Reliable and rapid identification of Listeria monocytogenes and Listeria species by artificial neural network-based Fourier transform infrared spectroscopy

Differentiation of the species within the genus Listeria is important for the food industry but only a few reliable methods are available so far. While a number of studies have used Fourier transform infrared (FTIR) spectroscopy to identify bacteria, the extraction of complex pattern information from the infrared spectra remains difficult. Here, we apply artificial neural network technology (ANN), which is an advanced multivariate data-processing method of pattern analysis, to identify Listeria infrared spectra at the species level. A hierarchical classification system based on ANN analysis for Listeria FTIR spectra was created, based on a comprehensive reference spectral database including 243 well-defined reference strains of Listeria monocytogenes, L. innocua, L. ivanovii, L. seeligeri, and L. welshimeri. In parallel, a univariate FTIR identification model was developed. To evaluate the potentials of these models, a set of 277 isolates of diverse geographical origins, but not included in the reference database, were assembled and used as an independent external validation for species discrimination. Univariate FTIR analysis allowed the correct identification of 85.2% of all strains and of 93% of the L. monocytogenes strains. ANN-based analysis enhanced differentiation success to 96% for all Listeria species, including a success rate of 99.2% for correct L. monocytogenes identification. The identity of the 277-strain test set was also determined with the standard phenotypical API Listeria system. This kit was able to identify 88% of the test isolates and 93% of L. monocytogenes strains. These results demonstrate the high reliability and strong potential of ANN-based FTIR spectrum analysis for identification of the five Listeria species under investigation. Starting from a pure culture, this technique allows the cost-efficient and rapid identification of Listeria species within 25 h and is suitable for use in a routine food microbiological laboratory.     

FT-IR microspectroscopy: a promising method for the rapid identification of Listeria species

This work presents a pilot study to investigate the potential of fourier transform infrared (FT-IR) microspectroscopy for rapid identification of Listeria at the species level. Using this technique, FT-IR spectra were acquired from 30 strains from five Listeria species. The FT-IR spectra were analysed using stepwise canonical discriminant analysis and partial least-squares regression in a stepwise identification scheme. The results showed that 93% of all the samples were assigned to the correct species, and that 80% of the Listeria monocytogenes strains were correctly identified. In comparison, 100% of the samples, including the L. monocytogenes samples, were correctly identified using spectra acquired by FT-IR macrospectroscopy. The results show that FT-IR microspectroscopy has potential as a rapid screening method for Listeria, which is especially valuable for the food industry.     

Fourier transform infrared and raman spectroscopy for characterization of Listeria monocytogenes strains

The purpose of this study was to characterize the variation in biochemical composition of 89 strains of Listeria monocytogenes with different susceptibilities towards sakacin P, using Fourier transform infrared (FTIR) spectroscopy and Raman spectroscopy. The strains were also analyzed using amplified fragment length polymorphism (AFLP) analysis. Based on their susceptibilities to sakacin P, the 89 strains have previously been divided into two groups. Using the FTIR spectra and AFLP data, the strains were basically differentiated into the same two groups. Analyses of the FTIR and Raman spectra revealed that the strains in the two groups contained differences in the compositions of carbohydrates and fatty acids. The relevance of the variation in the composition of carbohydrates with respect to the variation in the susceptibility towards sakacin P for the L. monocytogenes strains is discussed.     

Pulsed-field fingerprinting of listeriae: identification of genomic divisions for Listeria monocytogenes and their correlation with serovar.

Clamped homogeneous electric field (CHEF) electrophoresis was optimized for genomic analyses of Listeria monocytogenes. Various human, animal, food, and environmental isolates, as well as strains representing other Listeria species, were separately digested with rarely cutting endonucleases. Of 176 L. monocytogenes strains analyzed, the enzymes AscI and ApaI established 63 and 72 unique restriction endonuclease digestion profiles (REDP), respectively. The 22 non-L. monocytogenes strains exhibited 18 AscI and 19 ApaI unique REDP. Statistical analyses of REDP information using the Dice coincidence index and principal component analysis revealed two distinct genomic divisions of L. monocytogenes that also correlated with the flagellar (H) antigen type: division I contained serovar 1/2a, 1/2c, 3a, and 3c stains and division II contained serovar 1/2b, 3b, 4b, 4d, and 4e strains. Division I isolates digested with ApaI were further grouped into cluster IA (serovar 1/2c and 3c) and cluster IB (serovar 1/2a and 3a) strains. Likewise, division II isolates digested with ApaI were further grouped into cluster IIA (serovar 1/2b and 3b) and cluster IIB (serovar 4b, 4d, and 4e) strains. These data indicate that genotypic data generated by CHEF can be directly related to phenotypic data generated by serotyping for establishing the overall relatedness of isolates. Moreover, these data further substantiate that CHEF analysis is a reproducible and highly discriminating method for characterizing L. monocytogenes strains at the molecular level.     

Characteristics and frequency of detection of fecal Listeria monocytogenes shed by livestock, wildlife, and humans

Listeria monocytogenes is a facultative intracellular pathogen that can be carried asymptomatically in various animals and can be shed in feces. We investigated the prevalence and characteristics of L. monocytogenes isolated from livestock, wildlife, and human potential sources of contamination in 2 areas in Ontario, Canada. From February 2003 to November 2005, a total of 268 fecal samples were collected from different animals. Listeria monocytogenes was isolated using selective enrichment, isolation, and confirmation procedures, and 15 samples (6%) yielded to the isolation of 84 confirmed strains. Listeria monocytogenes was isolated from livestock (beef and dairy), wildlife (deer, moose, otter, and raccoon), and human (biosolids and septic) fecal sources. Thirty-two isolates were from serovar 1/2a, 34 from serovar 1/2b, 1 from serovar 3a, and 17 from serovar 4b. Listeria monocytogenes populations were resolved into 13 EcoRI ribotypes, and 18 ApaI and 18 AscI pulsotypes, with Simpson indexes of discrimination of 0.878 and 0.907, respectively. A majority (59%) of L. monocytogenes isolates exhibited potential virulence linked to the production of a functional internalin A, which was supported by higher entry into Caco-2 cells (9.3%) than isolates producing truncated and secreted internalin A (1.3% of entry). Listeria monocytogenes fecal isolates were on average resistant to 6.4 +/- 2.5 antibiotics out of 17 tested, and potentially virulent isolates exhibited an enhanced resistance to kanamycin, gentamicin, streptomycin, and rifampicin. Livestock, wildlife, and human L. monocytogenes fecal communities exhibited overlapping but distinct populations, and some genotypes and phenotypes were similar to those previously described for surface water isolates in the same area.     

Identification of foodborne bacteria by infrared spectroscopy using cellular fatty acid methyl esters

Identification of bacterial species by profiling fatty acid methyl esters (FAMEs) has commonly been carried out by using a 20-min capillary gas chromatographic procedure followed by library matching of FAME profiles using commercial MIDI databases and proprietary pattern recognition software. Fast GC (5 min) FAME procedures and mass spectrometric methodologies that require no lipid separation have also been reported. In this study, bacterial identification based on the rapid (2 min) infrared measurement of FAME mixtures was demonstrated. The microorganisms investigated included Gram positive bacteria Staphylococcus aureus, Listeria monocytogenes, Bacillus anthracis, and Bacillus cereus, and Gram negative bacteria from the family Enterobacteriacae: Yersinia enterocolitica, Salmonella typhimurium, Shigella sonnei, and Escherichia coli (four strains of E. coli), and non-Enterobacteriacae: Vibrio cholerae, Vibrio vulnificus, and Vibrio parahemolyticus. Foodborne bacterial mixtures of FAMEs were measured by using an attenuated total reflection (ATR)-Fourier transform infrared (FTIR) spectroscopic procedure and discriminated by multivariate analysis. Results showed that the Enterobacteriacae could be discriminated from the vibrios. The identification was at the level of species (for the Bacillus and Vibrio genera) or strains (for the E. coli species). A series of bacterial FAME test samples were prepared and analyzed for accuracy of identification, and all were correctly identified. Our results suggest that this infrared strategy could be used to identify foodborne pathogens.     

DNA fragments from regions involved in surface antigen expression specifically identify Listeria monocytogenes serovar 4 and a subset thereof: cluster IIB (serotypes 4b, 4d, and 4e).

Listeria monocytogenes serotype 4b has frequently been implicated in sporadic as well as epidemic listeriosis. On the basis of pulsed-field fingerprinting, serotype 4b strains, along with strains of serotypes 4d and 4e, constitute one genomic cluster (IIB). We have identified two genomic regions essential for the expression of surface antigens which previously were shown to be specific to cluster IIB strains. A DNA probe of 1.1 kb derived from one of the regions (probe 1) hybridized only with strains of serotypes 4b, 4d, and 4e in Southern blots and dot blots. A different DNA probe of 0.3 kb (probe 2), derived from the other region, hybridized with all serovar 4 strains (serotypes 4b, 4a, 4c, 4d, and 4e). All other L. monocytogenes serotypes were negative with probe 1 or 2. Use of probe 1 in Southern blots of EcoRI-digested genomic DNA revealed a restriction fragment length polymorphism in serotype 4b strains, with the hybridizing EcoRI fragments being 4.5 kb (strains of the epidemic clone) and either 4.5 or 5.0 kb (all other serotype 4b strains). Although the probes hybridized with a special group of Listeria innocua strains which also expressed the surface antigens, the latter could be readily distinguished by the size of the hybridizing EcoRI fragment with probe 1 (ca. 2.2 kb). These data suggest that the combined use of these probes with L. monocytogenes can readily and specifically identify cluster IIB strains as well as the entire serovar 4 complex.     

API Listeria, a new and promising one-day system to identify Listeria isolates.

API Listeria is a new 10-test strip for 24-h biochemical identification of Listeria isolates. With this commercial system, 85% of 646 Listeria strains, including atypical isolates selected for this study, were recognized at the species and subspecies level without a complementary test. A new test differentiates Listeria monocytogenes from L. innocua on the basis of the absence of arylamidase from the former. With this system, 97.7% (252 of 258) of the L. monocytogenes strains tested were correctly identified and differentiated from 99.4% (175 of 176) of the L. innocua strains also tested. Gram-positive bacteria other than Listeria spp. gave quite different biochemical patterns. This system considerably reduced the time needed for conventional identification, since results were available within 18 to 24 h.     

API Listeria, a new and promising one-day system to identify Listeria isolates.

API Listeria is a new 10-test strip for 24-h biochemical identification of Listeria isolates. With this commercial system, 85% of 646 Listeria strains, including atypical isolates selected for this study, were recognized at the species and subspecies level without a complementary test. A new test differentiates Listeria monocytogenes from L. innocua on the basis of the absence of arylamidase from the former. With this system, 97.7% (252 of 258) of the L. monocytogenes strains tested were correctly identified and differentiated from 99.4% (175 of 176) of the L. innocua strains also tested. Gram-positive bacteria other than Listeria spp. gave quite different biochemical patterns. This system considerably reduced the time needed for conventional identification, since results were available within 18 to 24 h.     

Evaluation of the Organon-Teknika MICRO-ID LISTERIA system.

The MICRO-ID LISTERIA system, designed to identify Listeria isolates to species level within 24 h, was compared with conventional biochemical identification. MICRO-ID LISTERIA used in combination with the CAMP test correctly identified 409 (98.8%) of 414 strains isolated from human, animal, food, and environmental sources belonging to the seven species currently defined within the genus Listeria. The kit was easy to use and simple to interpret. However, 8 of the 15 tests (i.e., phenylalanine deaminase, hydrogen sulfide, indole, ornithine decarboxylase, lysine decarboxylase, malonate, urease, and o-nitrophenyl-beta-D-galactopyranoside) were considered superfluous for the differentiation of Listeria spp. The CAMP test was indispensable when using the MICRO-ID LISTERIA system, in particular to differentiate CAMP test-positive L. monocytogenes from the nonhemolytic, rhamnose-positive L. innocua. The hemolytic L. seeligeri and L. ivanovii strains and the nonhemolytic, non-rhamnose-acidifying L. welshimeri strains could also be differentiated from one another only on the basis of their CAMP test results. The very few strains of L. grayi and L. murrayi were easily differentiated from the other nonhemolytic species. Catalase-negative cocci should not be tested, because 12 out of 19 catalase-negative strains (all enterococci) in our test were misidentified as Listeria spp. The MICRO-ID LISTERIA system identified strains within 18 to 24 h and is thus less time-consuming than conventional tests. The system could, therefore, be used together with correctly done CAMP tests for the rapid identification of Listeria isolates, especially food and environmental isolates, for which rapid species differentiation is important.     

Evaluation of the Organon-Teknika MICRO-ID LISTERIA system.

The MICRO-ID LISTERIA system, designed to identify Listeria isolates to species level within 24 h, was compared with conventional biochemical identification. MICRO-ID LISTERIA used in combination with the CAMP test correctly identified 409 (98.8%) of 414 strains isolated from human, animal, food, and environmental sources belonging to the seven species currently defined within the genus Listeria. The kit was easy to use and simple to interpret. However, 8 of the 15 tests (i.e., phenylalanine deaminase, hydrogen sulfide, indole, ornithine decarboxylase, lysine decarboxylase, malonate, urease, and o-nitrophenyl-beta-D-galactopyranoside) were considered superfluous for the differentiation of Listeria spp. The CAMP test was indispensable when using the MICRO-ID LISTERIA system, in particular to differentiate CAMP test-positive L. monocytogenes from the nonhemolytic, rhamnose-positive L. innocua. The hemolytic L. seeligeri and L. ivanovii strains and the nonhemolytic, non-rhamnose-acidifying L. welshimeri strains could also be differentiated from one another only on the basis of their CAMP test results. The very few strains of L. grayi and L. murrayi were easily differentiated from the other nonhemolytic species. Catalase-negative cocci should not be tested, because 12 out of 19 catalase-negative strains (all enterococci) in our test were misidentified as Listeria spp. The MICRO-ID LISTERIA system identified strains within 18 to 24 h and is thus less time-consuming than conventional tests. The system could, therefore, be used together with correctly done CAMP tests for the rapid identification of Listeria isolates, especially food and environmental isolates, for which rapid species differentiation is important.     

Division of Listeria monocytogenes serovar 4b strains into two groups by PCR and restriction enzyme analysis.

Altogether, 133 strains of Listeria monocytogenes serovar 4b were investigated. A segment of 2,916 bp containing parts of the two genes inlA and inlB in L. monocytogenes was amplified by the PCR technique. The PCR product obtained was cleaved with the restriction enzyme AluI, and the fragments generated were separated by gel electrophoresis, leading to two distinct groups: PCR-restriction enzyme analysis groups I and II, containing 37 and 96 strains, respectively. The PCR-restriction enzyme analysis method described in this paper could be a useful tool for the subtyping of L. monocytogenes serovar 4b strains.     

An improved amplified fragment length polymorphism (AFLP) protocol for discrimination of Listeria isolates

Nine restriction enzyme combinations and 108 different primer combinations were initially tested for suitability for amplified fragment length polymorphism (AFLP) analysis of Listeria monocytogenes; the combination of HindIII and HpyCH4IV showed consistently strong signals on gels, amplified an adequate number of DNA fragments and detected polymorphism among closely related strains based on AscI macrorestriction profiles. AFLP also distinguished between L. monocytogenes, L. innocua, L. ivanovii, L. seeligeri, L. welshimeri and L. grayi species. All Listeria species showed species-specific clusters, with less than 33% similarity between different species. A total of 34 L. monocytogenes strains were characterised by using both AFLP and pulsed-field gel electrophoresis (PFGE). The results of AFLP analysis of L. monocytogenes strains were in concordance with those obtained by PFGE. Both methods identified 29 different genotypes of L. monocytogenes and had a high discrimination index (> 0.999). By combining the results of AFLP and PFGE, subtype discrimination was further improved. Numerical analysis of both AFLP and PFGE profiles yielded three genomic groups of L. monocytogenes strains. AFLP was found to be faster and less labour-intensive than PFGE. We conclude that the AFLP protocol is a highly discriminatory, reproducible and valuable tool in characterisation of Listeria strains and may also be suitable for Listeria species identification.     

Characteristic of Listeria spp. bacteria isolated from food products

The frequency of occurrance of Listeria strains in different food products was determined. Biochemical characteristic of the isolated strains was achieved in accordance with procedure included in PNEN ISO 11290 standard, genus was determined byApiListeria (bioMéieux) test. Sensitivity to selected antibacterial medicines was investigated using disck method and Mueller-Hinton 2 Agar medium. From the 577 examinated food samples 126 strains of Listeria were isolated and among them: 34.1% L. monocytogenes, 36,5% L. welshimeri, 19.0% L. innocua, 3.17% L. grayi, 0.79% L. seeligeri, 0.79% L. seeligeri/welshinmeri and 5.56% L. ivelshimeri/innocua. L. monocytogenes strains most often were found in minced pork, culinary products and in frozen vegetables. On the base of ApiListeria (bioMéieux) test the isolated L. monocytogenes strains were qualified into 2 biochemical types. It was found that all L. monocytogenes were sensitive to sulphametaksazol/trimetoprim and ampicyllin, 25% of strains were moderatety sensitive to penicillin and only 2 L. monocytogenes strains were resistant to gentamicin. Presence of Listeria spp. microorganisms in food products may be an production hygiene indicator for critical control point and show the possibility of contamination with L. monocytogenes strains.     

Analysis of surface proteins of Listeria in relation to species, serovar and pathogenicity.

SDS extracts of whole bacteria, representing five species and 15 serovars of Listeria, were analysed by SDS-PAGE and by immunoblotting with serum directed against whole formalin-treated L. monocytogenes. Profiles of L. monocytogenes were very different from those of other species of Listeria (i.e. L. innocua,L.welshimeri, L. seeligeri and L. ivanovii). This low degree of similarity between species was found even in the case of common serovars. Within the species L. monocytogenes, protein patterns were characterized, on the one hand, by a high degree of homogeneity between all strains of the same serovar and, on the other hand, by large differences between serovars, especially between sv. 1/2 and 4b. Thus we have identified major, surface-located protein antigens, specific for L. monocytogenes, either common to all serovars (64 and 68 kDa) or characteristic of certain serovars: 98 kDa for sv. 1/2 and 3; 76 and 78 kDa for sv. 4b, 4d and 4e; and 80 and 100 kDa for sv. 4a and 4c. Moreover, some of these bands (68 and 98 kDa) might be related to virulence, since differences were noticed between the profiles of haemolytic L. monocytogenes vs. 1/2a differing only in their virulence for immunocompromised mice. All these results confirmed, for the first time, the classification of Listeria obtained previously by genomic studies. They should help in the identification of new virulence factors and the development of easier and more specific methods of detection and identification.     

Inhibition of Listeria monocytogenes by food-borne yeasts

Many bacteria are known to inhibit food pathogens, such as Listeria monocytogenes, by secreting a variety of bactericidal and bacteriostatic substances. In sharp contrast, it is unknown whether yeast has an inhibitory potential for the growth of pathogenic bacteria in food. A total of 404 yeasts were screened for inhibitory activity against five Listeria monocytogenes strains. Three hundred and four of these yeasts were isolated from smear-ripened cheeses. Most of the yeasts were identified by Fourier transform infrared spectroscopy. Using an agar-membrane screening assay, a fraction of approximately 4% of the 304 red smear cheese isolates clearly inhibited growth of L. monocytogenes. Furthermore, 14 out of these 304 cheese yeasts were cocultivated with L. monocytogenes WSLC 1364 on solid medium to test the antilisterial activity of yeast in direct cell contact with Listeria. All yeasts inhibited L. monocytogenes to a low degree, which is most probably due to competition for nutrients. However, one Candida intermedia strain was able to reduce the listerial cell count by 4 log units. Another four yeasts, assigned to C. intermedia (three strains) and Kluyveromyces marxianus (one strain), repressed growth of L. monocytogenes by 3 log units. Inhibition of L. monocytogenes was clearly pronounced in the cocultivation assay, which simulates the conditions and contamination rates present on smear cheese surfaces. We found no evidence that the unknown inhibitory molecule is able to diffuse through soft agar.     

Proteomic expression profiles of virulent and avirulent strains of Listeria monocytogenes isolated from macrophages

Listeria monocytogenes is able to survive and proliferate within macrophages. In the current study, the ability of three L. monocytogenes strains (serovar 1/2a strain EGDe, serovar 4b strain F2365, and serovar 4a strain HCC23) to proliferate in the murine macrophage cell line J774.1 was analyzed. We found that the avirulent strain HCC23 was able to initiate an infection but could not establish prolonged infection within the macrophages. By contrast, strains EGDe and F2365 proliferated within macrophages for at least 7 h. We further analyzed these strains by comparing their protein expression profiles at 0 h, 3 h, and 5 h post-infection using multidimensional protein identification technology coupled with electrospray ionization tandem mass spectrometry. Our results indicated that similar metabolic and cell wall associated proteins were expressed by all three strains at 3 h post-infection. However, increased expression of stress response and DNA repair proteins was associated with the ability to proliferate in macrophages at 5 h post-infection. By comparing the protein expression patterns of these three L. monocytogenes strains during intracellular growth in macrophages, we were able to detect biological differences that may determine the ability of L. monocytogenes to survive in macrophages.     

RAPD analysis, serotyping, and esterase typing indicate that the population of Listeria monocytogenes strains recovered from cheese and from patients with listeriosis in Belgium are different.

Populations of Listeria monocytogenes strains isolated in Belgium from cheese and from patients with listeriosis were characterised by randomly amplified polymorphic DNA (RAPD) analysis using two 10-mers primers (OPA-04 and OPA-13). High discrimination levels were obtained with each of these primers alone (discrimination indices (DI) of 0.899 and 0.935 for OPA13 and OPA04, respectively) or in combination (DI of 0.960). The clustering of strains obtained by RAPD was compared with a clustering previously made using serotyping and esterase typing. RAPD allowed the subdivision of each serovar cluster and of most of the clusters determined by the polymorphism of the bacterial esterases. Our analysis indicates that the population of strains of L. monocytogenes found in cheese differs from the one isolated from patients with listeriosis during the same period.     

Revision of the antigenic structure of genus Listeria

O-antigenic structure of genus Listeria was studied, using antisera (obtained from rabbits) against different O-antigens of reference strains of each serovar. The titres of sera were determined by agglutination using antigens of the same reference strains as well. Some differences from the actual scheme were found: serum antifactor-IX gave a lower titre than expected against antigens 4ab and 6b, while the titre observed against antigen 4b was higher than the expected in this case. Serum antifactor-VIII presented a higher titre than could be expected against antigen 6b. The strains of serovars 4d and 4e used in this experience were impossible to distinguish, and could have been classified in the same serovar. We could not obtain serum antifactor-XI from serovar 6b after several trials. From these differences we propose some modifications of the current antigenic scheme of genus Listeria.     

Amplified intergenic locus polymorphism as a basis for bacterial typing of Listeria spp. and Escherichia coli

DNA-based methods are increasingly important for bacterial typing. The high number of polymorphic sites present among closely related bacterial genomes is the basis for the presented method. The method identifies multilocus genomic polymorphisms in intergenic regions termed AILP (amplified intergenic locus polymorphism). For each locus, a pair of unique PCR primers was designed to amplify an intergenic sequence from one open reading frame (ORF) to the adjacent ORF. Presence, absence, and size variation of the amplification products were identified and used as genetic markers for rapidly differentiating among strains. Polymorphism was evaluated using 18 AILP sites among 28 strains of Listeria monocytogenes and 6 strains of Listeria spp. and 30 AILP markers among 27 strains of Escherichia coli. Up to four alleles per locus were identified among Listeria strains, and up to six were identified among E. coli strains. In both species, more than half of the AILP sites revealed intraspecies polymorphism. The AILP data were applied to phylogenetic analysis among Listeria and E. coli strains. A clear distinction between L. monocytogenes and Listeria spp. was demonstrated. In addition, the method separated L. monocytogenes into the three known lineages and discriminated the most common virulent serotypic group, 4b. In E. coli, AILP analysis separated the known groups as well as the virulent O157:H7 isolates. These findings for both Listeria and E. coli are in agreement with other phylogenetic studies using molecular markers. The AILP method was found to be rapid, simple, reproducible, and a low-cost method for initial bacterial typing that could serve as a basis for epidemiological investigation.