Evaluation of p-Naphtholbenzein-β-d-Galactoside as a Substrate for Bacterial β-Galactosidase

Abstract

We describe the synthesis of a new substrate for the detection of β-galactosidase and evaluate its performance in comparison with 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) and cyclohexenoesculetinβ-d-galactoside (CHE-Gal). Of 206 Enterobacteriaceae strains able to hydrolyze X-Gal, 194 (94.2%) hydrolyzed CHE-Gal and 192 (93.2%) hydrolyzed p-naphtholbenzein-β-d-galactoside (PNB-Gal). We conclude that PNB-Gal is an effective substrate for the detection of β-galactosidase.

The enzyme β-galactosidase has long been regarded as an important taxonomic marker in microbial identification, particularly among gram-negative species. The long-established ortho-nitrophenyl-β-d-galactopyranoside test relies on the hydrolysis of ortho-nitrophenyl-β-d-galactoside by β-galactosidase, releasing yellow ortho-nitrophenol (5). Fluorogenic galactosides such as those based on fluorescein, resorufin, and 4-methylumbelliferone are also well established (1, 8). When agar-based methods are used, chromogenic substrates that yield insoluble products are desired so that the aglycone released by hydrolysis does not diffuse widely (3, 6). Indoxylic galactosides such as indoxyl-β-galactoside and its halogenated derivatives, including 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal), have advantages when used in solid media, as the aglycone released is oxidized rapidly by air to produce an insoluble colored product which is restricted to the colony mass (4). Although these substrates are very effective when used aerobically, their synthesis is not straightforward. Although many substrates for β-galactosidase have been described, there are few alternatives to indoxyl-based substrates suitable for inclusion in agar-based media. One alternative is to use a galactoside derivative of cylcohexenoesculetin, a core molecule which forms an insoluble black chelate when released by hydrolysis (3). Such substrates require the inclusion of iron in the medium, which can be a disadvantage as deaminase activity may also generate colored products in the presence of peptone and iron (2, 7).

We describe the synthesis of a new chromogenic substrate for the detection of β-galactosidase, p-naphtholbenzein-β-d-galactopyranoside (PNB-Gal) (Fig. 1). The p-naphtholbenzein released by hydrolysis remains localized on bacterial colonies that appear pink. The effectiveness of this substrate was evaluated in direct comparison to cyclohexenoesculetin-β-d-galactoside (CHE-Gal) and X-Gal for the detection of β-galactosidase within the Enterobacteriaceae and other gram-negative species.

FIG. 1.

Structure of PNB-Gal.

p-Naphtholbenzein was obtained from Acros Organics, Geel, Belgium. CHE-Gal was synthesized by a method described previously (3). All other chemicals and materials were obtained from the Sigma-Aldrich Chemical Company Ltd., Poole, United Kingdom. The synthesis of PNB-Gal was as follows. Five millimoles of p-naphtholbenzein (1.87 g) was dissolved in 20 ml of acetone with vigorous stirring. A 5-ml solution containing 1.4 mol of potassium hydroxide/liter was added to this followed by an additional 10 ml of acetone. Approximately 5 ml of water was then added dropwise to generate a homogeneous deep-blue solution. Ten millimoles of α acetobromogalactose (4.1 g) was dissolved in 10 ml of acetone and added to this solution. In order to maintain the pH above 11, 2 ml of a 20-mol/liter potassium hydroxide solution was added after 30 min of stirring and again after 90 min. After 3.5 h, a further 2 ml of alkali was added followed by an additional 5 ml of α acetobromogalactose in acetone. After 4.5 h, another 2 ml of alkali was added and the solution was left, with stirring, overnight.

The acetone was removed under reduced pressure, and the residual solution was poured into 300 ml of a 0.06-mol/liter sodium carbonate solution at 0°C with good stirring. The brown precipitate was separated by vacuum filtration, washed with water, and air dried. This solid was dissolved in 100 ml of dichloromethane and washed thoroughly with 0.5 mol of potassium hydroxide/liter at 0°C to remove most of the free p-naphtholbenzein. Residual p-naphtholbenzein was removed by stirring with Dowex Marathon resin in 100 ml of water at pH 11.0 for 2 to 3 h. The purification was followed by thin-layer chromatography using ethyl acetate-toluene (3:1) as the solvent, with subsequent transient exposure of the chromatogram to ammonia.

The deep-yellow solution was dried overnight using anhydrous magnesium sulfate and was evaporated, reconstituted with methanol, and evaporated to produce a mousse. The mousse was directly dissolved in 50 ml of methanol and deprotected over 5 h using a 20-ml solution containing 0.4 mol of sodium methoxide/liter of methanol. The methanolic solution was adjusted to pH 6.5 using ion-exchange resin (120[H⁺]) and was separated by decantation and the solvent removed under reduced pressure. The glycoside formed a brownish yellow powder. This was removed, yielding 1.5 g of PNB-Gal.

All galactosides used in this study were added to the media prior to autoclaving, which is normal practice in our laboratory. PNB-Gal agar was prepared as follows: 41 g of Columbia agar was added to 1 liter of distilled water, along with 100 mg of PNB-Gal. A 30-mg sample of the gratuitous inducer isopropyl-β-d-thiogalactopyranoside (IPTG) was also included, to aid the induction of β-galactosidase. The agar was sterilized by autoclaving for 10 min at 116°C. The medium was then allowed to cool to 55°C before being poured into 20-ml volumes. Both CHE-Gal agar and X-Gal agar were prepared in identical fashion, except that 300 mg of CHE-Gal and 80 mg of X-Gal were substituted for PNB-Gal, respectively. Ferric ammonium citrate (500 mg/liter) was also included in CHE-Gal agar to allow formation of the metal chelate. A duplicate batch of X-Gal agar was prepared in which the X-Gal was added after autoclaving once the agar had cooled to 50°C. Each chromogenic medium was also prepared with IPTG excluded to assess the performance of each substrate in the absence of an enzyme inducer. In order to investigate the sensitivity of PNB-Gal, two more agars were prepared with a reduced substrate concentration (50 and 20 mg/liter); both were prepared as described above with IPTG included.

Strains (397) of a range of species, including 333 Enterobacteriaceae collected from a wide range of clinical and environmental samples, were identified using the API 20 E system (bioMérieux). These strains were cultivated on Columbia agar at 37°C for 24 h. Each strain was then inoculated into physiological saline to produce an inoculum of approximately 10⁸ organisms per ml (equivalent to a McFarland standard of 0.5). Using a multipoint inoculator (Denley), 1 μl of each suspension was then inoculated onto all three of the test media and Columbia agar as a growth control. Twenty strains were inoculated per plate.

All plates were incubated at 37°C for exactly 18 h. After incubation, PNB-Gal plates were examined for the presence of pink colonies, CHE-Gal plates for the presence of black colonies, and X-Gal plates for the presence of blue colonies; Columbia agar plates were examined for growth as well.

From the results shown in Table 1 it can be concluded that PNB-Gal showed good correlation with both X-Gal and CHE-Gal. Of 206 Enterobacteriaceae strains able to hydrolyze X-Gal, 194 (94.2%) hydrolyzed CHE-Gal and 192 (93.2%) hydrolyzed PNB-Gal. None of the strains which failed to hydrolyze X-Gal produced any coloration with the other substrates; i.e., there were no false-positive results. The addition of X-Gal prior to autoclaving had no detectable impact on its performance. The performance of X-Gal was least affected when IPTG was excluded, when it still detected 98.5% of β-galactosidase producers compared with 92.2% by CHE-Gal and 87.4% by PNB-Gal. Overall, there were only a small number of discrepancies between these three substrates; however, there was substantial variation in the ability of the substrates to detect β-galactosidase produced by Yersinia enterocolitica. For example, X-Gal detected β-galactosidase in all strains of Y. enterocolitica, whereas only 5 of 14 Y. enterocolitica strains (36%) were detected as weakly positive with both CHE-Gal and PNB-Gal.

TABLE 1.

Hydrolysis of different β-galactosidase substrates by strains of Enterobacteriaceae and related species

Gram-negative species	No. of strains	% Positive with substrate
		PNB-Gal with IPTG	PNB-Gal	X-Gal with IPTG	X-Gal	CHE-Gal with IPTG	CHE-Gal
Acinetobacter spp.	53	0	0	0	0	0	0
Aeromonas caviae	7	86	86	86	86	86	86
Aeromonas hydrophila	3	100	100	100	100	100	100
Citrobacter diversus	9	89	78	89	89	89	67
Citrobacter freundii	16	100	81	100	100	100	100
Enterobacter aerogenes	9	100	67	100	100	100	100
Enterobacter agglomerans	1	100	100	100	100	100	100
Enterobacter cloacae	21	100	95	100	100	100	95
Escherichia coli	71	83	82	83	83	83	82
Escherichia hermannii	1	100	0	100	100	100	100
Hafnia alvei	10	70	70	90	90	80	80
Klebsiella oxytoca	13	100	100	100	100	100	100
Klebsiella ozaenae	3	33	33	33	33	33	33
Klebsiella pneumoniae	19	100	89	100	89	100	100
Morganella morganii	12	0	0	0	0	0	0
Proteus mirabilis	16	0	0	0	0	0	0
Proteus penneri	1	0	0	0	0	0	0
Proteus vulgaris	4	0	0	0	0	0	0
Providencia alcalifaciens	3	0	0	0	0	0	0
Providencia rettgeri	3	0	0	0	0	0	0
Providencia stuartii	10	0	0	0	0	0	0
Salmonella spp.	64	0	0	0	0	0	0
Serratia odorifera	1	100	100	100	100	100	100
Serratia liquefaciens	6	83	83	100	100	83	83
Serratia marcescens	8	75	75	100	88	100	100
Shigella boydii	1	0	0	0	0	0	0
Shigella dysenteriae	2	0	0	0	0	0	0
Shigella flexneri	2	0	0	0	0	0	0
Shigella sonnei	10	100	100	100	100	100	100
Vibrio cholerae	1	100	100	100	100	0	0
Yersinia enterocolitica	14	36	36	100	100	36	36
Yersinia pseudotuberculosis	3	0	0	0	0	0	0
Total no. of strains	397
Overall % positive		48.3	45.4	51.9	51.1	48.8	47.9

In all cases, strains which hydrolyzed PNB-Gal produced a clearly visible pink precipitate that remained highly restricted to the bacterial colony. Reducing the concentration of PNB-Gal resulted in a reduced sensitivity of the substrate. For example, when PNB-Gal was used at 50 mg/liter, 91.3% of positives were detected, and when it was tested at 20 mg/liter, 88.8% of positives were detected.

We have shown that PNB-Gal is able to detect β-galactosidase activity in strains of Enterobacteriaceae and that its performance stands good comparison with that of both X-Gal and CHE-Gal. The substrate is highly sensitive at low concentrations, and its performance is almost identical to that of CHE-Gal when used at one-third the concentration. Even at 20 mg/liter, if strains of Y. enterocolitica are excluded, PNB-Gal detected 94.3% of all β-galactosidase producers. This is one quarter of the concentration used for X-Gal and 1/15 of the optimal concentration of CHE-Gal. The synthesis of PNB-Gal is very straightforward and cost-effective. This is because, unlike CHE-Gal and X-Gal, the core molecule required for derivatization is available commercially and is inexpensive, thus simplifying the synthetic process. This factor, combined with the high sensitivity of the substrate, makes the use of PNB-Gal highly economical.

A further advantage of PNB-Gal is that cofactors are not required for generation of the colored product, as the core molecule is naturally colored and relatively insoluble. This offers a greater flexibility when using such substrates in chromogenic media, whereas X-Gal and CHE-Gal require oxygen and metal ions, respectively, for the generation of color.