Home
About
Overview
Contact Us
Products
Application Data
PlanktonTrap - Ballast Water Treatment
Ballast Water Do You Know What's in it ?
Application Data

Development of a Revised Method for Increased Sensitivity for Recovery

of Total Coliform and E. coli in Drinking Water

Randi M. McCuin1, Thomas M. Hargy1, Jeffery Rosen2 and Jennifer L. Clancy1

Clancy Environmental Consultants, Inc. Saint Albans, VT 054781 and Scituate, MA 020662

Abstract

Although it is not a conclusive indicator, total coliform (TC) bacteria have been used for decades as indicators of fecal

contamination and/or distribution system breaches. Despite new data indicating TC are not useful as indicators of fecal

contamination, TC sampling is still seen by water quality managers as a useful indicator of the integrity of their distribution

system. A number of researchers have identified the difficulty in estimating the underlying spatial and temporal distributions of

TC in distribution systems based on presence/absence sampling. For TC to be a useful indicator there must be an underlying

assumption that monitoring will result in observing TC when present. However, the sampling plans currently employed by water

utilities are not based on optimizing sampling to identify TC when present but rather they are optimized by accessibility and

logistics. TC sampling of even large numbers of samples (480/month) tests an infinitesimal portion of the water flowing through

the system in a month; since TC are highly heterogeneous, mostly zeros and perhaps a few positives, result. In order to

optimize a sampling plan to detect TC when present, the plan, including locations, periodicity and sample volumes, should be

based on the underlying distribution. This distribution cannot be measured directly, but the distributions of the sampling results

can be evaluated. A recent study on sample volume determined that when taking 100mL and 2L samples side by side and

comparing the results, the 100mL samples significantly underestimated the occurrence of TC. While these results suggest that

the current sampling and analytical strategies may not be optimal, they do not resolve the issue of how many samples of what

size are needed to meet the objectives listed above. It may be that different objectives require different sampling and analytical

plans. To explore high volume sampling, we have developed a method using capsule filtration and enzyme substrate technology

that can accommodate samples up to 20L. We evaluated the ability of the process to filter, recover and identify total coliform and

E. coli at different concentrations and sample volumes ranging from 100 mL to 20L.

Material and Methods

Description of Filters Used

The MVL CT44M filter is a sterile hollow fiber module with a filtration area of 4000 mm2 and a 0.2

μm nominal pore size. The MVL module was able to filter 25 L in 90 min at 30 psi (2 bar). Each filter can operate at a

maximum pressure of 75 psi (5.1 bars) at ambient temperature.

Recovery of E. coli from MVL modules

Replicate MVL CT44M units were challenged with 5, 10, and 100 E. coli, respectively, as 1 mL suspensions. The

suspension was delivered directly to the feed port on the MVL module using a micropipette equipped with a sterile 1000 μL

tip. To elute the module, the end cap (opposite feed port) of each MVL module was removed, and 100 mL sterile DI

delivered through the feed port, with water passing through MVL and into an IDEXX coliform sample bottle (125 mL)

containing sodium thiosulfate. The entire 100 mL elution fluid from each test replicate was membrane filtered and plated to

tryptic soy agar (TSA) plates to assess the effectiveness of the elution procedure employed. In addition, 1 mL subsamples of the

spiking suspensions were membrane filtered and plated to TSA to verify E. coli concentration.

Recovery of E. coli from Filters at Increased Sample Volumes

MVL Modules

Replicate MVL modules were challenged with ~5 E. coli, respectively, as 1 mL suspensions. The suspension was delivered

directly to the feed port of the module using a micropipette equipped with a sterile pipette tip. After seeding, each module was

connected by flexible tubing to a pressurized dechlorinated (GAC) tap and valved on at full tap pressure (~45 psig). Three

replicates were prepared filtering each of the following volumes: 1, 2, 5, and 10 L. After filtering the desired volume, each unit

was valved off, and the end cap (opposite feed port) of each MVL was removed and the side discharge port capped. The

unit was then valved on with the far open end to flush 100 mL volume of the test water into a 125 mL IDEXX sample bottle

containing sodium thiosulfate. A second elution was made in the above manner for each of the 10 L samples. Each sample was

treated as a TC/E. coli sample by adding Colilert reagent to each bottle and incubating as prescribed by the manufacturer.

Recovery of low concentration of E. coli dispersed in drinking water by MVL with increasing sample volumes

A pressure tank containing 20 L dechlorinated tap water was seeded with 15 (± 6.0) E. coli to a concentration of 0.8 ± 0.3 cfu/L,

from a stock suspension enumerated by plating ten 1 mL replicate subsamples. The continuously stirred feed tank was

pressurized with compressed air, and the discharge was directed through a sterile tube to the MVL module. Triplicate units

were then allowed to filter 0.1, 1, 2, 5, and 10 L of the seeded feed water, receiving the theoretical E. coli challenge levels of 0.1,

0.6, 1.2, 3, and 6 E. coli per test filter, respectively. Additional E. coli feed batches were prepared as the 20 L tank was depleted.

After filtration, each module was connected by flexible tubing to a pressurized GAC tap, valved on, with the far open end and

flushed directly into a 125 mL IDEXX sample bottle containing sodium thiosulfate. For comparison purposes, three replicate 100

mL samples of the seeded feed water were delivered directly to IDEXX sample bottles. Each sample was treated as a TC/E. coli

sample by adding Colilert reagent to each bottle and incubating per the manufacturer’s instructions.

Results

Recovery of E. coli from MVL modules

E. coli recovered in the 100 mL MVL eluted samples were, in nearly all instances, >100% of the predicted concentrations

obtained in the spike dose enumerations. Results are shown in Table 1. An error in the quantifying the seeded suspension may

explain this phenomenon. Only five replicates of the spiking suspensions were performed in these trials and if more replicates

were analyzed a more accurate count of the spike dose suspension would theoretically result. However, these data do show that

the elution procedure and elution volume employed was highly effective in recovering low levels of seeded bacteria introduced

into the small modules. This elution method could be performed in the field by sampling a given volume with the MVL in

sampling mode (end cap closed), then removing end cap, directing flow to a 100 mL sample bottle, using site water pressure to

flush and elute sample.

Table 1. Recovery of E. coli from MVL Modules

When Spiked Directly into Filters

Recovery of E. coli from Filters at Increased Sample Volumes

Replicate

Target Dose (cfu)

Measured Dose (cfu)

Recovered (cfu)

Percent Recovery (%)

1

5

5

13

 

2

 

4

6

 

3

 

4

7

 

4

 

4

4

 

5

 

5

9

 

Average+/- SD

 

4.4+- 0.5

7.8+- 3.4

177

1

10

7

12

 

2

 

6

9

 

3

 

14

11

 

4

 

4

 

 

5

 

8

 

 

Average +/- SD

 

7.8 +- 3.8

10.7+- 1.5

137

1

100

106

109

 

2

 

83

85

 

3

 

90

104

 

4

 

106

 

 

5

 

95

 

 

Average +/- SD

 

96.0 +- 10.0

99.0 +- 12.7

103

Cfu colony forming unit SD standard deviation

MVL Modules

Sample results generated using one elution for all tests (1, 2, 5 and 10 L) indicated the presence of TC (yellow coloration) and E.

coli (fluorescence under UV). Where a second elution was performed results were negative for E. coli and TC (Table 2).

Table 2. Recovery of E. coli Seeded to MVL Modules

Followed by Filtration of Dechlorinated Tap Water

Test Volume Litres

E Coli results 3 replicates

Total Coliform results 3 replicates

1

+++

+++

2

+++

+++

5

+++

+++

10

+++

+++

10 (2nd Elution)

+++

+++

- = negative result

+ = positive result

E. coli = 5.0 ± 2.2 cfu

Results for direct samples and samples generated by elution of MVL modules are presented in Table 4. None of the

samples representing 100 mL volumes either directly sampled to IDEXX bottles or filtered and eluted from MVL filters, were

positive for the presence of TC or E. coli. Except for the single 2.5 L volume sample (intended as a 5 L sample but terminated as

the 20 L seeded feed supply was depleted) all MVL samples of 1 L volume or greater were positive for TC and E. coli.

Table 4. MVL Recovery Trials in Volumes Ranging From 0.1 – 10 L

Test Volume Litres

# Ecoli per test Volume

E Coli results 3 replicates

Total Coliform results 3 replicates

0.1 (direct Colilert)

0.08 +- 0.03

---

---

.1

0.08 +- 0.03

---

---

1.0

0.8 +- 0.3

+++

+++

2.0

1.5 +- 0.6

+++

+++

2.5 (one replicate only)

1.9 +- 0.8

-

-

5

3.8 +- 1.5

+++

+++

10 (two replicate only)

7.5 +- 3.0

++

++

Unseeded feed blank (0.1litre)

0

-

-

- = negative indication of analyte, + = positive indication

In the trials above, because negative results were obtained at 100 mL volumes and nearly all positive above 1 L, intermediate

volumes were tested to determine at what volume positive results began to predominate. Volumes tested included 100, 250,

500, 750 and 1000 mL. Triplicate tests for all volumes and direct IDEXX bottle samples (100 mL) were run. E. coli was seeded

at a concentration of 15 E. coli per 20 L, or 0.8 cfu/L.

Analyses of direct samples and low volume samples generated by elution of MVL modules are presented in Table 5. In

this trial, one sample representing a 100 mL volume delivered directly to an IDEXX bottle indicated the presence of total

coliforms and E. coli. Of all the MVL samples, one positive was observed for each of the volumes of 250, 500 and 750 mL,

none in the 100 mL samples, and two of the three 1 L volumes were positive for TC and E. coli (see Figure 1).

Table 5. MVL Recovery Trials in Volumes Ranging From 100 – 1000 mL

Total Volume Litres

#of E coli per test Volume

E coli results 3 replicates

Total Coliform Results 3 replicates

0.1 (direct Colilert)

0.08 +- 0.03

-+-

-+-

0.1

0.08 +- 0.03

---

---

0.25

0.2 +- 0.06

-+-

-+-

0.5

0.4 +- 0.1

--+

--+

0.75

0.6 +- 0.2

+--

+--

1.0

0.8 +- 0.3

+-+

+-+

Unseeded Feed Blank (0.1litre)

0

-

-

Positive Control (0.1litre)

38

+

+

- = Negative result + = Positive result

Discussion

These trials demonstrate that technology exists to improve detection of total coliform and E. coli in potable water by increasing

the sample volume collected. In two separate trials with MVL modules, 5 of 6 1 L samples were positive for E. coli when

seeded at a level of 0.8 cfu/L. With minor modifications to methods currently in practice to detect the target organisms, water

quality mangers may be able to respond more rapidly to water quality issues in the distribution system using these methods.

References

IDEXX Laboratories, Inc. Package Insert.

Standard Methods for the Examination of Water and Wastewater, 21st ed. 2005. Method 9223 B.





Enhanced Plasmon Bio-Detector (EPBD)

 

Testing the Suitability of Mem-teq Filters for the Capture of Cryptosporidium Oocysts from Water Samples for Contamination Testing

Progress report September 2006

The project comprises four phases of which the first two phases together with Phase 3 stage 1 have been completed and reported.

This report describes work undertaken in the second stage of Phase 3 (Optimisation of CellTrap™ Elution Method).

In outline a series of tests was conducted on CellTrap™ to ascertain the optimum elution strategy to recover Cryptosporidium oocysts from 10 ml aliquots of seeded sterile tap water using sterile distilled water as the eluant.

A series of elution steps was conducted in triplicate; the eluates from each step combined and oocysts recovered using the  Isolate™  immunomagnetic separation system, as described in the previous report on Phase 3 stage 1 of the project. Following dual-staining with a fluorescein-conjugated antibody (Crypto Cel)  and the DNA stain  DAPI, oocysts were enumerated using fluorescence microscopy.

Table 1. Recovery of Cryptosporidium oocysts from CellTrap™ using multiple backwashes with distilled water.

1 Mean result from three tests ± 1 standard deviation

 2 Corrected for loss occurring in the IMS recovery stage as determined during Phase 3 (stage 1) of the project

The trial results summarised in Table 1 show that improved recoveries were obtained

employing multiple backwashes of the CellTrap device. The highest recovery efficiency of 75.9% was obtained when 20 backwashes were performed. However, this value showed only a marginal increase over the recovery efficiency of 73.5% obtained when 10 elution steps were undertaken. Tests in which less than 10 elution steps were performed showed poorer recoveries and it may be concluded that 10 steps  appears to be the optimum from the work undertaken.

The results are encouraging as they demonstrate that Cryptosporidium oocysts are retained by CellTrap and can be recovered with a high efficiency into a relatively small volume (2 ml) of an easily-prepared  eluant (sterile distilled water).

The incorporation of detergents and/or adjustment of the pH and ionic strength of the eluting buffer could be investigated as approaches to further enhance the recovery efficiency of oocysts from CellTrap.

The next stage of the programme  (Phase 3, stage3)  in which the ability of CellTrap to recover a wide concentration  range of  Cryptosporidium oocysts is currently under investigation. 

 

 

 Ocean Waters treated with Planktontrap

 Ocean Studies using Planktontrap

 


HomeAboutOverviewContact UsProductsApplication Data