HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Torkko, P. Articles by Kontro, M. Search for Related Content PubMed PubMed Citation Articles by Torkko, P. Articles by Kontro, M. Agricola Articles by Torkko, P. Articles by Kontro, M.

Gas-chromatographic lipid profiles in identification of currently known slowly growing environmental mycobacteria

Pirjo Torkko¹², Marja-Leena Katila² and Merja Kontro^,1

¹Laboratory of Environmental Microbiology, National Public Health Institute, PO Box 95, FIN-70701 Kuopio, Finland ²Department of Clinical Microbiology, Kuopio University Hospital, PO Box 1777, FIN-70211 Kuopio, Finland

Correspondence Pirjo Torkko pirjo.torkko{at}ktl.fi

Received 1 November 2002 Accepted 18 December 2002

Cellular fatty acid analysis by GLC is widely used in the species identification of mycobacteria. Combining mycolic acid cleavage products with shorter cellular fatty acids increases the informative value of the analysis. A key has been created to aid in the identification of all currently known slowly growing environmental species. In this scheme, the species are classified into six categories, each characterized by a combination of fatty markers shared by those species. Within each category, individual species may be distinguished by the presence or absence of specific marker substances, such as methyl-branched fatty acids or secondary alcohols. This study also describes earlier unpublished GLC profiles of 14 rare, slowly growing, environmental mycobacteria, Mycobacterium asiaticum, Mycobacterium botniense, Mycobacterium branderi, Mycobacterium conspicuum, Mycobacterium cookii, Mycobacterium doricum, Mycobacterium heckeshornense, Mycobacterium heidelbergense, Mycobacterium hiberniae, Mycobacterium kubicae, Mycobacterium lentiflavum, Mycobacterium scrofulaceum, Mycobacterium triplex and Mycobacterium tusciae. Though no single identification technique alone, even sequencing of an entire single gene such as 16S rRNA, can identify all mycobacterial species accurately, GLC has proven to be both reliable and reproducible in the identification of slowly growing mycobacteria. In cases of earlier unknown species, it generates useful information that allows their further classification and may lead to the description of novel species.

Abbreviation: MACP, mycolic acid cleavage product.

Environmental mycobacteria are an emerging problem as causative agents of diseases, especially among immunocompromised hosts. Many mycobacterial species previously regarded as non-pathogenic to man are today classified as inducers of mycobacterioses, this occurring predominantly in AIDS patients (Carbonara et al., 2000; Luque et al., 1998; Mayo et al., 1998). The precise identification of species is of great importance for appropriate patient management. Environmental microbial flora, also rich in mycobacteria, represent a reservoir of previously unknown species (Torkko et al., 2002). In the 1980s, the genus Mycobacterium amounted to about 40 documented species. Today, the number has increased to over 90, and novel species, both saprophytic and potential pathogens, are continually being introduced (Tortoli et al., 2001b; Willumsen et al., 2001; Wilson et al., 2001; Torkko et al., 2002). The taxonomic classification of a new bacterial species is based on the genetic and biochemical characteristics of a group of identical or highly similar unclassifiable isolates. The chemical analysis of their lipid composition is regarded as essential for categorization of a new species (Vincent Lévy-Frébault & Portaels, 1992). Both fatty acid analysis by GLC and mycolic acid analysis by HPLC are commonly applied techniques (Luquin et al., 1991; Butler & Guthertz, 2001).

In the routine identification of isolates recovered from both clinical and environmental sources, GLC analysis has proven to be highly applicable for the identification of the increasing number of validly described mycobacterial species. It has also been found to be a helpful tool in classifying isolates that do not meet the identification criteria of currently accepted species (Brander et al., 1992; Koukila-Kähkölä et al., 1995; Torkko et al., 2000, 2001, 2002). If the GLC analysis is extended to cover both mycolic acid cleavage products (MACPs) and cellular fatty acids, the method is increasingly reliable for species identification (Chou et al., 1998; Müller et al., 1998; Torkko et al., 1998).

In the present work, the main focus was on GLC identification of recently described or rare slowly growing species and their differentiation from the well-established species. The identification scheme devised relied on visual assessment of the fatty acid profile. Three rapid biochemical tests, which were found to be useful for final confirmation of certain species, were incorporated into the scheme.

METHODS

Bacterial strains.

Strains used in this study were mostly type strains, originating from several type culture collections (Table 1). They were stored in Middlebrook 7H9 broth (Difco) at –80 °C and subcultured for GLC analyses on Middlebrook 7H11 agar supplemented with OADC enrichment (Difco) at 36 °C. Haemin and mycobactin J were respectively added for growth of Mycobacterium haemophilum and Mycobacterium genavense. Mycobacterium cookii, M. haemophilum and Mycobacterium marinum were incubated at 30 °C and Mycobacterium botniense, Mycobacterium heckeshornense and Mycobacterium xenopi at 42 °C. For this study, cells were harvested after 3–5 weeks incubation to ensure the production of all fatty acids.

GLC method.

In the direct transesterification procedure, the bound fatty acids were cleaved from cells by mild acid hydrolysis and derivatized to the corresponding methyl esters to convert them into volatile compounds suitable for GLC. The method was described previously in detail (Jantzen et al., 1989; Torkko et al., 1998) and was used with minor modifications. In practice, one loopful of bacteria was harvested into acidic methanol and incubated for 16 h at 85 °C. Fatty acid methyl esters were extracted with n-hexane. The analyses were performed with a Perkin-Elmer AutoSystem gas-liquid chromatograph, equipped with a fused-silica capillary column coated with methylpolysiloxane (NB-30, 25 m x 0.32 mm x 0.25 µm; HNU-Nordion). TurboChrom Workstation software (version 6.1.2.0.1; Perkin-Elmer) was used in the operation of sampling, analysis and integration of the chromatographic data. The carrier gas was helium at a pressure of 8 lb in^–2 (55.2 kPa). The injector and flame-ionization detector temperatures were 325 °C and the oven temperature was programmed from 125 °C (hold for 1 min) to 280 °C, rising by 10 °C min^–1 and holding at 280 °C for 7.5 min. Sample size was 1 µl in the splitless injection; the split was opened after 1 min. Individual fatty acids were identified by comparing their retention times with those of a standard mixture of straight-chain saturated fatty acid methyl esters between C₁₄ and C₂₆. Identification of methyl-branched fatty acids and alcohols was based on the appropriate retention time relationships observed in previously published profiles of known mycobacterial species. The identity of fatty marker substances was additionally determined by mass spectrometry (MS) using a Hewlett Packard G1800A GCD System chromatograph equipped with an electron ionization detector, an HP-5 (30 m x 0.25 mm x 0.25 µm) column and an HP 7673 automatic sampler, as described previously in detail (Torkko et al., 1998).

Biochemical methods.

Tests for Tween 80 hydrolysis and semi-quantitative catalase were performed as described previously (Vincent Lévy-Frébault & Portaels, 1992; Torkko et al., 1998, 2000). Urease activity was determined using commercial discs (Rosco), following the manufacturer's recommendations but using overnight incubation at 36 °C.

RESULTS AND DISCUSSION

The cellular fatty acids present in all of the mycobacterial species examined were hexadecanoic (16 : 0), tetradecanoic (14 : 0), hexadecenoic (16 : 1), octadecanoic (18 : 0) and octadecenoic (18 : 1) acids, as determined previously by Luquin et al. (1991). MACPs ranging from eicosanoic acid (20 : 0) to hexacosanoic acid (26 : 0) were detected in varying combinations and secondary alcohols and methyl-branched fatty acids were recovered in some of the species. On the basis of fatty acid, MACP and alcohol composition, the GLC profiles of slowly growing mycobacteria could be divided into six categories that formed the basis for our identification scheme. As shown in the key (Fig. 1), each main path was characterized by a combination of fatty markers regarded as being typical for that category (Table 2). Each category included several species, some with and others without species-specific lipid markers. Whenever more than one species had an approximately similar basic profile without any distinguishing species-specific markers, one to three auxiliary tests were used to confirm the identification. Only tests that were quick to perform, gave reproducible results and had a high discriminatory power in the separation of the species with similar combinations of lipid content were used (Table 2).

View larger version (19K): [in this window] [in a new window]   Fig. 1. A fatty acid/alcohol marker-based key for the identification of slowly growing environmental mycobacteria.

View larger version (21K): [in this window] [in a new window]   Fig. 2. GLC profiles of M. cookii, M. scrofulaceum, M. tusciae and M. hiberniae in the category ‘Avium'. FID, Flame-ionization detector. See Table 2 for explanation of other abbreviations.

View larger version (23K): [in this window] [in a new window]   Fig. 3. GLC profiles of M. asiaticum, M. doricum and M. kubicae in the category ‘Triviale'.

View larger version (26K): [in this window] [in a new window]   Fig. 4. GLC profiles of M. botniense, M. branderi and M. heckeshornense in the category ‘Xenopi'.

View larger version (24K): [in this window] [in a new window]   Fig. 5. GLC profiles of M. heidelbergense, M. lentiflavum and M. triplex in the category ‘Simiae'.

View larger version (23K): [in this window] [in a new window]   Fig. 6. GLC profile of M. conspicuum in the category ‘Interjectum'.

Category ‘Triviale'

The category ‘Triviale’ is composed of ten species, Mycobacterium asiaticum, Mycobacterium doricum, Mycobacterium gastri, Mycobacterium gordonae, M. haemophilum, Mycobacterium kansasii, Mycobacterium kubicae, M. marinum, Mycobacterium szulgai and Mycobacterium triviale. In the basic profile represented by M. triviale, the main MACP is 24 : 0 and no secondary alcohols are present. Species other than M. triviale contain combinations of species-specific markers that enable their correct identification (Table 2). M. asiaticum was placed in the ‘Triviale’ category on the basis of the GLC profile of the type strain (ATCC 25276^T). In a visual inspection, the GLC profile of another reference strain of M. asiaticum appeared to be identical. Unexpectedly, GLC-MS analysis indicated that the two peaks following 18 : 0 (Fig. 3) were not identical in the two strains. They were 2-methyloctadecanoic (2-Me-18 : 0) and nonadecanoic acid (19 : 0) in the M. asiaticum type strain (ATCC 25276^T), compared with tuberculostearic acid (10-Me-18 : 0) and 2-OH-20 : 0 alc in M. asiaticum ATCC 25274. The amounts of both 19 : 0 and 2-OH-20 : 0 alc were small. In 16S rDNA sequencing, M. asiaticum ATCC 25274 had a one base difference from the type strain (Turenne et al., 2001), at position 1020 of the 16S rDNA. M. doricum, also included in this category, contains only trace amounts of 2-OH-20 : 0 alc (Tortoli et al., 2001b).

Category ‘Xenopi'

The species placed in the category ‘Xenopi’ have a GLC pattern characterized by the presence of 2-OH-20 : 0 alc and a large amount of 26 : 0 (26 : 0 > 24 : 0). This category consists of M. botniense, Mycobacterium branderi, Mycobacterium celatum, M. heckeshornense and M. xenopi. The GLC profiles of M. branderi and M. celatum are characterized by their large amounts of 2-OH-20 : 0 alc. In M. branderi, the amount of 26 : 0 is always greater than that of 24 : 0, whereas the three subtypes of M. celatum (types 1, 2 and 3) differ in their relative amounts of 24 : 0 and 26 : 0. Though M. branderi and M. celatum are indistinguishable by both GLC and biochemical tests (Koukila-Kähkölä et al., 1995), M. branderi is always non-pigmented whereas, in our experience, M. celatum always produces variable amounts of yellow pigment. A recent report has indicated that the tellurite tolerance test could be used to differentiate between these two species (Wolfe et al., 2000). The GLC profiles of M. botniense, M. heckeshornense and M. xenopi contain large amounts of 2-docosanol (2-OH-22 : 0 alc). Of these species, M. botniense is separated by the presence of two specific markers (Table 2; Fig. 4), one of which co-elutes with 2-OH-22 : 0 alc (Torkko et al., 2000). Differentiation of M. heckeshornense from M. xenopi is problematic (Roth et al., 2000, 2001; Richter et al., 2001) and, to our knowledge, reliable separation between the two can only be attained by 16S rDNA sequencing at present.

Category ‘Simiae'

The category ‘Simiae’ consists of M. genavense, Mycobacterium heidelbergense, Mycobacterium lentiflavum, Mycobacterium malmoense, Mycobacterium simiae and Mycobacterium triplex. The major MACP is 26 : 0, and its amount is always greater than that of 24 : 0. The basic profile of this category is similar to that of Mycobacterium tuberculosis with the exception that the 14 : 0 peak is always substantial in this category, whereas it is a minor peak in M. tuberculosis. M. malmoense and M. heidelbergense have the same species-specific fatty markers. However, in the profile of M. malmoense, the amount of 2,4,6-trimethyltetracosanoic acid (2,4,6-triMe-24 : 0) is always greater than that of 24 : 0. This ratio is opposite to the profiles of the two isolates of M. heidelbergense available for our analyses so far. However, more isolates of M. heidelbergense should be analysed to confirm this observation. Fastidious growth and the requirement for mycobactin J separate M. genavense from the other species in this group (Böttger et al., 1993). Three species, M. lentiflavum, M. simiae and M. triplex, have highly similar profiles (Fig. 5) (Chou et al., 1998; Garcia-Barceló et al., 1993). M. triplex, with its variants (Suomalainen et al., 2001), does not produce pigment. The scotochromogenic M. lentiflavum is negative in most biochemical tests (Springer et al., 1996). It grows as pinpoint colonies, in contrast to the photochromogenic M. simiae, growth of which is more abundant. According to Chou et al. (1998), M. simiae can be differentiated from M. genavense and M. tuberculosis by the presence of cis-11-hexadecenoic acid (cis-11-16 : 1) and the lack of cis-10-hexadecenoic acid (cis-10-16 : 1) in the GLC pattern.

Categories ‘Interjectum’ and ‘Palustre'

The remaining two categories, ‘Interjectum’ and ‘Palustre', are characterized by the presence of several methyl-branched fatty markers and both 24 : 0 and 26 : 0 as MACPs. The differentiating feature of the category ‘Interjectum’ is the presence of the secondary alcohols 2-OH-18 : 0 alc and 2-OH-20 : 0 alc. Mycobacterium interjectum and Mycobacterium conspicuum differ in their specific fatty markers (Table 2; Fig. 6) and are therefore easily identifiable by GLC. In contrast, the category ‘Palustre’ lacks secondary alcohols. The two species in this category, M. palustre and Mycobacterium intermedium, have unique combinations of methyl-branched fatty acids (Torkko et al., 2002).

Conclusions

GLC and HPLC are widely used in the routine identification of mycobacteria, in spite of the increasing number of genetic methods designed for the identification of mycobacteria. In our experience, GLC is reliable and reproducible in the identification of both pathogenic and environmental slowly growing mycobacteria. In our identification scheme, we have added one to three biochemical tests to confirm the initial identification result.

The criteria for identification of a species should never be based on a single isolate, regardless of the method used. This also applies to GLC analysis. More than one isolate has to be analysed to ensure the significance of the marker substances that are being regarded as typical of this species. This principle has also been used as the basis of the scheme developed, though, for the present article, the GLC profiles of the type strains were selected. In this study, our major emphasis was placed on rare species whose fatty profiles have remained unavailable in the literature. Though the number of isolates of some species has remained small to date, more than one isolate has been included in our analyses.

Our experience has also verified the usefulness of GLC in the grouping of mycobacteria belonging to previously undescribed species (Koukila-Kähkölä et al., 1995; Torkko et al., 2000, 2002). This grouping can be used as a basis for an extended identification scheme, which may eventually lead to the description of a novel species. In our research studies and in the routine identification of clinical and veterinary isolates, all but the most recent species have been represented. Nonetheless, among mycobacteria recovered from a variety of environmental sources (Iivanainen et al., 1993; Katila et al., 1995), the majority of isolates represent species other than those fully characterized today. This leads us to assume that species identification will become an even more complicated procedure in the future.

According to our current experience, no single identification technique alone, including sequencing of a single gene such as 16S rRNA, allows infallible identification of all established mycobacterial species. Therefore, a rational battery of techniques should be available in all laboratories that perform mycobacterial identification.

REFERENCES