Development of a simple, rapid multiplex PCR tool kit by using the 16S rRNA gene for the identification of faecal and non-faecal coliforms in drinking water Open Access

Microbial safety of drinking water is assessed by the establishment of the presence of the microbial indicators like non-coli and faecal coliform bacteria such as Escherichia coli (E. coli). Faecal contamination in water is monitored by the existence of E. coli and entails that presence of pathogenic viruses, protozoa, bacteria, etc. (Environment Agency 2002). Faecal contamination of water bodies may produce diarrheal illness and cause human mortality (World Health Organization 1993). In the developing world, the major sources of human illness are water related diseases. These diseases to which humans are vulnerable are caused by the drinking of untreated and contaminated water. E. coli O157:H7 is the most common type of Shiga toxin-producing Escherichia coli (STEC), but other types exist, including non O157; H7 strains include: O26:H11, O103:H2, O145:H28, and O111:H8, producing shiga toxin in the environmental waters, which are the major reservoirs of infectious E. coli (Bonetta et al. 2011; Bej et al. 1991a).

Global health issues have arisen due to waterborne diseases. Hence rapid identification of indicators of faecal contaminants, possibly of human pathogens, is very critical in environmental samples. The identification and enumeration of bacteria from water contributes to healthy public consumption of safe drinking water. Therefore, numerous methods for rapid detection and quantification of water quality indicators and waterborne pathogens have been established. Traditional approaches based on cultures are laborious and time-consuming. A rapid, simple protocol for the direct detection of microbes therefore needs to be established.

Compared to the conventional approach, multiplex polymerase chain reaction (PCR) is a valuable, quick and secure substitute for microbiological identification of drinking water organisms. For the concurrent amplification of target sequences in a single PCR, multiplex PCR (MPCR) uses one or more template and several sets of primers. For the differentiation and recognition of bacterial species, multiplex PCR uses several templates and special primers. When multiplex PCR was first applied it was used to effectively detect Saccharomonospora strains by means of the 16sRNS gene and was found to be capable of being species-specific, quick and simple. Detection of 16S rRNA genes through PCR amplification is a widely accepted method for the identification of bacteria. Sequence analysis of 16S rRNA can be used for taxonomic studies and bacterial species identification (Hassan et al. 2014; Veloo et al. 2016). In this study 16S rRNA sequences were obtained from 64 strains from 7 genera representing both total and faecal coliforms. These sequences were used to develop a multiplex PCR assay targeting the 16S rRNA gene. The practicality of using multiplex PCR investigation for quick detection of total coliform strains in contaminated water was examined rapidly. Three primers were previously designed to target variable regions (V2, V3, V9) in 16S rRNA for the rapid detection of total coliform bacteria in drinking water and shortened the assay time significantly compared to traditional culture methods (Chakravorty et al. 2007).

Isolates used in the present study were procured from Microbial Type Culture Collection (MTCC), Chandigarh, India and are tabulated in Table 1.

Mahabubnagar is one of the drought districts in Telangana state, India. The geographical area of the district is 18,432 Sq. kms, consists of 64 mandals (sub-districts) with 1,549 villages. It lies between north latitudes 15° 55′ 00″ and 17° 20′ 00″ and east longitudes 77° 15′ 00″ and 79° 15′ 00″. It is bounded to the north by Rangareddy and Nalgonda districts, to the east by Guntur and Nalgonda districts, to the south by the Krishna and Tungabhadra rivers and to the west by Raichur and Gulbarga districts of Karnataka state.

The main sources of potable water in the township of Mahabubnagar are the Ramanpadu balancing reservoir and Koilsagar reservoir (Figure 5). Ramanpadu is a balancing reservoir under the left-hand canal of the jurala project. A 64 km Ramanpadu drinking water pipeline project started in 1999 for the town of Mahabubnagar to reduce its drinking water problem. With a population of more than 2 lakhs (200,000) the town requires 13 million liters of water every day, whereas the Ramanpadu water scheme was designed to lift 18 million liters of Krishna water every day.

Koilsagar dam is located at Koilsagar village of the Deverakadra mandal in Mahabubnagar district. Koilsagar dam is built on a minor tributary of the Krishna river. The proposal to construct a dam was originally put forward by the British rulers to collect and store the excess water in a catchment area of the Krishna river for irrigation purposes.

Two separate lakes, i.e., the Ramanpadu and Koilsagar, in Mahabubnagar district, Telangana state, India were the study areas and a total of 64 water samples were collected in 1 L sterile bottles. Details of the source of water, time and date of sample collection were labeled on sample bottles (Volokhov et al. 2007; Shiva Shanker et al. 2019).

Bacterial cells were collected by filtering 100 mL of water sample using vacuum manifold through a polycarbonate filter with 47 mm diameter, pore size 0.2 μm. To avoid possible contamination, it was performed within the laminar flow.

Genomic DNA was extracted from the trapped bacteria as described previously (Pindi et al. 2013) using commercial kits (DNeasy Plant Mini Kit (cat. nos. 69104 and 69106)).

The PCR primers used in the study are shown in Table 2. The species-specific primer pairs targeting 16S rRNA gene variable regions were designed by using Primer 3 software. Multiple sequence alignment was done by CLC sequence viewer 8.0 software to identify the high variable region to target the primer within the gene. The designed primers were tested insilico by software available from www.bioinformatics.org. Oligonucleotide primers were procured from Eurofins MWG Operon (Bangalore, India).

The MJ Mini Personal Gradient Thermal Cycler from Bio-Rad was used for all PCR analysis. The uniplex PCR reaction mixture consisted of 25 μl of KAPA SYBR^® FAST qPCR Master Mix (2X) Kit (Sigma Aldrich, India) or subsidiary in India. 1 μl 5 pmol of forward primer 194F, 1 μl 5 pmol of reverse primer 1436R, 1 μl 30–50 ng DNA template and finally 12 μl PCR grade water for a total volume of 50 μl were used for the amplification reaction.

The PCR gradient, with annealing temperature varying from 51 to 55 °C, was used to optimize the best annealing temperature (Fricker & Fricker 1994). The PCR loop, denaturation, annealing and extension were carried out for 5 min at 95 °C, 54.4 °C for 30 s, and 72 °C for 1 min, respectively. The final extension was carried out for 2 minutes at 72 °C. The PCR product was examined in 1.5% agarose gel Bio Rad (Jyothikumar et al. 2003). The molecular weight marker of the 100 bp (NEB # B7025, Biolabs) DNA ladder (Figure 1) was used to classify the product bands. Using a PCR purification kit (Bangalore Genei, India), the amplified product was purified and sequenced at Bioserve Biotechnologies Pvt, Hyderabad. Further, the 16S rRNA gene sequence was deposited in the National Center for Biotechnology Information (NCBI) GenBank and accession numbers were allocated in Table 3.

The multiplex PCR reaction mixture consisting of 25 μl of KAPA SYBR^® FAST qPCR Master Mix (2X) Kit (Sigma Aldrich or subsidiary in India). The PCR conditions are similar for 1 μl 5 pmol of forward primers 194F, 474F and 1 μl 5 pmol of reverse primer 1436R, 1 μl 30–50 ng DNA template and finally 12 μl PCR grade water for a total volume of 50 μl used for the amplification reaction (Figure 2). PCR product was examined by gel electrophoresis through a 1.5% agarose gel (Jyothikumar et al. 2003). Identification of the product bands was established by molecular weight marker of the 100 bp (NEB # B7025, Biolabs) DNA ladder. The fitness of the primer pairs for the multiplex PCR system was checked using MultiPLX (Kaplinski et al. 2005) and FastPCR.

Total coliform screening methods are mainly targeted to the 16S rRNA markers due to the presence of species-specific variable regions in total coliforms; V2 and V3 were most suitable for distinguishing all bacterial species to the genus level except for closely related enterobacteriaceae (Chakravorty et al. 2007). The efficiency of new designed primer pairs were tested separately to amplify their targets by uniplex PCR. Each uniplex PCR gave the expected one amplicon for non-faecal coliform and faecal coliform species of standard cultures. A 1,285 bp of PCR product was obtained when the primer set 194F & 1436R was used for all 7 genera of the total coliform group (Figure 1). No PCR fragment was observed for non-coliforms and water control (Figure 2) indicating that running a larger challenge set and actually calculating sensitivity and specificity using standardized methods was required. Moreover, the intensity of the DNA band increased correspondingly when the 194F & 1436R primer set was used, increasing the amount of specific template in the samples.

PCR amplification with two sets of oligonucleotide primers (reverse primers common to both) resulted only in the presence of their respective DNA template in a detectable molecular weight fragment (1,285 bp for non-faecal coliform bacteria and 1,009 bp for E. coli) of a predicted molecular weight. These results indicate each of the three primers, namely 194F is a degenerated primer that targets the 7 genera of total coliforms in combination with 1436R and in combination with 474F targets for E. coli. However, when both bands were observed (1,285 bp and 1,009 bp) it indicated that water was contaminated with faecal coliform bacteria, E. coli (Figure 2), thus by using these primers (194F, 474F, 1436R) we can differentiate the water contaminated with faecal or non-faecal coliform bacteria. However, no PCR products were observed or other bacteria found in the negative control (Figure 3) that confirmed the specificity of the assay.

Sixty-four different groups of bacterial 16S rRNA sequence and multiple sequence alignment were performed using CLC sequence viewer 8.0 software to identify the highly variable region to target primers from the V2, V3 and V9 variables regions of 16S rRNA. V2 and V9 conserved regions specific for the seven genera and these were grouped as total coliforms showing nucleotide sequence identity (Figures 4 and 5). The V3 region is specific for E. coli. Primer 474 has 11 nucleotide variations from the six genera of total coliforms, including: Citrobacter sp., Enterobacter sp., Klebsiella sp., Serratia sp., Yersinia sp. and Salmonella sp.

The samples from different places were collected and validation was conducted of the most probable number (MPN) positive sample for field level conformation and later checked for coliform contamination in the two major Ramanpadu and Koilsagar reservoirs, and total number of bacteria was estimated. The average total bacterial numbers of each sample ranged between 9.5 and 107 CFU/mL, and 8.0 to 105 CFU/mL respectively in Ramanpadu and Koilsagar (Figure 2). The microbial diversity of water samples in Ramanpadu and Koilsagar are represented in Figures 6 and 7.

The 64 isolates of faecal and non-faecal coliforms from drinking water were subjected to multiplex PCR in the present study. Each of the E. coli isolates showed positive amplification of 16S rRNA fragments. These findings showed that the multiplex PCR method, based on the co-amplification of three primary DNA target sequences derived from 16S rRNA genes, allowed the unmistakable identification from drinking water for both faecal and non-faecal coliforms.

In order to validate the designed primers, standard cultures of coliforms (7 genera) were procured from MTCC (Table 4). Later the 1,285 bp gene sequence was confirmed and deposited with NCBI with accession No. MW029934-MW029940. Meanwhile the non-coliforms by 16S rRNA and biochemical, phenotypic tests (Pindi et al. 2013) were found to be correctly identified with the newly designed primers. Thus accuracy of the primers for coliforms in the present study was found to be 99%.

Water-borne disease outbreaks continue to be a significant concern for global public health providers, claiming millions of lives each year (Jones et al. 2009; Cabral 2010; Gelting et al. 2011; Breathnach et al. 2012; Cheun et al. 2013; Pitkanen 2013; Walser et al. 2014). WHO reported in 2008 that 2.5 million people died from diarrheal disease, and that in 2011, the number of cases of cholera rose by 85% compared to 2010 (WHO 2013). The coliform group includes the members of genera Escherichia sp., Klebsiella sp., Citrobacter sp., and Enterobacter sp. and is termed as total coliforms regularly found in water. Faecal indicator bacteria are being used to predict the presence of pathogens in water intended for human consumption and bathing water by the European Commission (EC) (EC 1998, 2006, 2009), and the United States Environmental Protection Agency (USEPA) (USEPA 2002; Wade et al. 2003; Boehm et al. 2009). The conventional methods for the detection of pathogens takes a long processing period and is labor-intensive (Velusamy et al. 2010; Li et al. 2016; Pan et al. 2018). Newer techniques such as PCR are more adaptive, high-throughput, and many bacteria can be identified in a shorter time and with less reagent consumption without repeated steps with multiplex PCR technology (De Freitas et al. 2010; Sánchez-Parra et al. 2019). Multiplex PCR has been successfully applied to rapidly detect various pathogens from environmental waters (Bej et al. 1991b, 1991c; Nguyen et al. 2016; Varma-Basil et al. 2004). In this study, multiplex PCR was used for the identification of total coliform and non-faecal coliform bacteria in drinking water and the specificity of PCR was calculated for 64 bacterial strains. Only in the presence of their respective DNA templates did the findings of PCR yield a detectable DNA fragment of predicted molecular weight.

The input parameters for the Primer 3 design software include primer length of about 18–24 bp, G/C content of 40–60%, start/end with 1–2 G/C pairs, primer melting point (Tm) of 50–60°C, F and R primer Tm value within 5 °C, and that primers should not have complementary regions. One of the most critical parameters is the temperature for annealing. Since many individual loci could be precisely amplified at 56–60 °C, in the present study the annealing temperature for multiplex PCR is 56–60 °C (Untergasser et al. 2012; Maheux et al. 2014; Abada et al. 2019; Xie et al. 2020).

In the present study, the 16S rRNA gene was used to detect both faecal coliform and non-faecal coliforms. Bacterial 16S rRNA consists of 9 hypervariable areas, which display considerable diversity of sequences among various bacterial species and can be used to classify species (Van de Peer et al. 1996). The hyper-variable regions of most bacteria are flanked with conserved stretches allowing the PCR to amplify target sequences with a universal primer (McCabe et al. 1999; Lu et al. 2000; Baker et al. 2003; Munson et al. 2004). 16S rRNA hypervariable area sequences have been found in various studies that classify a single bacterial species or distinguish between a small number of different species or genera (Choi et al. 1996; Kataoka et al. 1997; Marchesi et al. 1998; Lu et al. 2000; Bertilsson et al. 2002; Rothman et al. 2002; Yang et al. 2002; Becker et al. 2004; Clarridge 2004; Maynard et al. 2005).

In the current study two primers of 194F and 1436R were designed which amplifies the 1,285 bp of PCR product specific for the V2 hypervariable region of the 16S rRNA gene and which is specific for the detection of coliform genera including Yersinia enteroclitica, Citrobacter freundii, Enterobacter aerogenes, Salmonella enteric, Serratia marcescens and Klebsiella pneumonia. 474F and 1436R primers are specific for V3 hypervariable region specific for Escherichia coli and amplifies the 1,009 bp of PCR product. (Chakravorty et al. 2007; Merkel et al. 2019; Sune et al. 2020). This method is a more sensitive method than traditional culture-based methods because it depends on stable genetic parameters and detects accurate results within 3 hrs compared to the classic method for regular monitoring that takes 3–4 days. The presence of total coliforms from the environmental samples from Ramanpadu and Koilsagar reservoirs shows the specific amplicon size. The results are similar to Shiva Shanker et al. (2019).

Total coliform organisms show high genetic diversity hence they can be diversified by using molecular tools for their detection in drinking water. Among various methods available, the present study utilizes the 16S rRNA gene which proved to be a simple and sensitive technique. The 194F, 474F and 1436R primer set is specific for amplification of V2, V3 and V9 hyper variable regions of the 16S rRNA gene. In multiplex PCR we are able to differentiate total coliforms from that of faecal coliform by specific amplicons including 1,285 bp of amplicon specific for 6 genera of non-faecal coliform and the 1,009 bp of amplicon specific for faecal coliform ie. E. coli. If the drinking water was contaminated with both faecal and non-faecal coliforms then we get two amplicons of 1,285 bps and 1,009 bps. This rapid method of molecular detection of total coliforms is able to differentiate both faecal and nonfaecal coliforms in the water. We can get accurate results within 3 hrs which compares well with the classic method of MPN (3–4 days).

All relevant data are included in the paper or its Supplementary Information.