INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Two acetate-utilizing methanogenic genera, Methanosarcina and Methanosaeta (formerly Methanothrix), have been identified as important methanogens in granular sludge from upflow anaerobic sludge blanket (UASB) reactors (7, 11, 14, 22). Methanosaeta spp. are filamentous organisms which are known to grow only on acetate (15). Methanosarcina spp. grow either as single coccoidal cells or in large clumps up to 1 to 3 mm in diameter. These clumps consist of large numbers of individual cells surrounded by a thick polymeric wall. Besides acetate, Methanosarcina spp. are also capable of growing on substrates such as methanol, methylamines, and sometimes hydrogen and carbon dioxide. Methanosaeta spp. have a lower growth rate at high acetate concentrations than do Methanosarcina spp., but their affinity for acetate is 5 to 10 times higher (15, 32). These kinetic data indicate that a low effluent concentration of acetate from a UASB reactor results in a selection for granules dominated by Methanosaeta spp. However, it should be noted that Methanosarcina spp., unlike Methanosaeta spp., can grow on several substrates; therefore, their role in UASB reactors cannot be based solely on their ability to use acetate. For example, Forster (9) investigated the correlation between the effluent concentrations of acetate from UASB reactors and the ratio of Methanosaeta to Methanosarcina. The results showed that there was no correlation and that other factors probably were involved in selection, such as macro and micro nutrients and the hydraulic loading of the reactor.

Both Methanosaeta and Methanosarcina spp. have been identified in granules from UASB reactors under stable conditions. It is generally assumed that Methanosaeta spp. improve granulation and result in more stable reactor performance; consequently, Methanosaeta spp. should be favored over Methanosarcina spp. Changes in morphology from clumps to single cells of Methanosarcina spp. cause the granules to disintegrate; i.e., granules dominated by Methanosarcina spp. disintegrate under conditions of high cation concentrations (1, 20). Methanosaeta spp. have also shown two morphological forms: a filamentous type consisting of long multicellular rod-shaped microorganisms and a rod-shaped microorganism consisting of fragments of up to five cells. Methanosaeta spp. typically grow in the filamentous form under substrate limitation (14, 30). The filamentous type has been found to result in bulking of sludge and hence to wash out from the reactor. However, most conclusions have been drawn from microscopic examinations of granular sludge from UASB reactors operated in an unsystematic manner.

Obviously, there is a need for more systematic experiments, e.g., in which granules are produced from specific known bacterial consortia, in order to obtain a better understanding of the granulation process and to obtain knowledge of the development of granular sludge after long-term operation of the UASB reactor. In the present study, we examined the immobilization patterns of two different species of acetate-utilizing methanogens and their interrelations by using acetate-fed UASB reactors with granules containing either immobilized Methanosaeta concilii GP-6, immobilized Methanosarcina mazeii S-6, or a combination of the two species.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Organisms. M. mazeii S-6 (DSM 2053) was kindly provided by E. Conway de Macario, Wadsworth Center for Laboratories and Research, Albany, N.Y. M. concilii GP-6 (DSM 3671) was obtained from the Deutsche Sammlung von Mikroorganismen (Braunschweig, Germany).

Granules. The granules came from a mesophilic full-scale UASB reactor that treats wastewater from a paper mill (Industriwater Eerbeek BV, Eerbeek, The Netherlands).

Medium and cultivation. The methanogens were grown on BA medium (3), modified as follows: 0.25 g of yeast extract and Casitone per liter were added, and NaS₂ · 9H₂O was omitted. Acetate was used as carbon source at a final concentration of 30 mM. Kinetic parameters were determined by using vials of modified BA medium inoculated with fresh wet granular sludge (2 to 5 ml) from the experimental UASB reactors or with an exponentially growing culture. The vials were acclimatized at 37°C for 1 h before sodium acetate was added to a final concentration between 0 and 30 mM. The degradation of acetate was monitored over time. The plot of Woolf was used for the determination of µ_max and K_s.

UASB reactors and sterilization of granules. Three glass UASB reactors with an active volume of 200 ml were used. The design of the reactors was as previously described (20). The reactors were run under sterile conditions during the experiments. Fifty-milliliter granules were added to each reactor as carrier material, and the reactors were autoclaved (40 minutes at 134°C) three times before use. The reactors were fed the medium described above, and the ratio of feed to recirculation was 1:4 during the whole experiment. The reactors were routinely checked for contamination by microscopic examination and by inoculation of 2 ml of effluent into vials containing BA medium supplemented with 2 g of yeast extract per liter and 1 g of glucose per liter.

Inoculation and start-up of UASB reactors. The reactors, containing sterile granules, were inoculated by the following scheme: the M. concilii reactor was inoculated with M. concilii GP-6, the M. mazeii reactor was inoculated with M. mazeii S-6, and the M. mazeii/M. concilii reactor was inoculated with both M. concilii GP-6 and M. mazeii S-6. Twenty-five milliliters of culture in late exponential phase was used for inoculation. The M. mazeii S-6 cells used as inoculum were growing as clumps. A hydraulic retention time (HRT) of approximately 18 h was used during start-up of the reactors. The HRT was decreased stepwise when the effluent concentration of acetate decreased to below 6 mM in the M. mazeii reactor and 3 mM in the M. concilii and M. mazeii/M. concilii reactors. The acetate concentrations were chosen based on the different kinetics of the two methanogens.

Analytical methods. Methane and volatile fatty acids were quantified by gas chromatography with flame ionization detection, as previously described (24). The different methanogens were quantified by an enzyme-linked immunosorbent assay (ELISA) using polyclonal antibodies raised against M. mazeii S-6 and M. concilii GP-6, as previously described by Sørensen and Ahring (25). Determination of the inner structure of the granules was done by immunohistochemical methods, as described by Schmidt et al. (23). Polyclonal antibodies against M. mazeii S-6 or M. concilii GP-6 were used. Microscopic observation was done with a Leica DMIRB/E microscope or a Leica DMIRB/E microscope equipped with a Leica True Confocal Scanner, model TCS 4D.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Initial immobilization of pure cultures. (i) Performance of the M. mazeii reactor. The initial HRT was approximately 18 h. After 10 days of operation, the effluent concentration of acetate decreased to below 6 mM, indicating growth inside in the reactor (Fig. 1A). Consequently, the HRT was decreased to approximately 14 h. The HRT was further decreased stepwise over the next 130 days of operation, resulting in an HRT of approximately 6 h, i.e., lower than the generation time of M. mazeii S-6, and the reactor still degraded acetate with an efficiency of approximately 95%. This indicates that M. mazeii S-6 was either immobilized in the granular sludge or was forming clumps which were retained in the reactor.

View larger version (18K): [in this window] [in a new window]   FIG. 1.   Acetate concentrations in the effluent () and HRT () as functions of time in the three UASB reactors. (A) M. mazeii reactor. (B) M. concilii reactor. (C) M. mazeii/M. concilii reactor.

(ii) Performance of the M. concilii reactor. After 90 days at an HRT of approximately 18 h, the effluent concentration of acetate dropped below 3 mM and the HRT was decreased to approximately 16 h (Fig. 1B). During the next 200 days, the HRT was stepwise lowered to approximately 6 h, i.e., much shorter than the generation time of M. concilii GP-6, and the reactor still degraded acetate with an efficiency of 98 to 100%. This indicates that M. concilii GP-6 was immobilized in the granular sludge.

(iii) Performance of the M. mazeii/M. concilii reactor. After approximately 40 days with an HRT of 18 h, the acetate concentration dropped below 3 mM and the HRT was lowered (Fig. 1C). The effluent concentration of acetate was less than 0.5 mM after 250 days of operation at an HRT of approximately 6 h, i.e., much shorter than the generation time of both M. mazeii S-6 and M. concilii GP-6. These data indicate that the methanogens were immobilized in the reactor.

Microbial composition of the granules. The ELISA using polyclonal antibodies raised against either M. concilii GP-6 or M. mazeii S-6 showed only M. mazeii S-6 in the M. mazeii reactor and only M. concilii GP-6 in the M. concilii reactor (Table 1). Furthermore, the highest number of M. concilii GP-6 organisms was found in the top of the granular sludge in the M. mazeii/M. concilii reactor with low acetate concentration, while M. mazeii S-6 dominated in the bottom of the granular sludge, where the acetate concentration was highest (Table 1).

Kinetic parameters. The kinetic parameters for acetate degradation in the granular sludge from the three UASB reactors were determined (Table 2). There was no significant difference between the kinetic parameters (µ_max and K_s) for acetate determined for the granular sludge in the M. concilii or M. mazeii reactor and the values determined for free suspended cultures. In the M. mazeii/M. concilii reactor, the kinetic parameters determined changed depending on the position of the granular sludge in the UASB reactor (Table 2). The highest obtained values of µ_max and K_s were found for methanogens immobilized in the bottom of the granular layer in the M. mazeii/M. concilii reactor. In the bottom of the reactor, there was no significant difference between the kinetic parameters found for immobilized methanogens and the kinetic parameters for free suspended cultures of M. mazeii S-6. In the top granular layer, the K_s value found for the immobilized methanogens was significantly different from the K_s values for free suspended cultures of both M. concilii GP-6 and M. mazeii S-6, while there were no significant differences between the µ_max values determined for methanogens immobilized in the top granular layer and free suspended cultures of M. concilii GP-6.

Immunohistochemistry. Thin sections of the granules from the different UASB reactors were examined with antibodies raised against M. mazeii S-6 or M. concilii GP-6. M. concilii GP-6 was found to be immobilized on the surface and in the center of the granules from the M. concilii reactor. Furthermore, M. concilii GP-6 was observed in what was assumed to be old channels in the granules (Fig. 2). M. concilii GP-6 was observed only as single rod-shaped cells and was never seen in the filamentous form.

View larger version (109K): [in this window] [in a new window]   FIG. 2.   Thin histologic sections of granules. Bright areas show positive reactions with the probe; dark areas show negative reactions with the probe. (A) M. concilii GP-6 immobilized on the surface, in the center, and in old channels of granules. (B) Single cells and small packets of M. mazeii S-6 on the surface and in channels of autoclaved granules. (C) Individual clumps of M. mazeii S-6. (D) Rods of M. concilii GP-6 immobilized on the surface of clumps of M. mazeii S-6.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The dynamics of M. mazeii S-6 and M. concilii GP-6 were studied after immobilization in sterile granules in UASB reactors. Immobilization was found to be fastest in the reactor with M. mazeii S-6 present, while the slowest immobilization was observed in the reactor inoculated with M. concilii GP-6. The autoclaved granules consist of dead cell material and extracellular material (21). The relatively fast start-up of the M. mazeii reactor could be due to a fast immobilization of this methanogen in the extracellular material, which probably facilitates incorporation in the granular layer. The long start-up period for the M. concilii reactor could, however, be due to the composition of the feed to the reactor, as the medium contained yeast extract and Casitone, which previously have been found to be inhibitory for M. concilii GP-6 (19). The data further indicate that even though growth of M. concilii GP-6 in the autoclaved granules was slow, the granules served as inert support material. The relatively easy immobilization of the methanogens in this study could be due to the fact that the liquid surface tension is high combined with the fact that M. mazeii S-6 and M. concilii GP-6 are both hydrophobic (5, 12). Under such conditions, both methanogens either adhere to the surface of the autoclaved granules or they self-immobilize, forming clumps of several cells (27-29).

Previously, other researchers have tried to form granules from pure cultures. Grotenhuis et al. (11) started UASB reactors with either Methanothrix soehngenii together with a propionate coculture or Methanosarcina barkeri to study the granulation process in UASB reactors. Using a similar approach, Grotenhuis et al. (13) also studied immobilization of an ethanol-degrading bacterium together with Methanobrevibacter arboriphilus AZ in a UASB reactor. In both cases, granula formation was reported but no further analyses were done. Wu et al. (31) studied the initial step during formation of granules from defined cultures degrading fatty acids. Again, however, no further examination of the granules was done. To our knowledge, this is the first report of pure cultures consisting of both Methanosarcina and Methanosaeta spp. immobilized in granular material, enabling us to study the dynamics of these methanogens in a "natural" environment. The same method has previously been used with success to study de novo dechlorinating activities in granular sludge (2, 4).

The immobilization patterns of the two methanogens were significantly different, indicating that the roles of the two methanogens are different in UASB granules. M. concilii GP-6 grew only on existing support material in the reactor, such as autoclaved granules, and did not self-immobilize. For start-up of UASB reactors without granular sludge (primary start-up), these results indicated that Methanosaeta spp. can only be immobilized if existing inert material or clumps of other microorganisms are present. M. mazeii S-6 showed a different immobilization pattern: in the bottom of the reactor where the acetate concentration was high, this organism formed large clumps. Independent of the location of the granular sludge, M. mazeii S-6 immobilized on the surface of the granules, making a structure of single cells and small clumps. Similar blanket-like structures have been reported for single cells of M. mazeii S-6 (18). These data indicate that the immobilization patterns of the two methanogens depend on the location in the granular layer and the availability of inert support material.

Several literature reports have shown that immobilized bacteria have different properties than free suspended cells. There are reports showing both increases and decreases in growth of different bacteria (8, 10, 14, 16, 17, 26). The nature of the substratum and the substrate utilized by the bacteria was further shown to influence the growth rate after immobilization. When Klebsiella oxytoca was immobilized on granular activated carbon, it was shown to have a 10-times-higher growth rate on glutamate than in free suspended culture. No differences in the growth rate were observed if the bacteria were grown on glucose, a substrate that did not absorb to the granular activated carbon (6, 10). In our study, we found no significant differences between the growth parameters obtained with suspended and immobilized cultures of M. mazeii S-6 or M. concilii GP-6.

There were significant differences in the performance of the two reactors depending on the methanogen immobilized in the reactor. The effluent concentration of acetate during steady state was lower for the M. concilii reactor than for the M. mazeii reactor. This is in accordance with the differences in growth parameters of M. mazeii S-6 and M. concilii GP-6. The data indicate that a better performance of the UASB reactor is obtained if Methanosaeta spp. are present in addition to Methanosarcina spp. However, in granules where Methanosarcina spp. were the only acetate-utilizing methanogen present, a syntrophic acetate-oxidizing system was found (20). This enabled effective degradation at low concentrations of acetate and consequently a low effluent concentration of acetate.

Start-up of the M. mazeii/M. concilii reactor was faster than that of the M. concilii reactor but still slower than that of the M. mazeii reactor. In the M. mazeii/M. concilii reactor, both methanogens were found to stick to autoclaved granules. However, growth of M. concilii GP-6 could be inhibited by the composition of the medium (19) so that mainly M. mazeii S-6 was metabolically active in the first phase. As the criterion used for decreasing the HRT in the M. mazeii/M. concilii reactor was more restrictive than that for the M. mazeii reactor, this could be the reason for the slower start-up of the M. mazeii/M. concilii reactor than the M. mazeii reactor. The effluent concentration of acetate was generally lower in the M. mazeii/M. concilii reactor than in the two other reactors; furthermore, the organic loading rate had only a slight effect on the concentration of acetate found in the M. mazeii/M. concilii reactor compared to the others, indicating that the performance of the M. mazeii/M. concilii reactor was the most stable.

The immobilization patterns in the M. mazeii/M. concilii reactor were similar to the patterns found in the two other reactors. M. concilii GP-6 was, in addition, found to have been immobilized on the surface of clumps of M. mazeii S-6. However, significant variations in the distributions of the two methanogens were observed in the M. mazeii/M. concilii reactor. The highest number of M. concilii GP-6 organisms was found in the top granular layer, indicating that the niche for M. concilii GP-6 was where the acetate concentration was lowest. In contrast, M. mazeii S-6 was primarily immobilized in the bottom of the reactor, with the highest acetate concentration. These results indicate that the two methanogens were immobilized according to the actual acetate concentration in the granular sludge. Whether Methanosarcina spp., Methanosaeta spp., or both dominate in the granular sludge from the UASB reactor depends on the wastewater composition and the actual operating conditions applied to the UASB reactor, i.e., organic loading rate and HRT.

The data presented in this paper clearly show that both methanogens can be immobilized in UASB systems, either separately or together in the same reactor. There are clear differences in the immobilization patterns and dynamics of the two aceticlastic methanogens corresponding to the expected behaviors based on kinetic parameters of the two methanogens. Our results further illustrate the advantage of having both methanogens present in granular sludge from UASB reactors: Methanosarcina spp. can be inert support material for the growth of Methanosaeta spp., and the reactor will be more stable and resistant to fluctuations in the loading and composition of the wastewater fed to the reactor.