1. Introduction
Water is a vital part of human life, through drinking, washing, bathing to use in injectables. Water quality testing is thus a prerequisite for protecting human health. Unfortunately, many people worldwide do not have access to clean and safe drinking water, which results in waterborne bacterial infections with many fatalities associated [
1].
At present, routine testing of water relies on the detection of indicator organisms as the only accepted means for assessing the sanitary quality of water supplies. Indicator organisms, such as
E. coli, are used as a sign of faecal contamination of water rather than having to test for a myriad of specific pathogens. Yet, current culture-based methods for faecal organisms have several drawbacks as these indicators may reproduce in water and the culture results are retrospective, taking at least 24–48 h [
2,
3,
4]. Moreover, although they indicate the presence of faecal contamination, other pathogens that are not necessarily of faecal nature, may go undetected [
5]. In addition, the recent review of the ISO 9308 redefined its scope to cleaner waters (mostly with less than 100 colonies on chromogenic coliform agar) due to the background growth of less clean waters, which can interfere with the reliable enumeration of E. coli and coliform bacteria [
6].
Although access to clean drinking water may be the norm in developed countries, in many developing countries people struggle to obtain access to safe water. According to the World Health Organisation (WHO) [
7], 2.5 billion people have no access to improved sanitation and more than 1.5 million children die each year from diarrheal diseases. Additionally, the mortality of water-associated diseases exceeds 5 million people per year.
For Potable water, the WHO recommend a threshold of 0 CFU (Colony Forming Unit)/100 mL of
E. coli, however, the WHO understand that achieving 0 CFU/100 mL
E. coli may be challenging in many developing countries and rural areas. Therefore, the WHO assigned a health risk score for
E. coli in drinking water: 0 CFU/100 mL (conformity), 1–10 CFU/100 mL (low risk), 10–100 CFU/100 mL (intermediate risk), 100–1000 CFU/100 mL(high risk), and >1000 CFU/100 mL (very high risk) [
7,
8].
A rapid and simple test of bacterial water quality should be a very useful tool, especially in disaster situations, such as floods or hurricanes, and in instances of treatment system failure. Several investigators [
9,
10,
11,
12,
13] have suggested that the
Limulus amebocyte lysate (LAL) assay for endotoxin may be a useful technique for rapidly determining bacterial biomass and quality of water. Endotoxin is the lipopolysaccharide (LPS) present in the outer membrane of Gram-negative bacteria and some cyanobacteria. While ingestion of large amounts of endotoxin (>1000 endotoxin units (EUs) can cause fever, diarrhoea, vomiting, acute respiratory illnesses, and lung inflammation, it can cause severe inflammatory reactions leading to shock, organ failure and death, if small amounts (<0.1 EU) enter sterile body cavities through inhalation or injection [
3]. Endotoxins are therefore of immediate concern in many pharmaceutical industry water systems producing parenteral products because of these effects [
3]. Endotoxin in pharmaceutical products for injection must be carefully monitored using the LAL assay which can detect picogram amounts of the molecule. Under these circumstances, the assay requires about 2 h to perform by skilled technicians in specialised laboratories.
Previous work by our group has shown the applicability of using endotoxin as a marker of faecal contamination of seawater [
14] and that measuring endotoxin correlates with inflammatory effects of contaminated water samples [
15]. The team at Molendotech have developed a near real-time assay (BacterisK) to detect endotoxin in water which can be conducted by non-specialist staff in situ. The advantage of measuring endotoxin as an indicator for contaminated water is that the test is specific for LPS, a compound which only naturally occurs in the cell walls of Gram-negative bacteria. LPS comprises a relatively constant proportion of a Gram-negative bacterial cell and Gram-negative bacteria account for 80 to 95% of the prokaryotes found in waters [
16]. Moreover, endotoxin could indicate the presence of Gram-negative pathogens not detected by current culture of total coliforms or
E. coli. Therefore, this novel assay would provide rapid testing in remote and disaster areas where access to laboratories and water testing facilities is challenging.
The present study was undertaken to test the applicability of the novel BacterisK assay as a rapid test of water quality and to correlate the amount of endotoxin in water with other measuring methods used in water quality assessment.
3. Results
A total of 64 water samples were analysed by both the Kinetic-QCL™ assay and the BacterisK assay to establish a correlation between Endotoxin unites (EU/mL) and ER. The BacterisK assay directly correlates (
p < 0.0001, R
2 = 0.929) with endotoxin activity (
Figure 1).
Water samples were classified into three groups based on the level of
E. coli, 0–10, 11–100 and >100 CFU/100 mL according to the WHO recommendations [
7,
8], and the mean ER for each group are shown in
Figure 2. The 0–10 CFU/100 mL group had a mean ER of 2118 with a standard error of the mean (SEM) of 550, the 11–100 CFU/100 mL group had a mean ER of 4356 with a SEM of 898 and the >100 CFU/100 mL group had a mean ER of 7796 with a SEM of 275. The 0–10 and 11–100 CFU/100 mL groups were significantly different (
p < 0.05) as were the 11–100 and >100 CFU/100 mL groups (
p < 0.001).
The BacterisK assay measures ER that can be used as a rapid determinant of water quality in terms of Gram-negative bacterial contamination. Using the data shown in
Figure 2, ER ‘cut-off’ values that would allow water samples to be graded as ‘low risk’, ‘intermediate risk’ or ‘high risk’ with respect to probable bacterial contamination were determined. The ‘low risk’, ‘intermediate risk’ and ‘high risk’ groups equated to ER cut-off values of <500, 500–7000 and >7000, respectively. The lower cut-off value of <500 ER was determined to ensure no false negative results and is lower than the 2118 mean ER for the 0–10 CFU/100 mL group (
Figure 2), while the upper cut-off value of >7000 ER was based on the 7796 mean ER for the >100 CFU/100 mL group (
Figure 2). A Chi-squared analysis using these groupings provides a significant correlation (
p < 0.0001) (
Table 1).
From the E. coli culture analysis, 23 samples gave a count of less than 10 CFU/100 mL. Among them, 9 samples (39%) gave an ER result less than 500 ER on the BacterisK assay. The other 14 samples gave a result of >500 ER units. For the 11–100 CFU/100 mL category, composed of 14 samples, 9 of them had an ER result between 500 and 7000 and 5 others had an ER result above 7000. Importantly, no samples with E. coli culture of >10 CFU/100 mL gave a result less than 500 ER. Finally, in the >100 CFU/100 mL category, 100% of samples (32) gave a result of >500 ER units and 29 of these (90%) gave an ER result over 7000.
Fifteen samples were assessed in duplicate by the BacterisK assay to determine the repeatability. The mean and standard deviation were calculated to obtain a variation coefficient (
Table 2). Global variation coefficient, obtained from the means of all variation coefficients (
n = 15), is equal to 3%. As this result is less than 10%, the repeatability of the BacterisK assay is shown to be acceptable.
4. Discussion
According to the World Health Organization, more than 1.8 billion people use drinking-water sources that are contaminated with faeces [
8]. Contaminated water is known to be linked to various diseases, such as cholera, diarrhoea, or dysentery. According to the WHO, the mortality of water-associated diseases exceeds 5 million people per year and 1.5 million children die annually from diarrhoea linked to unsafe drinking-water, sanitation, and hand hygiene [
7]. Bacterial contamination of drinking water is a major contributor to water-borne diseases in rural areas of most developing countries where water sources are communally shared and exposed to multiple faecal-oral transmission pathways [
7].
Since much of the microbiological water contamination is derived from human and animal faecal origin, monitoring water sources for faecal bacteria has become the practice to determine water quality. Detection of all possible pathogens in water would be too difficult, expensive and laborious; hence indicator organisms are used to infer the presence of faecal organisms which may include pathogens. The most used indicator bacteria is
E. coli. which is usually only found associated with faecal matter from warm-blooded animals and humans [
4,
19,
20].
Currently, the gold standard method to quantify these bacteria is the membrane filtration method. However, for shallow and surface waters the ISO 9308 is no longer applicable, due to a change in the growth medium described in the standard, from 2,3,5,-triphenyltetrazolium chloride (TTC) to Coliforms Chromogenic Agar (CCA). The latter is a more efficient medium, but this change rendered the ISO no longer applicable for use with recreational waters. This change caused some commotion on the enforcement of the current European Bathing Water directive because it was the reference method [
7] and it must now be reviewed. Nonetheless, for in-scope waters, the main drawback of this technique is that results are still retrospective, taking between 24–48 h to have results. In addition, it requires sending samples to be processed in a laboratory with technically skilled staff.
Unfortunately, in low-income countries, especially in rural areas, access to a laboratory and expensive equipment is often an issue. Therefore, there is a pressing need for faster and more convenient methods to test water samples and determine a rapid risk analysis of water to be used for drinking. Ideally, such a rapid risk assessment test should be portable and be able to be used in field at the site of water sources as well as at distribution networks.
The patented BacterisK technology, developed by Molendotech, is a rapid, in situ testing kit that targets the presence of endotoxins as a biomarker for bacterial contamination. Therefore, BacterisK can provide a risk assessment for the presence of common pathogens in waterborne diseases, such as E. coli, but also all Gram-negative bacteria, including Pseudomonas aeruginosa, Vibrio cholerae, Salmonella and Campylobacter. The BacterisK technology is based on a colorimetric assay that takes less than thirty minutes to obtain results. BacterisK directly reports an ER value: the higher the value the higher the water contamination risk. However, it is important to show that such a rapid risk assessment tool has a correlation with faecal bacterial contamination. The present study was conducted to compare the ER with CFU/100 mL E. coli measured by the membrane filtration method.
Results from the current study show that the BacterisK assay can be used as a rapid risk assessment tool to indicate the level of faecal bacterial contamination of water.
Levels of ER correlated with the presence of E. coli over a large range from very low levels (0–10 CFU/100 mL) to high levels (>100 CFU/100 mL). Moreover, the BacterisK assay can be used to generate risk assessment in saline bathing water samples (Good et al., manuscript in preparation).
Currently, water quality control is monitored following the WHO guidelines [
8]. Faecal indicator bacteria (FIB) are used as a proxy for the presence of faecal contamination in water resources. These indicators include
Escherichia coli, total coliforms, faecal coliforms and
Enterococci. Regarding the presence of
E. coli, WHO guidelines recommend a threshold of 0 CFU (Colony Forming Unit)/100 mL, However, the WHO understand that 0 CFU/100 mL
E. coli may not be easily achievable in many developing countries and has assigned a health risk score for
E. coli in drinking water: 0 CFU/100 mL (conformity), 1–10 CFU/100 mL (low risk), 10–100 CFU/100 mL (intermediate risk), >100 CFU/100 mL (high risk) [
7,
8]. Results using the BacterisK assay show it to be useful in predicting risk bands aligned with these WHO guidelines. Our results indicate potential thresholds or cut-offs for the risk evaluation of water <500 ER for low risk, 500–7000 for medium risk, and >7000 high risk (
Figure 2,
Table 1). While the ER values correlated well with
E. coli CFU, it should be remembered that these assays measure different targets (live bacteria vs. molecular endotoxin) and therefore we would not expect a strong correlation between their values. However, the results do indicate a good association and the predictive value of the ER assessment would suggest this as a useful measure of water quality.
It can be seen from
Figure 2 and
Table 1 that some samples produced a positive ER response above the 500 threshold yet their
E. coli content was low (<10 CFU/100 mL). In essence, these would be classified as ‘false positives’. This is expected due to the BacterisK assay detecting a molecule, endotoxin, present in all Gram-negative bacteria and would indicate a potential risk from these samples due to contamination with other species. Investigation of these samples indeed showed the presence of other coliform species (results not shown). False-positives (indicators present in the absence of detectable pathogens) also occur fairly often with
E. coli measurement. Although they cause false alarms, these false positives are tolerated. Public health officials are willing to accept reasonable numbers of false-positives because they err on the side of public health. These false positives alone will cause correlation coefficients to be low. Importantly, in the current study, none of the samples had an ER value below the 500 ER cut off and an
E. coli result of >10 CFU/100 mL. This suggests that the BacterisK method does not produce false negative results, which is more important for public health assessment. With a threshold of 500 ER and 10 CFU/100 mL
E. coli, the BacterisK assay has a true positive predictive value of 100%.
In general, coliform indicators alone cannot provide conclusive, non-site-specific and non-pathogen-specific information about the presence and/or concentrations of most important pathogens in surface waters suitable for irrigation [
19,
20]. Nonetheless,
E. coli concentrations are currently used by regulatory agencies to assess the presence of faecal contamination and other pathogens in freshwaters. These organisms are monitored primarily as an index of suspicion: if faecal contamination is present, pathogens may also be present. It should be noted that BacterisK will respond to endotoxin from all Gram-negative organisms present in the sample and might therefore give a risk assessment from non-faecal organisms. The presence of human pathogens that are not of faecal origin represent a significant concern in many areas. Pathogenic groups of Vibrio bacteria, such as
V. vulnificus and
V. parahaemolyticus, for example, are naturally occurring and cause numerous cases of gastroenteritis and wound infections [
21].
The results presented in this study highlight the advantage of the BacterisK assay as a rapid risk assessment of water contamination with faecal and other Gram-negative bacteria. The results obtained in 30 min provide a near real-time monitoring of the water contamination as compared with 24–48 h for the culture of E. coli as is the usual practice. In addition to the rapid results, the portability of the BacterisK assay mean it can be performed in situ at the water being tested. The assay is relatively simple to perform and does not require sterility needed for the membrane filtration method. The rapid results with low COV in repeatability coupled with the portability allow the BacterisK assay to be used for source monitoring programs. The BacterisK assay is also in line with the United Nations vision of ensuring availability and sustainable management of water and sanitation for all (SDG 6); where rapid testing could provide valuable information to the users and potentially reduce health risks.
Current drinking water compliance monitoring using indicator organisms may not provide effective protection of public health and waterborne outbreaks remain common, even in high-income countries [
22]. Delays resulting from culture methods also limit the ability to rapidly communicate the risks to local communities. To address some of these limitations with FIB monitoring, the WHO recommends a risk-based management approach to ensure water safety [
7].
Other rapid methods that have been used for water quality include fluorescence and ATP measurement. Intrinsic fluorescence from organic matter at excitation 280 nm and emission at 350 nm due to tryptophan like fluorescence (TLF) can be correlated with biochemical oxygen demand. Studies have suggested that TLF can be correlated with microbial contaminants and
E. coli. However, the number of compounds that can potentially give fluorescence in this region and the number of potential interferences to fluorescence limit the correlation with
E. coli numbers [
23]. Moreover, metals may quench the fluorescence while high dissolved organic carbon (DOC) and nitrate can give high TLF that do not correlate with high fecal indicator organisms [
24,
25]. The complexity and limitations coupled with the high cost of the instruments, particularly for online measurements, may limit their use in microbial water testing [
26].
ATP bioluminescence can measure cellular material through measuring ATP levels expressed in relative light units. However, the correlation between ATP and the presence of microorganisms in water systems was found not to be significant [
26]. Limitations to the ATP method include that it does not easily distinguish ATP from microorganisms, animals, and plants while luminescence from other sources can affect the actual ATP bioluminescence readings [
26].