Abstract
This study aimed to investigate phenotypic virulence factors and virulence associated enzymes, assess the biofilm forming capacity of human and animal origin isolates and to verify the presence of selected adhesion and regulation related genes. A total of 40 P. mirabilis isolates, comprising 23 human and 17 feline isolates. The isolates were screened for the production of several virulence-associated enzymes, including hemolysin, urease, protease, lecithinase, lipase, α-amylase, and gelatinase. Biofilm formation was evaluated using the microtiter plate assay, while molecular detection of the ompA and luxR genes was performed by conventional polymerase chain reaction (PCR). The results showed that urease was the most common virulence factor (100% of isolates). Other virulence-associated enzymes were produced to variable proportions: α-amylase was noted in 60% of isolates, lipase in 45%, and hemolysin and protease each were detected in 40% of isolates. Production of gelatinase was evident in 37.5 % isolates. In contrast, the least determined enzyme was lecithinase (10% of isolates). The α-amylase-positive among fecal isolates was more frequent than those recovered from ears, while ear isolates, especially those from cats, demonstrated higher gelatinase and protease activities than their fecal counterparts. In the biofilm assay, 34 out of 40 (85%) isolates were able to form a biofilm. Also the 13 isolates (32.5%) were classified as strong biofilm producers, 15 isolates (37.5%) moderate and six isolates (15%) weak producers and six non-adherent strains (15%). These results suggest that biofilm formation is a virulence-related phenotype of the majority of P. mirabilis clinical isolates tested in this study. Molecular analysis showed the ompA gene in 36 out of 40 isolates (90%; Fisher’s exact test, human 82.61% vs feline 100%, P = 0.123), including 19 human isolates and all those obtained from cats. The luxR gene, detected by conventional PCR, was found in only 13 isolates (32.5%), reflecting genetic diversity among the isolates examined. As only gene presence, and not gene expression, was assessed in this study, the detection of ompA may indicate a genetic potential for adhesion-related virulence, while the limited detection of luxR suggests that classical AHL-mediated quorum-sensing regulation is not uniformly present among the isolates. The study data indicate that biofilm formation in Proteus mirabilis is a multifactorial virulence factor which depends on the interaction between phenotypic and genotypic factors. Meanwhile, biofilm-forming ability was demonstrated by the great majority of isolates and co-occurred at the population level with a high prevalence of ompA gene detection but only partial luxR gene detection, suggesting that adhesion- and regulation-associated genes may contribute to persistence. These data support the hypothesis that virulence of P. mirabilis is consistent with a network of adhesive, motility-related, enzymatic, and regulatory factors that may contribute to environmental adaptation and persistent colonization in human or animal hosts.

Key words: Proteus mirabilis, Enzymatic virulence factors, Biofilm, OmpA, LuxR, Feline isolates