Pathogen tests for sludge

Each year, the UK generates about 15,000 tonnes of sewage sludge and, like many developed countries, has a commitment both to minimise the amount sent to landfill, and to reuse the sludge as a fertiliser or soil conditioner. The issue with this, however, is that in its raw state, sewage sludge contains many microbes of concern including human, animal and plant pathogens, so care must be taken to treat and test the sludge to reduce the risk of farmland being contaminated.

Treatments include anaerobic digestion, composting, lime treatment, storage and drying, but the effect on the survival of pathogens will vary according to the exact parameters of the process, and the pathogen under consideration.

It is both impossible and unnecessary to test sludge for the presence of all pathogens, so screening involves the identification of indicator organisms, and in some cases specific pathogens such as Salmonella spp. Utility companies, as part of a Hazard Analysis and Critical Control Point (HACCP) plan, will define parameters for testing  of sludge material to ensure that the treatment is both effective and consistent, and by far the most common microbiological analysis undertaken is the enumeration of Escherichia coli (E.coli).

Pathogen tests for sludgeThe UK Environment Agency publishes a series of good practice guidelines, entitled ‘The Microbiology of Sewage Sludge (2003)’, part 3 of which is ‘Methods for the isolation and enumeration of Escherichia coli, including verocytotoxigenic Escherichia coli’. This document lists three separate procedures for the enumeration of E. coli, however it gives no guidance to the user as to which method is more applicable to the sewage sludge matrix.

Method 1: Isolation and enumeration of E. coli by chromogenic membrane filtration

Sewage sludge is a difficult matrix to deal with and requires vigorous homogenisation to attempt to evenly disperse the bacteria present, but even after this has been done, the number and nature of the particulates present can cause problems with sample handling and, in particular, membrane filtration.

This procedure uses a medium (m-LGA) that relies upon the metabolism of a β-D-glucuronide (BCIG) which, when cleaved, yields an intensely blue coloured product. The medium also contains 30 g/L lactose and phenol red indicator. This combination means that organisms which ferment lactose will form yellow colonies. Thus, in combination with the cleavage product of the BCIG, E. coli colonies are green.

This method has limitations, as it cannot be used when there are relatively high concentrations of solids, as the solids themselves block the pores of the membrane. However, more significantly, m-LGA was designed for the examination of drinking water samples where typically only low numbers of organisms are encountered. Higher numbers of organisms, and particularly the presence of other coliforms, can reduce the pH of the medium and affect the ability of many E. coli strains to form green colonies because of the sensitivity of the β-D-glucuronidase to low pH.1,2 The consequence of this incompatibility between media containing glucuronidase substrates and high concentrations of fermentable carbohydrate means that the numbers of E. coli reported can be a considerable underestimate.

Method 2: Isolation and enumeration of E. coli by a multiple tube most probable number (MPN)

Pathogen tests for sludgeThe conventional MPN procedure described in the series is lactose-based, requires confirmation of results and is labour-intensive. It relies on a completely different procedure for identifying E. coli, the identification being based upon the ability to ferment lactose at 44oC, and to produce indole from tryptophan. Consequently, this procedure identifies a different subset of E. coli to the membrane filtration procedure since some E. coli fail to ferment lactose, some do not grow in brilliant green bile broth at 44oC and some do not produce indole. Conversely, other organisms such as Klebsiella oxytoca fulfil all of these criteria and can therefore result in false positive results. The MPN procedure has a wide counting range and this is extremely beneficial for the examination of biosolids where the likely content of E. coli is unknown. However, the need for confirmation of results increases laboratory hands-on time and delays the result to a minimum of 48 hours.

Estimating the number of target bacteria present can also give rise to problems since, particularly with samples from new sites or processes where the numbers of target bacteria likely to be present may be completely unknown, several dilutions may need to be analysed in order to obtain a valid result. These factors are discussed to some extent in Part 2 of the series, “Practices and procedures for sampling and sample preparation,” but no detailed guidance is given.

Method 3: Enumeration of E. coli by a defined substrate MPN

This method combines the best attributes of the two other systems: the detection of E. coli is on the basis of β-D-glucuronidase expression; and incorporates a large counting range. A suspension of the sample is made in a liquid medium which is poured into a sealable 97-well tray before incubation at 37°C for 18-22 hours. The multiple wells allow enumeration with a counting range of 1-2400 cells.

Although this method will exclude a small proportion of strains of E. coli that do not produce β-D-glucuronidase  it does not suffer from the false negative results encountered with m-LGA due to low pH and results require no confirmation. The procedure is extremely simple and a single tray gives a counting range over three orders of magnitude.


Each of the three methods described is very different and may be expected to give very different results. The International Standards Organisation now defines E. coli as a member of the Enterobacteriaceae that produces the enzyme β-D-glucuronidase. Thus, only the first and third methods are in accordance with this. Of these, the m-LGA method was intended for use with drinking water where high background counts do not interfere with the activity of β-D-glucuronidase through production of excess acid, whereas the defined substrate procedure is the simplest of all three methods, gives the highest dynamic counting range and the most accurate result.


  1. Fricker CR et al: 2008 ‘False-negative beta-D-glucuronidase reactions in membrane lactose glucuronide agar medium used for the simultaneous detection of coliforms and Escherichia coli from water.’ Letters in Applied Microbiology.  47(6):539-42
  2. Fricker CR et al: 2010 ‘Understanding the cause of false negative β-D-glucuronidase reactions in culture media containing fermentable carbohydrate’ Letters in Applied Microbiology 50 (6): 547–551.