TEXT

Top Abstract Text References

Bacterioplankton are the principal consumers of the dissolved organic carbon (DOC) pool in the sea, which besides soil humus represents the largest reactive organic carbon reservoir on earth (24). The chemically characterizable macromolecular compounds of the DOC pool are carbohydrates, proteins, and lipids, comprising all together about 20 to 40% of the total DOC, commonly with higher percentages of macromolecular compounds present in the euphotic layers of the ocean than in the deep waters (5). The turnover of this DOC pool is highly variable, ranging between 40 to 100 days in the euphotic layers and averaging 6,000 years in the deep sea (14, 25).

As shown by ultrafiltration techniques, only about 20 to 30% of the DOC present in the ocean is of >1,000 Da (2). Yet this high-molecular-size fraction, which is contemporarily produced (20), is turned over more rapidly than the <1,000-Da DOC fraction (1, 2). Gram-negative bacterioplankton, which usually constitute more than 90% of the total bacterioplankton in seawater (9), can only transport molecules of <600 Da through their cell walls (17). This hydrolysis of the large bioavailable molecules is mediated by surface-bound ectoenzymes located at the cell wall or in the periplasmic space (8, 18). Usually bacterial surface-associated ectoenzymatic activity dominates over freely dissolved extracellular enzymes (sensu Chróst [8]) in the marine environment (6).

Over the past two decades, a number of studies have focused on the determination of bacterial ectoenzymatic activity using substrate analogs linked to fluorophores which fluoresce upon cleavage from the substrate by the activity of the appropriate ectoenzymes (12, 13). The high fluorescence yield upon cleavage allows determination of the ectoenzymatic activity of bacterioplankton using incubation periods of only minutes to hours and in situ temperature conditions. This approach has been widely applied to determine the ectoenzymatic activity of bacterioplankton.

Biphasic or even multiphasic kinetics have been obtained not only for ectoenzymatic activity but also for uptake of organic substrates (4, 21, 23). The obvious question which arises from these studies is as follows: why should natural bacteria exhibit K_m values in the millimolar range if the substrate is present only in the nanomolar to micromolar range in the environment? Another question is whether a single bacterial species or even an individual cell can exhibit biphasic uptake and ectoenzymatic activity. To the best of our knowledge, there are no reports of bi- or multiphasic hydrolytic ectoenzyme kinetics available for single bacterial strains. The potential ecological implications of these observations have been discussed previously (3, 23). Without going into detail here, the problem of the apparently existing ectoenzymatic diversity in the natural environment has not been addressed yet. There is only one study which determined the diversity of a bacterioplankton ectoenzyme, -glucosidase, in the natural environment (19). These authors found only two different -glucosidases in the pycnocline layer of the Adriatic Sea, indicating a rather limited diversity of -glucosidase, especially if the large number of potential substrates is considered (19).

With the present methods available for assessing ectoenzymatic activity under near-natural conditions, the required resolution to determine ectoenzyme diversity in natural bacterioplankton communities cannot be achieved. We therefore modified a capillary electrophoresis (CE)-based method, originally described by Xue and Yeung (26) for detecting lactate dehydrogenase in mammalian cells, to separate and detect the different bacterial -glucosidases present in seawater. This modified method not only allows the determination of the ectoenzyme diversity present in a given sample but also allows the simultaneous determination of the kinetics of all the different ectoenzymes present in the sample. In this paper we present the outline of the method with the example of -glucosidase. (This work was performed in partial fulfillment of the requirements for a Ph.D. from the University of Groningen, Groningen, The Netherlands, by J.M.A.)

Bacterial ectoenzyme extraction. For developing this method, four -glucosidase-producing bacterial strains (two gram-positive strains, Planococcus citreus and Salinicoccus roseus, and two gram-negative strains, Vibrio sp. and Paracoccus alkenifer, belonging to the - and -proteobacteria, respectively) isolated from the northern Adriatic Sea (11) were used. The strains were cultured in ZoBell 2216 broth (5 g of peptone, 1 g of yeast extract, 1 liter of 0.2-µm-pore-size-filtered seawater). One-milliliter aliquots of exponentially growing cultures were harvested by centrifugation (3,200 × g, 4°C, 15 min) and washed three times with 0.2-µm-pore-size-filtered artificial seawater (16) before resuspension of the pellet in the buffer solution for ectoenzyme extraction (described below).

View this table: [in this window] [in a new window]   TABLE 1.   Mean extraction efficiency and range of -glucosidase activity measured in triplicate for the single bacterial strains and in five different bacterial communities from the coastal North Sea

CE separation and online detection of -glucosidase activity. Analyses were performed using a Biofocus 3000 CE system equipped with a Biofocus LIF2 laser-induced fluorescence detector (Bio-Rad, Hercules, Calif.). The enzyme separation and detection were performed in 50-µm-inside-diameter fused silica capillaries. To avoid electro-osmotic flow and to minimize protein interaction with the wall of the capillary, the inner surface of the capillary was coated with polyacrylamide by means of a siloxane bond (15). The total length of the capillary was 75 cm and the distance from the inlet to the detection window was 70.4 cm. Detection of -glucosidase activity in the capillary was based on the hydrolysis of the fluorogenic substrate analog resorufin--D-glucopyranoside (Rglu) (Sigma). The fluorophore released upon hydrolysis is resorufin and was measured using the 594-nm line of the helium-neon laser as the excitation source, with the emission measured at 630 nm in the LIF2 system (600DRLP02 beam splitter, 630DF30 discrimination filter; Bio-Rad).

View larger version (31K): [in this window] [in a new window]   FIG. 1.   Schematic outline of the method used to separate and detect different types of -glucosidases. Different symbols represent different enzymes in the sample (see text for a more detailed explanation of the four steps).

Separation of different -glucosidases. The separation conditions were optimized for -glucosidase suspensions extracted from two different strains of gram-negative bacteria (Vibrio sp. and P. alkenifer). Single peaks of -glucosidase activity were obtained when the extracts from the two bacterial strains were injected separately (Fig. 2A). The fluorescent signal increased linearly with increasing incubation times of 5, 10, 20, 30, and 40 min, without significant peak distortion due to diffusion of the fluorescent product. However, a tail of elevated fluorescence was found preceding the peaks (Fig. 2A) in all electropherograms. The intensity of this feature was proportional to the amount of sample injected but did not increase with increasing incubation times and did not appear in heat-denatured samples. Therefore, we believe that it is produced by the enzyme(s) during its migration in the capillary. A similar behavior was also reported for lactate dehydrogenase (27). This fluorescent tail, however, biases the determination of an accurate baseline, especially if dealing with more complex samples such as natural enzyme extracts. To accurately determine hydrolysis rates and evidence the enzyme peaks, repeated injections with increasing incubation periods are required in order to discriminate the fluorescence produced during the incubation period from the signal produced elsewhere. Therefore, incubation times of 10 and 20 min were chosen for the experiment with the natural community and the ectoenzymatic activity was calculated by integrating the peak area. Different pH values ranging between 7.50 and 8.50 were tested. The peak resolution decreased at pHs of 8.00. Therefore, pH 7.50 was chosen for an optimal separation of the -glucosidases.

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Electropherograms obtained after injecting enzyme extracts from two different strains (Vibrio sp. and P. alkenifer) separately (A) and mixed together (B); note that the peak height is reduced by 50% in the mixture (B) as it contains only 50% of each of the extracts. RFU, relative fluorescence units.

Separation of -glucosidases from natural bacterial communities. Once the method was optimized for the separation of -glucosidases from strains, the ectoenzymes from a whole bacterial community were extracted as described above and run in the CE under the same conditions described for the strains. Peaks were considered to represent -glucosidases if the integrated area of fluorescence increased systematically in replicate incubations of 10 and 20 min, as explained above. In contrast to previous observations on the variability (diversity) of -glucosidases in marine systems (19), the electrophoretic pattern we obtained (Fig. 3) revealed seven distinct peaks of -glucosidase activity, while Rath and Herndl (19) found only one to two different types of -glucosidase using chromatographic methods. The differences between the former study and the CE approach may be partly due to the superior resolution power of CE in combination with the sensitivity of laser-induced fluorescence detection. The differences in electrophoretic mobility of native proteins are caused by differences in the amino acid sequence and/or by the secondary structure of the protein (10). Therefore, these different forms of -glucosidase may have different functional properties such as reaction kinetics. Differences in the kinetic properties of ectoenzymes often reflect differences in the substrate supply and utilization among different members of the bacterioplankton community (22, 23).

View larger version (22K): [in this window] [in a new window]   FIG. 3.   Electrophoretic pattern obtained from enzymes extracted from a natural bacterial community. The different lines show the results obtained with different concentrations of Rglu. RFU, relative fluorescence units.

Determination of -glucosidase kinetics using CE. Standard solutions of resorufin ranging from 8 nM to 400 µM were injected in triplicate to establish the calibration curves necessary for converting the relative fluorescence units into resorufin concentrations. Therefore, the absolute amount of resorufin injected ranged from 0.90 to 58,800 fmol (1 fmol = 10¹⁵ mol). Despite the fact that the 594-nm wavelength provided by our laser is not optimal for the excitation of resorufin (excitation maximum, 571 nm), we successfully detected as few as 3.45 × 10¹⁰ molecules (57.33 fmol) of resorufin in the electrophoresis buffer. Below this threshold it was not possible to clearly distinguish the signal from background noise. The increase in fluorescence intensity of resorufin was linear (r² = 0.998) over the whole range from 57.33 up to 58,800 fmol of resorufin, making this method suitable for quantification of enzyme activity over a large range of enzyme concentrations and specific activities. At the lowest amounts of resorufin (<1 pM), the measured peak area was slightly lower than expected from the calculated regression line obtained over the whole concentration range. However, this had no effect on our measurements due to the high fluorescent readings obtained for both the strains and the natural samples.

1

View this table: [in this window] [in a new window]   TABLE 2.   Kinetic parameters of the different -glucosidases detected in a natural bacterioplankton sample

Characterizing the diversity of -glucosidases. For the characterization of the -glucosidase diversity it is necessary to measure both the number of different -glucosidases present in the environmental sample and the relative amount of each ectoenzyme (V_max is proportional to the amount of enzyme present). The relative amount of ectoenzyme is certainly biased due to the different extraction efficiencies observed. However, if we assume that the extraction efficiency is constant for a specific ectoenzyme produced by a bacterial species, it is possible to compare the relative abundance of a specific -glucosidase between related samples, i.e., by studying the succession of ectoenzyme activities in bacterioplankton communities. Different ectoenzymes might have the same electrophoretic mobility, thus potentially leading to an underestimation of the number of enzymes present. Therefore, the determination of the -glucosidase richness as obtained by CE has to be considered a conservative estimate. Nevertheless, we detected a higher number of -glucosidases in natural bacterioplankton with the CE approach than that reported hitherto.