TEXT

Top Abstract Text References

Most current assays for extracellular enzyme activity rely upon enzymatic degradation of a dissolved nonfluorescent substrate into a fluorescent soluble product. The activities of several enzymes associated with intact aquatic biofilms have been characterized with such techniques (see, e.g., references 2, 3, 5, 6, 12, and 13). An important limitation of these assays is that they are not spatially explicit. Enzymes located at all positions within the biofilm and in the overlying water hydrolyze the dissolved substrate, although enzymes deep within a biofilm might contribute less to this hydrolysis, because of limited transfer of dissolved substrates through the biofilm (2). Enzymes involved in the degradation of particulate organic matter (POM) to which a biofilm is attached are not distributed throughout the biofilm; only enzymes in contact with POM can hydrolyze it. This discrepancy between the hydrolysis of dissolved substrates and the hydrolysis of POM is a serious problem (15). Another limitation of dissolved-substrate techniques is the inability to identify microspatial patterns in enzyme activity (i.e., potentially enzyme-rich areas near microbes and the identification of individual enzyme-producing microbes). A recently developed phosphatase assay, in which an insoluble fluorescent product is deposited at the site of enzymatic action (i.e., ELF-97; Molecular Probes), has yielded new insights regarding extracellular phosphatase production by individual cells and microspatial patterns in phosphatase activity in biofilms (11, 24; E. M. Espeland, S. N. Francoeur, and R. G. Wetzel, unpublished data). Such information is critical for understanding functional roles and microbial interactions within biofilm communities. A comparable assay for protease activity is lacking.

Wheat germ agglutinin (WGA) is a lectin that selectively binds to N-acetylglucosamine and N-acetylneuraminic acid residues (8, 28) and certain plant tissues, particularly hemicellulose (9). Fluorescent molecules can be conjugated to WGA and thereby attached to substrata via WGA binding (8). Plant detritus, a considerable proportion of which consists of hemicellulose, is a common POM substratum upon which autotrophic and/or heterotrophic biofilms grow (18, 26). Extracellular proteases produced by algal-bacterial biofilms presumably contribute to the breakdown of plant detritus. Thus, a protease assay utilizing fluorescein-conjugated WGA (FWGA) bound to a cellulose substratum may be useful for detecting, localizing, and measuring protease activities located at the biofilm-substratum interface.

The objective of this study was to devise a spatially explicit method for quantifying protease activity at the biofilm-substratum interface. Procedures for binding FWGA to a substratum were developed, and the subsequent release of fluorochrome by proteolytic actvity was documented. The performance of the FWGA method was experimentally compared with a standard leucine 7-amido-4-methylcoumarin (LAMC) assay. Microspatial patterns of bacterial distribution and FWGA fluorescence were visualized by confocal microscopy.

FWGA binding and enzymatic release. Millipore HA filters (mixed cellulose esters, 0.8-µm pore size) were cut into 7-mm-diameter circles and stained with FWGA by a procedure modified from that of Hogetsu (9). Eighteen filters were washed in 50 mM sodium phosphate buffer (pH 7.0) containing 0.01% MP-40 detergent and then rinsed twice in ultrapure water (pyrogen-free Millipore Milli-Q). Nine filters were placed in a solution of FWGA (400 µl at 20 µg ml¹, dissolved in an aqueous solution of 10 mM HEPES buffer, 150 mM NaCl, 0.1 mM CaCl₂, and 0.08% NaN₃ [pH 7.5]) and phosphate buffer (500 µl). Nine filters (unstained controls) were washed and placed in phosphate buffer without FWGA. All filters were then placed on a shaker table (50 rpm) in a laboratory growth chamber (dark, 30°C) for 18 h. After the staining period, unbound FWGA was removed from the filters by alternating rinses of sodium phosphate buffer and ultrapure water (three rinses in each). FWGA-stained filters were placed in vials containing 1.2 ml of one of three solutions: ultrapure water, sodium phosphate buffer, or sodium phosphate buffer with protease (0.83 mg of protease [EC 3.4.24.31; Sigma] ml¹, from the bacterium Streptomyces griseus). Unstained filters were placed in vials containing 1.2 ml of either ultrapure water or phosphate buffer. All filters were then returned to the shaker table (50 rpm, dark, 30°C).

1

Control of FWGA binding. To examine the effects of concentration and staining duration on the amount of FWGA binding to filters, filters were stained as previously described, except that a range of concentrations of FWGA stock solution (0.2, 2, and 20 µg ml¹) and staining durations (4, 36, 360, and 3,600 s) were used. All combinations of FWGA concentration and staining duration were replicated three times. After being stained, the filters were placed into sterile vials filled with protease solution (1.2 ml, 0.83 mg ml¹ in 50 mM sodium phosphate buffer) and incubated as previously described. After sufficient time to ensure hydrolysis of FWGA had reached an asymptote (160 h), fluorescence in the overlying liquid was measured as previously described. Protease solution in 50 mM phosphate buffer served as a blank. Stained filters were also observed with confocal microscopy.

1

1

View larger version (16K): [in this window] [in a new window]   FIG. 1.   (A) Saturation curve (6 h of incubation) for the FWGA protease assay. Note the saturation indicated by the constant fluorescence with increasing substrate supply. The smoothing line was generated by LOWESS (tension = 0.7). (B) Effect of staining duration of the FWGA content of filters used in the saturation experiment.

Experimental comparison of FWGA and LAMC assays. Nitrogen availability can influence microbial extracellular protease activity (4; S. N. Francoeur and R. G. Wetzel, unpublished data). The ability of the FWGA technique and a standard LAMC assay to detect responses to altered N supply were compared experimentally. Millipore GS filters (mixed cellulose esters, 0.2-µm pore size) were stained with FWGA (20 µg ml¹, 23 h). Low-biomass biofilms (chlorophyll a = 11.8 ± 2.2 mg m², ash-free dry mass = 3.80 ± 0.07 g m², mean ± 1 standard error) were constructed by gentle (<13 kPa) filtration of stationary-phase cultures of Nitzschia palea, Scenedesmus basiliensis, and Anabaena flos-aquae and a log-phase culture of bacteria from the Talladega Wetland Ecosystem (TWE; a 15.1-ha wetland in west-central Alabama) onto stained filters. The bacterial inoculum was generated by the method of Wetzel et al. (27) and included Klebsiella pneumoniae, Enterobacter sakazakii, Serratia sp., and an unidentified, gram-positive, coccoid bacterium (bacterial identification by fatty acid analysis; Microbial ID, Inc., Newark, Del.). After filtration, 7-mm-diameter disks were cut and placed in vials with 1 ml of either +N Moss Medium (16) [containing both NH₄Cl and Ca(NO₃)₂, ~10,000 µg-atoms N liter¹] or N Moss Medium (without NH₄Cl and Ca(NO₃)₂, 0 µg-atoms N liter¹). Each treatment was replicated 6 times. Vials were incubated for 6 h (30°C, ~35 µmol m² s¹ PAR), and then the level of dissolved fluorescence from the FWGA hydrolysis was determined as previously described. Media from vials containing stained, sterile filters and vials containing unstained filters with biofilms served as blanks.

Potential for simultaneous use of FWGA and MUF techniques. Fluorometric assay of FWGA and 4-methylumbelliferone (MUF) can be done simultaneously because of the wide separation of excitation and emission wavelengths of FWGA (~494-nm excitation, 518-nm emission) and MUF (~360-nm excitation, 449-nm emission) (8). Combining FWGA and MUF techniques would allow simultaneous assay of two different enzyme systems (protease and a nonproteolytic enzyme specific for a particular MUF-labeled substrate) in a single sample. High variability among samples often hampers biofilm enzyme activity assays (6) and other measurements (18). Thus, simultaneously quantifying the activity of two biofilm enzymes in the same sample would be a considerable advantage. Simultaneous application of FWGA and MUF or AMC protease assays in a single sample is not advisable, because of potential competitive inhibition among the substrates.

Direct visualization of microspatial patterns. Millipore GS filters were stained with FWGA (20 µg ml¹, 1 h) and inoculated via gentle filtration with the TWE bacterial inoculum. Filters were then cut into 7-mm-diameter cores, placed in N Moss Medium enriched with 1.1 mM glucose, and incubated (3 days, dark, 30°C). Stained filters were also submerged in a large (2,000-liter) wetland mesocosm containing sediments and macrophytes originally collected from the TWE. Microorganisms were allowed to naturally colonize the mesocosm filters for 3 days, and then the filters were placed into N Moss Medium and incubated (8 days) in the dark. After incubation, biofilms were counterstained with either the nucleic acid stain Syto-64 (20 µM for 5 min; Molecular Probes) or a propidium iodide-Syto-9 mixture (constructed biofilms only, 15 min; Live/Dead Bacterial Viability Kit; Molecular Probes). Filters were then rinsed three times with ultrapure water, mounted in immersion oil, and imaged with dual-channel scanning laser confocal microscopy (Nikon PCM-2000; Ar laser [488-nm excitation, 515-nm/30-nm emission]; HeNe laser [543.5-nm excitation, 600-nm LP emission]).

View larger version (93K): [in this window] [in a new window]   FIG. 2.   Confocal micrographs of bacteria on FWGA-stained filters. Note the textured surface of the filter. (A) Bacteria (red) surrounded by area of reduced FWGA fluorescence (area 1) and bacteria fluorescing green, possibly as a result of FWGA binding (area 2). Note the dark regions caused by irregularities in the filter thickness (area 3). (B) Membrane-compromised ("dead") bacteria (red), membrane-intact ("live") bacteria (bright green), and FWGA stained filter (light green textured background). Note the reduction of FWGA fluorescence in the area surrounding some live bacteria (area 1) but not others (area 2). Note also the reduced brightness of FWGA on right side of micrograph, as filter surface begins to leave the focal plane. (C) Red channel-only image of highlighted area in Fig. 2A. Note the stained bacterial cell and textured filter surface. (D) Green channel-only image of area shown in panel C. Note the lack of FWGA fluorescence in the region surrounding the bacterium, despite the presence of the filter surface (as shown in panel C). Bars, 10 µm. FIG. 3.   Confocal micrographs of mesocosm biofilms on FWGA-stained filters displaying areas of reduced FWGA fluorescence near some large coccoid cells (area 1) but not others (area 2). Reduced FWGA fluorescence was not observed near small bacterial rods. Bar, 10 µm.