Enumeration of Respiring Pseudomonas spp. in Milk within 6 Hours by Fluorescence In Situ Hybridization following Formazan Reduction

Abstract

Respiring Pseudomonas spp. in milk were quantified within 6 h by fluorescence in situ hybridization (FISH) with vital staining. FISH with an oligonucleotide probe based on 16S rRNA sequences was used for the specific detection of Pseudomonas spp. at the single cell level. 5-Cyano-2,3-ditolyl tetrazolium chloride (CTC) was used to estimate bacterial respiratory activity. The numbers of respiring Pseudomonas cells as determined by FISH with CTC staining (CTC-FISH) were almost the same or higher than the numbers of CFU as determined by the conventional culture method.

Conventional culture methods are commonly used for the microbiological quality assurance of food and drink. These methods are simple but generally require more than 24 h to yield reliable results. Underestimation of bacterial numbers sometimes occurs, because the cells which are viable but no longer culturable by culture methods are difficult to detect. Therefore, rapid and simple culture-independent methods are required.

Several culture-independent methods are used for the detection of bacteria in food. PCR is one of the most useful techniques because of its high sensitivity. Reverse transcription-PCR, in particular, is a potentially valuable technique for the detection of viable bacteria (13, 27); however, these techniques require the extraction of nucleic acid from samples, and PCRs are sometimes inhibited by components of food, such as lipids, proteins, and salts (30). Enzyme-linked immunosorbent assay (ELISA) is also used frequently for the detection of bacteria in food (28); however, it is often used in conjunction with culture methods and incurs the limitations of culture techniques in enumerating bacterial cells. Fluorescence in situ hybridization (FISH) is widely used and is also suitable for the specific detection of targeted bacteria phylogenetically at species, genus, and family levels (2, 4), because the databases of rRNA sequences are publicly available.

We attempted to enumerate Pseudomonas spp. with physiological activity in milk by combining FISH with 5-cyano-2,3-ditolyl tetrazolium chloride (CTC) staining (CTC-FISH). Pseudomonas spp. are among the most important spoilage bacteria in milk (5, 6). They grow actively during refrigerated storage and produce enzymes, such as protease and lipase, which cause the degradation of milk compounds and reduce its shelf life. CTC is a redox dye widely used for the evaluation of bacterial respiratory activity (23, 24). CTC staining can be applied to anaerobic as well as aerobic bacteria (3). CTC is reduced to red fluorescent CTC-formazan crystals in actively respiring cells. However, the crystals are easily dissolved in organic solvents used in the main processes of FISH, and CTC-stained cells are no longer visible with the conventional FISH procedure.

We optimized the CTC concentration and the procedure for FISH in order to retain CTC-formazan crystals inside cells and detect bacteria with a specific rRNA sequence simultaneously. In addition, the numbers of viable Pseudomonas cells determined by CTC-FISH were also compared with the numbers of CFU determined by the traditional culture technique.

Bacterial strains and culture conditions.

Pseudomonas putida ATCC 12633, Pseudomonas fluorescens RIMD 1615005, and Escherichia coli O157:H7 ATCC 43888 cells in the stationary phase were obtained by incubating them at 30°C (P. putida and P. fluorescens) or 37°C (E. coli) in liquid medium (5 g Bacto tryptone [Difco], 1 g glucose, and 2.5 g yeast extract/liter of distilled water). Starved cells were obtained by the following procedure. P. putida cells in the stationary phase were collected by centrifugation and washed twice with sterile phosphate-buffered saline (PBS; 130 mM NaCl, 10 mM Na₂HPO₄, and 10 mM NaH₂PO₄ [pH 7.2]). Washed cells were collected by centrifugation, suspended in PBS (10⁶ cells/ml), and stored at 4°C without shaking in the dark for 4 weeks. The cells were considered to be in the starved state when the number of culturable cells on standard methods agar (5 g Bacto tryptone, 1 g glucose, 2.5 g yeast extract, and 15 g agar/liter of distilled water) dropped to less than 5 CFU/ml.

Bacterial cells in the stationary phase or in the starved state were inoculated into the milk (heat treated at 140°C for 2 s; fat ≥ 3.5%, protein ≥ 3.0%; pH 6.8) purchased from a retail store and used immediately. In this study, enzymes and surfactants were not used for removal of proteins and lipid in milk because these materials in milk were removed during the CTC-FISH procedure.

CTC-FISH procedure.

Yamaguchi et al. stained E. coli O157:H7 cells in food samples with 2 mM CTC (31). However, Pseudomonas spp. in milk were not effectively stained with this concentration by the CTC-FISH method; thus, the CTC concentration was optimized. PBS (400 μl) was added to 100 μl of milk samples, and the samples were stained with CTC at room temperature (approximately 25°C) for 1 h in the dark. Without the FISH process, the ratio of CTC-stained P. putida was approximately 90% of total P. putida cells with 2, 4, 6, 8, 10, 12, and 15 mM CTC, and nonspecific signals by CTC-formazan crystals outside the cells did not appear. CTC-stained cells, however, were barely detectable with less than 4 mM CTC, while approximately 90% of respiring cells were detected with more than 8 mM CTC after the whole CTC-FISH procedure. The optimal concentration was determined to be 8 mM. It has been reported that vital staining with CTC underestimates the number of active cells when combined with the common FISH procedure and that microautoradiography (MAR) combined with FISH (MAR-FISH) is superior for the detection of active cells (16). MAR-FISH is a useful technique which has been successfully applied to activated sludge samples (11); however, MAR detects active cells by incorporating radioactively labeled substrate. Utilization of these substrates is limited in some laboratories; thus, in this study, CTC, a common and easy-to-use vital stain, was chosen.

After CTC staining, the cells were fixed with paraformaldehyde (final concentration, 8%) at 4°C for 1 h. In an examination of fixation conditions, this condition gave the best results. A more rapid technique of fixation, using ethanol, was reported (18). However, if ethanol is used for CTC-FISH, CTC-formazan crystals will be dissolved and CTC-stained cells cannot be detected completely.

The fixed sample was centrifuged at 900 × g for 5 min at 4°C to remove paraformaldehyde, and the pellets were resuspended in 100 μl of PBS. A 10-μl sample of fixed cells was spotted in the hole of a polyester seal which was attached to an adhesive-coated glass slide (Matsunami Glass, Ltd., Osaka, Japan) and dried by a vacuum, according to the method described by Maruyama et al., for the quantification of cells (12). Bacterial cells in the hole were spread uniformly and can be quantified correctly and easily by microscopy.

The sample was coated by agarose, with a modification of the process described by Pernthaler et al. (19). Agarose (0.1%, Metaphor; Bioproducts, Rockland, ME) was dropped onto the dried sample, which was dried again at 35°C in a humid chamber, and the seal was peeled off the glass slide. Without this agarose coating, many of the CTC-formazan crystals inside the cells disappeared.

After 50 μl of 80% ethanol was dropped onto the dried cells, the ethanol was shaken off immediately, and the sample was kept at 35°C for 5 min in a humid chamber. Dehydration conditions were also examined. CTC-stained cells vanished completely after dehydration with an ethanol series (50, 80, and 100% ethanol) for 3 min each, which has often been used with the FISH technique (1). However, more than 90% of respiring cells were still detectable after dehydration with 80% ethanol if the ethanol was removed immediately. Without dehydration, more than 90% of respiring cells also were detected by CTC-FISH; however, Pseudomonas cells could not be detected selectively.

The dehydrated cells were hybridized. In this study, two kinds of Cy5-labeled oligonucleotide probes based on 16S rRNA sequences were used; one was the probe for the detection of P. putida, specifically designed by DuTeau et al. (5′-to-3′ sequence, TTG CCA GTT TTG GAT GCA GT) (7), and the other was the probe for the detection of Pseudomonas spp., specifically designed by Gunasekera et al. (5′-to-3′ sequence, GAT CCG GAC TAC GAT CGG TTT) (8). These hybridization and washing conditions were modified for the detection of CTC-stained cells by CTC-FISH. For the detection of P. putida specifically, hybridization buffer (0.9 M NaCl, 20 mM Tris-HCl [pH 7.5], and 0.01% sodium dodecyl sulfate [SDS]) containing 1 ng/μl probe for P. putida was prewarmed at 35°C in a chamber. After 20 μl of the prewarmed hybridization buffer was dropped onto the dehydrated cells, the cells were hybridized at 35°C for 2 h in the dark. After the buffer was shaken off, 50 μl of the prewarmed washing buffer (which is the same as the hybridization buffer) was applied, and the sample was kept at 37°C for 15 min in the dark and then rinsed with distilled water. The sample was counterstained with 50 μl of 4′,6′-diamidino-2-phenylindole (DAPI; final concentration, 1 μg/ml; Sigma) for the determination of total bacterial number.

For the detection of Pseudomonas spp. specifically, hybridization buffer containing 1 ng/μl probe for Pseudomonas spp. was used. The dehydrated cells were hybridized at 35°C for 1.5 h. These cells were rinsed with PBS, and the sample was counterstained with DAPI. Under the modified conditions, the specific detection of P. putida and of Pseudomonas spp. was achieved.

We also examined SDS concentrations in hybridization and washing buffers. SDS is one of the important elements for the specific detection of targeted cells by FISH (29). However, CTC-formazan crystals are easily dissolved in SDS solution. Removing the SDS completely failed to provide specific detection of Pseudomonas cells consistently, although it avoided the reduction of CTC-formazan crystals inside cells which occurs with SDS. Ultimately, 0.01% SDS enabled the detection of respiring cells and Pseudomonas cells simultaneously by CTC-FISH.

The cells were observed under UV excitation for viewing DAPI-stained cells (total bacteria), green excitation for CTC-stained cells (respiring cells), and red excitation for Cy5-labeled probe hybridized cells (Pseudomonas cells) on the same microscopic field, with an E400 epifluorescence microscope (Nikon). The following filter combinations were used: for UV excitation, UV-2A; for green excitation, G-2A; and for red excitation, CY5 HYQ. More than 1,000 cells under different microscopic field sites per sample were counted to determine bacterial numbers.

Discrimination of Pseudomonas spp. from other strains in milk.

In order to evaluate the specificity of CTC-FISH for Pseudomonas spp., E. coli O157:H7 cells in the stationary phase were inoculated as a negative control with Pseudomonas spp. (P. putida and P. fluorescens) in milk. E. coli O157:H7 is well-known as one of the most dangerous bacteria for food-borne disease, and it has been detected often in milk (14). Pseudomonas cells in the milk sample were detected specifically with the probe for Pseudomonas spp., and E. coli O157:H7 cells were detected specifically with fluorescein isothiocyanate-labeled anti-E. coli O157:H7 direct antibody. For the detection of E. coli O157:H7, 20 μl of fluorescent antibody (FA) solution (4 ng/ml; Kirkegaard & Perry Laboratories, Gaithersburg, MD) was spotted onto the sample and samples were incubated at 37°C for 30 min in the dark after the CTC-FISH washing process. The sample was counterstained with DAPI, and each cell was observed under each excitation as described above, combined with an FITC filter block (Nikon) for the observation of cells stained with FA (E. coli O157:H7) under blue excitation.

Total bacteria (stained with DAPI), respiring bacteria (stained with CTC), Pseudomonas spp. (hybridized with a probe specific for Pseudomonas spp.), and E. coli O157:H7 (stained with FA) were quantified by microscopy on the same field with each excitation (Fig. 1A). Total bacteria, respiring cells, Pseudomonas spp., and E. coli O157:H7 were observed specifically under each excitation. By the combination of three images (Fig. 1B to D), the specific detection of respiring Pseudomonas spp. was achieved (Fig. 1E). In addition, respiring P. putida and E. coli O157:H7 cells were mixed at different ratios (P. putida:E. coli ratios of 10:0, 7:3, 5:5, 3:7, and 0:10; n = 15). The sum of the number of P. putida and E. coli O157:H7 cells inoculated for each sample is 10⁶ cells/ml. It was confirmed that the targeted cells were detected at the same ratio observed in the initial mixture (Fig. 2) (R² = 0.99).

FIG. 1.

Discrimination of respiring Pseudomonas spp. (P. putida and P. fluorescens) from nonrespiring Pseudomonas spp. and E. coli O157:H7 in milk by CTC-FISH with fluorescent antibody. (A to D) Images observed under (A) UV excitation for the detection of total bacteria, (B) green excitation for respiring bacteria, (C) red excitation for Pseudomonas spp., and (D) blue excitation for E. coli O157:H7. (E) Combined image of the three images in panels B to D; (a) respiring Pseudomonas spp., (b) nonrespiring Pseudomonas spp., (c) respiring E. coli O157:H7, and (d) nonrespiring E. coli O157:H7.

FIG. 2.

Correlation between inoculated respiring P. putida cells and respiring P. putida cells detected by CTC-FISH.

Comparison of CTC-FISH counts with CFU as determined by the conventional plate counting method.

We attempted to detect cells in the stationary phase and in the starved state by CTC-FISH, and the bacterial numbers determined by CTC-FISH and those determined by the culture method were compared, since Pseudomonas spp. contaminating milk can exist in several growth cycle phases. Standard methods agar, which is recommended as a standard medium for the examination of dairy products by the American Public Health Association, was used to count the bacterial number. P. putida cells in the stationary phase or in the starved state were inoculated into milk. The samples were diluted serially, and 0.1 ml of each diluted or nondiluted sample was spread on the agar. After incubation at 30°C for 48 h, CFU were counted. The number of respiring P. putida cells in the same milk samples was also determined by CTC-FISH and was the same (in the stationary phase) or higher (in the starved state) than the number of CFU on agar medium (Table 1). CTC-FISH, then, seems better able to detect cells in various growth phases, particularly cells in the stages that are sometimes underestimated by culture methods.

TABLE 1.

Comparison of the numbers of viable P. putida cells in milk as determined by CTC-FISH and the numbers of CFUs as determined by the conventional plate counting method^a

Method	Result (± SD) at:
	Stationary phase	Starved stage
CTC-FISH (no. cells/100 μl of milk)	2.1 (± 1.7) × 106	1.9 (± 0.7) × 105
Plate counting (CFU)	1.6 (± 1.3) × 106	<5 (± 0.0)

^an = 3.

Discussion.

In this study, respiring Pseudomonas spp. in milk were quantified within 6 h by CTC-FISH. This method could be applied to other active spoilage bacteria in milk through the use of suitable rRNA-targeted probes. Peptide nucleic acid probes hybridize in a shorter time than oligonucleotide probes (26). CTC-FISH with peptide nucleic acid probes may be possible and could simplify further the rapid detection of specific bacterial contaminants in food. In this study, the detection limit of CTC-FISH is 10⁴ cells/ml by the use of the frame spotting method (12). Combination with concentration techniques, for example, filtration or immunomagnetic separation, which has been described for the detection of bacteria with a high recovery rate (22, 25), should improve the limit. In addition, epifluorescence microscopy is a major tool for the direct detection of bacteria and was also used in this study. However, epifluorescence microscopy is time and labor intensive, and the criteria for microscopic visual counting are different for different investigators (10, 15). Automated enumeration systems (17, 20, 21) and flow cytometry (9, 31) may be applicable to future studies of CTC-FISH.