Biodegradation of Metal-EDTA Complexes by an Enriched Microbial Population

Abstract

A mixed culture utilizing EDTA as the sole carbon source was isolated from a mixed inoculum of water from the River Mersey (United Kingdom) and sludge from an industrial effluent treatment plant. Fourteen component organisms were isolated from the culture, including representatives of the genera Methylobacterium, Variovorax, Enterobacter, Aureobacterium, and Bacillus. The mixed culture biodegraded metal-EDTA complexes slowly; the biodegradability was in the order Fe>Cu>Co>Ni>Cd. By incorporation of inorganic phosphate into the medium as a precipitant ligand, heavy metals were removed in parallel to EDTA degradation. The mixed culture also utilized a number of possible EDTA degradation intermediates as carbon sources.

EDTA, an aminopolycarboxylic acid chelating agent, has many applications. For example, it is used in washing powders as a substitute for polyphosphates, which have been implicated in the eutrophication of aquatic environments (11, 12). The large quantities of EDTA used commercially, and its low-level biodegradability (11, 12), allow it to remain at high levels in wastewater (11, 12, 14), from which it is not removed by conventional wastewater treatment, and also to persist in drinking water (10, 14, 36, 37). EDTA is also used for the decontamination of nuclear power plant equipment (3, 20, 23, 31) because it forms very strong complexes with heavy metals and makes them soluble and easy to remove from contaminated surfaces and soils (22, 24). Conversely, if EDTA is not removed from wastes before dumping, it can mobilize metals, e.g., the radionuclides in nuclear waste dumps (8, 31). At the Oak Ridge National Laboratory (Oak Ridge, Tenn.) and Maxey Flats (Kentucky) sites, radionuclides were detected up to 100 m from the burial site (8, 25), with EDTA-enhanced nuclide migration implicated in nuclide mobilization (24, 25). Radionuclides normally bind to humic and fulvic acids in soil and become immobilized due to the tight complexes formed (7). EDTA prevents this, having a higher affinity for the radionuclides than the humate and fulvate ligands (7). It is now illegal to dump complexing agents together with radionuclides in low-level burial sites in the United States (29); such mixed wastes require pretreatment (13). UV photodegradation of the ferric chelate of EDTA was observed in early studies (16, 17), but extensive photolytic mineralization does not occur (19) and this method of degradation might not be applicable to large volumes. Alternatively, biodegradation can release free metal, allowing this to be concentrated via chemical or biotechnological methods (13).

EDTA is degraded slowly in the environment (6, 35). Its biodegradation was proposed to comprise two pathways (4). The first involves the stepwise removal of acetate groups to leave ethylenediamine and the second involves the stepwise removal of an iminodiacetate (IDA) group, a glycine, and then an ammonium group with nitrilotriacetate (NTA)-aldehyde (4).

An Agrobacterium radiobacter strain able to biodegrade the ammonium ferric complex of EDTA was isolated (18) but was unable to degrade nickel, cobalt, or copper EDTA complexes or ferric complexes of NTA, IDA, or ethylenediamine-N,N′-diacetic acid (EDDA) (18, 30), despite their role as possible intermediates of EDTA degradation (4). EDTA biodegradation by a mixed culture and one of its component organisms isolated from sewage sludge was studied by Nortemann et al. (27, 28), using ferric ammonium EDTA as the sole nitrogen source initially and as the sole carbon source in later studies. A preliminary report suggested the possible growth of a mixed culture at the expense of heavy metal-EDTA complexes, but the loss of EDTA from the medium was not confirmed (21).

To date, the biodegradation of heavy metal complexes of EDTA, other than Fe, has not been demonstrated. The present study investigated the biodegradation of these complexes by an enriched microbial population and suggested the possible use of such populations in the remediation of wastes containing these complexes.

MATERIALS AND METHODS

Culture isolation and culture conditions.

Samples were taken from the River Mersey (United Kingdom) and from liquid effluent sludge from an industrial effluent treatment plant. The river water (8 ml) and sludge (2 ml) were inoculated into 90 ml of filter-sterilized minimal medium (MM) with the following components (grams per liter): CaCl₂, 0.025; MgSO₄ · 7H₂O, 0.2; NaCl, 0.1; (NH₄)₂SO₄, 0.5; disodium EDTA, 0.015; ZnSO₄ · 7H₂O, 0.0066; MnCl₂ · 4H₂O, 0.00171; FeSO₄ · 7H₂O, 0.0015; CoCl₂ · 6H₂O, 0.000483; CuSO₄ · 5H₂O, 0.000471; NaMoO₄ · 2H₂O, 0.000453; 3-(N-morpholino)propanesulfonic acid, 5.225; and KH₂PO₄, 0.272. The pH was adjusted to 7 with 1 M NaOH, and 7.34 g of FeNaEDTA (Sigma, Poole, United Kingdom) liter⁻¹ (20 mM) was added as the sole utilizable carbon source. The cultures were shaken (160 rpm, 30°C) for a month and then subcultured (10-ml inocula) into 90 ml of fresh MM with 1.83 g of FeNaEDTA liter⁻¹ (5 mM); this was repeated monthly for 2 years.

To test for growth on heavy metal-EDTA complexes, samples (10 ml) taken from cultures pregrown with FeNaEDTA as the sole carbon source (30 days) were harvested by centrifugation (7,000 rpm, MSE high-speed 18 centrifuge, room temperature), washed twice in sterile isotonic saline (8.5 g of NaCl liter⁻¹), and inoculated into 100 ml of MM containing 5 mM KH₂PO₄ and supplemented (to 5 mM) with CuNa₂EDTA (1.88 g liter⁻¹), CdNa₂EDTA (2.22 g liter⁻¹), CoNa₂EDTA (1.96 g liter⁻¹), NiNa₂EDTA (1.96 g liter⁻¹) (all from TCI, Tokyo, Japan), FeNaEDTA (1.83 g liter⁻¹), or disodium EDTA (1.86 g liter⁻¹; Sigma) as the sole carbon source. The EDTA compounds were of the highest grade commercially available (the NaEDTA was 99% pure; the others were at least 96% pure, with sulfate as the major contaminant, according to the manufacturer’s specifications), and the purity of each was confirmed by high-performance liquid chromatography (see below). Controls were samples without biomass (for each EDTA complex) and non-EDTA-supplemented cultures. Cultures were incubated as described above, with samples (1 ml) taken periodically and stored at −20°C for later analyses.

Growth on various substrates.

Samples (10 ml) were taken from cultures previously grown on FeEDTA (30 days), harvested, and washed as before and inoculated into MM (100 ml) supplemented with (5 mM) EDDA, ethylenediamine, trisodium acetate, trisodium NTA, IDA, or glycine, with growth tested as described above. Controls were uninoculated media and non-substrate-supplemented cultures.

Isolation and identification of microorganisms.

A 10⁻³ serial dilution was made in sterile isotonic saline from a month-old FeNaEDTA culture. A sample (10 μl) was spread plated onto MM plates containing 2 mM sodium acetate (BDH, Poole, United Kingdom) or 2 mM FeNaEDTA (30°C). Isolates were Gram stained by the Huckner method (9) with the Sigma Gram stain kit. Gram-negative rods were identified with the API 20E and 20NE kits (Biomerieux, Marcy l’Etoile, France). Isolates were also identified by the Microbial Identification System (MIDI-MIS, Newark, Del.) as described previously (33). Fatty acid methyl ester (FAME) profiles were matched with those in the MIDI-MIS Trypticase soy broth agar aerobic database (version 3.2).

Biomass immobilization.

Cultures (50 ml) previously grown on 1.83 g of FeEDTA liter⁻¹ (30 days), harvested, and washed as described above were inoculated into 2.5 liters of MM containing 1.5 g of ethylenediamine liter⁻¹ and 3.4 g of trisodium acetate liter⁻¹ (both 25 mM) in an airlift reactor constructed in the laboratory. Shale particles (60 g; average particle size, 3 by 3 mm; Thames Water) were washed with distilled water, sterilized (10% Chloros, 2 days), and washed several times with sterile distilled water. The particles were divided into two 30-g portions and placed into the lumens of two identical glass columns (each with a 90-ml capacity; 14.5 by 2.8 cm). The culture was cycled (Watson Marlow flow inducer) though both columns simultaneously at a flow rate of 1 ml min⁻¹ for 5 days. Random particles from each column were removed for protein analysis. After preparation, the columns were stored at 4°C until analyzed. For EDTA degradation experiments, 150 ml of MM containing 1 mM FeEDTA (0.367 g liter⁻¹) and 1 mM KH₂PO₄ was pumped through each column at a flow rate of 0.25 ml min⁻¹ (single pass). Column outflow samples, withdrawn hourly, were stored at −20°C for later analysis, together with samples of the input solution.

Analyses of biomass, residual EDTA, and metals.

Frozen samples (1 ml) were thawed, the biomass was removed by centrifugation (13,000 rpm in a Heraeus Sepatech Biofuge A at room temperature), and supernatant samples were removed, filtered (0.45-μm pore size), and diluted into fresh vacuum-filtered and degassed MM. EDTA analysis was based on a previously described method (5) using high-performance liquid chromatography with a 410 series UV detector (the pumps, injector, and detector were all from Waters, Milford, Mass.) and a Nucleosil 5-μm C₁₈ column, 250 by 4.6 mm (Phenomenex, Macclesfield, Cheshire, United Kingdom). The mobile phase was 0.03 M sodium acetate/acetic acid buffer (pH 4)/20 mM tetrabutyl ammonium hydroxide and 100 ml of methanol liter⁻¹, filtered and degassed as described above. EDTA was detected at wavelengths of 254 and 300 nm. Metals were assayed by atomic-absorption spectroscopy (Pye Unicam, Cambridge, United Kingdom), with an acetylene (9-lb/in²) and air (30-lb/in²) flame. Cd, Cu, Zn, Ni, and Co were detected at wavelengths of 228.8, 324.8, 213.9, 232.0, and 240.7 nm, respectively, against a blank of fresh MM.

For protein analysis, cell pellets were resuspended in 50 mM NaEDTA (1 ml), mixed for 10 min to remove bound metals, recentrifuged, and resuspended in MilliQ water. Protein was measured by the copper sulfate-bicinchoninic acid method (reagents were all from Sigma). For protein analysis of biomass-loaded shale, specimen shale particles were placed in a 1.5-ml tube containing MM with 50 mM NaEDTA (500 μl) and mixed vigorously for 10 min. This was repeated until no protein was detected in the wash solution. The dislodged cells were harvested and the protein concentration was estimated as described above.

Treatment of results.

All experiments were done in triplicate on three separate occasions, and the data were pooled and calculated as means ± standard deviations (SD) for three replicates. In general the SD was within ±10% of the mean and is only shown where this value was exceeded. Means ± SD for individual values are given below as appropriate.

RESULTS AND DISCUSSION

Microorganism isolation and identification.

Fourteen morphologically distinct colony types were isolated from the mixed culture after 18 subcultures. Four were gram-negative bacteria and were identified by the API 20E and 20NE systems as strains of Enterobacter and Pseudomonas, respectively. These, and the 10 gram-positive bacteria, were identified further on the basis of cellular FAME profiles. Of the culturable organisms within the culture, gram-positive bacteria predominated, comprising species of Methylobacterium (35%), Variovorax (17%), Bacillus (10%), and Aureobacterium (10%) (Table 1). Variovorax (Alcaligenes) paradoxus was shown previously to tolerate EDTA (15) or, possibly more correctly, the Fe starvation that could be imposed by extensive tight chelation of Fe. This organism can degrade a variety of xenobiotics and also hydroxybutyrate, an ability it shares with Aureobacterium saperdae (2, 26, 32, 37, 38), which was also found in the present culture (Table 1). The microorganisms identified in Table 1 represent only those that were culturable on plates. It is likely that a number of other microorganisms contributing to EDTA biodegradation but probably unculturable in isolation were also present. The total number of microorganisms found per milliliter was 10- to 100-fold less than expected on the basis of viable counts and compared with the determined amount of protein per milliliter. Therefore, it can be assumed that >90% of the microorganisms are not culturable in isolation by the method used here. This is supported by a published report which suggests that <1% of all microorganisms are culturable (1). Further studies using molecular probe methods are warranted.

TABLE 1.

Composition of the EDTA-degrading mixed culture

Isolate	Organisma	Composition of culturable isolates (approx. %)
EDTA 1	Methylobacterium mesophilicum	10
EDTA 2	Aureobacterium saperdae	11
EDTA 3	Unidentifiable gram-positive organism	3
EDTA 4	Bacillus globisporus	5
EDTA 5	Enterobacter agglomerans (99.9)b	5
EDTA 6	Pseudomonas putida (99.8)b	3
EDTA 7	Unidentifiable gram-positive organism	2
EDTA 8	Pseudomonas aeruginosa (79.1)b	3
EDTA 9	Bacillus globisporus	6
EDTA 10	Methylobacterium mesophilicum	10
EDTA 11	Variovorax paradoxus	12
EDTA 12	Methylobacterium mesophilicum	15
EDTA 13	Enterobacter agglomerans (81.8)b	10
EDTA 14	Variovorax paradoxus	5

^aThe culturable isolates were identified by FAME-Microbial Identification System analysis and compared against the MIDI-MIS database. 

^bAlso identified with the API 20 system. Numbers in parentheses are probabilities (in percent) obtained by the API test. 

Biodegradation of metal-EDTA complexes and likely metabolic intermediates.

Incubation of the culture with the metal-EDTA complexes resulted in growth, biodegradation of each complex, and removal of metal over 28 days (Fig. 1). Equimolar inorganic phosphate was incorporated to scavenge released metal by precipitation, avoiding metal toxicity and illustrating the potential for remediation of metal-EDTA wastes. Growth on Fe- and NaEDTA gave doubling times of 66 and 288 h, respectively, with corresponding levels of EDTA degradation of 3.01 ± 0.19 mM (60%) and 1.57 ± 0.26 mM (31%) after 28 days (Fig. 1B and A). Biodegradation of the FeEDTA complex also resulted in the removal of 44% ± 0.03% of Fe from solution, probably as the hydroxide or phosphate. The apparently poor growth with and low biodegradability of NaEDTA in comparison to FeEDTA are probably due to the replacement of the Na ligand with essential trace metals from the medium, effectively starving the cells of these metals. Growth at the expense of the recalcitrant heavy metal complexes, Cd-, Cu-, Co-, and NiEDTA, was observed. The respective extents of EDTA biodegradation for these were 0.95 ± 0.13 mM (19.3%), 1.54 ± 0.23 mM (30%), 1.3 ± 0.23 mM (25%), and 1.16 ± 0.24 mM (23.4%) (Fig. 1). The corresponding extents of metal removal were 19% ± 0.02% (Cd), 16% ± 0.01% (Cu), 14% ± 0.01% (Co), and 31% ± 0.00% (Ni) (Fig. 1C to F). Metal removal was stoichiometric or greater in the cases of Cd and Ni but only approximately 50% of that released from EDTA with Cu and Co. Insufficient metal was precipitated to permit X-ray powder diffraction analysis of the recovered precipitate.

FIG. 1.

Growth (⧫) of the mixed culture on MM supplemented with 5 mM inorganic phosphate and NaEDTA (A), Fe(III)EDTA (B), CdEDTA (C), CuEDTA (D), CoEDTA (E), and NiEDTA (F) complexes; EDTA consumption (•); and metal removal (▴). All SD are within 10% except where indicated. ▪, control (non-substrate-supplemented culture).

From the above data the biodegradability of the heavy metal-EDTA complexes was ranked as follows: Fe>Cu>Co>Ni>Cd. From previous studies it could be predicted that FeEDTA would be the most readily biodegraded (4, 18). Previous work (18) also demonstrated biodegradation of Fe(III)EDTA by an Agrobacterium sp. at a rate of approximately 275 μmol h⁻¹ μg of protein⁻¹, which was nearly 20-fold higher than that obtained with the present mixed culture (15 μmol h⁻¹ μg of protein⁻¹). With a mixed culture, not all of the component organisms will degrade EDTA per se and may instead scavenge low-molecular-weight products, contributing to the total biomass protein and lowering the specific activity against the parent compound. The above-mentioned Agrobacterium sp. was of limited potential, being unable to biodegrade EDTA at concentrations below 5 mM or to utilize other metal-EDTA complexes. However, in addition to being versatile, the present culture gave a residual EDTA concentration of 2 to 2.5 mM (FeEDTA), with continuing activity at termination of the experiment at 30 days. Degradation of the other heavy metal-EDTA complexes appeared to cease at 3 to 4 mM EDTA (Fig. 1). This could be due to a specific threshold (e.g., as for the Agrobacterium sp.) below which the complexes of metals other than Fe are unable to be transported into the cells; the biphasic nature of the removal of FeEDTA (Fig. 1B) could suggest the presence of two uptake processes, one nonspecific, low-affinity system and one Fe-specific, higher-affinity, system. The apparent K_m values for the various complexes have never been studied and would repay further investigation, as they may ultimately limit the usefulness of an industrial process. However, such K_m data would be difficult to interpret by using a mixed culture because individually, the isolates grew poorly at the expense of EDTA.

Other metal complexes (Mg-, Mn-, and CaEDTA) have been biodegraded completely by both a mixed microbial population and a single isolate. However, these metals are of relatively low toxicity and are essential micronutrients (27). A study of CoEDTA biodegradation using an Fe(III)EDTA-biodegrading Agrobacterium sp. was successful, but only because of displacement of the Co by Fe(III) (30). Biodegradation of the Co, Cu, Ni, and Cd complexes of EDTA has not been reported previously. Other workers have suggested the role of a specific Fe binding siderophore in the biodegradation of Fe(III)EDTA (18), and indeed, EDTA degradation did not cease as with the other metals but continued at a lower rate after 5 days (Fig. 1). It has been suggested that microorganisms isolated from an EDTA-rich environment have developed a specific Fe transport mechanism which is dependent on the binding of EDTA (18).

Previous studies have paid little attention to the likely breakdown intermediates of EDTA as potential “bottlenecks” which may limit the potential for its complete mineralization. The culture in the present study grew on substrates (EDDA, ethylenediamine, trisodium acetate, NTA, IDA, and glycine) previously proposed as degradation intermediates (4). Growth was more rapid than on FeEDTA for all substrates except EDDA, which, after a lag, gave a growth rate equal to that on EDTA (Fig. 2). Buildup of these intermediates was therefore unlikely to inhibit complete EDTA mineralization.

FIG. 2.

Growth of the culture on acetate (○), ethylenediamine (▾), IDA (▵), glycine (▿), NTA (•), EDDA (◊), and FeEDTA (▴). ▪, control (non-substrate-supplemented culture). All SD are within 10%.

EDTA biodegradation by immobilized cells.

The potential for developing a continuous-flow method of treating wastes bearing metal-EDTA complexes was assessed by using biomass samples of the mixed culture immobilized as a biofilm on shale particles. At a residence time of only 2.5 h (the fluid volume was 37 ml; the flow rate was 0.25 ml min⁻¹), the immobilized cells degraded 20% ± 0.02% of the supplied FeEDTA and 13% ± 0.03% of the Fe was removed (150 ml of 1 mM FeEDTA over 10 h). In contrast, the time required for 20% removal of FeEDTA in the batch system was 4.5 days (Fig. 1). The input concentration of EDTA here was 1 mM, below the concentration which could be degraded by Agrobacterium in previous studies (see above).

The generation of a natural biofilm on shale particles may not be the most suitable method for the continuous treatment of metal-EDTA wastes. As an alternative, immobilization onto microcarriers as described previously (34) could be a superior method, and this should be investigated in future studies.

In conclusion, we demonstrate for the first time the feasibility of biodegradation of several heavy metal-EDTA chelates and also identify some unusual microorganisms implicated in this process. Significant removal of EDTA in a continuous-flow system by immobilized biofilm justifies development of process engineering aspects of the system and further elucidation of the microbiological composition and biochemical processes within the culture.