ABSTRACT
The ultimate bacterial genotyping method must involve comparison of whole genome sequence data. Sequence data is now available for many bacterial genomes and, in some cases, for different isolates of the same species in various genome databases (e.g., www.tigr.org/tdb/mdb/mdbcomplete.html and www.ncbi.nlm.nih.gov). However, until the comparison of sequence data from every bacterial genome is technically and financially feasible, genome sampling will remain the method of choice. This sampling is usually in the form of some kind of restriction fragment length polymorphism (RFLP) analysis or sequencing of selected regions of the genome (multilocus sequence typing, MLST).1 One of these genome-sampling methods, FAFLP, has emerged as the method of choice for a variety of applications, including outbreak investigation and molecular epidemiological studies.2-6
B. FAFLP
Based on the radioactive AFLP method described by Vos et al.,7 the FAFLP method is a simple procedure that uses a specific PCR to select, amplify, and fluorescently label a subset of fragments generated by the restriction enzyme digestion of genomic DNA. These fragments are then detected and sized on an automated sequencer. The advantages of FAFLP include speed (compared with techniques such as pulsed-field gel electrophoresis [PFGE] and objectivity, as the fragments are precisely sized by machine against an internal size standard run together with the sample. As well as being precise, FAFLP data can also be checked against the sizes of FAFLP fragments predicted in silico using the complete genome of a sequenced bacterial strain of the same species. In silico predictions and in vitro FAFLP data compared surprisingly well (97 to 98%).8 Importantly, the fragments generated are derived from multiple sites throughout the genome,8,9 the results are not weighted in favor of particular genes such as those that code for an easily detectable phenotype (e.g., antibiotic resistance) or, as is the case for MLST, in favor of other selected genes (both “housekeeping” and more variable genes). Furthermore, FAFLP results have been shown to be reproducible between laboratories as long as generated with the same sequencer, the same run conditions, and therefore identical sizing.
