Molecular Characterization, Intra-Species Diversity and Abundance of Freshwater Plesiomonas shigelloides Isolates

Molecular signatures of Plesiomonas shigelloides strain specific to pathogenic and nonpathogenic variants are not well established till present. There is a need for intra-species barcoding of P. shigelloides to aid infection control. This study aims at characterizing and assessing intra-species diversity and abundance of P. shigelloides isolated from three freshwaters in the Eastern Cape Province. The study used a Plesiomonas-specific PCR to characterize the isolates. Intra-species (dis)similarities were assessed using ERIC-PCR and (GTG)5-PCR techniques. The DNA fingerprints produced were electrophoresed, digitized, and documented via computer-assisted pattern analysis. The fingerprints were analyzed using neighbor-joining clustering (NJC) based on Euclidean similarity index. Results revealed 80%, 83.64%, and 80% of the water samples from Tyhume, Kat, and Kubusie rivers, respectively, positive for P. shigelloides isolation. The prevalence of P. shigelloides from sites ranged from 13.5% to 88.9%. NJC delineated 48 isolates to 8 clades (ERIC-fingerprints) and 34 isolates into 7 clades ((GTG)5-fingerprints). The relative abundance of unique strains ranged from 6.3% to 22.9% via the two methods. Both fingerprinting approaches have strain-differentiating potential for P. shigelloides, however ERIC-PCR possessed higher resolution (D = 37.46) advantage over (GTG)5-PCR (D = 29.64). In conclusion, the study achieved intra-species diversity and abundance of P. shigelloides from aquatic milieu and provide further opportunity for intra-species-specific barcoding.


Introduction
Molecular signatures of Plesiomonas strains specific to virulent and nonvirulent strains still defy understanding and remain a diagnostic challenge. With a clinical obligation to differentiate pathogenic strains from nonpathogenic variants, there is a need to understand intra-species/strain diagnostic features that would aid discrimination of a pathogenic strain of P. shigelloides from nonpathogenic counterparts. Generally, pathogenic strains of a bacterium species have unique genetic signature(s)/trait(s) that are not present in its nonpathogenic strains, thus, these signature(s)/trait(s) form the basis for differentiating virulent strains from avirulent strains. Presently, strain-typing of P. shigelloides irrespective of pathogenic potential relies on somatic (O) and flagella (H) antigens serotyping [1]. Improved development of P. shigelloides O/H antigens serotyping was advanced by Aldova [2] and Aldova and Schubert [3]. Unfortunately, a good number of P. shigelloides strains are not serotypeable [4]. Aside from the inability to serotype some strains, P. shigelloides also cross-react P. shigelloides is a single-species genus in the family Enterobacteriaceae [23]. It is well known to cause infections such as travelers' diarrhea, gastroenteritis, to severe extraintestinal infections [9,[24][25][26][27]. Also, some foodborne and waterborne outbreaks have been solely credited to P. shigelloides with sound microbiological and epidemiological validation [28,29]). Particularly, the incidence of Plesiomonas travelers' diarrhea increased from 23.2% to 77.8% between 1987 and 1999 at Kansai Airport, Japan [9]. These and many more attest to the need to identify virulence signatures in P. shigelloides and the need for virulent strains' diagnostic features. Strain-specific or virulent strain-specific diagnostic features would give room for discriminating its virulent strains from avirulent variants. Presently, no suitable phenotypic or molecular methods have been described for P. shigelloides virulent strain diagnosis.
From the foregoing, there is a need to understand and identify diagnostic traits of virulent and avirulent strains of P. shigelloides. However, the present study aims at characterizing and assessing intra-species heterogeneity and abundance of P. shigelloides isolated from aquatic environments in the Eastern Cape Province using molecular methods. This was aimed at enabling grouping of the strains to facilitate comparative studies of different groups to uncover the groups' diagnostic traits. This approach will hopefully assist in extensive studies relevant for determining strain-specific molecular signatures of P. shigelloides that may provide insights for future endeavors in specific tagging of virulent (pathogenic) strains from avirulent (nonpathogenic) variants. For, we hypnotized that virulent strains of P. shigelloides have diagnostic trait(s) different from avirulent ones as it is common in other pathogens.

Materials and Methods
2.1. Cultural Isolation of P. shigelloides from River Water P. shigelloides strains were isolated from 165 river water samples collected from freshwater resources at various points where human activities were prominent (see detailed sampling point descriptions in Table 1) across three popular rivers, viz. Tyhume, Kat, and Kubusie, in the Eastern Cape, South Africa. Sampling was done monthly and sequentially from the same sampling locations throughout February to December 2017. Water samples were collected aseptically in 1 L sterile glass bottle. Standard serial dilution of the water samples was carried out according to the standard protocol of APHA (American Public Health Association) [30]. Then, 100 mL aliquots of the respective dilutions were filtered through a 0.45µ millipore filter (Ø 47 mm) [30]. Each membrane filter, according to the dilution, was plated aseptically on a pre-labeled dried plate of inositol brilliant green bile agar (IBGBA) (HiMedia Laboratories, Mumbai-400086, India) using sterile forceps. After a period of 24 h incubation at 39 • C, pink colonies on the plates were counted and recorded as presumptive P. shigelloides isolates. Some randomly selected pink colonies were further streaked on a fresh IBGBA to purify them, subsequently grown on nutrient agar, and assayed for oxidase enzyme production using oxidase strips [31]. Oxidase-positive isolates were stored on glycerol stocks (−80 • C) for further studies.

DNA Extraction and Molecular Characterization of P. shigelloides
The purified oxidase-positive isolates were re-streaked onto fresh nutrient agar plates and cultured overnight at 37 • C. Total DNA of the 24 h culture was extracted by direct boiling procedure [32]. Two to three single colonies of the 24 h culture were picked and reconstituted in 200 µL sterile distilled water by vortex using a vortex mixer (DiGiSystem Laboratory instruments INC, Taiwan, Republic of China (ROC). The reconstituted cells were washed repeatedly with three changes of sterile DH 2 O at 15,000 rpm/2 min. The final cell pellet was re-suspended in nuclease-free water and boiled (100 • C/10 min) using Dri-Block ® DB-3D (Bibby Scientific LTD, Staffordshire, UK). The boiled cells suspension was then centrifuged at 15,000 rpm/10 min in a micro-centrifuge (HEMLE Labrtechnik GmbH, Germany) to separate cell debris. The supernatant (DNA) was collected in a sterile Eppendorf tube and stored in a freezer (−20 • C) until further use.
All fingerprint images were digitized for a computer-assisted pattern analysis using GelJ version 2.0 software [34]. The DNA fingerprints occurrence matrices and molecular weight of ERIC2-PCR and (GTG) 5 -PCR bands were generated from their gel images using the unweighted pair group arithmetic mean algorithm at 1.0% tolerance level for quality control.

Assessment of Intra-Species/Strain Diversity of P. shigelloides Isolates
The band occurrence matrices of (GTG) 5 -PCR and ERIC-PCR fingerprints generated by the GelJ 2.0 software were imported into PAleontological Statistics Version 3.23 (PAST3.23) [35] for diversity studies. The absence/presence of a fingerprint across the isolates formed the basis for strain homogeneity or heterogeneity (associations) assessment. Firstly, dendrograms of the two matrices were created by neighbor-joining (NJ) [36] using a Euclidean similarity index (Equation (1)).
Here, NJ defines the distance between any strain pair i and j as d ij that minimizes Q criterion (7.2).
where r is the current number of bands (representing strains) and the sums run on the band (strain) set. Secondly, the abundance of strains that form a clade was computed. Then, the study compared the resolution power of each fingerprinting technique using the Shannon's index (H) and Simpson's index based on the number of clades from the neighbor-joining clustering (NJC).

Results
Fifty-five samples were processed for each of the three sampled rivers during the eleven months Seven hundred and forty-eight randomly selected presumptive colonies of P. shigelloides were oxidase-positive ( Table 2). Out of these, two hundred and eleven isolates (n = 211, 28.21%) were identified as P. shigelloides by P. shigelloides-specific 23S rRNA polymerase chain reaction. These 211 P. shigelloides isolates yielded a band of 284 bp that confirmed their identity as P. shigelloides [4]. The sites 2TY and 5KT contributed 1.42% (lowest) and 18.01% (highest) of the total P. shigelloides confirmed from the rivers, respectively. A representative P. shigelloides-specific 23S rRNA gel image is shown in Figure 2.  Seven hundred and forty-eight randomly selected presumptive colonies of P. shigelloides were oxidase-positive ( Table 2). Out of these, two hundred and eleven isolates (n = 211, 28.21%) were identified as P. shigelloides by P. shigelloides-specific 23S rRNA polymerase chain reaction. These 211 P. shigelloides isolates yielded a band of 284 bp that confirmed their identity as P. shigelloides [4]. The sites 2TY and 5KT contributed 1.42% (lowest) and 18.01% (highest) of the total P. shigelloides confirmed from the rivers, respectively. A representative P. shigelloides-specific 23S rRNA gel image is shown in Figure 2.  A representative ERIC-PCR fingerprint image is presented in Figure 3. The ERIC-PCR banding patterns of isolates ranged from 0 to 10 bands. Molecular weight of bands also varied from 380 to 5665 bp (Figure 3). Certain isolates did not produce any band and appeared not typeable by the ERIC-PCR. The ERIC-PCR fingerprints dendrogram constructed by NJ using a Euclidean similarity index is shown in Figure 4. All the isolates clustered together. However, eight clades of strains were observed. The clades were as follows, numbering from origin (0)   A representative ERIC-PCR fingerprint image is presented in Figure 3. The ERIC-PCR banding patterns of isolates ranged from 0 to 10 bands. Molecular weight of bands also varied from 380 to 5665 bp (Figure 3). Certain isolates did not produce any band and appeared not typeable by the ERIC-PCR. The ERIC-PCR fingerprints dendrogram constructed by NJ using a Euclidean similarity index is shown in Figure 4. All the isolates clustered together. However, eight clades of strains were observed. The clades were as follows, numbering from origin (0)    A typical digitized (GTG)5-PCR fingerprint image is presented in Figure 5. (GTG)5 fingerprints produced band patterns that ranged from 4 to 14. The (GTG)5 fingerprint bands ranged in size from 147.76 to 5304.98 bp. All the isolates produced (GTG)5 fingerprint bands and clustered together. The NJ clustering of (GTG)5 fingerprint resulted in seven clades. The composition of each clade, numbering from origin (0) along the horizontal axis to 35 scale, was as listed ( Figure 6)    A typical digitized (GTG)5-PCR fingerprint image is presented in Figure 5. (GTG)5 fingerprints produced band patterns that ranged from 4 to 14. The (GTG)5 fingerprint bands ranged in size from 147.76 to 5304.98 bp. All the isolates produced (GTG)5 fingerprint bands and clustered together. The NJ clustering of (GTG)5 fingerprint resulted in seven clades. The composition of each clade, numbering from origin (0) along the horizontal axis to 35 scale, was as listed ( Figure 6)  A typical digitized (GTG)5-PCR fingerprint image is presented in Figure 5. (GTG)5 fingerprints produced band patterns that ranged from 4 to 14. The (GTG)5 fingerprint bands ranged in size from 147.76 to 5304.98 bp. All the isolates produced (GTG)5 fingerprint bands and clustered together. The NJ clustering of (GTG)5 fingerprint resulted in seven clades. The composition of each clade, numbering from origin (0) along the horizontal axis to 35 scale, was as listed ( Figure 6)

Discussion
High detection (positive) rates of P. shigelloides were observed in the sampled waters in this present study. The P. shigelloides cultural prevalence rates of 80% (44/55), 83.64% (46/55), and 80% (44/55) observed in river water from Tyhume, Kat, and Kubusie, respectively, were higher than those in earlier reports from river water in literature.
The observed high recovery rate compared with previous studies in part could be attributed to a combined use of filtration technique and Inositol Brilliant Green Bile Agar for the isolation of P. shigelloides in this study. Most previous studies used some other media and culture techniques [37,38] that allowed P. shigelloides to compete with a host of other bacteria, which, in most cases, have higher growth advantages in the media compared with P. shigelloides. Some rates of positive detection of P. shigelloides from river, well, or pond water in other works include 40.7% [37], 16.67% (4/24) in Rio de Janeiro City in Brazil [38], 0.6% (well) and 7.4% (pond) in Zaria, Nigeria [39], 12.8% in Japan [40], and 13.3% in Dhaka, Bangladesh [41].
The overall P. shigelloides yield of 81.21% (134/165) from river water observed in this study could have arisen from probably organic and inorganic contaminants in the rivers that support the growth of P. shigelloides. For instance, herbicides, fertilizers, and pesticide input into the rivers occurs frequently at 3TY, 5TY, 2KT, 4KT, 3KB, and 5KB. More so, anthropogenic resuspension of the riverbeds, as frequently observed in the area during animal watering in the sites, might increase the levels of P. shigelloides in the overlying water because sediment resuspensions release P. shigelloides trapped in riverbeds, biofilms, and other matrices into the water column. P. shigelloides has been isolated from pond sediment, hydrophytes, and phytoplankton [41]. Some authors have reported isolation rates of P. shigelloides from freshwater sediment, hydrophytes, and phytoplankton as 29.2%, 20.8%, and 18.3%, respectively [41]. Islam et al. [41] noted that matrices associated with pond such as soil, sediment, phytoplankton and hydrophytes had 62.5%, 41.7%, and 33.3% P. shigelloides positive isolation rate in their study.

Discussion
High detection (positive) rates of P. shigelloides were observed in the sampled waters in this present study. The P. shigelloides cultural prevalence rates of 80% (44/55), 83.64% (46/55), and 80% (44/55) observed in river water from Tyhume, Kat, and Kubusie, respectively, were higher than those in earlier reports from river water in literature.
The observed high recovery rate compared with previous studies in part could be attributed to a combined use of filtration technique and Inositol Brilliant Green Bile Agar for the isolation of P. shigelloides in this study. Most previous studies used some other media and culture techniques [37,38] that allowed P. shigelloides to compete with a host of other bacteria, which, in most cases, have higher growth advantages in the media compared with P. shigelloides. Some rates of positive detection of P. shigelloides from river, well, or pond water in other works include 40.7% [37], 16.67% (4/24) in Rio de Janeiro City in Brazil [38], 0.6% (well) and 7.4% (pond) in Zaria, Nigeria [39], 12.8% in Japan [40], and 13.3% in Dhaka, Bangladesh [41].
The overall P. shigelloides yield of 81.21% (134/165) from river water observed in this study could have arisen from probably organic and inorganic contaminants in the rivers that support the growth of P. shigelloides. For instance, herbicides, fertilizers, and pesticide input into the rivers occurs frequently at 3TY, 5TY, 2KT, 4KT, 3KB, and 5KB. More so, anthropogenic resuspension of the riverbeds, as frequently observed in the area during animal watering in the sites, might increase the levels of P. shigelloides in the overlying water because sediment resuspensions release P. shigelloides trapped in riverbeds, biofilms, and other matrices into the water column. P. shigelloides has been isolated from pond sediment, hydrophytes, and phytoplankton [41]. Some authors have reported isolation rates of P. shigelloides from freshwater sediment, hydrophytes, and phytoplankton as 29.2%, 20.8%, and 18.3%, respectively [41]. Islam et al. [41] noted that matrices associated with pond such as soil, sediment, phytoplankton and hydrophytes had 62.5%, 41.7%, and 33.3% P. shigelloides positive isolation rate in their study.
Another possible explanation for the high recovery rate of P. shigelloides from the sampled rivers without enrichment in this study could be due to favorable water temperature. The average temperature of the rivers ranged from 4.7-25.8 • C during the sampling period. The relative high water temperatures, known to be favorable for P. shigelloides multiplication, may have resulted in the higher rate of detection. However, isolation of P. shigelloides from freshwater water samples in the temperate and colder regions of the world, such as Czech Republic [48], Hungary [37]; the Netherlands [49], Slovakia [50], subpolar region of Sweden [47,50], and Serbia [37], has been reported.
Also, insanitary activities such as in-stream flow of domestic wastewater (3KB), poultry wastewaters (2KB, 5KB), slaughterhouse wastewaters, piggery wastewater (2KB, 5KB), livestock manure and litters (throughout the sites except 1TY), wastewater treatment plant effluents (1KT and 5KB), leachates from manhole and community dumpsite (5KT, 5TY), and fertilizers applications (2KT, 5KB, 5TY) offer conditions that could have led to direct input and proliferation of microorganisms, including P. shigelloides, along the sampled waters. All sampled sites are livestock watering sites coupled with several other uses with the exception of 1TY (a swimming/recreational hotspot) [51].
The computer-assisted analysis of the ERIC-PCR and (GTG)5-PCR fingerprints of the P. shigelloides clustered the isolates into eight and seven clades, respectively, thus suggesting that the P. shigelloides from the freshwater resources possessed intra-species or strain diversity. The clustering of strains from different sampling sites together is suggestive of an evolutionary relationship. This is in agreement with [8], who reported that clustered strains originated from different matrices (human and animal sources) or geographical location depicts clonal association. Uniqueness of P. shigelloides strains from diarrheic travelers among the Japanese was also demonstrated by Shigematsu et al. [9] using PFGE and DNA macrorestriction digests. Identical DNA-based profile of two pairs of P. shigelloides strains from human and animal origin has also been reported [8]. While González-Rey et al. [8] used DNA-based techniques for comparative delineation of a population of single/the same serovar strains, our study did not consider serotypes of the isolates since it has been previously reported that some strains are not serotypeable. Also, subclade (strain clade) diversity was observed in some clades. This explains potential within-clade strain dissimilarity. Notable in this study is the inability of ERIC-PCR to type some strains. Other authors have reported inability of ERIC-PCR to type certain E. coli strains [52,53]. In the study of Prabhu et al. [52], 13 out of 40 E. coli isolates were not ERIC-PCR typeable, and Ramazanzadeh et al. [54] observed 25 of 230 E. coli isolates not typeable by ERIC-PCR. The González-Rey et al. [8] study involved strains that belonged to the same serovar and this might be accounted for the differences observed in our study that involved strains from undifferentiated serovars. (GTG)5-PCR yielded bands for all the strains and thus allowed the differentiation of all P. shigelloides strains more efficiently compared with ERIC-PCR.
The relative abundance of the isolates that made up each clade of strain varied significantly. This connotes differences in the occurrence of P. shigelloides strains in nature. Heterogeneity in P. shigelloides has been reported by many studies, even at the same serovar level [8,9,37]. However, establishing unique identities that discriminate pathogenic strains from nonpathogenic strains has been a challenge. A combination and integrated fingerprinting techniques will be needed to fully identify the intra-species signatures of P. shigelloides strains. Besides, strain-specific signatures are now being sought for P. shigelloides delineation [54]. Although the ability to cause diarrheal illness has been hypothesized to be universal among P. shigelloides' strains from DNA-based fingerprints [9], the resolution of such procedures is not sufficient to support the assumption.

Conclusions
This study reports intra-species genetic diversity of P. shigelloides isolated from freshwaters in the Eastern Cape province, South Africa. The high detection rate of P. shigelloides was similar in the three sampled waters (Tyhume, Kat, and Kubusie). The overall P. shigelloides' positive isolation rate observed in the study could be attributed to the high level of pollution along the sampled rivers' courses. Also, abundance and occurrences of P. shigelloides strains vary in the aquatic milieu. Both fingerprinting approaches have intra-species/strain characterizing potential for P. shigelloides' preliminary grouping purposes. However, ERIC-PCR possessed potentially superior resolution merits over (GTG)5-PCR in P. shigelloides intra-species typing.