In many cases, assigning a name to bacterial isolates can be a difficult task. A wide range of bacterial species, including those that cause concern to the food industry (e.g. pathogenic bacteria), may pose serious problems in terms of identification. This has led to development of molecular identification methods, especially those based on PCR. The automation of many techniques, coupled with development of statistics and bioinformatics for microbiology, have led to a modification or replacement of conventional procedures in food microbiology laboratories.
The use of DNA probes for genes coding for rRNA offers a great potential in microbial identification. As rRNA (or other gene) sequences have become increasingly available, comparisons have revealed oligonucleotide stretches which are specific for different microbial taxa. These oligonucleotides can be labeled and used as probes in hybridization experiments with DNA of unknown isolates. Currently, oligonucleotide probes for the identification of almost all food-associated LAB are available (Schleifer, et al. 1995). A very useful tool is probeBase - an online resource for rRNA-targeted oligonucleotide probes (Loy, et al. 2003). The site (www.microbial-ecology.net/probebase) contains all the necessary information for probe sequences and protocols (even for FISH applications), as well as references concerning development and applications of the taxa-specific probes.
Different formats can be used for probe assays. For the dot-blot assay, the target nucleic acid has to be extracted from the cell and immobilized on a membrane. Then, either radioactively or non-radioactively labeled probes can be used for hybridization with the immobilized nucleic acid. The introduction of non-radioactive labeling methods (e.g. those based on chemiluminescence) has greatly facilitated the application of probes in food microbiology. A variation of this approach is the use of colony hybridization using group-specific probes. The advantage of this
Table 1.2 Advantages and Limitations of Using the Principal Molecular Techniques for Microbial Identification and Typing
- DNA-DNA hybridization
(e.g. rRNA-targeted oligo-nucleotide probes; dot-blot; colony hybridization, etc.)
ARDRA-PCR, ITS (or RISA)-PCR; metabolic genes amplification)
- DNA sequencing
B. Typing RFLP Methods
High discrimination level; easy interpretation; high reproducibility
Rapidity; high reproduc-ibility; easy to use and interpret; low or moderate costs
High discrimination level; best accuracy and repro-ducibility; automated platforms available; public databases available
High discrimination level; easy interpretation; high reproducibility; automated platforms available Excellent discrimination level; excellent reproduc-ibility; easy interpretation; public databases available
High discrimination level; rapidity; easy use and interpretation; low costs Rapidity; easy use and interpretation; high reproduc-ibility; low costs Moderately easy to use and interpret; high discrimination power; high repro-ducibility; automated platforms available
Cumbersome; time-consuming; expensive
Moderate discrimination level (to be raised by ampli-con restriction, e.g. by ARDRA-PCR) High technical competence needed; very expensive
Cumbersome; time-consuming; expensive
Cumbersome; difficult to use; long time to get result; moderate to high costs
Moderate reproducibility; no public databases available
Moderate discrimination level; no public databases available
High costs; no public databases available
Acronyms legend: ARDRA-PCR, Amplification Ribosomal DNA Restriction Analysis-PCR; ITS-PCR, Internal Transcribed Spacer-PCR; RISA-PCR, rRNA gene Internal Spacer Analysis-PCR; RFLP, Restriction Fragment Length Polymorphism; REA-PFGE, Restriction Endonuclease Analysis-Pulsed Field Gel Electrophoresis; RAPD-PCR, Randomly Amplified Polymorphic DNA-PCR; Rep-PCR, Repetitive element sequence-based-PCR; AFLP, Adaptor Fragment Length Polymorphism.
technique is that it allows the specific differentiation and quantification of target population(s) without the need of colony isolation and subculturing. In colony hybridization, bacteria are plated on membranes layered on appropriate agar media and allowed to form colonies. After lysis of the colonies, hybridization with a labeled probe will show which and how many of the colonies contain the target sequence. The most recent development of colony hybridization is the in situ detection and identification (or ISH and its variant FISH) of whole cells with fluorescently labeled nucleotides.
Ribosomal RNA-targeted oligonucleotides have been used for the specific identification of LAB and yeasts (Schleifer, et al. 1995; Ampe, et al. 1999a). The use of DNA probes in colony or dot-blot hybridization experiments allowed the LAB community to be controlled in wine at different stages of wine-making, and to monitor the evolution of thermophilic lactobacilli belonging to Lactobacillus helveticus, Lb. delbrueckii, and Lb. fermentum during the early phases of Grana Padano cheese-making (Lonvaud-Funel, et al. 1991; Sohier and Lonvaud-Funel 1998; Giraffa, et al. 1998). Comparison of the use of rRNA probes and conventional methods also enabled identification of strains of Lb. sakei and Lb. curvatus isolated from meat (Nissen and Dainty 1995). Erlandson and Batt (1997) described a method using hydrophobic grid membrane filter colony hybridization for quantitative strain-specific detection of lactococci in bacterial populations. More recently, dot-blot hybridization has been applied to detect yogurt LAB in total fecal DNA (del Campo, et al. 2005). Colony hybridization has been applied for LAB identification (Betzl, et al. 1990; Hertel, et al. 1991), to characterize the microflora of Fontina cheese (Senini, et al. 1997), and to search for the presence of virulence genes related to diarrheal pathogenesis in Escherichia coli strains isolated from Pozol, an acid-fermented maize beverage consumed in Mexico (Sainz, et al. 2001).
Although species identification can be obtained with a high level of discrimination and reproducibility, hybridization techniques are being abandoned for taxonomic purposes. These methods are not particularly suited to the laboratory environment because protocols are generally cumbersome, time-consuming and expensive. A big obstacle is that multiple hybridizations for simultaneous identifications of more than one species are not possible. Another problem can be mis-identification, which may result from the presence of identical probe target sequences in phylogenetically diverse organisms. This has led to the development of commercial kits for specific microbial groups (e.g. food-associated pathogenic species), which result in the significant reduction of costs for those laboratories performing such tests. Another recent improvement has been the introduction of the "multiple probe concept," which is based upon the assumption that the problem of mis-identification can be reduced by the simultaneous application of multiple probes targeting independent sites (Behr, et al. 2000).
The on-line availability of DNA sequences of ribosomal RNA (rRNA) genes and rRNA gene spacers of practically all the known microbial species, and the presence of taxa-specific oligonucleotide stretches within the ribosomal locus has enabled these genes (or portions of them) to be routinely PCR-amplified and examined for differences indicative of genus and species identity. The rRNA gene sequences of the target taxa are aligned with the sequences of phylogenetically associated organisms and, on the basis of the presence of both conserved and variable regions within the ribosomal (or other well-conserved) genes, genus- or species-specific oligonucle-otide primers are designed. Amplified products, which can range from the single ribosomal genes to part or all the ribosomal locus, can be obtained by either simplex or multiplex PCR. A variant of this approach is ITS (or RISA)-PCR, whose principle has been explained above. Using primers published by Jensen, et al. (1993), several authors showed that ITS-PCR between the 16S and the 23S rRNA genes can produce amplicon profiles which are characteristics for each bacterial species when examined with high resolution non-denaturing acrylamide-bisacrylamide gel electrophoresis. Similar approaches have been applied for yeast identification (Arroyo-Lopez, et al. 2006). Amplified products are then examined as a whole or subjected to restriction endonuclease analysis (such as in the case of amplified ribosomal DNA restriction analysis, ARDRA), which evidences RFLP of amplified genes within and between taxa and allows increasing the taxonomic resolution.
Several PCR amplification protocols are presently available for practically all food-associated Lactobacillus spp. (Giraffa and Neviani 2000), Leuconostoc spp. (Ward, et al. 1995; Moschetti, et al. 2000), and Pediococcus spp. (Mora, et al. 1997). Recently, the amplification by PCR of the intergenic spacer region (IGS) of rRNA gene followed by restriction RFLP analysis was evaluated as a potential method for distinguishing the 16 species belonging to the genus Debaryomyces (Quiros, et al. 2006). Moreover, the increasing availability of non-ribosomal (metabolic) gene sequences has further revolutionized the PCR-based diagnostics. For example, the pepIP and lacZ genes can be respectively used to distinguish Lb. del-brueckii subsp. lactis from Lb. delbrueckii subsp. bulgaricus (Torriani, et al. 1999) and to identify St. thermophilus (Lick, et al. 1996). Similarly, Lactococcus lactis subsp. lactis and L. lactis subsp. cremoris can be distinguished on the basis of primers designed on the histidine biosynthesis operon (Corroler, et al. 1998). Fortina, et al. (2001) described a multiplex PCR based on pepC, pepN and htrA targeted primers to identify Lb. helveticus. Jackson, et al. (2004) optimized a genus and species-specific multiplex PCR based on the sodA gene for identification of enterococci. Clearly, the use of non-ribosomal genes for taxonomic purposes is opening new possibilities to study the ecological evolution of microorganisms on the basis of the polymorphism of metabolic genes.
DNA sequencing is considered the gold standard for microbial identification. The introduction in the early '90s of automated DNA sequencing machines and the development of bioinformatics (see later) have allowed individual laboratories to increase their output of DNA sequences from a few thousand base pairs per week to millions of base pairs per week, with much less effort and greater accuracy and reproducibility. DNA sequencing generally begins with PCR amplification of DNA (or RNA) directed at genetic regions of interest, followed by sequencing reactions, which can be performed either by use of DNA sequencing or capillary gels of the amplified products. During electrophoresis, these fluorescently-labeled products are excited by an argon laser and are automatically detected. The resulting data are stored in digital form for subsequent processing into the final sequence with the aid of specialized software.
A number of sequence-based identification systems have been used to analyze the rRNA operon genes as well as other conserved genes in bacteria. Concerning the rRNA operon, the 16S rRNA gene is usually amplified for bacterial identification, whereas the 26S rRNA gene is generally used for yeast identification. Once the whole gene sequence is determined, it is compared to sequences from known microorganisms by the aid of specialized software programs and/or on-line tools. The programs use powerful algorithms to construct a phylogenetic tree (dendrogram) of how closely the sequences match and, hence, how closely the microorganisms are related. The accumulating set of information on rRNA sequences has proved to be effective for comparative identification of microorganisms, leading to recognition of thousands of microbial species. However, with regard to food-associated bacteria, non-ribosomal genes such as the recA gene (Felis, et al. 2001; Torriani, et al. 2001b) and the rpoB gene (Rantsiou, et al. 2004; Renouf, et al. 2006), are increasingly being used as phylogenetic markers for taxonomic purposes.
On the other hand, DNA sequencing is generally expensive and requires a high degree of technical competence to perform. Furthermore, sequencing all of the rRNA genes is not a practical method for routine microbial identification. This stimulated the sequencing of the hypervariable region in the 5'-end of the 16S rRNA gene (approx 500 bp), which is sufficient for specific identification of most bacterial species (Patel, et al. 2000). Finally, automated DNA sequencers are still very expensive, with some costing in excess of $100,000 (Olive and Bean 1999). To meet the increasing needs of the food industry, a number of private companies or associations (e.g. Belgian Coordinated Collections of Microorganisms [BCCM-LMG], Campden and Chorleywood Food Research Association [CCFRA]) provide a service to reliably and definitively characterize and identify bacterial isolates in a few hours at reduced costs (Dawson 2001).
Was this article helpful?