Virulence and Antimicrobial Resistance Characterization of Glaesserella parasuis Isolates Recovered from Spanish Swine Farms

Abstract

Glaesserella (Haemophilus) parasuis, the causative agent of Glässer's disease, is present in most pig farms as an early colonizer of the upper respiratory tract. It exhibits remarkable variability in virulence and antimicrobial resistance (AMR), with virulent strains capable of inducing respiratory or systemic disease. This study aimed to characterize the virulence and the AMR profiles in 65 G. parasuis isolates recovered from Spanish swine farms. Virulence was assessed using multiplex leader sequence (LS)-PCR targeting vtaA genes, with all isolates identified as clinical (presumed virulent). Pathotyping based on ten pangenome genes revealed the virulent HPS_22970 as the most frequent (83.1%). Diverse pathotype profiles were observed, with 29 unique gene combinations and two isolates carrying only potentially non-virulent pangenome genes. AMR phenotyping showed widespread resistance, with 63.3% classified as multidrug resistant, and high resistance to clindamycin (98.3%) and tylosin (93.3%). A very strong association was found between certain pathotype genes and AMR phenotypes, notably between the virulent HPS_22970 and tetracycline resistance (p < 0.001; Φ = 0.58). This study reveals the wide diversity and complexity of G. parasuis pathogenicity and AMR phenotype, emphasizing the need for the targeted characterization of clinical isolates to ensure appropriate antimicrobial treatments and the implementation of prophylactic measures against virulent strains.

Keywords: Glaesserella parasuis; Haemophilus parasuis; TbpB clustering; antimicrobial susceptibility profiling; pathotyping; pig; porcine respiratory disease complex; vtaA genes.

Figure 1

Pathotype characterization of 65 G. parasuis isolates from Spanish swine farms: (a) pathotype clustering based on the presence of ten pangenome genes, using the unweighted pair group method with arithmetic mean (UPGMA) as the hierarchical clustering method; (b) frequency (%) of each pathotype gene; and (c) principal component analysis (PCA) of the ten evaluated pangenome genes, showing grouping based on pathologic process and TbpB cluster of each G. parasuis isolate.

Figure 2

Antimicrobial resistance (AMR) characterization at the class level of 60 G. parasuis isolates from Spanish swine farms: (a) AMR phenotype clustering based on antimicrobial classes, using the unweighted pair group method with arithmetic mean (UPGMA) as the hierarchical clustering method, and (b) principal component analysis (PCA) of the AMR patterns, showing grouping based on the TbpB cluster of each G. parasuis isolate.

Figure 3

Network associations between pathotype genes and antimicrobial resistance (AMR) at the class level in 60 G. parasuis isolates from Spanish swine farms. Node size is determined by the percentage occurrence of the pathotype gene or AMR class. Edge size is proportional to the magnitude of the association based on the Φ coefficient. The network was constructed using significant associations (p < 0.05).