Simultaneous Immunofluorescent Detection of Coentrapped Cells in Gel Beads

Abstract

An immunofluorescent method involving double color labeling and confocal microscopy was reported to specifically detect lactic acid bacteria and probiotic cells coimmobilized in gels beads. The method described is rapid (4 h) and sensitive and may be useful for studying cell dynamics during mixed-culture starter production using immobilized cells in gel beads. Microscopic observations were perfectly correlated to cell counts obtained using a sandwich enzyme-linked immunosorbent assay.

Immobilized cell technology with lactic acid bacteria (LAB) has been proposed for different industrial applications such as continuous prefermentation of milk for yogurt production (11) and cheese manufacture (14, 15) and production of concentrated lactic starters in single (10) or mixed (7) culture. Stable and reproducible mixed-strain starters in the effluent of a continuous reactor were obtained using this technology, and very high productivity resulted from the high cell density retained in the immobilized cell reactor (7, 8). However, a large cross-contamination of beads, initially entrapping pure cultures, was observed during continuous cultures over long fermentation times of 6 to 8 weeks in supplemented whey permeate (7, 8) or in milk (14). A theoretical model of cell release from cavities located near the gel bead surfaces has been recently proposed to explain this cross-contamination phenomenon (6). To experimentally validate this hypothesis and to identify factors responsible for this cross-contamination phenomenon, a method for specifically detecting the different strains in beads is needed. A model system with a probiotic strain (Bifidobacterium longum) as the noncompetitive strain and an LAB (Lactococcus lactis subsp. lactis biovar diacetylactis) as the competitive strain was chosen for this study. Bifidobacteria are increasingly used in fermented dairy products in combination with LAB strains because of their perceived importance in human health (9).

Single (13) and dual (1) labeling with green fluorescent protein has been reported to detect free LAB cells and gram-negative bacteria in mixed free-cell culture, respectively. Fluorescent polyclonal antibodies were used to specifically detect genetic variants of Streptococcus cremoris in mixed free-cell culture, using a direct and indirect fluorescence labeling method with fluorescein isothiocyanate (FITC) as a differential cell detection strategy (3). Nitrosomonas europaea and Nitrobacter agilis coimmobilized in gel beads were detected separately using a two-step fluorescent-labeling method with FITC-labeled anti-rabbit antibody (4). Hence, green fluorescent colonies of either N. europaea or Nitrobacter agilis were observed with this strategy. Dual immunofluorescent labeling has never been reported for the simultaneous and specific detection of probiotic and LAB cultures coimmobilized in gel beads.

The Lactococcus lactis subsp. lactis biovar diacetylactis strain (Rhone Poulenc, Brampton, Ontario, Canada) was grown at 30°C in M17 broth (Difco Laboratories, Detroit, Mich.) supplemented with 1% (wt/vol) lactose. The Bifidobacterium longum ATCC 15707 strain (Rosell Institute Inc., Montreal, Quebec, Canada) was cultivated at 37°C in MRS broth (Rosell Institute Inc.) supplemented with 0.5 g of cysteine per liter, 0.2 g of Na₂CO₃ per liter, and 0.1 g of CaCl₂ per liter (12).

Polyclonal antibodies against both strains were raised in rabbits using cell wall suspensions as immunogens. Cross-reactivities of anti-B. longum antibody on L. lactis subsp. lactis biovar diacetylactis and anti-L. lactis subsp. lactis biovar diacetylactis antibody on B. longum were removed using a cross-adsorption protocol. All operations were carried out at 4°C. Anti-B. longum antibody used at a final concentration of 5 μg/ml was mixed with 10 ml of an L. lactis subsp. lactis biovar diacetylactis cell suspension (10¹⁰ CFU/ml) containing protease inhibitors for 24 h in a rotary shaker at 4 rpm. The pH was adjusted to 7.5 ± 0.1 with 1 N NaOH before adsorption. After adsorption, free immunoglobulin G (IgG) was recovered on a protein A/G column (Pierce, Rockford, Ill.), dialyzed against phosphate-buffered saline (PBS), and concentrated to 2 mg/ml using centricon (Millipore, Bedford, Mass.). The same technique was used for eliminating anti-L. lactis subsp. lactis biovar diacetylactis IgG cross-reacting with B. longum cells.

The specificities of purified IgG (before and after adsorption) were determined by dot blot immunoassay on nitrocellulose membranes (Micron Separation Inc., Westboro, Mass.) using peroxidase-labeled antibodies (5).

Two fluorescent dyes, ALEXA 488 and ALEXA 568, were used to label the adsorption-purified anti-B. longum and anti-L. lactis subsp. lactis biovar diacetylactis antibodies, respectively, using an ALEXA protein labeling kit (Molecular Probes, Inc., Eugene, Oreg.), according to the manufacturer's instructions. The ALEXA 488-labeled anti-B. longum IgG and the ALEXA 568-labeled anti-L. lactis subsp. lactis biovar diacetylactis IgG have excitation maxima at 488 and 568 nm, respectively, and emission maxima at 517 and 603 nm, respectively (2).

The immobilization procedure for κ-carrageenan and locust bean gum gel beads (2.75 and 0.25% [wt/wt], respectively) was based on a two-phase dispersion technique (7) modified as follows. A 1% (vol/vol) mixed inoculum made of 90% (vol/vol) B. longum and 10% (vol/vol) L. lactis subsp. lactis biovar diacetylactis with cultures standardized at an absorbance of 0.5 at 550 nm, was used to favor the growth of the less competitive B. longum strain. Beads immobilizing pure cultures of B. longum and L. lactis subsp. lactis biovar diacetylactis strains were also prepared using the same procedure but with an absorbance-standardized inoculum of 2% (vol/vol) in the polymer solution.

All operations were then carried out with 0.1 M KCl to keep the bead structure. Beads coentrapping B. longum and L. lactis subsp. lactis biovar diacetylactis strains were incubated in supplemented MRS medium during six successive pH-controlled batch cultures for 16, 12, 8, 6, 4, and 4 h at 37°C in a 500-ml bioreactor (BioFlo model C30; New Brunswick Scientific Co., Edison, N.J.), with CO₂ injections in the headspace. Beads entrapping pure cultures were incubated separately for only two successive fermentations of 16 and 8 h in appropriate medium. The bioreactor was inoculated with 20% (vol/vol) gel beads, pH was kept at 6 by addition of 6 M NH₄OH, and mixing was set at 200 rpm. Beads coentrapping B. longum and L. lactis subsp. lactis biovar diacetylactis strains were randomly sampled from the reactor during the six successive pH-controlled batch cultures at different times corresponding to total fermentation times of 0, 6, 14, 17, 20, 24, 28, 39, and 50 h. The specific immobilized cell concentration was measured by enzyme-linked immunosorbent assay (ELISA) and confocal microscopy. Beads immobilizing pure cultures were collected at the end of the first 16-h batch.

A sandwich-type ELISA was developed to measure L. lactis subsp. lactis biovar diacetylactis and B. longum cell concentrations in beads. Flat-bottom microplates (Dynex, Chantilly, Va.) were coated overnight (4°C) with specific anti-L. lactis subsp. lactis biovar diacetylactis or anti-B. longum antibody (1 μg) diluted in 0.1 M sodium carbonate-bicarbonate buffer (pH 9.6). Microplates were washed four times with Tris-buffered saline supplemented with 0.1% (vol/vol) Tween 20, blocked for 1 h at 25°C in PBS–1% (vol/vol) blocking reagent (Roche Diagnostics, Laval, Quebec, Canada), and washed again before the addition of serial dilutions (1:10; 1:50; 1:100, and 1:500) of B. longum and L. lactis subsp. lactis biovar diacetylactis cultures to determine standard curves. Approximately 0.5 g of each bead sample was dissolved with an Ultra-Turrax (Janke and Kunkel, Staufen, Germany) on ice for 30 s at 13,500 rpm prior to the addition of appropriate dilutions (100 μl) to wells. Bacterial cells were detected by the addition of adsorption-purified anti-B. longum or anti-L. lactis subsp. lactis biovar diacetylactis antibody diluted 1:200 in PBS–0.5% blocking reagent previously labeled with horseradish peroxidase by using a peroxidase labeling kit as recommended by the manufacturer (Roche Diagnostics) and with an orthophenylene diamine solution (Sigma, St. Louis, Mo.) as the substrate. The absorbance was measured at 450 nm on an ELISA plate reader (Molecular Devices, Sunnyvale, Calif.).

Bead preparation for microscopic observation involved solutions supplemented with 0.1 M KCl. Beads were cut in half with a razor blade, washed with PBS, blocked for 30 min at 37°C in 10% (vol/vol) horse fetal serum, washed with PBS–0.1% Tween 20, and then incubated in the dark for 2 h at 37°C with 4 μg of each antibody (ALEXA 488-labeled anti-B. longum and ALEXA 568-labeled anti-L. lactis subsp. lactis biovar diacetylactis antibodies). After four washes with PBS–0.1% Tween 20, beads were ready for confocal microscopic observations. Beads immobilizing pure cultures were incubated separately with one fluorescent antibody. The confocal laser scanning microscope (LSM 310; Carl Zeiss, Oberkochen, Germany) was equipped with an Ar-ion laser (488 nm) and an He-Ne laser (543 nm) as the excitation source and with two photomultipliers (2 and 1) which selected emission signals from 515 to 565 nm and from 575 to 640 nm, respectively. Immunostained half-beads were placed onto cover slides (with the cut surface facing the cover slide) mounted on 1-mm-thick concave microslides. ALEXA 488-labeled anti-B. longum antibody was first excited at 488 nm, followed by the excitation of ALEXA 568-labeled anti-L. lactis subsp. lactis biovar diacetylactis antibody at 543 nm. Fluorescent signals emitted were detected separately with photomultipliers 2 and 1, respectively. The two images recorded were colored in green and red for anti-B. longum and anti-L. lactis subsp. lactis biovar diacetylactis antibody signals, respectively, and then overlaid using laser scanning microscopy version 3.95 software (Carl Zeiss) for the simultaneous observation of both strains in the beads.

A strong immunological response was obtained with both B. longum and L. lactis subsp. lactis biovar diacetylactis strains. Before adsorption, both antibodies showed nonspecific signals when high bacterial concentrations were used (Fig. 1A and C). These nonspecific signals were eliminated after the adsorption step even when a high cell concentration of 10⁸ CFU/ml was used (Fig. 1B and D).

FIG. 1.

Specificities of anti-B. longum and anti-L. lactis subsp. lactis biovar diacetylactis antibodies tested on washed whole-cell suspensions by dot immunoblot assay. Lanes 1, whole cells of the B. longum strain; lanes 2, whole cells of the L. lactis subsp. lactis biovar diacetylactis strain. (A) Anti-B. longum antibody before adsorption; (B) anti-B. longum antibody purified by adsorption to L. lactis subsp. lactis biovar diacetylactis cells; (C) anti-L. lactis subsp. lactis biovar diacetylactis antibody before adsorption; (D) anti-L. lactis subsp. lactis biovar diacetylactis antibody purified by adsorption to B. longum cells. Cell concentrations (in CFU/ml) are shown at the left.

A strong fluorescence signal was obtained with B. longum beads and ALEXA 488-labeled anti-B. longum antibody (Fig. 2A1), while no signal was detected with L. lactis subsp. lactis biovar diacetylactis beads and ALEXA 488-labeled anti-B. longum antibody (Fig. 2B1). When ALEXA 568-labeled anti-L. lactis subsp. lactis biovar diacetylactis antibody was used, detectable fluorescence signals were obtained only with L. lactis subsp. lactis biovar diacetylactis beads (Fig. 2A2 and B2), with a very low background signal.

FIG. 2.

Micrographs from confocal laser-scanning microscopy of gel beads immobilizing pure cultures of B. longum (A) and L. lactis subsp. lactis biovar diacetylactis (B) strains detected with ALEXA 488-labeled anti-B. longum antibody (panels 1) and ALEXA 568-labeled anti-L. lactis subsp. lactis biovar diacetylactis antibody (panels 2) after 16 h of batch culture. ALEXA 488-labeled anti-B. longum and ALEXA 568-labeled anti-L. lactis subsp. lactis biovar diacetylactis antibodies were excited with an Ar laser at 488 nm and an He-Ne laser at 543 nm, respectively.

Cell distributions in gel beads coentrapping L. lactis subsp. lactis biovar diacetylactis and B. longum strains are shown in Fig. 3. Individual cells were detected in beads immediately after immobilization (Fig. 3A). After 6 and 14 h, small-sized B. longum (green) and L. lactis subsp. lactis biovar diacetylactis (red) colonies were observed (Fig. 3B and C). Cell growth was readily visible after 14 h of culture (Fig. 3C) and after 17 h (Fig. 3D). Sizes and distributions of colonies in gel beads were easily monitored during the 50-h fermentation (data not shown).

FIG. 3.

Fluorescence images by confocal laser-scanning microscopy of B. longum (green) and L. lactis subsp. lactis biovar diacetylactis (red) cultures coentrapped in gel beads after 0 h (A), 6 h (B), 14 h (C), and 17 h (D) of incubation.

Detection limits of 5 × 10⁶ CFU/ml and 1 × 10⁶ CFU/ml were obtained for B. longum and L. lactis subsp. lactis biovar diacetylactis strains, respectively, by the ELISA. A high linear correlation (r² > 0.989) between 5 × 10⁶ and 1 × 10⁹ CFU/ml for B. longum and between 10⁶ and 10⁹ CFU/ml for L. lactis subsp. lactis biovar diacetylactis was observed. The ELISA was used to specifically estimate the growth of L. lactis subsp. lactis biovar diacetylactis and B. longum strains coimmobilized in gel beads (Fig. 4). After a rapid growth of B. longum and L. lactis subsp. lactis biovar diacetylactis cells in beads during the first 16 h of batch culture, the cell concentrations slowly increased in beads during the subsequent batch cultures and reached 2.5 ± 0.15 × 10¹¹ CFU/g and 1.2 ± 0.2 × 10¹¹ CFU/g at the end of the sixth batch for B. longum and L. lactis subsp. lactis biovar diacetylactis, respectively. Microscopic observations were also correlated with cell counts obtained by ELISA.

FIG. 4.

Change of biomass concentrations in gel beads co-entrapping B. longum (●) and L. lactis subsp. lactis biovar diacetylactis (○) strains during six successive batch fermentations of 16, 12, 8, 6, 4, and 4 h using ELISA. The arrows indicate the start times of batch cultures.

In this paper we described a method for specifically and simultaneously detecting two strains (B. longum and L. lactis subsp. lactis biovar diacetylactis) coentrapped in gel beads by immunofluorescence. Cell concentrations measured in gel beads by this method were in agreement with those obtained by the sandwich ELISA. Beads immobilizing pure cultures were used as controls to validate the specificities of the antibodies in confocal microscopy. The low background level obtained (Fig. 2A2 and B1) might be explained by the high specificities of both antibodies, the immunostaining, and the visualization protocol developed. The strong fluorescence signals emitted by both ALEXA 488-labeled anti-B. longum antibody and ALEXA 568-labeled anti-L. lactis subsp. lactis biovar diacetylactis antibody (Fig. 2A1 and B2) are partly due to the strong levels of fluorescence emitted by ALEXA 488 and ALEXA 568, which are brighter and more stable than conventional fluorophores, such as FITC (2). Moreover, because ALEXA 488 and ALEXA 568 fluorescent dyes have emission and excitation spectra that do not overlap, they can be separately and specifically detected using appropriate excitation source and emission filter sets. Until now, the use of the multicolor labeling has not been reported for detection of probiotic strains and LAB in mixed cultures. The approach developed here is a more appropriate tool for studying microbial populations in gel beads than the standard method (conventional epifluorescence microscopy), which is limited by a high level of autofluorescence interference (4). One difficulty associated with in situ detection of bacteria in gel beads is the potential disruption of the peripheral dense cell layer during preparation. The gel bead surface is indeed altered by cell growth and cell release, although the inner core structure of beads is unaffected (6, 14). This new method does not require any fixation step or thin slicing of beads, which risks degrading the bead structure.

In this study, polyclonal antibodies were successfully produced to detect B. longum and L. lactis subsp. lactis biovar diacetylactis strains in mixed cultures and coentrapped in gel beads. This approach was also successfully applied for the specific detection of two closely related bacteria such as Lactococcus lactis subsp. cremoris and L. lactis subsp. lactis biovar diacetylactis (data not shown). It might be applied for the detection of other LAB in mixed culture. The immunofluorescent method described in this paper is easy, rapid (4 h), and sensitive (individual cells can be detected) and involves a gentle nondestructive preparation of beads. This method combined with ELISA can be used to detect simultaneously free cells in mixed culture, to monitor microbial dynamics in gel beads, and to study the cross-contamination phenomenon during the production of mixed starters by immobilized-cell technology. This original approach may be applied to other mixed-strain starter systems when specific antibodies are available.