Histidine Decarboxylase in Enterobacteriaceae Revisited

In 1961, Møller's decarboxylation broth was used by Havelka for the detection of histidine decarboxylase (HDC), but only 21 strains of Enterobacteriaceae were tested, and their taxonomic position assigned at that time did not allow interpretation of the results reported (5). Later the presence of HDC was investigated by Enjalbert et al. as an additional test to differentiate Enterobacter aerogenes from Klebsiella pneumoniae and other related species (3). These authors examined a large number of strains belonging to the Voges-Proskauer-positive group of Enterobacteriaceae using a method based on histamine extraction from a broth culture and its detection by the addition of a diazonium salt of sulfanilic acid. All E. aerogenes strains were positive, whereas other species were negative.

In this study, we looked at the presence of HDC in 693 strains of Enterobacteriaceae belonging to the different genera and species encountered in clinical specimens. Identification was performed according to Farmer (4). Dubious or atypical strains were confirmed by sequencing of the 16S rRNA gene carried out as described previously (10). The strains are listed in Table 1.

A simple method was used, based on the medium described by Taylor for the detection of lysine decarboxylase (9) modified as follows: l-histidine hydrochloride (Sigma-Aldrich, St. Louis, Mo.), 7 g; yeast extract, 3 g; glucose, 1.5 g; bromcresol purple (alcoholic solution, 1%), 2 ml; and distilled water, 1,000 ml. The pH was adjusted to 6.5 ± 0.1. The broth was dispensed in 1.5-ml amounts in 9-mm-diameter tubes and sterilized. After inoculation, the medium was overlaid with paraffin oil. The tubes were incubated at 37°C (25°C for Yersinia sp.), and a deep purple color was observed when histamine was produced by decarboxylation of histidine. Negative reactions turned yellow. The majority of positive results were obtained after 24 h of incubation, but the tests were observed for 2 days.

Our results confirmed those obtained by Enjalbert et al. (3). All E. aerogenes strains were positive. In addition, we found HDC in all Morganella morganii strains tested and also in Raoultella ornithinolytica. Apart from these two species, no other Enterobacteriaceae were positive except for three out of nine strains of Citrobacter youngae and one out of seven strains of Raoultella planticola. These results are reported in Table 1.

To confirm the validity of these results, cloning and sequencing of the HDC gene were performed in 10 study strains, according to Takahashi et al. (8). The HDC gene was detected in two E. aerogenes strains but not in two E. cloacae strains. The three C. youngae strains positive in the HDC test were also positive for the HDC gene, whereas a negative strain of this species did not possess the gene. Similarly, in the positive R. planticola strain, the HDC gene was present, and it was absent in a negative control strain.

HDC can be detected with Møller's decarboxylase broth, as performed by Havelka (5). However, the color shift is not obvious and is often delayed. The medium turns grayish yellow in positive reactions, which are therefore not so clearly defined as in the Taylor modified broth, which usually turns purple within 24 h. Although the method described by Enjalbert et al. (3) is presented as a simple test, the extraction of histamine followed by the use of reagents for its detection is more tedious and time-consuming in the routine laboratory than the use of our HDC broth.

E. aerogenes has become an important nosocomial pathogen in the last few decades (2). Although strains with a typical profile are easy to identify by phenotypic characteristics, some strains, especially those with poor motility and/or delayed ornithine decarboxylase activity, may be misidentified as K. pneumoniae. Commercial systems do not always correctly identify E. aerogenes. In a comparative study of the BBL Crystal E/NE and API 20 E systems, 10 strains of E. aerogenes were examined. Three strains failed to be correctly identified by the BBL Crystal system, and one failed to be correctly identified by the API 20 E (11). In another study on the evaluation of the Microscan Rapid Gram-Negative ID Type 3 panel, 2 strains of E. aerogenes out of 10 were not identified and 1 was identified only by additional tests (7). It has been reported that Enterobacter species are more often misidentified by commercial systems than other Enterobacteriaceae (1). HDC provides a valuable confirmatory test for the identification of E. aerogenes, since no negative results were found in more than 50 clinical isolates. It could be helpful in strains with atypical characteristics, such as poor motility or delayed ornithine decarboxylase activity, or whenever results of identification by commercial systems are dubious or not discriminatory.

Production of histamine from histidine was reported by Kanki et al. in R. ornithinolytica and R. planticola (6), but using our method, decarboxylation was detected in only one out of seven R. planticola strains.

The medium proposed here is simple to prepare, and the test can readily be performed in any routine laboratory.

• Copyright © 2004 American Society for Microbiology