Identification of MAP in Cultures

When colonies that consist of AFB are recog­nized on solid media or when growth is reported in broth culture, the next challenge is to iden­tify MAP. A presumptive identification can be made based on slow growth - colonies develop after more than 2 weeks - and host tissue pre­dilection, as cases of granulomatous enteritis associated with AFB in livestock are most likely to be paratuberculosis.

18.15.1 Colony morphology

Twort and Ingram were the first to describe MAP colonies. They were initially round, smooth and white then tended to heap up slightly and become dull light yellow with wrinkling of the surface; pigmentation of colonies was influ­enced by the colour of the M. phlei or egg in the medium (Twort and Ingram, 1912). Colony morphology is dependent on the medium, and addition of supplements can dramatically alter it: Tween compounds cause otherwise irregular granular colonies on 7H9 agar to appear instead as entire, smooth and domed (van Boxtel et al., 1990). Colonies of a MAP-S strain growing on 7H10 agar reach of cul­ture media (Whittington et al., 2011). MAP cultured from the intestinal tissues of sheep in Spain were not dependent on mycobactin for primary isolation on 7H11 agar (Aduriz et al., 1995) and it was not required for growth of a variety of MAP strains on egg-free 7H11 agar but did stimulate their growth (Whittington et al., 2011). Mycobactin dependency should not be assessed on egg-free Middlebrook 7H11 agar (Whittington et al., 2011) but can be un­dertaken on Middlebrook 7H10 agar, HEYM or LJ agar (Whittington et al., 2011).

18.15.3 Molecular confirmation using IS900

The discovery of IS900, an insertion element thought to be specific for MAP (Green et al., 1989) provided a molecular basis for identifi­cation, and PCR is now used alone or in com­bination with mycobactin dependency for the identification of MAP in cultures. Arguably this is sufficient for livestock on endemically infected farms but is insufficient for index cases given the potential adverse consequences of a positive diagnosis.

IS900 is a member of a family of insertion sequences, some of which closely resemble IS900 and so positive results can be obtained from other mycobacterial species using com­mon probes and primers (Cousins et al., 1999; Englund et al., 2002; Kim et al., 2002). Strategies to resolve closely related IS900-like insertion sequences include restriction en­donuclease analysis of the PCR product, se­quencing of the PCR product, use of internal probes in assays based on technology such as Taqman and high-stringency real-time PCR protocols with specific primers and the assess­ment of melting temperatures (Cousins et al., 1999; Englund et al., 2002; Kim et al., 2002; Kawaji et al., 2007). Because the identification of MAP has been shown to be less than 100% specific on several counts, and regardless of assurances from laboratories, microbiological diagnoses should be questioned if they do not make epidemiological sense. This is how the or­ganisms that carry IS900-like sequences were discovered (Cousins et al., 1999).

For confirmation of an index case in any species, it is critical that IS900-based PCR not serve as a stand-alone test for culture­confirmation. Additional microbiological data to confirm the identification of MAP are need­ed, such as sequencing of the IS900 amplicon, identification of MAP-specific elements such as F57, demonstration of mycobactin dependency, colony morphology and growth rate, acid-fast staining and microscopic morphology of bacte­rial cells.

PCR analysis of colonies of AFB can be conducted after release of DNA from bacte­rial cells using simple methods such as boiling a suspension made from the colony (Whittington et al., 1999). However, it can be quite difficult to obtain suitable samples for identification of MAP from liquid cultures such as BACTEC, because egg yolk inhibits PCR amplification of IS900. The first protocols for PCR confir­mation required subculture to BACTEC 12B medium without egg yolk and incubation for a few weeks, which delayed the diagnosis (Cousins et al., 1995). Later, a simple extrac­tion of the BACTEC 12B broth in ethanol was developed to remove the egg yolk, followed by heating of the supernatant to lyse MAP cells and release DNA (Whittington et al., 1998); this method was used in diagnostic laboratories in Australia. Where the liquid culture contained other types of bacteria as well as MAP, purifica­tion of DNA from the lysate using a silica col­umn was shown to remove residual inhibitors of PCR (Whittington, 2009). To address the problem of identification of MAP from liquid cultures, Okwumabua et al. (2010) compared three methods for MAP nucleic acid extraction (magnetic bead, silica spin column and phenol­chloroform) after growth in Trek broth. The recovery of amplifiable MAP DNA was greatest for the magnetic bead method. After widespread adoption of M7H9C liquid medium in Australia, the manual ethanol extraction method was re­placed by a semi-automated purification sys­tem based on bead beating and magnetic bead purification of nucleic acid (Plain et al., 2015). Others have also tackled this problem. Simply boiling the broth to release DNA was almost as successful as a commercial column kit in one study (Sweeney et al., 2006). Salgado et al. (2014) found that an in-house sequential so­dium dodecyl sulphate/proteinase K lysis, bead beating and ethanol extraction method was superior to a commercial bead beating, protein­ase K, column purification kit and a cethyltri- methylammonium bromide (CTAB) chloroform purification method, but only 12 samples were evaluated.

All of the semi-automated methods for DNA extraction (such as magnetic bead) and those involving commercial kits are generally multi­step procedures. There are interactions between machine platforms and kits. Kits from different manufacturers may not be interchangeable, their composition is often not known and com­ponents can change over time without notifi­cation. Consequently, methods need to be fully documented, properly validated and monitored closely over time using process controls.

18.15.4 Other identification approaches

Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry (MS) is widely used in medical microbiology for rapid identification of clinical isolates, typically by analysis of bacterial colonies. Conceptually MALDI-TOF MS produces a molecular finger­print that can be compared against a database of fingerprints from known taxa. This method has been used to differentiate MAP from other mycobacterial species and to differentiate strains of MAP (Ravva et al., 2017). Similarly, volatile emissions can be collected from the headspace above a viable culture and analysed by gas chromatography mass spectrometry (GC-MS). Happily the resulting fingerprint is known as a bouquet (Nawrath et al., 2012). Volatile emis­sions are influenced by culture media and growth conditions, but MAP can be differentiat­ed from other mycobacterial species (Trefz et al., 2013; Kuntzel et al., 2016). Kuntzel et al. (2019) showed that there is a group of 28 volatiles com­mon to MAP even when there is variation in cul­tivation conditions. However, specificity of these new approaches for MAP is unproven.

18.15.5 Effect of contaminants on the identification of MAP

Many cultures of MAP are mixed cultures, i.e. they also contain irrelevant microorganisms. The latter confound the identification of MAP in liquid culture by inhibiting IS900 PCR, in­creasing the number of samples for which purification of DNA extracts is required prior to PCR (Whittington, 2009). On solid medium, contaminants can inhibit the growth of MAP or obscure MAP colonies (Secott et al., 1999; Whittington, 2009). Thus contaminants com­plicate, delay and increase the costs of culture of MAP. There are several studies in which PCR has been used to reveal the presence of MAP in contaminated or overgrown culture media (Whittington, 2009; Arango-Sabogal et al., 2016b; Rangel et al., 2017). Contaminated cultures may need to be redone, either from stored sample material or following collection of replacement samples.

18.16