Abstract
Hurricane-caused stormwater runoffs transport diverse terrestrial pollutants, adversely impact microbiological water quality, and introduce fecal and other pathogens to coastal water environments. This study investigated the genotypic diversity, phylogenetic composition, antibiotic resistance patterns, and virulence gene repertoire of the Enterococcus population in the Hilo Bay coastal water after the immediate impact of Hurricane Lane. DNA fingerprinting of Enterococcus isolates exhibited large genotypic diversity, while 16S rRNA gene sequencing identified four major species, including E. faecalis (34.7%), E. faecium (22.4%), E. hirae (22.4%), and E. durans (18.4%). Four common enterococcal virulence genes (cylA, esp, asa1, and gelE) were detected in the Enterococcus population, with significant portions of E. durans (33.3%), E. faecalis (41.2%), E. faecium (36.4%), and E. hirae (27.3%) isolates possessing two or more virulence genes. Considerable antibiotic resistance to rifampin, erythromycin, tetracycline, and nitrofurantoin was detected in the Enterococcus population, with one E. durans isolate showing vancomycin resistance. The results indicate considerable health implications associated with Enterococcus spp. in the hurricane-impacted tropical coastal water, illustrating the needs for more comprehensive understanding of the microbiological risks associated with storm-impacted coastal water.
HIGHLIGHTS
The study investigated the Enterococcus population in a tropical coastal water immediately after the impact of a major hurricane.
Genotypic diversity and phylogenetic composition of the Enterococcus population were determined.
A significant portion of the Enterococcus population contains various virulence factors.
Considerable antibiotic resistance was detected in the Enterococcus population.
Graphical Abstract
INTRODUCTION
Hilo town area (Hawaii island, Hawaii), which has a tropical rainforest climate (NOAA 2021), is populated with dense sewer networks and a large number of on-site wastewater treatment systems (e.g. 8,700 cesspools; Silvius et al. 2005) and is also located next to the mouth of the Wailuku River. Hurricane Lane, a category five hurricane that affected the area during August 15–29, 2018, caused heavy rainfall over the Hilo town area and resulted in its wettest 3-day period on record in precipitation (80.9 cm) (Beven II & Wroe 2019). This unprecedented rainfall resulted in significant stormwater runoff and flooding, and sanitary sewer overflows due to rainfall-derived infiltration and inflow into the aging sewer infrastructures in the area. As a result, the microbiological quality of coastal water in Hilo Bay was deteriorated according to the high levels of fecal indicator bacteria (FIB) Enterococcus observed in the water samples collected after the heavy rains (Saingam et al. 2021).
The Enterococcus genus is a group of Gram-positive bacteria with 58 known species (Parte 2014), which are commensals in human and various animal guts (Layton et al. 2010) and are used as an important FIB in monitoring coastal marine water quality (USEPA 2012). Certain Enterococcus spp. are pathogenic and cause endocarditis, bacteremia, neonatal, central nervous system, urinary tract, abdominal, and pelvis infections (Lewis & Zervos 1990), which is estimated to cause more than $539 million per year in healthcare costs (CDC 2019). E. faecalis and E. faecium are the two major species of opportunistic pathogens prevalent in health clinical, whereas E. durans, E. gallinarum, E. avium, E. hirae, and E. mundtii can also be responsible for infections (Blaimont et al. 1995). Pathogenicity of enterococci is contributed by several common virulence factors encoded by the asa1, cylA, gelE, esp, and hyl genes (Jett et al. 1994). The asa1 gene encodes aggregation substance proteins that assist in attachment and colonization during renal tissue infection (Waters et al. 2004). The cylA gene encodes cytolysin that aggravates the endocarditis infection (Huycke & Gilmore 1995). The gelE encodes extracellular gelatinase which mediates the virulence by modulating host immune response and degrading tissues (Park et al. 2008). The esp gene encodes the enterococcal surface proteins that are involved in biofilm formation and urinary tract infection (Shankar et al. 1999). The hyl gene encodes hyaluronidase that is linked to increased gastrointestinal colonization in mouse models (Rice et al. 2003).
The presence of antibiotic-resistant genes (ARGs) could further increase and complicate health risks caused by the pathogenic enterococci in coastal water (Kristich et al. 2014). Enterococci were reported to be inherently tolerant to several antibiotics within the classes of penicillin and aminoglycosides (Robbins & Tompsett 1951). Enterococci from environmental water also showed resistance to several other antibiotics, including tetracyclines, erythromycin, or gentamicin, likely through horizontal gene transfer (Macedo et al. 2011). Among them, vancomycin-resistant Enterococcus (VRE) are resistant to one of the last-resort antimicrobial agents (i.e. vancomycin) and have contributed to approx. 30% of enterococcal infections and about 5,400 mortalities each year in the USA (CDC 2019).
Previous studies on enterococcal isolates in coastal waters after the impact of major storms were largely focused on the abundance of Enterococcus population (Roca et al. 2019; Jiang et al. 2020). The phylogenetic composition of the population in Lake Pontchartrain, a fresh water lake, after the impact of Hurricane Katrina was also assessed by sequencing the 23S rRNA gene of the enterococcal isolates (Bae & Hou 2013). In a previous study, Saingam et al. (2021) reported that the Hilo Bay Lighthouse (HBL) site was the most contaminated site in Hilo Bay after the stormwater impact of Hurricane Lane, which had a mean enterococcal density of 144 CFU/100 mL and exceeded the water quality criteria statistical threshold value. Bae & Hou (2013) reported a large portion of classical fecal origins E. faecalis, E. faecium, E. durans, E. hirae, and E. mundtii were detected in freshwater as the impact of Hurricane Katrina-derived urban runoff. A study done by Jiang et al. (2020) reported that E. faecalis were detected in 60% of coastal waters and 67% of street surface runoff at a concentration of <10 MPN/L by the most probable number - loop mediated isothermal amplification (MPN-LAMP) assay after the back-to-back hurricane strikes by Irma and Maria. High concentrations of enterococci (mean concentration of gene marker of 5.05 × 105 copies/100 mL) were detected in 40–100% of surface water samples collected from flooded regions of the river impacted by Hurricane Harvey which was also reported by Kapoor et al. (2018). However, the pathogenic potentials, genotypic diversity, and ARG repertoire of the Enterococcus population after major storms in coastal marine water have not been adequately investigated.
The aim of this study was to understand microbial risks of the Enterococcus population in coastal water immediately after the impact of a major hurricane. Enterococcal isolates were obtained from near-shore coastal water at the HBL site, where the high level of FIB Enterococcus after Hurricane Lane was detected (Saingam et al. 2021). The Enterococcus spp. distribution was determined by 16S rRNA gene sequencing of the enterococcal isolates. The Enterococcus spp. genotypic diversity was determined by BOX-PCR fingerprinting and then compared with their 16S rRNA gene-based phylogeny. The prevalence of virulence genes in the Enterococcus population was determined by multiplex PCR detection, and their antimicrobial susceptibility was determined by microbroth dilution assays.
MATERIALS AND METHODS
Water sampling and enterococcal isolation
Coastal water sampling immediately after Hurricane Lane and sample processing were described in detail in a previous publication (Saingam et al. 2021). Briefly, approximately 2 L of the water samples (n = 7) were collected at knee-depth at the HBL sampling site from Day 5 to Day 11 when field sampling became feasible immediately after the peak hurricane impact (i.e. Day 0). The HBL site (GPS coordinates: 19°43′38.0″N 155°05′11.0″W) is located at the shoreline of Downtown Hilo area and approximately 100 and 2,000 m from the mouths of the Wailuku River and Wailoa River, respectively. Water samples (100 mL) were filtered through 0.45 μm GN-6 filter membranes, which were then incubated on the mEI agar to selectively grow enterococci colonies according to the EPA Method 1600 (USEPA 2006). Presumptive enterococcal colonies (i.e. colonies with blue halo) were picked and further isolated via streak plating on tryptic soy agar. The presumptive enterococcal isolates were subsequently confirmed based on growth in both Brain Heart Infusion Broth (BHIB) (35 °C ± 0.5 °C, 48 h) and BHIB supplemented with 6.5% NaCl (45 °C ± 0.5 °C, 48 h). A total of 117 confirmed enterococcal isolates were obtained and cultured in BHIB at 35 °C for 24 h and then stored at −80 °C in the presence of 18% glycerol.
16S rRNA gene sequencing
Forty-nine of the 117 enterococcal isolates were randomly selected and subjected to 16S rRNA gene sequencing to determine the phylogenetic composition of the Enterococcus population. Total genomic DNA from the enterococcal isolates was prepared by using the freeze–thaw method (Ran et al. 2013). The universal 16S rRNA gene primers (27F: 5′-AGAGTTTGATCMTGGCTCAG-3′ and 1492R: 5′-GGTTACCTTGTTACGACTT-3′) were used for PCR amplification. The reaction mixture contained 200 μM dNTPs, 0.125 U/μL Taq polymerase, and 0.4 μM primers in a final volume of 25 μL. The amplification program included initial denaturation at 94 °C for 5 min, 30 cycles of denaturation (94 °C for 30 s), annealing (53 °C for 45 s), and extension (72 °C for 1 min), and a final extension step (72 °C for 10 min). PCR products were purified using a Wizard PCR Preps DNA Purification System (Promega, Madison, WI, USA). The sequencing reaction was performed using the ABI 3730XL sequencer at the Advanced Studies in Genomics, Proteomics and Bioinformatics (ASGPB), University of Hawaii at Manoa. The sequence reads were quality trimmed and checked manually against the chromatograms by using the Sequence Scanner (Version 2.0; Thermo Fisher, Waltham, MA, USA), and the species identification was confirmed by using BLASTn. The 49 16S rRNA gene sequences were deposited in the GenBank database (accession no. MZ668452–MZ668500).
Multiplex PCR for virulence gene detection
The presence of enterococcal virulence genes, including asa1, gelE, cylA, esp, and hyl, were tested on the 49 Enterococcus spp. by following a multiplex PCR assay (Vankerckhoven et al. 2004). The reaction mixture contained 0.16 mg/mL BSA, 200 μM dNTPs, 1 mM MgCl2, 0.1 U/μL Taq polymerase, 0.1 μM of asa1 (F: 5′-GCACGCTATTACGAACTATGA-3′; R: 5′-TAAGAAAGAACATCACCACGA-3′), hyl (F: 5′-ACAGAAGAGCTGCAGGAAATG-3′; R: 5′-GACTGACGTCCAAGTTTCCAA-3′), and gelE (F: 5′-TATGACAATGCTTTTTGGGAT-3′; R: 5′-AGATGCACCCGAAATAATATA-3′) primer pairs each, and 0.2 μM of esp (F: 5′-AGATTTCATCTTTGATTCTTGG-3′; R: 5′-AATTGATTCTTTAGCATCTGG-3′) and cylA (F: 5′-ACTCGGGGATTGATAGGC-3′; R: 5′-GCTGCTAAAGCTGCGCTT-3′) primers each. The amplification steps included the initial denaturation at 95 °C for 7 min, 30 cycles of 94 °C for 1 min, 1 min annealing at 56 °C, 1 min extension at 72 °C, and 10 min final extension at 72 °C. PCR amplicons were separated by 3.0% agarose gel electrophoresis for 3 h at 80 V in 1× TBE. Gel with PCR amplicons was stained in 1× GelRed solution (Biotium; Hayward, CA, USA) and visualized using a GelDoc imager (Bio-Rad, Hercules, CA, USA). E. faecalis MMH594 (gelE+asa1+cylA+esp+) and E. faecium U0317 (esp+hyl+) were used as positive controls. The Enterococcus isolates that carried more than two virulence genes were considered multi-virulence gene carriers.
Antibiotic susceptibility testing
Antimicrobial susceptibility of the Enterococcus isolates (n = 49) was determined by the microbroth dilution method according to the Clinical and Laboratory Standards Institute (CLSI) protocol (CLSI 2012). Briefly, stock cultures were revived in the cation-adjusted Mueller Hinton broth (CAMHB) (BD; MD, USA) in 96-well microtiter plates at 35 °C until OD620 reached 0.4, which was the equivalent of 0.5 McFarland standard (∼5 × 108 cells/mL). Approximately 5 × 105 cells were transferred to wells of CAMHB supplemented with different antibiotics at different concentrations. The antibiotics tested included ciprofloxacin (Cip) (0.5–8 μg/mL), ampicillin (Amp) (4–32 μg/mL), chloramphenicol (Chl) (4–64 μg/mL), tetracycline (Tet) (2–32 μg/mL), vancomycin (Van) (2–64 μg/mL), erythromycin (Ery) (0.25–16 μg/mL), rifampin (Rif) (0.5–8 μg/mL), and nitrofurantoin (Nit) (16–256 μg/mL). The microbroth dilution plates were incubated at 35 °C for 20 h. The minimal inhibitory concentrations were recorded, and the strains that showed growth at the resistant breakpoints were recorded as ‘Resistant’ per CLSI criteria (CLSI 2012). E. faecalis ATCC 29212 was used as a quality control for the test media and antibiotics. Isolates that showed resistance to equal or more than two antibiotics were categorized in a group named as multi-antibiotic resistance (MAR) carrier.
Co-occurrence of virulence factors and AR
Co-occurrence of virulence genes and antibiotics resistance in the Enterococcus isolates were calculated using the Jaccard index, which is the number of isolates carrying both entities (A ∩ B) divided by the total number of isolates carrying either entity (). The Jaccard index ranged from 0 (no co-occurrence) to 1 (always co-occurrence). The analysis and plotting were performed in the R environment (Version 3.6.1).
BOX-PCR fingerprint clustering
The DNA fingerprinting of environmental Enterococcus isolates from the HBL site followed the procedures described previously (Saingam et al. 2021). Briefly, BOX-PCR fingerprints of the 49 Enterococcus isolates that had their 16S rRNA gene sequenced were generated and analyzed through clustering analysis in BioNumerics v.7.01 (Applied Maths Inc., Sint-Martens-Latem, Belgium). The level of similarity between fingerprints was calculated using the Dice coefficient at 1.0% optimization and 1.0% band position tolerance. Bands of up to 6 kb were included in the analysis. A representative dendrogram was constructed in BioNumerics using the unweighted pair group method with arithmetic mean. The relationship between phylogenetic clusters of DNA fingerprinting and species identification of enterococci isolates was evaluated by using Fisher's exact test. Comparisons of species among cluster groups were achieved by the post-hoc test, and the p-value was adjusted by the Benjamini–Hochberg false discovery rate method. The analysis was performed by using the rcompanion package (Mangiafico & Mangiafico 2017) in R.
RESULTS
Diversity of Enterococcus in hurricane-impacted coastal water
Forty unique genotypes out of a total 49 Enterococcus isolates were identified based on the similarities of the BOX A2R fingerprinting patterns, which were grouped into four major clusters (Figure 1). The 16S rRNA gene sequencing showed that the Enterococcus isolates belong to five different species, including E. faecalis (n = 17, 34.7%), E. faecium (n = 11, 22.4%), E. hirae (n = 11, 22.4%), E. durans (n = 9, 18.4%), and E. mundtii (n = 1, 2.0%) (Figure 2). Fisher's exact test showed that there was a significant difference (p < 0.05) between clusters and Enterococcus spp. proportion in each cluster. In cluster A, the proportion of E. hirae is significantly different to other clusters, especially clusters B (p < 0.001) and C (p < 0.001). In cluster B, the proportion of E. faecalis is significantly different (p < 0.001) from all other clusters. There were 11 genotypes of E. faecalis (n = 13), all of which clustered in group B and possessed gelE gene and were resistant to rifampin. In cluster C, the proportions of E. faecium (p = 0.006) and E. durans (p = 0.012) were significantly different from other clusters, and the post-hoc test showed that E. faecium and E. durans in cluster C were specifically different from cluster B (E. faecium: p = 0.019; E. durans: p = 0.012). Eleven genotypes comprised of E. faecium (n = 8), E. durans (n = 7), and E. faecalis (n = 1) were grouped in cluster C, where one strain of E. durans which was resistant to vancomycin and one strain of E. faecium which was resistant to ampicillin were grouped together in this cluster.
Cluster analysis of Enterococci isolates (n = 49) based on the similarity between BOX A2R-PCR fingerprints of each strains, cutoff threshold of 89%. The detailed information of assigned genotype, species, virulence genes, AR profiles, and date of isolation were shown next to the dendrogram.
Cluster analysis of Enterococci isolates (n = 49) based on the similarity between BOX A2R-PCR fingerprints of each strains, cutoff threshold of 89%. The detailed information of assigned genotype, species, virulence genes, AR profiles, and date of isolation were shown next to the dendrogram.
Relative abundance of virulence genes detected among 49 isolates of Enterococcus spp. identified. Four virulence genes of cylA, esp, asa1, and gelE were identified among E. durans (n = 9), E. faecalis (n = 17), E. faecium (n = 11), E. hirae (n = 11), and E. mundtii (n = 1). The numbers of each virulence gene detected in the specific Enterococcus spp. were shown on top of the bars.
Relative abundance of virulence genes detected among 49 isolates of Enterococcus spp. identified. Four virulence genes of cylA, esp, asa1, and gelE were identified among E. durans (n = 9), E. faecalis (n = 17), E. faecium (n = 11), E. hirae (n = 11), and E. mundtii (n = 1). The numbers of each virulence gene detected in the specific Enterococcus spp. were shown on top of the bars.
Genotypes 4, 6, 18, 19, 21, 24, and 28 were found to have more than one Enterococcus isolate grouped in the same genotype, with Genotype 21 which has the highest number of strains (E. faecium, n = 2; E. durans, n = 2) and the strains in this genotype were isolated on Day 7 after Hurricane Lane. Three of the strains in this genotype contained asa1 and gelE, but only one strain contained an esp gene. One E. faecium in this genotype was found to be resistant to ampicillin and MAR (Amp–Nit–Ery–Rif). Genotypes 4, 6, 24, and 28 contained two different Enterococcus species in each genotype, but both species have different virulence genes and AR gene structures. Although genotypes 18 and 19 contained the two strains of the same species that is E. faecalis, both strains in each genotype possess different characteristics of virulence genes and AR gene profiles. Genotype 23 has a single strain of E. durans which was found to be resistant to vancomycin and also to MAR (Van–Nit–Chl–Ery–Rif).
Detection and distribution of virulence genes
Four virulence genes (cylA, esp, asa1, and gelE) were detected in the Enterococcus species isolated from the HBL site (Figure 2). Most of the Enterococcus isolates contained the asa1 gene (E. durans, 3/9; E. faecalis, 7/17; E. faecium, 4/11; E. hirae, 3/11; E. mundtii, 1/1) and/or gelE gene (E. durans, 3/9; E. faecalis, 17/17; E. faecium, 5/11; E. hirae, 4/11; E. mundtii, 1/1). Interestingly, the cylA gene was only detected in E. faecalis (3/17) and the esp gene in E. durans (4/9) and E. faecium (4/11), respectively. Overall, there were 33.3% of E. durans, 41.2% of E. faecalis, 36.4% of E. faecium, and 27.3% of E hirae isolates possessing two or more virulence genes. Among the strains, three E. faecalis isolates possess cylA–asa1–gelE and one E. durans possess esp–asa1–gelE simultaneously. The rest of the species which possess more than one virulence genes were comprised of asa1–gelE that was distributed among the five species.
Antibiotic resistance
AR profiles of the Enterococcus isolates to eight different antibiotics are shown in Figure 3. E. durans, E. faecalis, and E. faecium exhibited AR to a maximum of six different kinds of antibiotic. The first, second, third, and fourth highest numbers of isolates resistant to rifampin (31/49 or 63.3%; E. faecalis = 16, E. faecium = 6, E. durans = 4, E. hirae = 4, E. mundtii = 1), erythromycin (20/49 or 40.8%; E. faecium = 9, E. durans = 7, E. faecalis = 3, E. hirae = 1), tetracycline (16/49 or 32.7%; E. hirae = 8; E. faecalis = 5; E. faecium = 2, E. durans = 1), and nitrofurantoin (15/49 or 30.6%; E. faecium = 8, E. durans = 3, E. hirae = 3, E. faecalis = 1) were enumerated. Only one E. durans isolate was detected to confer resistance to vancomycin, which, however, indicates significance due to VRE being an urgent threat of AR designed by the US CDC (CDC 2019). This E. durans isolate exhibited multiple resistance to five antibiotics (Van–Nit–Chl–Ery–Rif). Only one E. faecium was found to be resistant to ampicillin and had multiple resistance to four antibiotics (Amp–Nit–Ery–Rif). Two isolates were also shown to be resistant to five different antibiotics, including one E. faecalis isolate (Cip–Tet–Chl–Ery–Rif) and one E. faecium isolate (Cip–Tet–Nit–Ery–Rif). In total, 44.4% of E. durans, 52.9% of E. faecalis, 72.7% of E. faecium, 45.5% of E. hirae, and 100% of E. mundtii were found to be resistant to MAR (i.e. resistant to at least two antibiotics).
AR profile of 49 Enterococcus spp. to eight different kinds of antibiotic tested (ciprofloxacin, tetracycline, vancomycin, ampicillin, nitrofurantoin, chloramphenicol, erythromycin, and rifampin). The numbers of each Enterococcus spp. resistant to each antibiotic were shown by the number on top of the bars.
AR profile of 49 Enterococcus spp. to eight different kinds of antibiotic tested (ciprofloxacin, tetracycline, vancomycin, ampicillin, nitrofurantoin, chloramphenicol, erythromycin, and rifampin). The numbers of each Enterococcus spp. resistant to each antibiotic were shown by the number on top of the bars.
Co-occurrence frequency of virulence genes and AR
A co-occurrence pyramid based on the Jaccard index value (JIV) was constructed to illustrate the correlation of virulence genes and AR in the Enterococcus isolates (Figure 4). Most of the E. faecalis isolates possessed the gelE gene (JIV: 0.6) while resistant to rifampin (JIV: 0.5). E. hirae isolates have a high JIV of 0.4 to tetracycline that is shown in Figure 3 where this Enterococcus species has the highest number of isolates resistant to tetracycline. E. faecium isolates also have a high JIV to nitrofurantoin (0.4) and erythromycin (0.4) in which the highest number of isolates from this species were found to be resistant to nitrofurantoin (n = 8) and erythromycin (n = 9). Overall, there were high JIV numbers of the co-occurrence of the gelE gene (0.9) when the presence of rifampin resistance (0.5) and the asa1 gene (0.5) was detected simultaneously.
Matrix showing the co-occurrence frequency of virulence genes (n = 4) and the AR profile (n = 8) with the five Enterococcus spp. calculated by the JIV. A score near 1.0 indicates that the variables are perfectly correlated, whereas a score near 0 indicates that the variables are perfectly inversely correlated. The higher the number, the higher the frequency of co-occurrences.
Matrix showing the co-occurrence frequency of virulence genes (n = 4) and the AR profile (n = 8) with the five Enterococcus spp. calculated by the JIV. A score near 1.0 indicates that the variables are perfectly correlated, whereas a score near 0 indicates that the variables are perfectly inversely correlated. The higher the number, the higher the frequency of co-occurrences.
Temporal changes of pathogenic Enterococcus spp. after hurricane impact
A temporal change of Enterococcus spp. from the fifth day to the tenth day right after Hurricane Lane impact was observed (Figure 5). Overall, E. faecalis, E. faecium, and E. durans were commonly detected in the Enterococcus population after the impact of Hurricane Lane. E. hirae, on the other hand, was not detected from Day 5 to Day 8, but became more predominant starting on Day 9 (11.1%) and Day 10 (76.9%). The isolates detected on Day 11 were removed from this analysis because the sample number was below 5 (n = 2) and was considered as the lack of statistical significance.
Temporal changes of Enterococcus spp. after the hurricane impact. Sampling started on the fifth day after the heaviest impact of Hurricane Lane where the stream flow was recorded as the highest.
Temporal changes of Enterococcus spp. after the hurricane impact. Sampling started on the fifth day after the heaviest impact of Hurricane Lane where the stream flow was recorded as the highest.
DISCUSSION
E. faecalis and E. faecium are the two most common Enterococcus species causing approximately 87.7 and 8.6% enterococcal infections in clinical settings (Ruoff et al. 1990). Other Enterococcus species, including E. durans (Kenzaka et al. 2013), E. hirae (Bourafa et al. 2015), and E. mundtii (Higashide et al. 2005), typically contribute to much smaller percentage of enterococcal infections. In this study, 57.1% of the Enterococcus isolates obtained in the Hilo Bay water immediately after the Hurricane Lane impact were E. faecalis and E. faecium, and the remaining 42.9% were E. durans, E. hirae, and E. mundtii. A similar distribution of Enterococcus species were also isolated from fresh river water after the storm event, including E. faecalis (43%), E. faecium (18%), E. mundtii (14%), E. casseliflavus (12%), E. durans (6%), E. hirae (7%), and E. avium (2%) (Sidhu et al. 2014). Furthermore, E. faecalis, E. faecium, E. casseliflavus, and E. mundtii have been identified as the most commonly isolated enterococcal species from urban runoff and receiving water (Moore et al. 2008).
Stormwater has been shown to increase the number of enterococcal isolates in the runoff, and the sources of the Enterococcus spp. were related to fecal sources. A study done by Hou et al. (2006) showed that the discharged floodwaters after the impact of Hurricane Katrina had an elevated number of enterococci counts (170 CFU/100 mL) at stations near the mouth of a canal compared to stations 5 km from the mouth of the canal where the number of enterococci was 5 times lower. Sinigalliano et al. (2007) found that after the impacts of Hurricane Katrina and Rita, the number of enterococci was 380 CFU/100 mL at the middle of the main canal discharging water into a nearby lake. They also found that the source of enterococci was from within the city where the number of enterococci in the canal water increased from 93 to 1,400 CFU/100 mL before and after the pumping of the effluent water. Bae & Hou (2013) determined that the sources of enterococci in the Hurricane Katrina floodwater discharges were from fecal contamination by the sequencing results of typical fecal Enterococcus spp. isolated from the impacted water.
Enterococcus spp. are an important group of opportunistic pathogens capable of spreading AR at intra- and interspecies levels as demonstrated by the high prevalence of acquired AR (Da Silva et al. 2006). The dissemination of AR in environmental Enterococcus strains has increased concerns when AR gene transfer was reported in sewage water treatment plants (Tejedor Junco et al. 2001). The abundance of enterococcal isolates from different sources has been found to possess AR, including stormwater (Saingam et al. 2021), bay and ocean (Moore et al. 2008), municipal wastewater (Da Silva et al. 2006; Abriouel et al. 2008; Moore et al. 2008), vegetables and produce (Johnston & Jaykus 2004; Abriouel et al. 2008; Igbinosa & Beshiru 2019), and various other water bodies (Tejedor Junco et al. 2001; Abriouel et al. 2008; Macedo et al. 2011). Particularly, VRE is of concern because it is a leading cause of healthcare-associated infection that particularly affects critically ill and immunocompromised patients (Gouliouris et al. 2019). The detection of VRE is not limited to the clinical settings, but has been widely found in wastewater treatment plants (Beier et al. 2008; Varela et al. 2013; Rosenberg Goldstein et al. 2014; Oravcova et al. 2017; Gouliouris et al. 2019). In this study, an isolate of E. durans was found to be resistant to vancomycin and four other antibiotics. Vancomycin-resistant E. durans was not as common as E. faecalis and E. faecium, but it has been isolated from stool samples from patients (Cercenado et al. 1995) and sewage (Torres et al. 1994). E. durans were commonly found in human and animal fecal sources (Layton et al. 2010), which suggests that the source of the vancomycin-resistant E. durans isolated from Hilo Bay could be originated from non-environmental sources impacted by stormwater from Hurricane Lane.
While enterococci in the environment are generally considered to be benign, they can become opportunistic pathogens due to the possession or acquisition of virulent genes and antibiotic-resistant determinants through horizontal gene transfer (Ferguson et al. 2016). E. faecalis and E. faecium in human waste foster higher numbers of virulence genes compared to other sources (Ferguson et al. 2016). In this study, 75.5% of Enterococcus isolates were found to possess at least one virulence gene. Studies have shown that the most frequently detected virulent genes among environmental enterococcal species impacted by storm events were gelE (gelatinase) and asa1 (aggregation substance) (Macedo et al. 2011; Sidhu et al. 2014), with gelE commonly associated in E. faecalis (Creti et al. 2004; Sidhu et al. 2014; Ferguson et al. 2016). However, gelE can also be detected in clinical isolates of E. faecium and E. mundtii (Biswas et al. 2016; Shokoohizadeh et al. 2018). Our findings indicated that E. durans, E. faecalis, E. faecium, E. hirae, and E. mundtii isolated from hurricane-impacted water possessed both gelE and asa1. The asa1 gene was detected in various clinical enterococcal isolates (Biswas et al. 2016) and environmental isolates, except for E. hirae (Sidhu et al. 2014). The asa1 gene, which was carried by a sex-pheromone-containing plasmid, could spread among enterococci through conjugative transfer (Galli et al. 1990), and suggests the potential occurrence of horizontal gene transfer in the environment or when there are input of virulence genes from other sources through natural impacts.
The cytolysin gene cylA, which can be carried by either chromosome or plasmid elements, was reported primarily in E. faecalis isolates from clinical and food origins (Semedo et al. 2003; Sidhu et al. 2014), but rarely detected in environmental enterococcal isolates or only present in very low abundance (Semedo et al. 2003; Lanthier et al. 2011; Ahmad et al. 2014). Similarly, in this study, the cylA gene was only detected in E. faecalis isolates from the water environment (17.6%). The common clinically identified enterococcal surface protein, esp, was mostly found in E. faecalis and E. faecium (Willems et al. 2001; Biswas et al. 2016). However, in this study, the esp gene was not found in E. faecalis, but instead in E. faecium (36.4%) and E. durans (44.4%). A study revealed that esp genes were more frequently identified in E. faecalis (45%) than E. faecium (21%) among clinical isolates (Di Rosa et al. 2006). On the contrary, as much as 84% of clinical E. faecium were confirmed to possess the esp gene as compared to E. faecalis (33%) (Ferguson et al. 2016). A study done by Ferguson et al. (2016) found that the esp gene was not detected in any of the E. faecium isolated from beach water, but the genes are detected in 12% of E. faecalis (Ferguson et al. 2016). Since the esp genes were typically detected in the clinical samples, the E. faecium and E. durans isolated in this study could suggest the possibility of clinical input via excessive runoff as a result of hurricane impact.
CONCLUSIONS
In this study, we were able to isolate a sufficient number of putative pathogenic Enterococcus spp. from coastal water immediately impacted by hurricane events, especially those responsible for human enterococcal infections such as vancomycin-resistant E. durans. The Enterococcus spp. isolated were found to possess prevalent virulence genes and multiple AR profiles that are related to non-environmental sources, suggesting considerable input from urban runoff nearby and from diffused sources of pollution. Temporal changes of Enterococcus spp. days after the hurricane impact suggested the ability of these bacteria to survive in the environment for a sustained period of time. Further monitoring of the coastal water quality was needed as the presence of these opportunistic human pathogenic Enterococcus spp. could adversely impact the human community health.
ACKNOWLEDGEMENT
This material is based upon work supported by the National Science Foundation under Grant No. CBET-1855128.
DATA AVAILABILITY STATEMENT
All relevant data are available from GenBank database https://www.ncbi.nlm.nih.gov/genbank/ (accession no. MZ668452 - MZ668500).
REFERENCES
Author notes
Prakit Saingam and Doris Y. W. Di contributed equally to the manuscript.