We have decided that for Spring 2013, the students at EPFL will focus on less time- and resource-consuming processes to detect coliform bacteria than polymerase chain reaction (PCR). However, how can we resist the chance to investigate this in parallel, when we have Urs of Gaudi Labs, the person himself who documented the Wild OpenPCR is coming with the device to Lausanne? The answer is… we can’t.

While this is not a portable field-device, nor is it an instantaneous reading, the idea would be that places/groups with a bit more funding and infrastructure could set this up in, for example, a community center.

We will have this workshop over 2 week-ends in April. This is a 3-way collaboration between Gaudilabs, hackteria, and BIO-DESIGN for the REAL WORLD, and the process will be documented on this site.

Specifically, the main objectives are:

  1. test the Wild OpenPCR machine against a commercially available one (Gaudilabs)
  2. hack the Taq isolation protocol from Open Biotechnology, and see if we can make and purify the enzyme with alternative reagents from the consumable reagents sold there (following the ethos of hackteria)
  3. test out some PCR primers to detect coliform bacteria (specific to the BIO-DESIGN course)

I just ordered from Open Biotechnology, and the shipping fee to Switzerland of the 3 items (pOpenTaq, mastermix, and protein extraction kit) was close to the same of the 3 items. This ups the motivation for making and purifying our own Taq.

More details on the nitty-gritty of the workshop plans will be up on the hackteria wiki. Here, I’d like to take this opportunity to explore the PCR primers.

Maheux et al summarizes the criteria for PCR assays

  • rapidity (this also includes sample preparation, PCR, and results analysis)
  • specificity (ability to target only the desired species)
  • ubiquity (ability to detect all strains of the targeted species)
  • analytical detection limit

Rather than designing our own primers, we decided to check the literature and use Primers described that have already been tested for specificity and ubiquity, and the analytical detection limit.

Specificity for Coliforms:

To increase the specificity for E. coli, increased target gene numbers can be considered. for example, Horakova et al. used 4 of the biochemical hall mark genes for E. coli:

  • uidA (codes for beta-D-glucuronidase, GUS, observed in ~94% of E. coli)
  • lacZ (codes for beta-D-galactosidase – beta-gal, which cleaves lactose into glucose and galactose, necessary for lactose fermentation)
  • lacY (codes for lactose permease – for lactose transport across cytoplasmic membranes)
  • cyd (codes for cytochrome bd complexes, for respiration under low aeration conditions)

The three genes correspond to biochemical activity of lactose fermentation also earlier described in a functional analysis (spread plate technique) of Yogyakarta river water. In the Horakova et al. paper, primers were designed for all 4 genes to be detected in 1 PCR reaction tube. This is called multi-plexing, a multiplex assay. An agarose gel analysis with samples with all 4 genes expressed will have 4 bands (PCR products are designed to have different sizes). They found that in the test strains, only E. coli and not other bacteria would have 4 PCR products (4 bands on the gel).

We decided to be a bit more simple (“simplex assay”), and will use 1 pair of primers described by Maheux et al. that detect both E. coli and Shigella. They are both gram-negative bacteria, which are genetically quite similar, and Shigella is disease-causing. These authors argue that they should be both detected for water testing. The tuf gene (house keeping gene) will is the target gene amplified as described in the 2009 paper by Maheux. This primer (called D in this paper) passed the specificity (0.5% non-E. coli and shigella detected out of 192 tested) and ubiquity (all E. coli and shigella detected) criteria, and the analytical detection limit is reported to be 10 to 1 genome copies of E. coli. Annealing temperature  is 58oC, and the PCR product (amplicon size) is 258 base pairs (bps).



One thing I still need to understand is what this tuf gene is doing – a quick search makes me wonder why this would be specific to E. coli and Shigella only, although the list of other gram negative bacteria they used to test the specificity is extensive. More will be followed up later.

For applications to detect pathogenic strains of bacteria, PCR primers should be designed to detect genes encoding virulence factors or bacterial toxins.

Reaction inhibition due to enzyme inhibition depending on the water sample – One can imagine that as the Taq polymerase (PCR) reaction is sensitive to, for example to MgCl2 concentrations, and other inhibitors in the environmental sample. To not have false negative samples, the water can be spiked with known amounts of positive control (E. coli) DNA, just to make sure that the solution itself is not inhibiting.

Concentrating Bacteria in Sample: such as capture by filtration, is also described in the article by Rompré et al (see references below). PCR has been extensively explored as a potential method to detect coliform bacteria, also in drinking water. However, from field samples, DNA extraction methods were important in PCR reliability, and concentratng dilute samples seems to be necessary much of the time.

Quantification: Generally, end-point PCR is not suitable for quantification of the PCR product. In other words, it will be difficult to quantify the amount of coliform bacteria in a water sample. This is more a threshold indicator, and a relative qualitative analysis.

This is why quantitative real-time PCR (qRT PCR) was developed. Most current DIY PCR “thermocycler” machines have not been adapted* for the on-line fluorescence detection of the qRT PCR method (accumulation of product > more fluorescence).

If not quantitative, ratios of beta-D-galactosidase (total coliform bacteria) to beta-glucuronidase (non-pathogenic E. coli) may give a relative indication of what types of coliform bacteria may be present —this goes back to multiplexing.

*However, the Open Biotechnology master mix, with the addition of a dye that fluoresces proportionately when bound to DNA (SYBR green), can be used as a qRT-PCR reagent.

See what happened on the hackteria site.

Reference papers:

Horakova et al. Specific detection of Escherichia coli isolated from water samples using polymerase chain reaction targeting four genes: cytochrome bd complex, lactose permease, beta-D-glucuronidase, and beta-D-galactosidase. Journal of applied microbiology, 105(4), (2008) pp.970–976.

Maheux et al. Analytical comparison of nine PCR primer sets designed to detect the presence of E. coli/Shigella in water samples Water Research 43 (2009) pp. 3019-3028.

Rompré et al. Detection and enumeration of coliforms in drinking water: current methods and emerging approaches. J. Microbiological Methods 49 (2002) pp. 31-54.

One thought on “coliform bacteria detection by PCR

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.