Short Communication Comparison of two biochemical methods for identifying Corynebacterium pseudotuberculosis isolated from sheep and goats Belén Huerta a,1 , Lidia Gómez-Gascón a,1 , Ana I. Vela b,c , José F. Fernández-Garayzábal b,c , Almudena Casamayor b,c , Carmen Tarradas a , Alfonso Maldonado a,⇑ a Departamento de Sanidad Animal, Facultad de Veterinaria CEIA3, Campus Universitario de Rabanales, 14071 Córdoba, Spain b Departamento de Sanidad Animal, Facultad de Veterinaria, Universidad Complutense, 28040 Madrid, Spain c Centro de Vigilancia Sanitaria Veterinaria (VISAVET), Facultad de Veterinaria, Universidad Complutense, 28040 Madrid, Spain article info Article history: Accepted 10 December 2012 Keywords: Corynebacterium pseudotuberculosis 16S rRNA gene sequencing Biochemical identification abstract The biochemical pattern of Cowan and Steel (BPCS) was compared with a commercial biochemical strip for the identification of Corynebacterium pseudotuberculosis isolated from small ruminants. On 16S rRNA gene sequencing, 40/78 coryneform isolates from the lymph nodes of sheep and goats with lesions resembling caseous lymphadenitis were identified as C. pseudotuberculosis. The sensitivities of the BPCS and the commercial biochemical strip relative to 16S rRNA sequencing were 80% and 85%, and their specificities were 92.1% and 94.7%, respectively; the level of agreement between the BPCS and the commercial biochemical strip was high (j = 0.82). Likelihood ratios for positive and negative results were 10.0 and 0.22 for the BPCS, and 16.0 and 0.16 for the commercial biochemical strip, respectively. These results indicate that the BPCS and the commercial biochemical strip are both useful for identifying C. pseudotuberculosis in veterinary microbiology laboratories. Ó 2012 Elsevier Ltd. All rights reserved. Corynebacterium pseudotuberculosis causes caseous lymphade- nitis (CL) in small ruminants. Diagnosis of CL is largely based on clinical signs and isolation of the bacterium from abscesses. However, C. pseudotuberculosis must be differentiated from other abscess-forming bacteria, such as Trueperella pyogenes and Pasteu- rella multocida (Dorella et al., 2006). Molecular techniques, such as 16S rRNA sequencing (Çetinkaya et al., 2002) and PCR-restriction fragment length polymorphism (Pavan et al., 2012) are reliable, but expensive; therefore, most laboratories rely on biochemical methods (Gavin et al., 1992; Soto et al., 1994; Funke et al., 1997; Adderson et al., 2008). Several studies have compared different biochemical methods with the API Coryne Strip (API CS; bioMérieux) for identification of coryneform bacteria (Gavin et al., 1992; Soto et al., 1994; Funke et al., 1997; Adderson et al., 2008), but there have been no studies to assess the diagnostic usefulness of these methods for identification of C. pseudotuberculosis in veterinary microbiology laboratories. The aim of this study was to compare the usefulness of the biochemical pattern of Cowan and Steel (2004) (PBCS) and the API Coryne Strip for identification of C. pseudotuberculosis among bacterial isolates recovered from the lymph nodes of sheep and goats with lesions resembling CL using 16S rRNA sequencing as a reference. Seventy-eight coryneform bacteria isolated from sheep (n = 52) and goats (n = 26) from January to May 2010 were evaluated. Col- onies of bacteria with morphological characteristics compatible with C. pseudotuberculosis were tested using the BPCS and API CS methods; 16S rRNA gene sequencing was used to confirm their identity (Vela et al., 2003). BPCS tests included production of cata- lase and urease, acidification of trehalose and maltotriose, syner- gistic haemolysis with Rhodococcus equi and antagonistic haemolysis with Staphylococcus aureus. Isolates positive for cata- lase and urease, which enhanced haemolysis in conjunction with R. equi, but which inhibited the b-haemolysin of S. aureus and were negative for trehalose and maltotriose, were identified as C. pseu- dotuberculosis. Isolates that failed in at least one of these tests were deemed not to be C. pseudotuberculosis. The API CS method involves a battery of 20 tests and the results are interpreted using the API- web system. 2 The sensitivity (Se), specificity (Sp), likelihood ratio (LR), Kappa value (j) and difference between the proportion of isolates identi- fied as C. pseudotuberculosis by each method (McNemar’s test; P < 0.05) were determined using the Diagnostic Test Calculator (Winepiscope for Window 2.0). 3 Positive and negative predictive values (PV+ and PVÀ, respectively) and LR values were used as mea- sures of diagnostic usefulness. 1090-0233/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tvjl.2012.12.008 ⇑ Corresponding author. Tel.: +34 957 218728. E-mail address: sa1magaa@uco.es (A. Maldonado). 1 These authors contributed equally to this work. 2 See: https://apiweb.biomerieux.com/servlet. 3 See: http://www.winepi.net/. The Veterinary Journal 196 (2013) 552–554 Contents lists available at SciVerse ScienceDirect The Veterinary Journal journal homepage: www.elsevier.com/locate/tvjl