Vol. 50, No. 2APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 1985, p. 342-3490099-2240/85/080342-08$02.00/0Copyright C 1985, American Society for Microbiology

Effects of Chemical Speciation on the Mineralization of OrganicCompounds by Microorganisms

E. L. MADSENt AND MARTIN ALEXANDER*Laboratory of Soil Microbiology, Department ofAgronomy, Cornell University, Ithaca, New York 14853

Received 25 March 1985/Accepted 28 May 1985

The mineralization of 1.0 to 100 ng each of four complexing compounds-oxalate, citrate, nitrilotriacetate(NTA), and EDTA-per ml was tested in media prepared in accordance with equilibrium calculations by acomputer program so that the H, Ca, Mg, Fe, or Al complex (chemical species) was predominant. Sewagemicroorganisms mineralized calcium citrate more rapidly than iron, aluminum, or hydrogen citrate, andmagnesium citrate was degraded slowest. Aluminum, hydrogen, and iron oxalates were mineralized morerapidly than calcium oxalate, and magnesium oxalate was decomposed slowest. Sewage microorganismsmineralized calcium NTA but not aluminum, magnesium, hydrogen, or iron NTA or any of the EDTAcomplexes. Pseudomonas sp. mineralized calcium and iron citrates but had no activity on hydrogen, aluminum,or magnesium citrate. Pseudomonas pseudoalcaligenes mineralized calcium, iron, hydrogen, and aluminumcitrates but had little activity on magnesium citrate. Pseudomonas alcaligenes used calcium, iron, hydrogen,and aluminum oxalates readily, but it used magnesium oxalate at a slower rate. Listeria sp. destroyed calciumNTA but had no effect on hydrogen, iron, or magnesium NTA. Increasing the Ca concentration in the mediumenhanced the breakdown ofNTA by Listeria sp. The different activities of the bacterial isolates were not a resultof the toxicity of the complexes or the lack of availability of a nutrient element. NTA mineralization was notenhanced by the addition of Ca to Beebe Lake water, but it was enhanced when Ca and an NTA-degradinginoculum were added to water from an oligotrophic lake. The data show that chemical speciation influences themineralization of organic compounds by naturally occurring microbial communities and by individual bacterialpopulations.

Many organic compounds that have amino or carboxylgroups spontaneously form coordinate bonds with metalcations (2, 16) and hence form metal-organic complexes.Aquatic chemists use the term "chemical speciation" inrecognition of the multiplicity of forms assumed by com-pounds that undergo complexation reactions (23). Althoughmany organic substrates may exist in a variety of chemicalspecies, little is known of the effects of chemical speciationon the mineralization of organic compounds.The complexation of humic substances by metal cations

has been proposed to explain the persistence of constituentsof humus (22). Insoluble metal-humate or metal-polysacch-aride complexes added to soil are mineralized at differentrates or to different extents or both, depending on the metalinvolved in complexation, but interpretation of these find-ings is sometimes difficult because of the possible toxicity ofthe metal (1, 13, 17). The chemistry and microbiology ofnitrilotriacetate (NTA) biodegradation have been reviewedby Warren (28). The mineralization of complexes of NTAand EDTA has been measured in samples from naturalenvironments, but interpretation of these data is difficultbecause of possible metal toxicity or because the speciationof the test complex might have changed during the experi-ments (3, 25, 26). A Pseudomonas sp. strain able to miner-alize sodium NTA could not mineralize nickel NTA (8).Moreover, different metal-citrate complexes are transportedat different rates into membrane vesicles of Bacillus subtilis(4, 30), and dissimilar transport rates could be reflected indifferent mineralization rates.The present study was designed to provide data that would

* Corresponding author.t Present address: Department of Agronomy, Pennsylvania State

University, University Park, PA 16802.

permit interpretation of the effects of chemical speciation onmicrobial mineralization. For this purpose, a computer pro-gram that calculates simultaneous chemical equilibria wasused as a guide in preparing defined media whose composi-tions favored particular chemical species of oxalate, citrate,NTA, and EDTA.

MATERIALS AND METHODS

Defined media. The compositions of the media weredetermined with information provided by MINEQL, acomputer program for the calculation of chemical equilibriaof aqueous systems (29). Computer programs for predictingcomplex chemical equilibria in aqueous solutions have beendeveloped (18), critically evaluated (19), and applied to theinterpretation of biological processes (9, 12, 21). Allglassware and polypropylene containers were cleaned inNoChromix solution (Godax Laboratories, Inc., New York,NY) and rinsed six times in distilled water and then six timesin reagent-grade water. Reagent-grade water was used in themedia. Inorganic nutrients were added to the media to obtainthe following elemental ratios (weight/weight): C/N, 4 orlower when the carbon source contained N; C/P, 8.5; C/S, 25;C/K, 5 or lower when KCI was added to adjust ionic strength;C/Ca, 50 or lower when Ca complexes were prepared; C/Mg,50 or lower when Mg complexes were prepared; C/Fe, 125 orlower when Fe complexes were prepared; C/Mn, 100; C/Cu,1,000; C/Zn, 1,000; C/Co, 1,000; and C/Mo, 1,000. Solutionscontaining K2HPO4, KH2PO4, (NH4)2S04, NH4Cl,FeSO4 * 7H20, MnSO4 * H20, CaCl2 2H20,MgCl2 * 2H20, MgSO4 * 7H20, NaMoO4 2H20,CaSO4 * 5H20, ZnSO4 * 7H20, CaCl2 * 6H20, KCl,FeCl3 - 6H20, AlCl3 * 6H20, disodium NTA, oxalic acid,citric acid, formic acid, EDTA, sodium acetate, and glucosewere added to 125-ml Erlenmeyer flasks to obtain the final

342

on February 14, 2020 by guest

http://aem.asm

.org/D

ownloaded from

EFFECTS OF CHEMICAL SPECIATION 343

ratio of these medium components. The ionic strength of allmedia was adjusted to 0.1 M. When complexes containing Caor Mg were used, the ionic strength was adjusted solely by theaddition of their chloride salts. When complexes containingFe or Al were used, sufficient amounts of Fe or Al were addedas determined by MINEQL, and the ionic strength was madeup to 0.1 M by the addition of KCI. When protonated oranionic forms of nonchelated organic compounds were used,only KCI was added.The media were then adjusted to pH 6.05 by the addition

of 0.01 M HCI or KOH, and 60-, 40-, or 30-ml portions wereplaced in the flasks. Radioactive compounds were added toyield between 300 and 1,200 dpm/ml, and the media weresterilized. Because the labeled compounds were added insmall volumes of dilute solutions, their effect on the pH wasconsidered negligible. The pH of the media was assumed tohave fallen to 6.00 as a result of autoclaving.The compositions of the media were designed to allow for

the dominance of individual chemical species. For example,the following were present in media to yield an organic anionconcentration of 10 ng of C per ml.

(i) Citrate (C6H507). The compositions were 44%H(C6H507)2-, 54% H2(C6H507) , 1% H3(C6H507), 1%FeC6H507; 95% CaH(C6H507), 5% CaH2(C6H507)+; 95%MgH(C6H507), 2% MgH2(C6H507)+, 1% H(C6H507)2-, 2%H2(C6H507)-; 91% FeC6H5O7, 4% H(C6H507)2-, 5%H2(C6H507) ; and 100% Al(OH)C6H507 when thecomplexing cations were designated as H; Ca; Mg; Fe; andAl, respectively.

(ii) Oxalate (C204). The compositions were 98% C2042,1% KC204-, 1% H(C204)-; 99% CaC204, 1% C2042-; and93% MgC204, 7% C2042- when the complexing cations weredesignated as H, Fe, or Al; Ca; and Mg, respectively. Thespeciation in the treatments containing H, Fe, and Al wasidentical because the added Fe and Al reacted with thehydroxyl, rather than the oxalate, anion.

(iii) NTA (CWH6N06). The compositions were 98%H(C6H6NO6)2-, 2% Fe(OH)C6H6NO6; 98% CaC6H6NO6,2% H(C6H6NO6)27; 85% MgC6H6NO6j, 15%H(C6H6NO6)2-; 85% Fe(OH)C6H6NO6-, 15%H(C6H6NO6)2-; and 8% AlC6H6NO6, 45%AI(OH)C6H6NO6-, 46% H(C6H6NO6)2-, 1%Fe(OH)C6H6NO67 when the complexing cations were des-ignated as H; Ca; Mg; Fe; and Al, respectively.

(iv) EDTA. The compositions were 32% hydrogenEDTA3, 65% dihydrogen EDTA2, 1% iron EDTA-, 2%manganese EDTA2-; 98% calcium EDTA2-, 2% ironEDTA-; 97% magnesium EDTA2-, 3% iron EDTA-; 79%iron EDTA-, 21% iron hydroxide EDTA2-; and 43% alumi-num EDTA-, 56% aluminum hydroxide EDTA2, 1% ironEDTA- when the complexing cations were designated as H,Ca, Mg, Fe, and Al, respectively.

In some cases, the preparation of mixtures of solublespecies with insoluble hydroxides was unavoidable. How-ever, although the calculations indicated that precipitatesshould have appeared in the media, they were never visibleto the eye. When the calculations showed that it wasimpossible to prepare the desired chemical species of ironand aluminum oxalates, the experiment was performed byadding iron and aluminum in equimolar ratios to the oxalate.Media identical to those described above but with the

organic compounds present at 1.0 to 100 ng of C per ml wereinoculated with sewage or pure cultures, and mineralizationwas determined. The inorganic components of the mediaused in measurements of the mineralization of nonspeciatingcompounds (i.e., organic compounds that do not react with

the test cations) were the same as those used for studying thedecomposition of citrate, oxalate, and NTA. The calculateddistributions of organic anions between chemical speciesprovided an estimate of the state of each medium immedi-ately before the addition of an inoculum. Changes in thedistribution of chemical species during carbon mineraliza-tion were considered to be unimportant because the initialstate of the medium governs microbial activity and becausethe equilibria were insensitive to a 10-fold reduction in theconcentration of organic anions (which corresponded to 90%mineralization of "'C).

Inocula. Raw sewage taken before secondary settling inthe Ithaca, N.Y., sewage treatment plant was passed se-quentially through Whatman no. 541 and no. 42 filter papersand a 5-,um membrane filter (Millipore Corp., Bedford,Mass.) The particulates that passed through the filter werecollected by centrifugation at 4°C, washed once in a cold0.74% KCI solution, and suspended in the KCI solution. Thesuspension was then passed through a 5-,um Millipore mem-brane filter to remove any clumps of material. Portions (0.4ml) of the resulting preparation, which contained ca. 106cells per ml, were added to each flask of medium.

Bacteria acting on the test compounds were isolated byinoculating portions of filtered sewage into an inorganic saltssolution (pH 6.1) supplemented with 0.02% Na2H(C6H6NO6), 0.08% oxalic acid plus 0.74% KCl, 0.08% oxalicacid plus 0.71% MgCl2 * 6H20, 0.02% citric acid plus 0.51%CaCl2 * 2H20, or 0.02% citric acid plus 0.71%MgCl2 6H20. The flasks containing the solutions wereincubated at 30°C on a shaker, the enrichments werestreaked on agar media of the same composition, and indi-vidual colonies were restreaked on the same media. Isolatesfrom these enrichments were identified by standard methods(6, 15) and designated Listeria sp. strain Nl, Pseudomonasalcaligenes 01, Alcaligenes sp. strain 02, Pseudomonas sp.strain Cl, and Pseudomonas pseudoalcaligenes C2, respec-tively.

Incubation. Tests of mineralization were conducted intriplicate flasks for inocula derived from environmentalsamples and in duplicate flasks for pure cultures. Oneanalysis per flask was performed at each sampling time. Theflasks containing defined media or sewage were incubated at25°C on a rotary shaker operating at 80 rpm. For incubationsof longer than 14 days, reagent-grade water was added toeach flask to compensate for evaporation. The flasks con-taining lake water were incubated at 18°C on a reciprocatingshaker (Eberbach Corp., Ann Arbor, Mich.) operating at 60cycles/min. No mineralization was detected in uninoculatedmedia.

Analytical methods. Radioactivity was measured with aliquid scintillation counter (Beckman Instruments, Inc.,Irvine, Calif.). The samples were processed as described bySubba-Rao et al. (24). The samples were placed in test tubes,and the liquid was acidified to pH 2 with H2SO4 and thenflushed with air for 10 min. Samples (1.0 ml) were introducedinto 20-ml scintillation vials containing 8.0 ml of Liquiscintsolution (National Diagnostics, Somerville, N.J.). The vialswere sealed, shaken vigorously, placed in a stationaryposition for 8 h, and then assayed for radioactivity.

Statistical analysis. The data were analyzed with a nestedtwo-way analysis of variance. Possible differences in miner-alization were evaluated with F tests at the 95% confidencelevel.

Special chemicals. All inorganic salts were analytical rea-

gent grade (Mallinckrodt, Inc., Paris, Ky.). TetrahydrogenEDTA, oxalic acid, glucose, citric acid, formic acid, and

VOL. 50, 1985

on February 14, 2020 by guest

http://aem.asm

.org/D

ownloaded from

344 MADSEN AND ALEXANDER

o ~~~~~~Fe-j

Ca40 - CaUI.0

20 A

1.0 ng/mi 10 ng/mi20 40 60 20

FIG. 1. Mineralization by sewage microorganisms

sodium acetate (Fisher Scientific Co., Fairlawn, N.J.) werecertified ACS grade. Disodium NTA (ca. 99.5% pure) waspurchased from Sigma Chemical Co., St. Louis, Mo. Rea-gent-grade water was obtained from a Milli-Q reagent-gradewater system (Millipore Corp.) supplied with singly distilledwater. Ethylenediaminetetra[2-14C]acetic acid (2.76 mCi/m-mol, >98% radiopurity) and [1,5-'4C]citric acid (8 mCi/m-mol, >99% radiopurity) were from ICN PharmaceuticalsInc., Irvine, Calif. Nitrilotri[1-_4C]acetic acid (53 mCi/mmol)and D-[U-14C]glucose (333 mCi/mmol) of 99% radiopuritywere purchased from Amersham Corp., Arlington Heights,Ill. [U-14C]oxalic acid (5 mCi/mmol, 98% radiopurity) and['4C]formic acid (52 mCi/mmol, 99% radiopurity) were ob-tained from New England Nuclear Corp., Boston, Mass.

RESULTSMineralization by sewage populations. Sewage microorga-

nisms rapidly mineralized some but not all forms of citrate(Fig. 1). Thus, magnesium citrate was not mineralized or wasslowly mineralized when provided at 1.0, 10, or 100 ng of Cper ml, and the rates were statistically significantly slowerthan for the other forms of citrate. The rates of mineraliza-tion of iron, aluminum, and hydrogen citrates were faster butwere not significantly different from one another. The ratesof disappearance of carbon from calcium citrate were alwaysgreatest, but these rates were only statistically different atthe two higher concentrations. It is noteworthy that morethan 90% of the carbon was often converted to CO2. Asexpected, the rates increased as the concentration of sub-strate increased.The sewage microflora also had different effects on the

various forms of NTA. Calcium NTA was mineralized atconcentrations of 1.0, 10, and 100 ng/ml (Fig. 2). As theconcentration of calcium NTA increased, the rate of itsbreakdown increased. No mineralization of aluminum, mag-nesium, hydrogen, or iron NTA was detected at concentra-tions of 1.0, 10, and 100 ng/ml in 585, 585, and 419 h,respectively.

DAYS,of citrate at levels of 1.0, 10, and 100 ng of C per ml.

Quite different results were noted in studies of the decom-position of oxalates. Each of the compounds was apprecia-bly mineralized in the test period (Fig. 3). Little magnesiumoxalate was mineralized in 130 h when added at 1.0 ng/ml,but the compound was readily mineralized at 10 and 100ng/ml. The rates of mineralization were greatest for alumi-num, hydrogen, and iron oxalates, but the rates of theirbreakdown were not significantly different from one another.Calcium oxalate was also mineralized readily but at rates ateach concentration that were significantly slower than those

100

80

z

w 60-

z4000 40 ng00U.~ ~ ~ OR

0

20

40 1.0 ~~~~~ng/mI

100 ng/m 0_0

200 400 600HOURS

FIG. 2. Mineralization by sewage microorganisms of calciumNTA at levels of 1.0, 10, and 100 ng of C per ml.

APPL. ENVIRON. MICROBIOL.

on February 14, 2020 by guest

http://aem.asm

.org/D

ownloaded from

EFFECTS OF CHEMICAL SPECIATION 345

0-w Fe~~~~~~~~~~(O)80 t 0 Fe(tJ))gx ~~~~~Cao AKA)

-0 140

0

AICA)20

H(O)Fe(o)

1.0 ng/ml 10 ng/mI30 60 90 120 30

FIG. 3. Mineralization by sewage microorganisms

for aluminum, hydrogen, and iron oxalates and faster thanthat for magnesium oxalate.Sewage microorganisms were incubated with the various

forms of EDTA supplied at C concentrations of 1.0, 10, and100 ng/ml. Half of the flasks also received 1.0 ,ug of sodiumacetate per ml each week to stimulate microbial activity.However, no EDTA mineralization was noted in 72 days.

Mineralization by pure cultures. Pseudomonas sp. and P.pseudoalcaligenes were grown in a glucose-inorganic saltsmedium, washed and suspended in 0.74% KCI, and added ata density of 105 cells per ml to the flasks. Pseudomonas sp.did not mineralize hydrogen, magnesium, or aluminum cit-rate in 90 h. In contrast, iron and calcium citrates were bothmineralized, the conversions being similar for 30 h, afterwhich time the breakdown of the former but not the latterstopped (Fig. 4). P. pseudoalcaligenes acted quite differ-ently, mineralizing hydrogen, calcium, iron, and aluminumcitrates at identical rates but converting only 15% of thecarbon of magnesium citrate to CO2 in 90 h. These data areanomalous because the first bacterium was isolated withcalcium citrate as the carbon source and the latter wasisolated with magnesium citrate as the carbon source.To test whether the abilities of these bacteria to mineralize

only some of the citrates resulted from toxicity of the cationsor the chelates or from a nutrient deficiency, we added thefive unlabeled citrates (100 ng of C per ml) to separate flaskscontaining a defined medium. A mixture of labeled (800dpm/ml) and unlabeled glucose was added to a final concen-tration of 100 ng of C per ml, and glucose mineralization bythe two bacteria was measured. Glucose was mineralized atidentical rates by Pseudomonas sp. in the presence of four ofthe citrates and was degraded at indistinguishably differentrates by P. pseudoalcaligenes in solutions containing all fivecitrates. Only in the case of calcium citrate was there aneffect, the mineralization by Pseudomonas sp. being delayedca. 20 h before the onset of a rapid breakdown. Such datasuggest that the patterns of citrate decomposition by the twobacteria are not the result of the toxicity of cations orchelates or of nutrient deficiencies that might occur becauseof the binding by the citrate of some essential cation. It is

HOURSof oxalate at levels of 1.0, 10, and 100 ng of C per ml.

possible that the delay in glucose mineralization in thepresence of calcium citrate resulted from the preferential useby Pseudomonas sp. of calcium citrate.

Listeria sp. was isolated in a medium containing a prepon-derance of hydrogen NTA as the carbon source. It wassuspended in 0.74% KCI and inoculated at a density of 103cells per ml into a defined medium with one of four forms ofNTA added at a C concentration of 10 ng/ml. The bacteriumhad no detectable activity on hydrogen, iron, or magnesiumNTA in 148 h. However, it did mineralize calcium NTA (Fig.5A). The lack of activity on hydrogen NTA is surprisingbecause this was the form of NTA used for the isolation ofthe bacterium. When the four forms ofNTA (unlabeled at 10ng of C per ml) were added to media containing 10 ng oflabeled glucose per ml and inoculated with Listeria sp., therates of glucose mineralization were found to be unaffectedby the presence of the various forms of NTA. This indicatesthat the use of only one of the four forms of NTA did notresult from toxicity of the other three or from a nutrientdeficiency in the media. Glucose metabolism was also notinfluenced if the media contained 0.035 M Ca.The previous data suggest that, of the chemical species of

NTA tested, only calciutn is susceptible to mineralization.To test whether the proportion of NTA as calcium NTAgoverns decomposition, mineralization by Listeria sp. wasmeasured in media with various proportions of NTA ascalcium NTA. The concentration ofNTA as C was 10 ng/ml,and the ionic strength was held constant with appropriateadditions of KCl. No mineralization occurred in 237 h inmedia with no calcium NTA but with a Ca concentration of5 nM or 0.5 ,um or in media with 6% calcium NTA.Mineralization was evident in solutions with 40, 87, and 97%calcium NTA (Fig. SB), and the differences in mineralizationwere statistically significant. As the calcium NTA concen-tration increased, the apparent lag period diminished and theextent of mineralization increased. The effect is not a resultof Ca limitation in the media, as shown by the lack of effectof Ca addition and of added Ca on glucose breakdown.

P. alcaligenes and Alcaligenes sp. were grown in nutrientbroth, washed and suspended in 0.74% KCl, and added to

VOL . 50j 1985

on February 14, 2020 by guest

http://aem.asm

.org/D

ownloaded from

346 MADSEN AND ALEXANDER

oxalate-containing media at an initial density of 5 x 105 cells Aper ml. P. alcaligenes used all five oxalates readily, and therates of mineralization of hydrogen, calcium, aluminum, and 75iron oxalates were statistically indistinguishable (Fig. 6A).However, mineralization of magnesium oxalate was slowerthan that of the other compounds, the difference being 50statistically significant. Alcaligenes sp. mineralized calcium 3oxalate most rapidly, iron oxalate more slowly, and magne- \sium and aluminum oxalates still less rapidly, and hydrogen s 25woxalate was the last to be mineralized (Fig. 6B). Except for crmagnesium and aluminum oxalates, which were destroyed at ° U 1 1

similar rates, the differences in the patterns of metabolism by BAlcaligenes sp. were statistically significant. The patterns 40% CaNTAwith the two bacteria do not reflect the carbon source used z 75for their isolation, hydrogen oxalate being used for P. 0alcaligenes and magnesium oxalate being used forAlcaligenes sp. 50

Radioactive formate (700 dpm/ml) and unlabeled formate 87% CaNTA

25- 97CaNTA100 Pseudomonas ap. Ca.

80 160 240HOURS

80 \ FIG. 5. Mineralization by Listeria sp. of (A) calcium NTA and(B) NTA in media containing various proportions of NTA as calcium

cD NTA, the remainder of the NTA in the latter instances beingz Ct \ hydrogen NTA.

wz 60 Fe at a C concentration of 100 ng/ml were used to test whethertoxicity of the cations or the oxalates accounted for thedifference in utilization of the several oxalates. The oxalates

Z 40 were added to a concentration of 100 ng of C per ml.U. \Calculations based on the stability constants for complexa-o tion between formic acid and inorganic components of the

media indicated that at most 8% of the formic acid was as

20 \ magnesium formate and calcium formate; thus, 92 to 100% ofthe formate in all the treatments was HCOO-. Formate waschosen because these bacteria did not mineralize glucose. P.alcaligenes mineralized formate in the presence of the fiveunlabeled oxalates at rates that could not be distinguished

100 . statistically. Alcaligenes sp. mineralized formate fastest inthe presence of calcium oxalate, somewhat later in mediacontaining iron oxalate, still later in solutions with aluminum

Mg or magnesium oxalate, and slowest in the presence of80 hydrogen oxalate. These differences were statistically signif-

icant. These data suggest that the pattern of mineralizationCD lllby P. alcaligenes is not the result of toxicity or nutrienta3 lll deficiency in the defined media but that that of Alcaligenes2 60 - FeC) sp. is.a: lll P. alcaligenes was grown in nutrient broth, and the cellsO lll were washed and suspended in a solution containing 0.035 M

MgCI2 and 100 ng of C as oxalic acid per ml, a solution inz 40 which 93% of the oxalic acid was as magnesium oxalate.U. After 4 h, the cells were collected by centrifugation and

o Ca(A) added to defined media containing the five oxalates at aconcentration of 100 ng of C per ml. Under these conditions,

20 magnesium oxalate was mineralized slowly at a rate that wasik\Al(-) statistically distinct from those of calcium, hydrogen, alumi-

H(o) num, and iron oxalates (Fig. 6C). The rates of degradation ofthe latter four complexes were not significantly different

P. pseudoalpaligenes from one another. Hence, exposing the cells to magnesium30 60 90 oxalate did not result in its being mineralized more quickly

HOURS than the other oxalates.FIG. 4. Mineralization of various citrate complexes (100 ng of C Mineralization in lake water. Because the data presented

per ml) by Pseudomonas sp. and P. pseudoalcaligenes. above suggest that mineralization of NTA by Listeria sp. is

APPL. ENVIRON. MICROBIOL.

on February 14, 2020 by guest

http://aem.asm

.org/D

ownloaded from

EFFECTS OF CHEMICAL SPECIATION 347

0DzzR Mg2 60

o ~~~~~~~~~~~~FeH(O)

!40-MgO Mg ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~FeC

Fe(O) C

20-CaCA Ca

I ~~AI(A)

6 12 18 12 24 36 6 12 1HOURS

FIG. 6. Mineralization of five oxalates by (A) P. alcaligenes, (B) Alcaligenes sp., and (C) cells of P.alcaligenes that had been preincubatedfor 4 h with magnesium oxalate.

governed by the proportion ofNTA as calcium NTA, a studywas conducted to determine mineralization of NTA in waterfrom Beebe Lake, Ithaca, N.Y., that received no Ca or Ca atthree concentrations. The ionic strength was not held con-stant, but the pH (7.9) of all the treatments was the same.The data indicate that the addition of Ca to Beebe Lakewater had no effect on NTA mineralization, the patterns ofmineralization being statistically indistinguishable (Fig. 7A).The absence of an effect suggests that Ca did not governNTA mineralization or that Beebe Lake had sufficient Ca to

10040 mjg/ml

0

80 0 jjg/nz 120 jig/mIII

o 360 jig/mI-J

40-U.0

prevent additions of the cation from causing a change inNTA speciation. The concentration of Ca in Beebe Lake,which is often ca. 40 jig/ml (R. T. Oglesby, personal com-munication) was sufficient to convert ca. 60% of the addedNTA to calcium NTA.White Lake, Old Forge, N.Y., is oligotrophic and has a Ca

concentration of ca. 5 ,ug/ml (20), a concentration sufficientto convert ca. 10% of the NTA at 10 ng of C per ml tocalcium NTA. No mineralization of NTA was detected in 22days in water from White Lake that was unamended or

HOURS HOURS

FIG. 7. Mineralization of NTA in the presence of various Ca concentrations in water from (A) Beebe Lake and (B) White Lake inoculatedwith a water sample from Beebe Lake.

VOL. 50, 1985

on February 14, 2020 by guest

http://aem.asm

.org/D

ownloaded from

348 MADSEN AND ALEXANDER

amended with 5, 20, or 100 ,ug of Ca per ml. BecauseNTA-mineralizing organisms were apparently absent, 40 mlof water from White Lake amended with 10 ng of C as NTAper ml was inoculated with 1.0 ml of water from a BeebeLake sample that contained microorganisms able to metab-olize NTA. NTA was mineralized in the White Lake waterunder these conditions (Fig. 7B). The apparent stimulationby 5 jig of Ca per ml was not statistically significant.However, the stimulation by 20 and 100 ,ug of Ca per ml wasstatistically significant, although the differences between 20and 100 ,ug/ml were not significant. Ca enhanced the rate ofmineralization but also decreased the apparent lag periodbefore the onset of mineralization. These data support theview that NTA mineralization is governed by the proportionof NTA as calcium NTA.

DISCUSSIONMicroorganisms mineralized different chemical species of

the same organic anion at different rates. Although thedefined media were at a constant ionic strength and pH,chemical speciation is not the only possible explanation forthe differences. The added cations may have influenced themineralization process independent of their speciation ef-fects or the chelating agents may have rendered an essentialcation unavailable. However, the data presented indicatethat the differences in mineralization among four of the fivebacteria tested were not related to toxicity or nutrientlimitation.Konings et al. (14) stressed the importance of the cyto-

plasmic membrane as a diffusion barrier between the bac-terial cytoplasm and the environment of the bacterium.Assuming that cell permeability controls the rate of miner-alization of the test compounds, two explanations for theinfluence of chemical speciation on microbial mineralizationmay be proposed. First, different chemical species of onecompound may be transported at dissimilar rates by a singlemembrane-transport system. This explanation is supportedby studies of membrane transport in B. subtilis (4). Second,several transport systems may be present, each acting onone chemical species. Because the isolated bacteria did notact on just one of a variety of chemical species, the validityof this hypothesis requires that more than one species-specific transport system be present in the isolates.The finding that calcium NTA but no other form of NTA

was mineralized is noteworthy. Vashon et al. (27) reportedthat NTA biodegradation was more rapid in hard water thanin soft water, and mineralization of NTA may be enhancedafter the addition of either Ca and Mg or only Ca to sewage(5, 11). Firestone and Tiedje (8) suggested that one or a fewchemical species may be acted on by the membrane-transport systems of bacteria that mineralize NTA. Thedependence of NTA mineralization on Ca may have beenoverlooked by earlier investigators because equilibrium cal-culations were not used in defining the compositions ofmedia, and the solutions that were used therefore may havecontained enough Ca to make a pool of calcium NTA inequilibrium with other chemical species. On the other hand,bacteria may have one substrate-uptake system that acts atlow concentrations and one that acts at high concentrations(10), and only the uptake system acting at low concentra-tions, such as those used in this study, may specificallyrecognize calcium NTA. It is also possible that microorga-nisms used in this and other studies differed in their ability touse various forms of NTA.Bergsma and Konings (4) demonstrated that the rate of

uptake of citric acid in vesicles from B. subtilis varies with

the particular metal complexed by citrate. The findings in thepresent study that different citrate complexes are mineral-ized at dissimilar rates are consistent with the observationsof Bergsma and Konings (4). Dijkhuizen et al. (7) suggestedthat the oxalate anion is accompanied across the cell mem-brane by at least two protons during the active transport ofoxalate by Pseudomnonas oxalaticus. The present data alsosuggest that discrimination among species of citric acid andoxalic acid occurs in nature and that the magnesium complexof both chemicals may be acted on more slowly than othercomplexes.

Cations and compatible organic compounds in aqueousenvironments spontaneously form coordination complexes.The particular complex formed at a given time and place willvary with the chemical composition of the region immedi-ately surrounding the organic molecule. This study showsthat different metal-organic complexes are mineralized atdifferent rates or to different extents in samples from naturalenvironments. Evaluation of the impact of chemicalspeciation on biodegradation in natural ecosystems requiresfurther investigation.

ACKNOWLEDGMENTThis research was supported by funds provided by the U.S.

Environmental Protection Agency under assistance agreementCR809735-02-0, by the U.S. Department of the Interior, and by agrant in aid of research from Sigma Xi.We thank C. T. Driscoll of Syracuse University for providing the

MINEQL computer program.

LITERATURE CITED

1. Andreaux, F. 1981. Utilisation de molecules modeles desynthese dans l'etude des processus d'insolubilisation et debiodegradation des polycondensats humiques. Bull. Assoc. Fr.Etude Sol, p. 271-291.

2. Basolo, F., and R. G. Pearson. 1967. Mechanisms of inorganicreactions: a study of metal complexes in solution. John Wiley &Sons, Inc., New York.

3. Belly, R. T., J. J. Lauff, and C. T. Goodhue. 1975. Degradationof ethylenediaminetetraacetic acid by microbial populationsfrom an aerated lagoon. Appl. Microbiol. 29:787-794.

4. Bergsma, J., and W. N. Konings. 1983. The properties of citratetransport in membrane vesicles from Bacillus subtilis. Eur. J.Biochem. 134:151-156.

5. Bjorndal, H., H. 0. Bouveng, P. Solyom, and J. Werner. 1972.NTA in sewage treatment. 3. Biochemical stability of somemetal chelates. Vatten 28:5-16.

6. Buchanan, R. E., and N. E. Gibbons (ed.). 1974. Bergey'smanual of determinative bacteriology, 8th ed. The Williams &Wilkins Co., Baltimore.

7. Dijkhuizen, L., L. Groen, W. Harder, and W. N. Konings. 1977.Active transport of oxalate by Pseudomonas oxalaticus OX1.Arch. Microbiol. 115:223-227.

8. Firestone, M. K., and J. M. Tiedje. 1975. Biodegradation ofmetal-nitrilotriacetate complexes by a Pseudomonas species:mechanism of reaction. Appl. Microbiol. 29:758-764.

9. Giblin, A. E., and R. W. Howarth. 1984. Porewater evidence fora dynamic sedimentary iron cycle in salt marshes. Limnol.Oceanogr. 29:47-63.

10. Harder, W., and L. Dijkhuizen. 1983. Physiological responses tonutrient limitation. Annu. Rev. Microbiol. 37:1-23.

11. Huber, W., and K. H. Popp. 1972. Der biologische Abbau derNitrilotriessigsaure in Gegenwart von Cadmium-lonen. FetteSeifen Anstrichm. 74:166-168.

12. Jackson, G. A., and J. J. Morgan. 1978. Trace metal-chelatorinteractions and phytoplankton growth in seawater media: the-oretical analysis and comparison with reported observations.Limnol. Oceanogr. 29:47-53.

APPL. ENVIRON. MICROBIOL.

on February 14, 2020 by guest

http://aem.asm

.org/D

ownloaded from

EFFECTS OF CHEMICAL SPECIATION 349

13. Juste, C., J. Delas, and M. Langon. 1975. Comparaison de lastabilit6 biologique de differents humates metalliques. C. R.Hebd, Seances Acad. Sci. Ser. D Sci. Nat. 281:1685-1688.

14. Konings, W. N., K. J. Hellingwerf, and G. T. Robillard. 1981.Transport across bacterial membranes, p. 257-283. In S. L.Bonting and J. J. H. H. M. de Pont (ed.), Membrane transport.Elsevier/North-Holland Publishing Co., Amsterdam.

15. Krieg, N. R., and J. G. Holt (ed.). 1984. Bergey's manual ofsystematic bacteriology, vol. 1. Williams & Wilkins, Baltimore.

16. Martell, A. E. 1971. Principles of complex formation, p.239-263. In S. J. Faust and J. V. Hunter (ed.), Organic com-

pounds in aquatic environments. Marcel Dekker, Inc., NewYork.

17. Martin, J. P., J. 0. Ervin, and S. J. Richards. 1972. Decompo-sition and binding action in soil of some mannose-containingmicrobial polysaccharides and their Fe, Al, Zn, and Cu com-plexes. Soil Sci. 113:322-327.

18. Morel, F., and J. Morgan. 1972. A numerical method forcomputing equilibria in aqueous chemical systems. Environ.Sci. Technol. 6:58-67.

19. Nordstrom, D. K., L. N. Plummer, T. M. L. Wigley, T. J.Wolery, J. W. Ball, E. A. Jenne, R. L. Bassett, D. A. Crerar,T. M. Florence, B. Fritz, M. Hoffman, G. R. Holdren, Jr., G. M.Lafon, S. V. Mattigod, R. E. McDuff, F. Morel, M. M. Reddy, G.Sposito, and J. Thrailkill. 1979. Comparison of computerizedchemical models for equilibrium calculations in aqueous sys-tems, p. 857-892. In E. A. Jenne (ed.), Chemical modeling inaqueous systems. American Chemical Society, Washington,D.C.

20. Pfeiffer, M. H., and P. J. Festa. 1980. Acidity status of lakes inthe Adirondack region ofNew York in relation to fish resources.Bureau of Fisheries, Division of Wildlife, New York StateDepartment of Environmental Conservation, Albany, N.Y.

21. Rubin, M., and A. E. Martell. 1980. The implications of trace

metal-nitrilotriacetic acid speciation on its environmental im-pact and toxicology. Biol. Trace Element Res. 2:1-19.

22. Stout, J. D., K. M. Goh, and T. A. Rafter. 1981. Chemistry andturnover of naturally occurring resistant organic compounds insoil, p. 1-73. In E. A. Paul and J. N. Ladd (ed.), Soil biochem-istry, vol. 5. Marcel Dekker, Inc., New York.

23. Stumm, W., and J. J. Morgan. 1981. Aquatic chemistry. JohnWiley & Sons, Inc., New York.

24. Subba-Rao, R. V., H. E. Rubin, and M. Alexander. 1982.Kinetics and extent of mineralization of organic chemicals attrace levels in freshwater and sewage. Appl. Environ. Micro-biol. 43:1139-1150.

25. Swisher, R. D., T. A. Taulli, and E. J. Malec. 1973. Biodegra-dation ofNTA metal chelates in river water, p. 237-263. In P. C.Singer (ed.), Trace metals and metal-organic interactions innatural waters. Ann Arbor Book Co., Ann Arbor, Mich.

26. Tiedje, J. M. 1975. Microbial degradation of ethylenediamine-tetraacetate in soils and sediments. Appl. Microbiol. 30:327-329.

27. Vashon, R. D., W. J. Jones, and A. G. Payne. 1982. The effect ofwater hardness on nitrilotriacetate removal and microbial accli-mation in activated sludge. Water Res. 15:1429-1432.

28. Warren, C. B. 1974. Biodegradation of nitrilotriacetic acid andNTA-metal ion complexes. p. 473-496. In M. A. Q. Khan andJ. P. Bederka, Jr. (ed.), Survival in toxic environments. Aca-demic Press, Inc., New York.

29. Westall, J. C., J. L. Zachary, and F. M. M. Morel. 1976.MINEQL: a computer program for the calculation of chemicalequilibrium composition of aqueous systems. Department ofCivil Engineering, Massachusetts Institute of Technology,Cambridge.

30. Willecke, K., E.-M. Gries, and P. Oehr. 1973. Coupled transportof citrate and magnesium in Bacillus subtilis. J. Biol. Chem.248:807-814.

VOL. 50, 1985

on February 14, 2020 by guest

http://aem.asm

.org/D

ownloaded from