence type, Virulence factors

a b s t r a c t Median enterococci counts of beach water samples gradually increased at statistically significant levels (v2: 26.53, df: 4; p < 0.0001) with increasing proximity to river influx. The difference in proportion of antibiotic resistant enterococci in beach water and river water samples was statistically significant (p < 0.05) for the tested antibiotics with river isolates generally presenting higher resistance frequencies. Virulence genes cyl, esp, gelE and asa were detected at varying frequencies (7.32%, 21.95%, 100% and 63.41% respectively) among river isolates. On the other hand, the prevalence of these genes was lower (0%, 20%, 67.27% and 41.82% respectively) among beach water isolates. Multi-Locus-Sequence-Typing analysis of Enterococcus faecalis presented four sequence types (ST) one of which shared six out of seven tested loci with ST6, a member of the clonal complex of multi-drug resistant strains associated with hospital outbreaks. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction The impact of external influence on surface water quality is well documented (Bhaduri et al., 2001; Schippmann et al., 2013; Mallin et al., 2000). Anthropogenic influences (urbanization, industrial and agricultural activities) as well as complex natural processes (changes in precipitation inputs, erosion, and weathering of crustal materials) are all implicated in the degradation of surface water quality (Zhang et al., 2009). These characteristically impair their use for recreation or other purposes (Peng et al., 2005). Surface water contamination in tropical and subtropical environments has been associated with waterborne diseases (Peng et al., 2005). Specifically, rivers in Malaysia have been reportedly treated as open sewers with attendant waste discharges which ultimately reach the sea (DailyExpress: Rubbish Ending Up in the sea, 2013). Pollutants from municipally-influenced sources may carry diverse bacteria which could directly pose threats to the health of recreational beach users (Barrell et al., 2000; Hamilton et al., 2010). Enterococci are commonly used as indicators of faecal contamination in recreational waters (Barrell et al., 2000). Studies ⇑ Corresponding authors. Address: School of Biosciences and Biotechnology, Faculty of Science & Technology, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Malaysia. Tel.: +60 149390730. E-mail addresses: [email protected] (A. Ahmad), [email protected] (A.C. Dada). http://dx.doi.org/10.1016/j.marpolbul.2014.03.028 0025-326X/Ó 2014 Elsevier Ltd. All rights reserved.

that highlight enterococci diversity may play critical epidemiological roles in source pollution monitoring of surface waters. To date, there is no published study on the genetic variability of enterococci recovered from recreational beach water in Malaysia. Enterococcus spp. are documented to be intrinsically resistant to a number of antibiotics including cephalosporins, penicillinaseresistant penicillins, and clinically available concentrations of lincosamides and aminoglycosides (Fisher and Phillips, 2009). Apart from the notoriety of members of this genus for antibiotic resistance, enterococci may possess a number of virulence factors which are associated with the severity and duration of infections caused by them. These include gelatinase, enterococcal surface protein (Esp), aggregation substance (asa), cytolysin (cyl) and hyaluronidase (hyl) Jett et al., 1994; Vergis et al., 2002; Semedo et al., 2003. While the occurrence of virulence strains has been extensively studied among clinical enterococci (Padilla and Lobos, 2013), studies are only recently beginning to emerge on the elucidation of virulence genes among enterococci recoverable from recreational beach waters is (Brownell et al., 2007; Santiago-Rodriguez et al., 2013; Layton et al., 2009; Rathnayake et al., 2012). The scanty published information partly explains why the ecology of antibiotic resistance among environmental strains of enterococci is still not well understood (Santiago-Rodriguez et al., 2013; Rathnayake et al., 2012). Furthermore, there is a dearth of information on the molecular

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38

determination of virulence markers among enterococci recoverable from tropical recreational beach waters (Santiago-Rodriguez et al., 2013). The aim of the current study is, thus, to determine via molecular characterization, the species prevalence, diversity, antimicrobial resistance and virulence markers among enterococci recovered from a tropical recreational beach in the East Coast of Malaysia. 2. Materials and methods 2.1. Study area description The coast of northern Pahang, Malaysia is composed of the three prominent bays (Cendur, Cerating and Beserah) and the minor bays of Pelindung and Chempedak (Tiong, 2001). Teluk Chempedak is Kuantan’s most popular beach located 5 km east from Kuantan town centre, Pahang, Malaysia. Teluk Chempedak is usually crowded during the weekends with people engaging in a number of sporting activities. Along the Chempedak bay, River Chempedak drains into the sea just overhead the main bathing area at Teluk Chempedak. There is also a stormwater drainage system that empties into this river at the brink of influx into the sea (Fig. 1). 2.2. Sample collection Bathing water samples for both beaches were taken from areas close to the swash zone of the bathing beach at locations that cover the length of the beach (Fig. 1). Water samples were collected also from the river draining into the sea at various locations and at the area of influx into the seawater. Sterile glass bottles (1000 ml) were used to collect water samples in triplicates. Sand samples were collected in sterile plastic containers. Faecal samples were collected at toilets proximate to the beach. 2.3. Isolation and enumeration of enterococci Bacterial densities of enterococci from seawater samples were determined by membrane filtration method (A.P.H.A, 1999), using Slanetz and Bartley (S + B) culture media. Plates were incubated at 37 °C for 24–48 h. Preliminary tests on presumptive enterococci were performed as previously described (Dada et al., 2013). All isolates were designated as members of the genus Enterococcus by the PCR detection of the tuf gene as described by Creti et al. (2004). Isolates were also confirmed by the amplification of sodA and ddl genes (Enterococcus faecalis and Enterococcus faecium) (Jackson et al., 2004; Dutka-Malen et al., 1995) and 16SrRNA sequencing (for identification of other Enterococcus spp.). Primers are listed in Table 1. 2.4. Multi Locus Sequence Typing Multi-drug resistant E. faecalis isolates with virulence traits were selected for Multi Locus Sequence Typing (MLST). MLST analysis involved a total of seven housekeeping genes: gdh (glucose-6-phosphate dehydrogenase), gyd (glyceraldehyde-3-phosphate dehydrogenase), pstS (phosphate ATP binding cassette transporter), gki (putative glucokinase), aroE (shikimate 5-dehydrogenase), xpt (shikimate 5-dehydrogenase), and yiqL (acetyl-coenzyme A acetyltransferase) (Table 1) (Ruiz-Garbajosa et al., 2006). PCR conditions for all amplification reactions were as follows: initial denaturation at 94 °C for 5 min; 30 cycles at 94 °C for 30 s, 52 °C for 30 s, and 72 °C for 1 min; and final extension at 72 °C for 7 min. Reactions were performed in 25-ll volumes using Taq polymerase (FirstBase Sdn Bhd). The resulting PCR amplicons were purified with a kit (Qiagen) and sequenced using an ABI 3130XL 20

27

genetic analyzer (Applied Biosystems). For each locus, a distinct allele number was assigned to the obtained nucleotide sequences in accordance with the E. faecalis MLST database (http://efaecalis.mlst.net/). 2.5. Nucleotide sequence accession numbers Nucleotide sequences determined in this study were submitted to GenBank (http://www.ncbi.nlm.nih.gov/genbank/). The deposited information was simultaneously made available to EMBL in Europe and the DNA Data Bank of Japan. sodA gene sequences were submitted under GenBank accession numbers KC603859KC603862; ddl gene sequences were submitted under GenBank accession numbers KC594678-KC594682; 16S rRNA gene sequences were submitted under GenBank accession numbers KC890838-KC890842, KC707577-707586. 2.6. Detection of antibiotics resistance and resistance determinants among Enterococci Antibiotic resistance among enterococci isolated from the beaches was determined by the single disc Kirby-Bauer diffusion method using Mueller–Hinton Agar (Oxoid, UK) NCCLS: Performance Standards for Antimicrobial Disk Susceptibility Testing, 2003. A total of 96 enterococci isolates recovered from Teluk Chempedak beach water and beach sand, River Chempedak water and sand were tested for resistance to seven (7) antibiotic groups. Antibiotics tested include vancomycin (V) (glycopeptides) (30 lg), kanamycin (K) (30 lg), streptomycin (S) (25 lg) (aminoglycoside), ampicillin (A) (10 lg) (B-lactam), tetracycline (T) (30 lg) (tetracycline), chloramphenicol (C) (30 lg) (chloramphenicol) and nitrofuranoin (N) (50 lg) (nitrofuran) (Oxoid, UK). Diameters of zones of inhibition were recorded in mm and interpreted as sensitive or resistant using breakpoints for enterococci as proposed by the National Committee for Clinical Laboratory Standards (NCCLS: Performance Standards for Antimicrobial Disk Susceptibility Testing, 2003). 2.7. Determination of virulence-markers distribution in enterococci Polymerase chain reaction (PCR) assays were applied to amplify virulence determinants {aggregation substance (asa), cytolysin (cylA), enterococcal surface protein (esp), gelatinase (gelE)} (Vankerckhoven et al., 2004). Details of primers are listed in Table 2. For each primer, initial optimization experiments were conducted to ascertain optimal pcr conditions for MgCl2 and annealing temperatures. Multiplex PCR conditions used in this study included an initial activation step at 95 °C for 4 min followed by 30 cycles of denaturation at 94 °C for 1 min, annealing at 55 °C for 1 min, extension at 72 °C for 1 min and a single cycle of 7 min at 72 °C. Reference strains were gratefully provided by Prof. Shankar of the Department of Medicinal Chemistry and Pharmaceutics, University of Oklahoma and Dr. Fatimah Lopez, Institute of Bioscience, Brazil. PCR amplicons of asa, cylA, esp, gelE genes were confirmed by DNA sequencing with an ABI 3130XL 20 genetic analyzer (Applied Biosystems). The DNA sequences were blasted for sequence similarity with annotated sequences at http:// www.ncbi.nlm.nih.gov. 2.8. Phenotypic assays All isolates were subjected to gelatinase assay as described by Cariolato et al. (2008) using Brain–Heart Infusion agar plates supplemented with 10 g/L peptone and 30 g/L gelatine. Assay for proteolytic activity was conducted by detecting casein hydrolysis in MHA containing 3% (w/v) skimmed milk. Biofilm formation was determined by crystal violet assay in polystyrene microtiter

28

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38

Fig. 1. (a) Map of sampled area (b) Teluk Chempedak showing the bathing beach and the combined loadings of river water and stormwater discharge.

plates and subsequent examination at an optical density of 570 nm (OD570) (Mohamed et al., 2004). 2.9. Statistical analysis StrataÒ, gratefully provided by the Harvard School of Public Health in collaboration with Strata Inc. (USA), was used for the analyses. Each of the responses from the phenotypic and genotypic tests were fed into the software program as conscripted variables in a binary coding system. These were thereafter used

to generate two-way tables with measures of association for all variables considered. In each case, an enumeration of the within-row, within-column frequencies and Fisher’s exact analysis were performed. In some instances, Spearman’s rank of correlation was used for hypothesis testing using unadjusted significant levels. Using this test, all possible pair-wise correlation coefficients were calculated and the results presented in a matrix table output. Wilcoxon’s tests were conducted using GraphPad Prism. For all analyses, statistical significance was inferred when P values were less than 0.05.

29

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38 Table 1 Concentration of enterococci recovered from Kuantan beach (n = 15) and urban flow influenced riverwater samples (n = 15). Site

BW BWNTR BS RC RS

p-Valuea

Sterile distilled water CFU/100 ml water

Beach water

No prior rainfall event before sampling

Prior rainfall event before sampling

N.D N.D N.D N.D N.D

125 (120–245) 110 (90–240) 125 (95–160) 100 (101–155) 70 (60–75)

1340 (980–1560) 680 (1160–1500) 760 (380–880) 680 (540–880) 640 (620–700)

Median CFU/100 ml water (Range)

p-Valueb

Riverwater Median CFU/100 ml water (Range)

<0.0001***

No prior rainfall event before sampling

Prior rainfall event before sampling

2.55  105 (1.55  105–3.2  105) 1.2  105 (1.05  105–3.05  105) 2.55  105 (1.5  105–2.3  105) 2.55  105 (1.75  105–3.1  105) 2.55  105 (1.05  105–3.55  105)

3.4  107 5.0  107 3.4  107 5.4  107 7.2  107

(2.8  107–4.2  107) (4.8  107–6.2  107) (3.4  107–5.8  107) (5.2  107–6.2  107) (5.0  107–9.0  107)

0.0079***

BW-Beach water (bathing area), BWNTR-point where River Chempedak empties into recreational beach water, BS-Beach soil, RC-River Chempedak water, RS-River Chempedak soil. *** Significant at p < 0.01.

Table 2 Primers used during the study.

Molecular characterization

Gene

Primer name

Sequence (50 –30 )

Product

Superoxide dismutase gene (sodA)

FL1 FL2 Ddlfm1

ACTTATGTGACTAACTTAACC TAATGGTGAATCTTGGTTTGG GCAAGGCTTCTTAGAGA

360

Ddlfm2 Ddlfl1 Ddlfl2 B27F U1492R gdh-1 gdh-2 gyd-1 gyd-2 pstS-1 pstS-2 gki-1 gki-2 aroE-1 aroE-2 xpt-1 xpt-2 yiqL-1 yiqL-2 ASA 11 ASA 12 GEL 11 GEL 12 CYT I CYT IIb ESP 14F ESP 12R

CATCGTGTAAGCTAACTTC ATCAAGTACAGTTAGTCTT ACGATTCAAAGCTAACTG AGA GTT TGATCC TGG CTC AG GGT TAC CTT GTT ACG ACT T GGCGCACTAAAAGATATGGT CCAAGATTGGGCAACTTCGTCCCA CAAACTGCTTAGCTCCAATGGC CATTTCGTTGTCATACCAAGC CGGAACAGGACTTTCGC ATTTACATCACGTTCTACTTGC GATTTTGTGGGAATTGGTATGG ACCATTAAAGCAAAATGATCGC TGGAAAACTTTACGGAGACAGC GTCCTGTCCATTGTTCAAAAGC AAAATGATGGCCGTGTATTAGG AACGTCACCGTTCCTTCACTTA CAGCTTAAGTCAAGTAAGTGCCG GAATATCCCTTCTGCTTGTGCT GCACGCTATTACGAACTATGA TAAGAAAGAACATCACCACGA TATGACAATGCTTTTTGGGAT AGATGCACCCGAAATAATATA ACTCGGGGATTGATAGGC GCTGCTAAAGCTGCGCTT AGATTTCATCTTTGATTCTTGG AATTGATTCTTTAGCATCTGG

D-alanine:D-lactate

and D-alanine:D-alanine ligase (ddl)

16S rRNA Multi Locus Sequence Typing

Glucose-6-phosphate dehydrogenase Glyceraldehydes-3-phosphate dehydrogenase Phosphate ATP binding cassette transporter Glucokinase Dhikimate-5-dehydrogenase Xanthine phosphoribosyltransferase Acetyl-CoA acetyltransferase

Virulence markers

Aggregation substance (asa+) Gelatinase (gel+) Cytolysin (cyl+) Enterococcal surface protein (esp+)

3. Results and discussion 3.1. Distribution of enterococci Enterococci counts obtained from the water quality analysis of samples collected from the beach water and an urban-flow-influenced river is presented in Table 2. Four main hypotheses tested in the study were rejected following statistical analysis based on the non-parametric Wilcoxon–Mann–Whitney tests. Statistical analysis was conducted to ascertain if the distribution of enterococci (CFU/100 mL) in (i) the absence of rainfall was the same across the various sampling locations, (ii) the absence of rainfall was the same in the beach water and river water samples, (iii) samples collected following rainfall event was the same across the various sampling locations and (iv) samples collected following rainfall event did not differ from those collected in the absence of pre-sampling rainfall event.

550 941 1300 530 395 583 438 459 456 436 375 213 688 510

Enterococci counts in beach water were generally lower than was observed for those observed for samples collected from the river (p < 0.0079) (Table 2). The enterococci counts obtained for the beach water samples following rainfall event were significantly higher than the ones collected without prior rainfall events. With or without prior rainfall events, the median enterococci counts observed in beach water samples gradually increased at statistically significant levels (v2: 26.53, df: 4; p < 0.0001) with increasing proximity to the river influx. The observation of gradient wise increase is consistent with previous reports on faecal indicators across surface water gradients (Lata et al., 2009; Sapkota et al., 2007). This increase was also consistent with the outcome observed for the Wilcoxon’s matched pair analysis for each of the locations tested, implicating the urban flow influenced river water that drains into the recreational beach as being responsible for elevated loads of enterococci in the beach. The pollution influx is further heightened particularly during and after rainfall events.

30

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38

This possibly explains the 2log and 1log increase in enterococci densities observed in river and beach water respectively when samples were collected following a prior rainfall event. Given the higher counts associated with the river water following discharge of stormwater though the stormwater conveyance system (Fig. 1), there is considerable evidence that the stormwater conveyance system in addition to the river flow may also be acting as a reservoir, or breeding basin for indicator bacteria. During rainfall events, high bacterial loads in urban influenced storm water are thus flushed through the system. This resuspends bacteria living in the sediments of the stormwater system further elevating the load discharged to the beach. Without the influence of rainfall, enterococci counts of the considered bathing beach water were still found to be above the decisive action level of 104 enterococci per 100 ml in salt water as recommended by USEPA (Bacterial Water Quality Standards for Recreational Waters (Freshwater and Waters), 2003). The values were also greater than the 640 per 100 mL water (95th percentile) benchmark stipulated by the WHO (2003) for intestinal enterococci in category B water. The counts obtained for the river water samples too were above the international single-sample advisory limit of 61 CFU/100 mL for enterococci in fresh water (Bacterial Water Quality Standards for Recreational Waters (Freshwater and Waters), 2003). There is no indicator bacteria standard for bathing water in Malaysia (Hamzah et al., 2011). 3.2. Species diversity of enterococci isolates There was also a marked difference in the diversity of enterococci species recovered from beach sand (BS) and beach water (BW). Among the beach water isolates, Enterococcus casseflavus and E. faecalis were detected at the highest proportions (47.37% and 42.11% respectively) while the least occurring species were Enterococcus hirae and E. faecium (both 5.26%). On the other hand, beach sand had a different species diversity profile, E. hirae being the most predominant (65.39%) followed by E. faecalis (26.92%). Notably, a high proportion of Enterococcus gallinarum was detected in the river water. While it was the most predominant species in the population tested (45.45%), E. gallinarum was not detectable in beach water and beach sand samples (Table 3). E. hirae and E. faecalis isolates were also present at varying proportions (27.27% and 22.72%, respectively) among the tested population of enterococci recovered from river chempedak water samples. Analysis of 2  2 contingency tables of E. faecalis and E. faecium frequencies in beach water and beach sand showed no association in the occurrence of the two species (p = 0.212). Chi square analysis of the diversity data however revealed significant differences (X2 = 70.86, df = 16, p < 0.0001) in the frequencies of occurrence of enterococci diversity from the various sites sampled during the study. The diversity of enterococci species recovered was thus significantly associated with the sampling sites.

In the current study, E. faecalis and E. casseliflavus dominated beach water isolates. In a similar study conducted at a tropical location, the dominant enterococci were identified as E. casseliflavus in both river and beach samples regardless of the season or location suggesting their adaptation to this environment when compared to the other enterococci. Also, among the non-pigmented, E. faecalis was the dominant enterococci (Hernández, 2011). This observation is in concert with our findings. Preliminary information on the beach where samples were collected for enterococci isolation revealed the presence of a zoo near to the beach. Previous studies have highlighted E. hirae as commonly associated with the intestinal flora of several animals (Kolbjørnsen et al., 2011; Devriese et al., 1994). This probably explains the predominance of E. hirae in beach sand isolates. A number of studies have attempted tracing movements of sediments in the coastal zone. In a pioneering study by Trask (1952) on the investigation of the movement of beach sand along the southern California coast, it was demonstrated that sand at the considered beach may have come from a distance of more than 160 km up the coast. More recently, variations in the degree of grain rounding have been used to trace sand movements, or to obtain additional information concerning the history of the sediment particles (Allan and Hart, 2007). The markedly differing characteristics observed among beach water and beach sand isolates in this study may be useful for application in microbial source tracking studies that aim to decipher if faecal enterococci loadings in adjacent waters are attributable to beach soil. 3.3. Antibiotic resistance among encountered enterococci Two-tailed t-test analysis revealed differences in the levels of antibiotic resistance demonstrated by isolates recovered from the beach area and the river area. In this case, beach area was taken as the combination of isolates from beach water (BW), beach soil (BS) and the river influx region on the beach (BWNTR). On the other hand, river area was defined as the combination of river water (RC) and river soil (RS) isolates. The differences in proportion of antibiotic resistance were statistically significant (p < 0.05) for most of the antibiotics tested with the river area isolates generally presenting higher resistance proportions (Table 4). The level of antibiotic resistance was highest for kanamycin (90.2% and 85.45% for river area and beach areas, respectively). However, none of these isolates were found to be resistant to high levels of aminoglycoside (500 mcg/ml and 2000 mcg/ml). Tetracycline and chloramphenicol resistance levels were at 44.79% and 37.59% respectively. Fisher’s exact test revealed significant associations at p-values less than 0.05 for ampicillin and vancomycin resistance. Nitrofurantoin resistance among the pool of tested enterococci (n = 96) was found to be associated significantly with resistance to chloramphenicol. Similar associations were also observable for chloramphenicol and streptomycin, kanamycin and

Table 3 Frequency of the distribution of Enterococcus species diversity among the five sampled sites. Species

E. casseflavus E. hirae E. gallinarum E. faecium E. faecalis Total Enterococci per site

Number of isolates (%) from each isolation site (n = 5) BW

BWNTR

BS

RC

RS

9 (9.38) 1 (1.04) 0 1 (1.04) 8 (8.33) 19 (19.79)

2 (2.08) 1 (1.04) 0 0 7 (7.29) 10 (10.42)

2 (2.08) 18 (18.75) 0 0 6 (6.25) 26 (27.08)

1 (1.0 4) 6 (6.25) 10 (10.42) 0 5 (5.21) 22(22.92)

0 5 (5.21) 8(8.33) 0 6(6.25) 19 (19.79)

Enterococci spp distribution

p-Value

14 (14.58) 31 (32.29) 18 (18.75) 1 (1.04) 32 (33.33) 96 (100)

0.0001

Pearson chi2 (v2) = 70.8603, df = 16, Statistically significant at alpha <0.05. BW-Beach water (bathing area), BWNTR- point where River Chempedak empties into recreational beach water, BS-Beach soil, RC-River Chempedak water, RS-River Chempedak soil.

31

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38 Table 4 Differences in antibiotic resistance and virulence factor carriage among beach water and river water emptying into the recreational beach water. River area (n = 41) Resistance to antibiotics

Possession of virulence factors

a,b,c,d

Beach area (n = 55) a

Two-tailed t-value

p-Value

V T C A K N S

8 (19.51) 28 (68.29)a,e,f 27 (65.85) b,g,h 10 (24.39)e,g,I,j 37 (90.24) c,I,k,l 11 (26.83) f,h,k 21 (51.22) d,j,l

15(27.27) 15(27.27)b 9(16.36)c 14(25.45)d 47(85.45)a,b,c,d,e,f 4(7.27)e 4(7.27)f

0.8902 4.3032 5.4797 0.1176 0.7138 2.4923 5.0758

0.3757 <0.0001 <0.0001 0.9067 0.4771 0.0155 <0.0001

cyl+ esp+ gel+ asa+

3 (7.32) 9 (21.95) 41 (100.0) 26 (63.41)

0 (0.00) 11(20.00) 37(67.27) 23(41.82)

N.D 0.2291 ND 2.1266

N.D 0.8194 ND 0.0363

*** ***

* ***

*

ND-not determined because one of the variables compared had standard deviations equals to zero. V, Vancomycin (glycopeptides), K, kanamycin, S-Streptomycin (aminoglycoside), A-ampicillin (B-lactam), T, tetracycline (tetracycline), C-chloramphenicol (chloramphenicol), N, nitrofurantoin (nitrofuran). Same letters within column are statistically significant based on Turkey Crammers multiple mean comparison test. * Significant at p < 0.05 *** Significant at p < 0.01

nitrofurantoin, and between kanamycin and streptomycin resistance levels. Nitrofurantoin is a common drug of choice for the treatment of urinary tract infections. However, the rate at which pathogens are developing resistance to the drug is exerting its toll on the reduction of the usefulness of this drug in the treatment of uncomplicated UTIs. While resistance to nitrofurantoin was low for beach isolates (7.27%), a significantly higher level of resistance (26.83%) to this drug was observed among river area isolates. Chloramphenicol resistance was not detected among beach water isolates although a high proportion of RC isolates (72.73%) were resistant to the antibiotics. Previous studies on the prevalence of antibiotic resistance among environmental strains of enterococci in Malaysia are few. Notwithstanding, Toosa et al. (2001) documented high chloramphenicol resistance levels (61%) among poultry enterococci isolates in Malaysia. In addition to testing for high level gentamicin resistance at 500 lg/ml and 2000 lg/ml concentrations, streptomycin resistance was also evaluated since enterococcal resistance to gentamicin and streptomycin occur by different mechanisms. Although gentamicin resistance may be a good predictor of resistance to a number of aminoglycosides, this may not be the case with streptomycin (Cetinkaya et al., 2000). Streptomycin-resistant enterococci, particularly due to the production of streptomycin adenyltransferase, are susceptible to gentamicin. Two types of streptomycin resistance are observable among enterococci. The first type is generally limited to moderate-level resistance to 62–500 lg/ml concentration of the drug simply because of low permeability and drug uptake. On the other hand, the second type of resistance is usually either ribosomally mediated or as a result of the production of aminoglycoside-inactivating enzymes that allow such strains to demonstrate high-level resistance at concentrations at or higher than 2000 lg/ml (Cetinkaya et al., 2000). On the other hand, gentamicin resistance is predominantly the result of the presence of the inactivating enzyme 200 -phosphotransferase-60 acetyltransferase. Although all strains in our study were found to be susceptible to gentamicin, 7.27% and 51.22% of enterococci strains recovered from beach area and river area respectively were resistant to streptomycin. Of all tested beach area isolates (n = 45), 27.27% were found to be vancomycin resistant. This proportion was however not significantly different (p = 0.3757) from the proportion of river area isolates (19.51%) that were resistant to the drug. Apart from a few studies that have reported on the occurrence of VRE in samples from clinical patients, chickens, pigs and poultries in Malaysia (Son et al., 1999; Raja et al., 2005; Getachew et al., 2013), there is a dearth of published information on the prevalence of VRE in

coastal beach water used for recreational purposes. The occurrence of VRE in the environment presents threats to public health. Besides the commonly reported nosocomial acquisition and subsequent colonization of vancomycin-resistant enterococci (VRE) in hospital settings, colonization among non-hospitalized persons and individuals outside the health-care setting have been documented to occur frequently (McDonald et al., 1997). There is thus the need for further studies on the prevalence and community transmission of VRE in Malaysia. If transmission with VRE from hitherto unrecognized community sources could be identified and controlled, increased incidence of colonization and infection both in the community and among hospitalized patients may be prevented. Notably, there was an extremely significant difference (p < 0.0001) in the observable levels of tetracycline resistance with the river isolates presenting higher levels (68.29%) as compared to the beach area isolates. Cheong et al. (1994), in a study on antimicrobial resistance among some hospital enterococci isolates reported a prevalence level of 60.6% for tetracycline resistance. There is no published information on the prevalence of tetracycline resistance among enterococci recovered from recreational beaches on Malaysia apart from a recent report (Dada et al., 2013) that reported a prevalence level of 3.3% at a separate beach. Consistent with our study, the occurrence of tetracycline resistant beach enterococci was also reported by Rathnayake et al. (2012). The levels of ampicillin resistance recorded in this study (24.39% and 25.45% respectively for the river and beach area) were much lower than was previously reported (Dada et al., 2013). In another study on public beaches (Turgeon et al., 2012), tetracycline and ampicillin were among the three antimicrobials with the highest frequency of resistance. Pronounced resistance to ampicillin has also been reported by previous studies and they appear common in samples from untreated sewage or in isolates from the human and animal intestines (Silva et al., 2006; Adria et al., 2011). Several enterococci isolates were observed with the ability to grow on oxacillin screen agar. However, based on PCR detection, all of these isolates were found to be devoid of mecA gene carriage associated with MRSA that are resistant to oxacillin antibiotics in the agar. Lata et al. (2009) also highlighted enterococci with ability to grow on oxacillin agar recovered from a polluted river. mecA gene codes for a protein with a low affinity to penicillin (PBP2a) and confers methicillin resistance is located on a mobile genomic element, the staphylococcal cassette chromosome (SCCmec). Although it was previously reported only among Staphylococcus genus from clinics, some studies (Turgeon et al., 2012; Tsubakishita et al., 2010) have also reported the carriage of mecA

32

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38

gene in non-staphylococcal isolates recovered from surface waters and other environmental sources. Acquisition by transduction of heterologous genes, particularly of antibiotic resistance genes, might represent an important mechanism of horizontal gene transfer in water bodies (de Vera and Simmons, 1996). 3.4. Occurrence of virulence determinants among tested bacteria Biofilm formation was highest among beach soil isolates (61.54%) followed by beach water (31.58%) and river water (22.73%). Biofilm production was significantly associated with the carriage of either asa or gelE genes. Conversely, fisher’s exact test revealed significant associations between caseinase and carriage of either esp or gelE gene. Virulence genes cyl, esp, gelE and asa were detected in 7.32%, 21.95%, 100% and 63.41% respectively among river isolates but at lower proportions of 0%, 20%, 67.27% and 41.82% respectively among beach water isolates. Notably, cyl positive isolates were detected only among river area isolates and not among beach water (BW) isolates thus implicating urban

flow as potential source of cyl carrying enterococci influx into recreational bathing water. This is the first study that reports this observation. Some beach water isolates (57.89%) also possessed the esp gene which was notably absent among all beach soil isolates. Unlike the beach water isolates, only the river isolates were found to carry the esp gene (22.73%). This observation also presents evidence to implicate the river emptying into the beach water along with other faecal sources as responsible for the preponderance of esp gene carrying enterococci in beach water. In developing nations, urban rivers often serve as sewers for communities who simply discharge waste into the flowing water body. Some of these may be rich in faecal contamination and may explain for the esp strains that eventually reach the sea. In terms of comparative levels of prevalence, asa gene was detectable in all faecal isolates, followed by the beach water (68.42%), river water (63.64%) and least among beach soil isolates (11.54%). GelE gene carriage was always present in all isolates recovered from BW (beach water), BWNTR (beach water at river influx region) and RC (river water). However, only 30.77% of beach

Table 5 Correlation observed between the enterococcus species isolation source and prevalence of resistance to single/multiple classes of antibiotics. S/ no

Combination of single/multiple antimicrobial resistance (groups/class)

No. of enterococci

(%)

Spearman correlation (rs)

p-Value

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Va T C A K N S Va–K K–S A–K T–K C–K T–C Va–T T–A C–N C–K–S T–K–S Va–C–A–K Va–C–A–K–S T–K–N T–A–K T–C–K Va–T–C–A–K T–C–A–K C–A–K–N C–A–K–N–S Va–T–A–N A–K–N Va–C–A T–C–A T–A–K–N–S T–C–A–K–S T–A–N–S–E C–A–N A–K–N–S Va–T–C–A–K– S Va–T–C–K–S Va–T–C–A–K– N–S T–C–K–S Va–T–C–K–N– S T–C–K–N–S

Glycopeptide Tetracycline Chloramphenicol B-lactam Aminoglycoside Nitrofuran Aminoglycoside Glycopeptide–aminoglycoside Aminoglycoside B-lactam–aminoglycoside Aetracycline–aminoglycoside Chloramphenicol–aminoglycoside Tetracycline–chloramphenicol Glycopeptide–tetracycline Tetracycline–B-lactam Chloramphenicol–nitrofuran Chloramphenicol–aminoglycoside Tetracycline–aminoglycoside Glycopeptide–chloramphenicol–B-lactam–aminoglycoside Glycopeptide–chloramphenicol–B-lactam–aminoglycoside Tetracycline–aminoglycoside–nitrofuran Tetracycline–B-lactam–aminoglycoside Tetracycline–chloramphenicol–aminoglycoside Glycopeptide–tetracycline–chloramphenicol–B-lactam–aminoglycoside Tetracycline-chloramphenicol–B-lactam–aminoglycoside Chloramphenicol–B-lactam–aminoglycoside–nitrofuran Chloramphenicol–B-lactam–aminoglycoside–nitrofuran Glycopeptide–tetracycline–B-lactam–nitrofuran B-lactam–aminoglycoside–nitrofuran Glycopeptide–chloramphenicol–B-lactam Tetracycline–chloramphenicol–B-lactam Tetracycline–B-lactam–aminoglycoside–nitrofuran Tetracycline–chloramphenicol–B-lactam–aminoglycoside Tetracycline–B-lactam–nitrofuran–aminoglycoside Chloramphenicol–B-lactam–nitrofuran B-lactam–aminoglycoside–nitrofuran–aminoglycoside Glycopeptide–tetracycline–chloramphenicol–B-lactam–aminoglycoside

23 43 36 24 84 15 25 20 24 18 38 32 23 11 13 9 19 17 6 4 8 10 21 4 7 5 3 2 7 7 9 2 4 3 5 4 3

(23.95833) (44.79167) (37.5) (25) (87.5) (15.625) (26.04167) (20.83333) (25) (18.75) (39.58333) (33.33333) (23.95833) (11.45833) (13.54167) (9.375) (19.79167) (17.70833) (6.25) (4.166667) (8.333333) (10.41667) (21.875) 4.166667) (7.291667) (5.208333) (3.125) (2.083333) (7.291667) (7.291667) (9.375) (2.083333) (4.166667) (3.125) (5.208333) (4.166667) (3.125)

0.08994 0.408 0.5057 0.01216 0.07163 0.2664 0.4953 0.02809 0.4742 0.01686 0.4207 0.4616 0.5021 0.2183 0.08911 0.3003 0.4168 0.4821 0.03806 0.1361 0.273 0.1192 0.511 0.1361 0.1628 0.1767 0.08699 0.0215 0.1628 0.0008437 0.3003 0.1689 0.1361 0.208 0.1767 0.1361 0.208

0.1918 <0.0001 <0.0001 0.4532 0.244 0.0043 <0.0001 0.3929 <0.0001 0.4352 <0.0001 <0.0001 <0.0001 0.0163 0.194 0.0015 <0.0001 <0.0001 0.3564 0.093 0.0036 0.1237 <0.0001 0.093 0.0565 0.0425 0.1997 0.4176 0.0565 0.4967 0.0015 0.0499 0.093 0.021 0.0425 0.093 0.021

Glycopeptide–tetracycline–chloramphenicol–aminoglycoside Glycopeptide–tetracycline–chloramphenicol–B-lactam– aminoglycoside–nitrofuran Tetracycline–chloramphenicol–aminoglycoside Glycopeptide–tetracycline–chloramphenicol–aminoglycoside– nitrofuran–aminoglycoside Tetracycline–chloramphenicol–aminoglycoside–nitrofuran– aminoglycoside

8 1

(8.333333) (1.041667)

0.3492 0.1188

0.0002 0.1244

14 2

(14.58333) (2.083333)

0.4189 0.1689

<0.0001 0.0499

2

(2.083333)

0.1689

0.0499

38 39 40 41 42

*

Correlation significant at p  0.05 Correlation significant at p  0.01 *** Correlation significant at p  0.001 **** Correlation significant at p  0.0001 **

**** ****

** ****

****

**** **** **** *

** **** ****

**

****

*

** *

* *

*

***

**** *

*

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38

sand isolates possessed the gene. Fisher’s exact test revealed significant associations between carriage of esp and asa genes, esp and gelE genes and between asa and gelE genes. Examining species variation in the carriage of virulence genes, no notable difference however was observable between virulence factors detectable among E. faecalis as compared with other enterococci species apart from the case of asa gene carriage. Also, Wilcoxon paired test did not reveal any significant correlation of virulence marker prevalence with species diversity (Table 5). However, statistically significant correlations were observable between the carriage of virulence markers and the site of isolation. This was extremely significant for gelE gene as carriage of this factor was mostly associated with strains from the river area. A multivariate typing approach which involved clustering of 18 different parameters covering resistance to antibiotics, demonstration of virulence phenotypes and carriage of virulence genes were analysed using conglomerate hierarchical clustering (Fig. 2). The dendogram obtained successfully clustered more than 70% of each of the isolates recovered from various sources together on a source basis. Interestingly when datasets of enterococci isolates recovered from the portion of the beach proximate to the influx of river chempedak were included in the analysis, they were found to cluster with the river chempedak isolates, predominantly; river chempedak water isolates (Fig. 2). Using a 75% cut-off point, analysis of the obtained dendogram revealed two main clusters of beach area and river area strains. 3.5. Evidence of co-selection of virulence markers and antibiotic resistance The highest levels of resistance to combination of antibiotics was observed for T–K (tetracycline–aminoglycoside) (39.58%) followed closely by C–K (chloramphenicol–aminoglycoside) while the least levels was observed for combined resistance to chloramphenicol and nitrofurantoin (C–N) (9.38%) (Table 4). For resistance to multiple antibiotic groups (P3), the highest proportion was observed for T–C–K (tetracycline–chloramphenicol–aminoglycoside) (21.88%). The lowest proportion of combined resistance observed was that of V–T–C–A–K–N–S, i.e. resistance to a combination of all 7 classes of antibiotics tested (1.04%). Considering co-selection of two groups of antibiotics at the species level (Fig. 3a), the highest level of combined resistance was observed for T–K–S (tetracycline–aminoglycoside) among 64.71% of E. gallinarum isolates. On the other hand, the least prevalence of combined resistance to two groups of antibiotics was observed among E. casseflavus isolates (2.63%). E. faecalis isolates recovered in the study were found to demonstrate resistance to a combination of two groups of antibiotics at prevalence levels that ranged from 15% (as in V–K {glycopeptide–aminoglycoside combination}) to a maximum of 53.85% (as in T–A {tetracycline–aminoglycoside}). For a combination of three groups of antibiotics (Fig. 3b), E. hirae strains presented the highest prevalence levels as 75% of the strains were observed to be resistant to a combination of A–K–N–S (B-lactam–aminoglycoside–nitrofuran). For combinations of more than three groups of antibiotics (Fig. 3c), E. gallinarum topped the list presenting a variety of 15 combination possibilities. It is critical to affirm at this juncture that E. gallinarum strains dominated the species diversity of strains recovered from the river. Notably, combined resistance to V–T–C–A–K–N–S, i.e. resistance to a combination of all 7 classes of antibiotics tested (1.04%) was only identified among the river isolates. The current study provides empirical evidence that multiple-resistant strains are being disseminated along with river flow into the considered recreational beach water. Evidently, an analysis of resistance to combination of groups of antibiotics vis-à-vis the considered isolation sites points also in this direction. For a combination of two groups (Fig. 3d), three groups (Fig. 3e) and four

33

or more groups of antibiotics (Fig. 3f), river water isolates were notably the highest in terms of prevalence. Considering the whole library of environmental enterococci (n = 96) tested in this study, Fishers exact analysis revealed evidence of co-selection of vancomycin resistance and a number of resistance to combinations of groups of antibiotics. These include resistance against V–K (P = 0.000), V–T (=0.000), V–C–A–K–S (P = 0.003), C–K–S (P = 0.011), V–T–C–A–K (P = 0.003), V–C–A (P = 0.000), T–C–A–K–S (P = 0.042), V–T–C–A–K–S (P = 0.012) and T–C–K–S (P = 0.007). Similarly evidence of co-selection of antibiotic resistance and virulence was found between the carriage of gelE and resistance to a number of combinations of antibiotic groups. The correlation observed between prevalence of resistance to single/multiple classes of antibiotics was also explored using a binary coding system that designated 0 and 1 for antimicrobial susceptibility and resistance respectively. Answers to two main questions were sought in the analysis. Could it be inferred by correlation analysis that at every instance, co-selection of antibiotic resistance (coder = 1) was always associated with E. faecalis or E. faecium species (coded = 1) and not the rest of the enterococci species (coded = 0)? At every instance resistance was observed, was it always associated with enterococci recovered from a particular source (codes = 1) and not from the other sources (codes = 0)? No statistically significant correlation was observable for the first hypotheses. However, for the second hypothesis that aimed to correlate resistance observed with source, several extremely significant correlations (p < 0.0001) were observed (Table 5). Resistance to tetracycline, chloramphenicol, B-lactam, aminoglycoside were linked with strains from the river draining into the recreational beach water (p-value < 0.0001). Furthermore, co-selection of resistance to T–K (tetracycline–aminoglycoside), C–K (chloramphenicol–aminoglycoside), T–C (tetracycline–chloramphenicol), C–K–S (chloramphenicol–aminoglycoside), T–K–S (tetracycline– aminoglycoside), T–C–K (tetracycline–chloramphenicol–aminoglycoside), T–C–K (tetracycline–chloramphenicol–aminoglycoside) and T–C–K–N–S (tetracycline–chloramphenicol–aminoglycoside– nitrofuran) were observed to be linked to strains from the river area at extremely significant probability levels (p < 0.0001). These observations are in concert with Thong et al. (2011) that reported chloramphenicol resistance as linked always with resistance to streptomycin (aminoglycoside), tetracycline and ampicillin (B-lactam). In attempting to provide similar answers for virulence, it was observed that the carriage of virulence markers was not statistically correlated with species diversity. However, an extremely significant correlation (p < 0.0001) was observed for gelE carriage and membership of the RA class (pooled population of river water and river sand) (Table 6). Relatively lower but significant correlations were also observed for cyl (p = 0.021) and asa (p = 0.018) gene carriage. With carriage of three of four tested virulence markers in this study found to correlate with membership of the RA class (pooled population of river water and river sand), our study presents evidence again to implicate the river in the preponderance of enterococci with virulence traits in the tested recreational beach water. Added to our observations of potential recombination events from sequence-based recombination detection analysis (data not included), suggestively these strains are able to exchange genetic materials with indigenous beach water strains. 3.6. Multi-Locus Sequence Typing The aim of this part of the study was to determine if any of selected strains with multidrug resistance and multiple virulence gene carriage could be linked to a globally recognised clonal complex that has been previously implicated in enterococcal infec-

34

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38

Fig. 2. Strain typing based on a combintion of phenotypic characteristics and the carriage of virulence genes.

tion outbreaks. Based on a combination of their harbourage of virulence genes and antibiotic resistance profiles, four E. faecalis isolates were selected for multilocus sequence typing analysis (MLST) (Table 7). The first strain tested (RC19) was observed to be ST181. ST181 shared six out of the seven tested loci with the popular ST6, a member of the clonal complex 2 (CC2). CC2 corresponds to a clone of widely distributed multi-drug resistant strain that has been previously associated with hospital outbreaks. It was interesting to note that the isolate was recovered from the low tide river which drains into the seawater available for recreational purposes. Our observation of recovery of CC2 strains in non-hospital settings corroborate a recent report that also recovered sequence type strains from this clonal complex from liquid manure and sewage (Freitas et al., 2009). A single representative of CC2 was re-

ported also from a farm animal (pig) (Freitas et al., 2011). Enrichment of STs belonging to CC2 among hospital-associated isolates has been previously documented in a number of countries (Freitas et al., 2009; Ruiz-Garbajosa et al., 2006; Kawalec et al., 2007). Notably, the ST181 strain was observed to be resistant to tetracycline, chloramphenicol, kanamycin and streptomycin. The observation of multi drug resistance is in concert with a previous study (Freitas et al., 2009). In addition to asa and gelE genes, ST181 possessed two other putative virulence genes, cyl and esp. These observations support previous findings from other authors that highlight the role of cyl and esp genes in the pathogenicity of enterococci (Freitas et al., 2009; Shankar et al., 2006; McBride et al., 2007). Furthermore, full genome sequencing of a representative strain of this CC (ST6) revealed the presence of several mobile ele-

35

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38

BW

100

BW

RS

60

40

20

100

RC

60

RS

40

RS 40

20

-A C -K -A T- K-S C C A-K -A C -K -A -N -K V- N-S TA A -N -K T- -N A- -T K T- -N C V- -A- S T- KC - S T- K-N C -S -K V- -NTS V- C T- A C -K -A -K V V- -T T- -C S C -A -K-S -K -N -S

-S

C

-K

V-

V-

N

-S

T-C

A-

KN

C-

A-

A

-S -N

TC-

TA

K

-A VC

K

CT-

A-

N

40

100

E.gallinarum E.faecium

60

E.faecalis 40

E.gallinarum E.faecalis 40

VV- C-A C -K -A T- K-S C C A-K C A-K -A -K N -N V- -S TA A -N -K T- -N A- -T K T- NV- C-A S T- -K C -S T- K-N C -S -K V- -N T- -S V- C T- -A C -K -A V V- -T K-S T- -C C -A K-S -K -N -S

AN KNS T-C -K -S

C-

A-

A

-S -N

TC-

TA

V-

C-

A

K CT-

AT-

K-

K

N

S K-

T-

C

T-

-K

-S

-N

A T-

C

T

C

V-

T-

T-

C

K

0

-K

0 K

20

-K

20

A

E.faecium

60

0 V-

E.hirae

80

20

Combination of two groups of antimicrobials along with diversity

E._casseflavus

E.hirae

% Resistance

E.faecalis

E._casseflavus

80

% Resistance

E.faecium

Combination of six groups of antimicrobials along with isolation site

(f)

100

E.gallinarum

60

T-

KT-

T

A T-

C

V-

T-

K T-

C -K

K V-

A -K

C -N C -K -S TKS

(e)

E.hirae

80

0

Combination of three groups of antimicrobials along with isolation site

E._casseflavus

100

RC

60

0

(d)

BS BWNTR

BWNTR

20

Combination of two groups of antimicrobials along with isolation site

BW

80

BS

80

% Resistance

RC

0

% Resistance

(c)

100

BWNTR

80

% Resistance

(b)

BS

% Resistance

(a)

Combination of three groups of antimicrobials along with diversity

Combination of four to six groups of antimicrobials along with diversity

Fig. 3. (a–c) Distribution of single or multiple antimicrobial resistance in different Enterococci species isolates based on isolation source and (d–f) species diversity. V, Vancomycin (glycopeptides), K, kanamycin, S-Streptomycin (aminoglycoside), A-ampicillin (B-lactam), T, tetracycline (tetracycline), C-chloramphenicol (chloramphenicol), N, nitrofuranoin (nitrofuran).

Table 6 Correlation of the prevalence of single/multiple virulence markers with species diversity and isolation site. Beach area (n = 55) proportions (%)

River area (n = 41) proportions (%)

S/no

Virulence markers

Ec

Eh

Eg

Efm

Efs

Ec

Eh

Eg

Efm

Efs

1 2 3 4 5 6 7 8

cyl+ esp+ asa+ gelE+ asa+gelE+ esp+gelE+ esp+asa+gelE+ cyl+esp+asa+gelE+

0 46.15 69.23 92.31 69.23 46.15 46.15 0

0 5 20 40 20 5 5 0

– – – – – – – –

0 0 0 100 0 0 0 0

0 19.04 47.61 76.19 47.62 19.05 19.05 0

0 0 0 100 0 0 0 0

9.09 27.27 72.72 100 72.73 27.27 27.27 9.09

11.11 27.78 61.11 100 61.11 27.78 27.78 11.11

– – – – – – – –

0 9.09 63.64 100 63.64 9.09 9.09 0

Total Enterococci (%)

3 (3.13) 20 (20.83) 49 (51.04) 78 (81.25) 49 (51.04) 20 (20.83) 20 (20.83) 3 (3.13)

Correlation with species diversity

Correlation with isolation site

(rs)

p-Value

(rs)

p-Value

0.01464 0.1255 0.0149 0.1034 0.0149 0.1255 0.1255 0.01464

0.4437 0.1116 0.4427 0.158 0.4427 0.1116 0.1116 0.4437

0.208 0.02377 0.2137 0.4148 0.2137 0.02377 0.02377 0.208

0.021* 0.4091 0.0183* <0.0001*** 0.0183* 0.4091 0.4091 0.021*

Notes: Ec-Enterococcus casseflavus, Eh-Enterococcus hirae, Eg- Enterococcus gallinarum, Efs-Enterococcus faecalis, Efm-Enterococcus faecium, cyl+: strains carrying gene coding for cytolysin; esp+: strains carrying gene coding for enterococcal surface protein; asa+: strains carrying gene coding for aggregation substance; gelE+: strains carrying gene coding for gelatinase.

Table 7 Characteristics of the four E. faecalis isolates selected for multilocus sequence typing analysis (MLST). Strain E. E. E. E.

faecalis faecalis faecalis faecalis

RC9 BW15 BW17 RC19

gdh

gyd

pstS

gki

aroE

xpt

yqil

Sequence Type (ST)

Colonal complex (CC)

Antibiotic resistance phenotype

Virulence marker

14 14 1 12

2 1 1 5

18 18 9 3

10 57 6 7

16 16 1 6

2 35 1 1

12 12 1 5

59 474 117 181

N/A N/A CC21 CC2

Tet, Chl, Kan, Nit Kan Kan Tet, Chl, Kan, Str

asa, gelE asa, gelE esp, asa, gelE esp, asa, cyl, gelE

ments, a pathogenicity island and an operon for capsular polysaccharide biosynthesis which have been demonstrated to aid resistance to phagocytic killing (Hancock and Gilmore, 2002). Analysis of the MLST data revealed another isolate (RC59) as being a ST59 strain. As with RC19, the isolate was recovered from the low tide river which drains into the seawater available for recreational purposes. The occurrence of ST59 has been previously reported among chickens, faeces sample from a healthy pig and from hospitalised patients (Ruiz-Garbajosa et al., 2006; Kawalec et al., 2007). The strain was found to harbour asa and gel genes. ST59

was found to be resistant to a number of antibiotics; tetracycline, chloramphenicol, kanamycin and nitrofurantoin. This is in agreement with a recent study that reported the isolation of ciprofloxacin resistant ST59 strains from Australian waterways (Rathnayake et al., 2012). The past years of abuse of chloramphenicol and other antibiotics in agricultural settings might have played a role in the spread of antibiotic resistance among these STs in Malaysia. A South East Asian study (Huys et al., 2007) reported that due to the decrease in therapeutic efficiency of inexpensive antibiotics such as oxytetracycline, many Asian aquaculture activities have

36

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38

gradually switched during the past decade to the use of other broad-spectrum antibiotics such as chloramphenicol and the synthetically derived florfenicol. Although the lack of exact data on usage pattern of antimicrobials in human and veterinary medicine limits the study in potentially exploring the quantitative nature of cause-effect relationships, it remains generally accepted that the selection of resistant bacteria in agricultural settings might be contributing to the spread of resistance (Cabello, 2006). One of the four isolates (BW17) recovered from recreational bathing water in the current study was found to be of the sequence type 117. A previous study (Freitas et al., 2009) isolated members of this clonal complex from both hospital patients and from piggery. CC21 isolates have also been frequently observed in the global E. faecalis collection among isolates of diverse origin, including hospitalized patients, non-hospitalized individuals, meat and farm animals (Ruiz-Garbajosa et al., 2006). This seems to highlight the dispersion of this CC in different ecological settings. The ST117 strain recovered from recreational bathing water in this study exhibited resistance to low level kanamycin resistance. Quinones et al. (2009) also observed the absence of multi drug resistance among ST117 strains tested. It was also reported that ST117 is a double-locus variant of ST21 in CC21 (one of the biggest CCs detected in Spain), which include strains from animals, hospitalized patients, and apparently healthy volunteers in the community (Quinones et al., 2009). The ST117 strain in our study harboured the esp, asa and gel genes. Although previously characterized CC21 strains usually harboured fewer antibiotic resistance and virulence traits than isolates of other CCs (Ruiz-Garbajosa et al., 2006; Kawalec et al., 2007; McBride et al., 2007), the possibility of recombination in beach water settings could herald successful acquisition of diverse genetic elements that might facilitate their persistence and spread in environments under selective antibiotic pressure. Among STs reported in this study, ST117 and ST181 were only previously reported as novel STs in a study conducted in Cuba (Quinones et al., 2009) and another study on waterways in Australia (Rathnayake et al., 2011). The regional geographical proximity of the latter to Malaysia may be a useful premise to present a ST181 clonal regional spread hypothesis. Notwithstanding, the local availability of virulence genes is a crucial determinant on the extent to which these factors are acquired by epidemic clones through horizontal gene transfer (Novais et al., 2004; Nilsson et al., 2009). It is difficult however to suggest a clonal regional spread hypothesis for the occurrence of the novel ST117 in beach environment settings in Malaysia. Only three studies have previously reported on the genetic variability of enterococci in Malaysia. All of these studies focused on clinical and animal strains. Weng et al. (2012) reported the occurrence of ST18 and ST596 strains among isolates recovered from hospital patients. Also, another study (Weng et al., 2013) reported ST17, ST78, ST203 and ST601 strains isolated from clinical specimens. In a most recent study, Getachew et al. (2013) reported ST4, ST6, ST87, ST108, ST274 and ST244 from a total of 11 isolates recovered from humans, chickens and pigs. To the best of our knowledge, this is the first report on the use of MLST to delineate strains recovered from beach water environment in Malaysia into STs and clonal complexes. Considering the growing evidences that clinical resistance is intimately associated with environmental bacteria (de Vera and Simmons, 1996; Abriouel et al., 2008), research activities need to be expended to include environmental microorganisms. In the current study, the isolation of a strain discovered to be a member of the hospital adapted clonal complex (CC2) is an issue of concern. A recent study also reported the occurrence of hospital adapted clonal complex 17 E. faecium from a beach environment. It could be that both the river from which the strain was isolated along with the storm water drainage which empties into it are acting as reservoirs for this ST and other potentially

virulent strains. Subsequent rainfall events dislodge them and ultimately herald their dispersal into recreational beach water. Suggestively, the findings of our study emphasize the need for more studies on the characterization of enterococci strains from beaches, particularly those recovered from beaches receiving significant influx from polluting rivers and storm drainage systems. Arguably, there is also the need for inclusion of these environments in the global discourse of epidemiological research. Some limitations were observed during the study. First, there is a dearth of precise data on usage pattern of antimicrobials in human and veterinary medicine in Malaysia. This arguably makes it a challenge to explore the quantitative nature of cause-effect relationships. This observation however continues to remain a global challenge for environmental health researchers particularly in less developed nations where relevant regulatory institutions are frail (Dada et al., 2012). Another limitation in this study is the financial restrictions that limited the number of isolates sequenced. In an ideal situation, a very large library of sequences obtained from various potential contributory sources of pollution ought to have been analysed in this study. This could have increased the confidence levels of our analysis. Despite this restriction however, our study provides useful insight into the prevalence and dispersion of antibiotic resistance and virulence makers in the studied beach. Financial restrictions also limited the number of strains subjected to MLST, it is thus difficult to generalise on an overall representation of the E. faecalis population based on the prevalence of certain clonal complexes neither is it feasible to identify a possible clonal expansion of a particular strain in the study location. 4. Conclusion The presence of antibiotic resistant enterococci with virulence traits in surface recreational water could be a public health risk. Conflict of Interest All authors declare that they have no competing interests. Acknowledgements The corresponding author (ACD) is in receipt of an Education Tax Fund PhD research grant. Thanks to Encik Alias for assistance during sample collection. Financial support for sampling and laboratory analysis from the Universiti Kebangsaan Malaysia Marine Pathogen Program Grant (Science Fund 04-01-02-SF0754) under the auspices of the School of Bioscience and Biotechnology, Faculty of Science and Technology is appreciated. Institute of Ecology and Environmental Studies, Obafemi Awolowo University, Ile-Ife, Nigeria gratefully granted study leave for the research. References A.P.H.A., 1999. Standard Methods for the Examination of Water and Wastewater, 20th ed. American Public Health Association (Apha), Washington, DC. Abriouel, H., Omar, N.B., Molinos, A.C., López, R.L., Grande, M.J., Martínez-Viedma, P., Ortega, E., Cañamero, M.M., Galvez, A., 2008. Comparative analysis of genetic diversity and incidence of virulence factors and antibiotic resistance among enterococcal populations from raw fruit and vegetable foods, water and soil, and clinical samples. Int. J. Food Microbiol. 123, 38–49. Adria, R.I., Harry, H., Steve, N., 2011. Occurrence and antimicrobial drug resistance of potential bacterial pathogens from shellfish, including Queen Conchs (Strombus Gigas) and Whelks (Cittarium pica) in Grenada. WebMedCentral Article ID: WMC001943. . Allan, J.C., Hart, R., 2007. Profile dynamics and particle tracer mobility of a cobble berm constructed on the oregon coast. In: Proceedings of the Sixth International Symposium on Coastal Engineering and Science of Coastal Sediment Processes, New Orleans, USA, pp. 1–14.

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38 USEPA: Bacterial Water Quality Standards for Recreational Waters (Freshwater and Marine Waters), 2003. . webcite EPA-823-R-03-008. U.S. Environmental Protection Agency, Washington, DC. Barrell, R., Hunter, P., Nichols, G., 2000. Microbiological standards for water and their relationship to health risk. Commun. Dis. Public Health 3, 8–13. Bhaduri, B., Minner, M., Tatalovich, S., Harbor, J., 2001. Long-term hydrologic impact of urbanization: a tale of two models. J. Water Resour. Plan. Manage. 127, 13– 19. Brownell, M., Harwood, V., Kurz, R., McQuaig, S., Lukasik, J., Scott, T., 2007. Confirmation of putative stormwater impact on water quality at a Florida beach by microbial source tracking methods and structure of indicator organism populations. Water Res. 41, 3747–3757. Cabello, F.C., 2006. Heavy use of prophylactic antibiotics in aquaculture: a growing problem for human and animal health and for the environment. Environ. Microbiol. 8, 1137–1144. Cariolato, D., Andrighetto, C., Lombardi, A., 2008. Occurrence of virulence factors and antibiotic resistances in Enterococcus faecalis and Enterococcus faecium collected from dairy and human samples in North Italy. Food Control 19, 886– 892. Cetinkaya, Y., Falk, P., Mayhall, C.G., 2000. Vancomycin-resistant enterococci. Clin. Microbiol. Rev. 13, 686–707. Cheong, Y., Lim, V., Jegathesan, M., Suleiman, A., 1994. Antimicrobial resistance in 6 Malaysian general hospitals. Med. J. Malaysia 49 (4), 317. Creti, R., Imperi, M., Bertuccini, L., Fabretti, F., Orefici, G., Di Rosa, R., et al., 2004. Survey for virulence determinants among Enterococcus faecalis isolated from different sources. J. Med. Microbiol. 53, 13–20. Dada, A.C., Asmat, A., Gires, U., Heng, L.Y., Deborah, B.O., 2012. Bacteriological monitoring and sustainable management of beach water quality in malaysia: problems and prospects. Global J. Health Sci. 4, 126–138. Dada, A.C., Ahmad, A., Usup, G., Heng, L.Y., 2013. Speciation and antimicrobial resistance of Enterococci isolated from recreational beaches in Malaysia. Environ. Monit. Assess. 185, 1583–1599. DailyExpress: Rubbish ending up in the sea, 2013. News published on Thursday, August 15. . (accessed 02.09.2013). de Vera, M.E., Simmons, R.L., 1996. Antibiotic-resistant enterococci and the changing face of surgical infections. Arch. Surg. 131, 338. Devriese, L., Hommez, J., Pot, B., Haesebrouck, F., 1994. Identification and composition of the streptococcal and enterococcal flora of tonsils, intestines and faeces of pigs. J. Appl. Microbiol. 77, 31–36. Dutka-Malen, S., Evers, S., Courvalin, P., 1995. Detection of glycopeptide resistance genotypes and identification to the species level of clinically relevant enterococci by PCR. J. Clin. Microbiol. 33, 24–27. Fisher, K., Phillips, C., 2009. The ecology, epidemiology and virulence of Enterococcus. Microbiology 155, 1749–1757. Freitas, A.R., Novais, C., Ruiz-Garbajosa, P., Coque, T.M., Peixe, L., 2009. Clonal expansion within clonal complex 2 and spread of vancomycin-resistant plasmids among different genetic lineages of Enterococcus faecalis from Portugal. J. Antimicrob. Chemother. 63, 1104–1111. Freitas, A.R., Coque, T.M., Novais, C., Hammerum, A.M., Lester, C.H., Zervos, M.J., Donabedian, S., Jensen, L.B., Francia, M.V., Baquero, F., 2011. Human and swine hosts share vancomycin-resistant Enterococcus faecium CC17 and CC5 and Enterococcus faecalis CC2 clonal clusters harboring Tn1546 on indistinguishable plasmids. J. Clin. Microbiol. 49, 925–931. Getachew, Y., Hassan, L., Zakaria, Z., Aziz, S.A., 2013. Genetic variability of vancomycin-resistant Enterococcus faecium and Enterococcus faecalis isolated from humans, chickens and pigs in Malaysia. Appl. Environ. Microbiology. Hamilton, M.J., Hadi, A.Z., Griffith, J.F., Ishii, S., Sadowsky, M.J., 2010. Large scale analysis of virulence genes in Escherichia coli strains isolated from Avalon Bay, CA. Water Res. 44, 5463–5473. Hamzah, A., Kipli, S.H., Rahil Ismail, S.I.I., 2011. Microbiological study in coastal water of Port Dickson, Malaysia. Sains Malaysiana 40, 93–99. Hancock, L.E., Gilmore, M.S., 2002. The capsular polysaccharide of Enterococcus faecalis and its relationship to other polysaccharides in the cell wall. Proc. Nat. Acad. Sci. 99, 1574–1579. Ríos Hernández LA, 2011. Report as of FY2011 for 2011PR129B. The population dynamics of the dominant Enterococci in the water systems of Puerto Rico.
37

Lata, P., Ram, S., Agrawal, M., Shanker, R., 2009. Enterococci in river Ganga surface waters: propensity of species distribution, dissemination of antimicrobialresistance and virulence-markers among species along landscape. BMC Microbiol. 9, 140. Layton, B., Walters, S., Boehm, A., 2009. Distribution and diversity of the enterococcal surface protein (esp) gene in animal hosts and the Pacific coast environment. J. Appl. Microbiol. 106, 1521–1531. Mallin, M.A., Williams, K.E., Esham, E.C., Lowe, R.P., 2000. Effect of human development on bacteriological water quality in coastal watersheds. Ecol. Appl. 10, 1047–1056. McBride, S.M., Fischetti, V.A., LeBlanc, D.J., Moellering Jr, R.C., Gilmore, M.S., 2007. Genetic diversity among Enterococcus faecalis. PLoS ONE 2, e582. McDonald, L.C., Kuehnert, M.J., Tenover, F.C., Jarvis, W.R., 1997. Vancomycinresistant enterococci outside the health-care setting: prevalence, sources, and public health implications. Emerg. Infect. Dis. 3, 311. Mohamed, J.A., Huang, W., Nallapareddy, S.R., Teng, F., Murray, B.E., 2004. Influence of origin of isolates, especially endocarditis isolates, and various genes on biofilm formation by Enterococcus faecalis. Infect. Immun. 72, 3658– 3663. NCCLS: Performance Standards for antimicrobial disk susceptibility testing. NCCLS document M2–A8. Wayne, PA: National Committee for Clinical Laboratory Standards. Nilsson, O., Greko, C., Top, J., Franklin, A., Bengtsson, B., 2009. Spread without known selective pressure of a vancomycin-resistant clone of Enterococcus faecium among broilers. J. Antimicrob. Chemother. 63, 868–872. Novais, C., Coque, T.M., Sousa, J.C., Baquero, F., Peixe, L., 2004. Local genetic patterns within a vancomycin-resistant Enterococcus faecalis clone isolated in three hospitals in Portugal. Antimicrobial Agents Chemother. 48, 3613–3617. Padilla, C., Lobos, O., 2013. Virulence, bacterocin genes and antibacterial susceptibility in Enterococcus faecalis strains isolated from water wells for human consumption. SpringerPlus 2, 1–6. Peng, X., Zhang, G., Mai, B., Hu, J., Li, K., Wang, Z., 2005. Tracing anthropogenic contamination in the Pearl River estuarine and marine environment of South China Sea using sterols and other organic molecular markers. Mar. Pollut. Bull. 50, 856–865. Quinones, D., Kobayashi, N., Nagashima, S., 2009. Molecular epidemiologic analysis of Enterococcus faecalis isolates in Cuba by multilocus sequence typing. Microbiol. Drug Resist. 15, 287–293. Raja, N., Karunakaran, R., Ngeow, Y., Awang, R., 2005. Community-acquired vancomycin-resistant Enterococcus faecium: a case report from Malaysia. J. Med. Microbiol. 54, 901–903. Rathnayake, I., Hargreaves, M., Huygens, F., 2011. SNP diversity of Enterococcus faecalis and Enterococcus faecium in a South East Queensland waterway, Australia, and associated antibiotic resistance gene profiles. BMC Microbiol. 11, 201. Rathnayake, I., Hargreaves, M., Huygens, F., 2012. Antibiotic resistance and virulence traits in clinical and environmental Enterococcus faecalis and Enterococcus faecium isolates. Syst. Appl. Microbiol. 35, 326–333. Ruiz-Garbajosa, P., Bonten, M.J., Robinson, D.A., Top, J., Nallapareddy, S.R., Torres, C., Coque, T.M., Cantón, R., Baquero, F., Murray, B.E., 2006. Multilocus sequence typing scheme for Enterococcus faecalis reveals hospital-adapted genetic complexes in a background of high rates of recombination. J. Clin. Microbiol. 44, 2220–2228. Santiago-Rodriguez, T.M.R.J., Coradin, M., Toranzos, A., 2013. Antibiotic-resistance and virulence genes in Enterococcus isolated from tropical recreational waters. J. Water Health 11, 387–396. Sapkota, A.R., Curriero, F.C., Gibson, K.E., Schwab, K.J., 2007. Antibiotic-resistant enterococci and fecal indicators in surface water and groundwater impacted by a concentrated swine feeding operation. Environ. Health Perspect. 115, 1040. Schippmann, B., Schernewski, G., Gräwe, U., 2013. Escherichia coli pollution in a Baltic Sea lagoon: A model-based source and spatial risk assessment. Int. J. Hyg. Environ. Health. Semedo, T., Santos, M.A., Lopes, M.F., Marques, J.J.F., Crespo, M.T., Tenreiro, R., 2003. Virulence factors in food, clinical and reference enterococci: a common trait in the genus? Syst. Appl. Microbiol. 26, 13–22. Shankar, N., Baghdayan, A.S., Willems, R., Hammerum, A.M., Jensen, L.B., 2006. Presence of pathogenicity island genes in Enterococcus faecalis isolates from pigs in Denmark. J. Clin. Microbiol. 44, 4200–4203. Silva, J., Castillo, G., Callejas, L., López, H., Olmos, J., 2006. Frequency of transferable multiple antibiotic resistance amongst coliform bacteria isolated from a treated sewage effluent in Antofagasta, Chile. Electron. J. Biotechnol. 9. Son, R., Nimita, F., Rusul, G., Nasreldin, E., Samuel, L., Nishibuchi, M., 1999. Isolation and molecular characterization of vancomycin-resistant Enterococcus faecium in Malaysia. Lett. Appl. Microbiol. 29, 118–122. Thong, K.-L., Hoe, C.H., Koh, Y.T., Yasin, R.M., 2011. Prevalence of multidrugresistant Shigella isolated in Malaysia. J. Health Popul. Nutr. (JHPN) 20, 356– 358. Tiong, S.T., 2001. Integrated shoreline management plan of North Pahang. Coast. Geomorphol. 1, 1–112, . Toosa, H., Radu, S., Rusul, G., Latif, A.R.A., Rahim, R.A., Ahmad, N., Ling, O.W., 2001. Detection of Vancomycin-Resistant Enterococcus Spp.(VRE) from Poultry. Malaysian J. Med. Sci. MJMS 8, 53. Trask, P.D., 1952: Source of beach sand at Santa Barbara, California, as indicated by mineral grain studies. In: Book Source of Beach Sand at Santa Barbara, California, City: Dtic Document.

38

A. Ahmad et al. / Marine Pollution Bulletin 82 (2014) 26–38

Tsubakishita, S., Kuwahara-Arai, K., Sasaki, T., Hiramatsu, K., 2010. Origin and molecular evolution of the determinant of methicillin resistance in staphylococci. Antimicrob. Agents Chemother. 54, 4352–4359. Turgeon, P., Michel, P., Levallois, P., Chevalier, P., Daignault, D., Crago, B., Irwin, R., McEwen, S.A., Neumann, N.F., Louie, M., 2012. Antimicrobial-resistant Escherichia coli in public beach waters in Quebec. Can. J. Infect. Dis. Med. Microbiol. 23, e20. Vankerckhoven, V., Van Autgaerden, T., Vael, C., Lammens, C., Chapelle, S., Rossi, R., Jabes, D., Goossens, H., 2004. Development of a multiplex PCR for the detection of asa1, gelE, cylA, esp, and hyl genes in enterococci and survey for virulence determinants among European hospital isolates of Enterococcus faecium. J. Clin. Microbiol. 42, 4473–4479. Vergis, E.N., Shankar, N., Chow, J.W., Hayden, M.K., Snydman, D.R., Zervos, M.J., Linden, P.K., Wagener, M.M., Muder, R.R., 2002. Association between the

presence of enterococcal virulence factors gelatinase, hemolysin, and enterococcal surface protein and mortality among patients with bacteremia due to Enterococcus faecalis. Clin. Infect. Dis. 35, 570–575. Weng, P.L., Hamat, R.A., MPath, Y.K.C., Zainol, N., Aziz, M.N., Shamsudin, M.N., 2012. Vancomycin-resistant Enterococcus faecium of multi locus sequence type 18 in Malaysia. Med. J. Malaysia 67, 639. Weng, P.L., Ramli, R., Shamsudin, M.N., Cheah, Y.-K., Hamat, R.A., 2013. High genetic diversity of enterococcus faecium and Enterococcus faecalis clinical isolates by pulsed-field gel electrophoresis and multilocus sequence typing from a hospital in Malaysia. BioMed Res. Ints. Zhang, Q., Li, Z., Zeng, G., Li, J., Fang, Y., Yuan, Q., Wang, Y., Ye, F., 2009. Assessment of surface water quality using multivariate statistical techniques in red soil hilly region: a case study of Xiangjiang watershed, China. Environ. Monit. Assess. 152, 123–131.