**4.2.3 Ribotyping**

The Fingerprinting of rRNA coding sequences, termed ribotyping, describes the hybridization of restriction-digested DNA fragments with probes specific for rDNA.

Multiple copies of the rRNA operon are present within the *Salmonella* chromosome (Mendoza & Landeras, 1999). The rRNA genes themselves are quite homologous among these copies and between isolates, but the intervening sequences vary in length and nucleotide composition.

Ribotyping begins with separating endonuclease-digested chromosomal DNA on agarose gels, DNA then is transferred to a membrane and fragments are hybridized to a probe that recognizes 16S and 23S rRNA. Analysis of multiple restriction endonucleases can improve the discriminatory powers of ribotyping (Millemann et al., 1995).

Ribotype analysis is clearly able to subtype some of the isolates that fall within some common serotypes and phage types (Landeras et al., 1996). Lin et al. (Lin et al., 1996) detected 7 different ribotypes among 17 *S. enteritidis* PT 8 isolates when chromosomal DNA was digested with SphI. Using rRNA gene restriction patterns to investigate the relatedness of S. Enteritidis strains isolated in São Paulo, from 1975 to 1995; Fernandes et al. showed that ribotyping is a genomic profiling method that is reproducible and suitable for tracing the spread of S. Enteritidis. They found that the restriction endonuclease SphI discriminated best between subtypes of this serotype. Dambaugh et al. presented evidence suggesting that the ribotyping of Salmonella using the restriction enzyme PvuII increased the incidence of discreet ribotype patterns for the most common Salmonella serovars. This study evaluates the potential of PvuII to generate serotype-specific DNA fingerprints. However, studies have identified isolates that belong to different phage types yet demonstrate identical ribotypes (Fontana et al., 2002). Therefore, ribotyping is considered not suitable for local epidemiological studies or surveillance studies in a restricted region (Riley, 2004).

Comparisons of ribotyping with PFGE have been somewhat unpredictable and often depend on the enzymes used for digestion as well as the nature of the population being tested. Several studies have found PFGE to be more discriminating than ribotype analysis (Fontana et al., 2002) while others have found the two procedures equivalent (Navarro et al., 1996) or ribotype analysis superior (Liebana et al., 2001). Ribotype analysis using two restriction enzymes, Pst I -SphI or HindIII - EcoRV, can improve discrimination (Liebana et al., 2001). Particular care must be taken when analyzing chromosomal patterns of *S. typhi.* The rapid genomic reassortment that occurs in *S. typhi* can affect ribotype analysis (Ng et al., 1999).

Though most laboratories continue to perform ribotyping manually, machinery has been developed to perform this entire procedure in an automated fashion. Data is stored electronically and the banding pattern from a particular organism can be compared to the entire databank stored in the computer. In contrast to PFGE, the time required to perform automated ribotyping is minimal; hybridization results can be obtained within 4 hours. A recent study tracking the rise of a multi-drug resistant, cephalosporin-resistant *S. newport*  proposes to use automated ribotyping as a way to rapidly identify the *newport* serotype and PFGE to further evaluate strain associations (Fontana et al., 2002). The major drawbacks of automated ribotyping are the high reagent costs per isolate and the cost of the automated riboprinter itself.

Laconha et al. and Ridley et al. investigated the genotypic differences between strains of *Salmonella* by plasmid analysis, ribotyping and pulsed-field gel electrophoresis (PFGE). The results obtained by those researchers indicated that PFGE may offer a better level of discrimination of *S. Enteritidis* types than other genotypic methods. Conversely, other epidemiological studies of *S. Enteritidis* have demonstrated that PFGE methodology has a lower discriminatory capacity than ribotyping (Olsen et al. 1994; Thong et al. 1998).
