May 19, 1983 - Hexavalent chromium [Cr(VI)] is a known carcinogen and mutagen; however, ... River sediment samples was predominantly hexavalent.
APPLIED
AND
ENVIRONMENTAL MICROBIOLOGY, OCt. 1983, p. 846-854
Vol. 46, No. 4
0099-2240/83/100846-09$02.00/0
Hexavalent Chromium-Resistant Bacteria Isolated from River Sediments GREGORY W. LULW,' JOSEPH W. TALNAGI,2 WILLIAM R. STROHL,' AND ROBERT M. PFISTER'* Department of Microbiology, The Ohio State University, Columbus, Ohio 43210,1 and Nuclear Reactor Laboratory, The Ohio State University, Columbus, Ohio 432122 Received 1 April 1983/Accepted 19 May 1983
Hexavalent chromium [Cr(VI)] is a known carcinogen and mutagen; however, the actual mechanisms of Cr toxicity are unknown. Two approaches were used to isolate Cr(VI)-resistant bacteria from metal-contaminated river sediments. Diluted sediments were plated directly onto a peptone-yeast extract (PYE) medium containing 0 to 100 ,ug of Cr(VI) ml-'. Approximately 8.4 x 105 CFU g-1 were recovered on 0 ,ug of Cr(VI) ml-', whereas 4.0 x 102 CFU g-1 were recovered on PYE plus 100 ,ug of Cr(VI) ml-1. Alternatively, continuous culture enrichment techniques were employed using PYE and 100 ,ug Cr(VI) ml-' input at dilution rates of 0.02 and 0.10 h-1. After six residence periods, 109 CFU were recovered on PYE agar containing 0 p,g of Cr(VI) ml- 1 and 107 CFU on PYE agar plus 100 jig of Cr(VI) ml-1. Of 89 isolates obtained by direct plating onto PYE, 47% were resistant to 100 ,ug of Cr(VI) ml-', and 29% were resistant to 250 ,ug of Cr(VI) ml-'. When the same isolates were plated onto PYE containing Cr(III), 88% were resistant to 100 ,ug ml-' but only 2% were resistant to 250 ,ug ml-'. Cr, Co, Sb, and Zn were found in significantly higher concentrations at an industry-related contaminated site than at a site 11 km downstream. Total Cr in the sediments at the contaminated site averaged 586 ,ug (dry weight) g-1, and the downstream site averaged 71 p.g (dry weight) g-1. The Cr recovered from acid-digested Ottawa River sediment samples was predominantly hexavalent. Five acid digestion procedures followed by atomic absorption spectroscopy were compared and found to be 30 to 70% efficient for recovery of Cr relative to neutron activation analysis. A population of aerobic, heterotrophic bacteria was recovered from sediments containing elevated levels of Cr. However, there was no apparent difference in the number of bacteria resistant to Cr(VI) at the contaminated site as compared to the downstream site. Recovery of Cr-resistant bacteria from river sediments was improved by using continuous culture enrichment as compared to direct plating.
Chromium (Cr) is an essential micronutrient involved in the peripheral action of insulin (18), normal glucose utilization (18), stimulation of enzyme systems (18), and possibly in the stabilization of nucleic acids (16). Despite such biological requirements, high concentrations of Cr are toxic (6, 24), mutagenic (17, 22, 24, 33), carcinogenic (6, 28, 33), and teratogenic (15). The cytotoxic effects of Cr are known to be valence dependent (24-26). Of the two stable Cr valences, hexavalent (VI) Cr is about 100-fold more toxic than the trivalent (III) form (24). Moreover, Cr(VI) is the predominant species involved in mutagenicity, carcinogenicity, and teratogenicity (24). Cr is widespread in the aquatic environment, primarily as the result of mining and industrial waste (9, 14, 34). Cr is used in the manufacture of alloys, corrosion-inhibitory paints, wood pre846
servatives, photographic sensitizers, mordants and fixatives for dyes and tanning, and pigments for rubber and ceramics (6). The chemistry of Cr has been studied in marine environments, fresh water, and soil (3, 4, 13, 19-21, 29). Nakayama et al. (20) have proposed a model for the reactions and interactions of Cr in marine water and sediments. The calculated Keq for oxidation of Cr(III) to Cr(VI) predicts that Cr(VI) is the predominant species under normal marine conditions. However, large amounts of Cr(III) are found associated with organic material and are not available for redox equilibria. Cr(VI) can be reduced by chemical (20, 25, 26) or by indirect biological (30) processes. In addition, one report has indicated direct enzymatic reduction (L. H. Bopp, and H. L. Ehrlich, Abstr. Annu. Meet. Am. Soc. Microbiol. 1980, Qlll, p. 212). To our
VOL. 46, 1983
knowledge, no direct microbiological oxidation of Cr(III) has been reported. In addition to the bacteria involved in Cr(VI) reduction, many microorganisms have been isolated which are resistant to high concentrations of Cr(VI) but which have not been shown to alter the Cr redox state (12, 23, 31). Cr resistance has been correlated with the presence of plasmid DNA in several isolates (12, 23, 31). However, the actual mechanisms of Cr toxicity and resistance in bacteria have not been identified. Because of the lack of knowledge about the mechanisms of Cr toxicity and resistance, we chose Cr-contaminated sediments in the Ottawa River in northeastern Ohio as a source for isolating Cr-resistant bacteria for future investigations of Cr toxicity. Neutron activation analysis (27, 35) was used to determine the concentrations of a variety of metals, with the primary interest on Cr. Because continuous culture enrichment has been shown to be applicable for the isolation of various groups of microorganisms, the effectiveness of direct plating of Ottawa River sediments onto various concentrations of Cr(VI) was compared to continuous culture enrichment for Cr-resistant bacteria. MATERIALS AND METHODS Sampling sites. Two sites were chosen in the Ottawa River in northeastern Ohio near the city of Lima. Site 1 (contaminated site) was located approximately 12 m downstream from an industrial point source. Site 2 was chosen approximately 11 km downstream from site 1 to represent reduced (but not necessarily background) levels of heavy metal contamination. Sample collection. Sediment samples were collected by using a piston-type corer made from a 50-ml plastic syringe barrel from which the closed end was removed. The barrel was fitted to an aluminum tube of the same outside diameter as the inside diameter of the barrel. The syringe shaft was connected to a rod which fit inside the aluminum tube. The syringe barrel and rubber plunger were autoclaved separately and fitted at the sampling site. Approximately 3.0 cm3 of sediment was collected and transferred directly into 50 ml of magnesiumphosphate buffer (0.9 mM MgCl2, 0.6 mM KH2PO4 [pH 7.2]) (5) in 100-ml wide-mouth bottles. Core samples were also collected in sterile Whirl-Pak (Fisher Scientific, Cincinnati, Ohio) plastic bags for neutron activation analysis. The preweighed sample bottles were then weighed to obtain the wet weight of the added sediments. Dry weight determinations were done after drying sediments at 85 + 2°C for 24 h. The sediment-buffer solution was further diluted in buffer for viable plate counts. In addition, 10 ml of the sediment-buffer solution was added to 90 ml of peptone yeast extract (PYE) broth (see below) and transported to the laboratory for use as inocula for continuous flow enrichments. Media. The medium used throughout this study was PYE broth and agar (15 g liter-1), which was modified
CHROMIUM-RESISTANT BACTERIA
847
from the yeast extract salts medium described by Dye (11). PYE contained per liter: 1.0 g of Bacto-Peptone (Difco Laboratories, Detroit, Mich.), 1.0 g of yeast extract (Difco), 0.5 g of NaCl, 0.5 g of NH3H2PO4, 0.5 g of KH2PO4, and 0.2 g of MgSO4 * 7H20. The final pH was adjusted to 7.2 with 1 N NaOH. Antifoam C (Sigma Chemical Co., St. Louis, Mo.) was autoclaved separately and added to PYE at a final concentration of 0.002% for use in the continuous flow systems. Chromium was added as K2Cr2O7 (VI) or as CrCl3 (III). The chromium salts were sterilized separately by either autoclaving Cr(VI) in distilled, demineralized water (ddH2O) or by filtering Cr(III) or Cr(VI) in PYE through membrane filters (pore size, 0.22 ,um; Millipore Corp., Bedford, Mass.). For heavy metal toxicity tests, the following metals were dissolved in ddH2O (final metal concentration given): CdCl2 (50 p.g ml-'), Co(NO3)2 * 6H20 (50 p.g ml-'), Cu(NO3)2 * 3H20 (50 ,ug ml-'), Pb(NO3)2 6H20 (100 ,ug ml 1), NiC12 * 6H20 (50 pg ml-'), and ZnSO4 (400 ,ug ml-'). Absorbant discs (12.7 mm in diameter) were soaked in each metal solution overnight, dried at 37°C for 1 h, and placed onto lawns of each isolate. Resistance to the metal was positive if growth occurred up to the edge of the disk after 48 h (2). To determine the antibiotic sensitivities of the bacterial isolates, antibiotics were added to freshly plated lawns of each isolate on PYE plates as Sensi-discs (BBL Microbiology Systems, Cockeysville, Md.) containing the following: ampicillin (10 ,ug), bacitracin (10 U), erythromycin (15 ,ug), gentamicin (10 pig), neomycin (30 pig), novobiocin (30 ,ug), penicillin (10 U), polymyxin B (50 U), rifampin (5 pig), streptomycin (10 ,ug), sulfathiazole (1 mg), and tetracycline (30 ,ug). Resistance to the antibiotics was positive if growth occurred up to the edge of the disk after 48 h (2). Neutron activation analysis. The sediment samples were dried at 100°C for 48 h, crushed with a motor and pestle, and passed through a nylon sieve having a 1mm2 exclusion value. The sediments were redried overnight in nitric acid-cleaned scintillation vials and taken to the Nuclear Reactor Laboratory, The Ohio State University, Columbus, Ohio. At the reactor laboratory, the moisture content was determined to be less than 0.55% by drying at 100°C to a constant weight (10). The moisture content was used to correct sample masses to a dry weight basis and was included in the computer analysis. Each sample was then subdivided into five different 0.1-ml polyethylene snap-top vials. River sediments obtained from the National Bureau of Standards (Standard Reference Material no. 1645) were used as standards for quantitative analysis of the metals. All samples were irradiated for 4 h at an average flux of 3.97 x 101" neutrons cm-' s-1. Gamma ray spectra were collected by using a Princeton Gamma-Tech Ge(Li) detector and a Canberra Model 8180 multichannel analyzer (4,096 channels). The spectra were analyzed with the aid of a Digital Equipment Corp. model PDP 11/05 minicomputer with IPSNAP programs (Nuclear Reactor Laboratory, The Ohio State University). Chromium speciation. A colorimetric assay (1) for hexavalent chromium was used. Ten milliliters of ddH2O added to 4 g of wet sediments and incubated at ambient temperature on a rotary shaker (160 rpm) for 24 h. The sediments were filtered through dry, -
848
LULI ET AL.
preweighed Whatman no. 44 filters (American Scientific Products, Obetz, Ohio). The filters were dried for 24 h at 85 + 2°C and weighed to obtain the dry weights of the sediments. The filtrate was assayed for total (atomic absorption spectroscopy [AAS]) and hexavalent (colorimetric assay) Cr. Because variations of recovery of metals from sediments is known, five different methods were used, and the efficiencies of Cr recovery were compared. Three standard digestion procedures (nitric acid, nitric-sulfuric acid, and nitricperchloric acid [1]) were compared to a modified nitric acid digestion and a nitric-hydrochloric acid (1:3) digestion. In all procedures, 0.1 to 1 g of dried sediments (subsamples of which had previously been subjected to neutron activation analysis) was used. For the modified nitric acid and nitric-hydrochloric acid procedures, 10 ml of each was added to sediments, boiled under gentle refluxing for 3 h, and diluted to 50 ml. All acid digests were filtered through Whatman no. 44 filters and assayed for total and hexavalent Cr. Comparison of the binding of Cr(VI) and Cr(III) to sediments was carried out by resuspending 0.5 g of dry sediments in 20 ml of ddH2O containing 0, 100, 200, 400, or 800 ,ug of Cr. After 24 h of incubation on a rotary shaker (160 rpm), the mixtures were filtered and assayed for total Cr by using AAS. Controls containing 20 ml of ddH2O plus 100,lg of Cr(VI) or Cr(III) were used to determine losses of Cr due to binding by glass and (or) the filters. Solutions containing 50 ,ug ml-' each of Cr(VI) and Cr(III) in ddH2O were analyzed for the amount of interference for the detection of Cr(VI) by Cr(III). To determine the efficiency of oxidation of Cr(III) to Cr(VI), the solutions were oxidized by the [-_rmanganate azide method (1). To determine potential volatility and (or) reduction of Cr(VI) by medium components, Cr(VI) (50 ,ug ml-' final concentration) was added to sterile PYE broth and incubated at 160 rpm for 48 h. After incubation, the medium was assayed for total and hexavalent Cr. Viable plate counts. All plate counts were made on PYE agar containing 0, 10, 25, 50, 75, or 100 ,ug of Cr(VI) ml-'. Samples of the appropriate dilution of sediments were spread onto predried plates. The total number of colonies and the number of different colony morphologies were recorded after incubation of the plates for 1 week at ambient temperature (24 ± 20C). Chromium resistance potential. Isolates from control plates [0 ,ug of Cr(VI) ml-'] were tested for resistance to Cr(III) and Cr(VI). The isolates were replica plated onto plates containing various amounts of added Cr(III) and Cr(VI) from 10 to 250 jig of Cr ml-1. Plates were incubated for 4 days, and growth was visually determined as positive or negative. Continuous flow enrichment. Continuous culture experiments were done in 500-mI systems similar to those described by Summers et al. (32). The vessels were prepared by the addition of 450 ml of PYE broth to 1-liter Kimax tall form beakers (American Scientific Products) fitted with a size 15 rubber stopper. Nitric acid-cleaned glass tubing was used for aeration via fritted gas dispersion tube, air out, medium addition, medium withdrawal by aspiration, and a sample port. A 47-mm magnetic stir bar was used for constant mixing. The vessels were inoculated by the aseptic addition of 50 ml of the sediment-PYE mixture pre-
APPL. ENVIRON. MICROBIOL.
pared at the site (see above). All experiments were done at ambient temperature (24 ± 2°C). After 24 h of incubation as batch cultures, fresh medium containing 100 ,ug of Cr(VI) ml-' was added continuously at dilution rates (D) of 0.1 or 0.02 h-1. The feed lines were passed through a heat trap (85°C) to prevent back contamination of the reservoir. Tenmilliliter samples were aseptically withdrawn at each residence time (10 and 50 h for D = 0.1 and 0.02 h-1, respectively). One milliliter was used for dilution for plate counts, 5.0 ml was used for the quantitation of Cr, and 4.0 ml was used to determine the turbidity of the mixed cultures by using a Klett-Summerson colorimeter (no. 66 red filter). The total Cr concentration in the continuous culture vessels was determined by using AAS. The 5.0-ml sample was centrifuged at 11,000 x g for 10 min, and the cell-free supernatant was either diluted or assayed directly. The wet cell pellet was digested with 1.0 ml of concentrated HNO3 at 60°C for 30 min and then diluted to 5.0 ml with distilled water. All determinations were performed on a Perkin-Elmer Model 403 Atomic Absorption spectrophotometer (Perkin-Elmer, Norwalk, Conn.), using standard operating conditions for chromium. Growth curves. Growth of several Cr-resistant and Cr-sensitive isolates at 25°C in PYE medium was monitored by using a Klett-Summerson colorimeter. Chromium(VI) was added at mid-log phase to give final Cr(VI) concentrations of 50 and 100 p.g ml-'. After 5 h of exposure to Cr(VI), Cr-sensitive (as determined by lack of growth) isolates were harvested by centrifugation (11,000 x g for 10 min), washed, and resuspended in PYE without Cr. Blanks containing 50 and 100 ,ug of Cr ml- 1 were used to compensate for the coloring of the medium by chromium. Biochemical testing. Isolates were identified to their generic level based on Gram reaction, colony morphology, cell morphology, and standard biochemical analyses. The analysis scheme followed was that of Colwell and Wiebe (8).
RESULTS Neutron activation analysis. Analysis of dried river sediments showed that there were elevated levels of Cr, Co, Sb, and Zn at the industrycontaminated site (site 1) as compared to the downstream site (site 2). Table 1 lists the concentrations of several elements readily detectable from the gamma ray spectra of neutronactivated sediments. Only Na was found in a greater concentration (twofold) at the downstream site than at the contaminated site. The concentration of total Cr at the contaminated site was found to average ninefold higher than at the downstream site. Chromium recovery and speciation. After incubation of dried sediments containing solutions of Cr(III) and Cr(VI), essentially all of the added Cr(III) was (ad)sorbed, whereas only 50% of the added Cr(VI) was (ad)sorbed over the entire range of Cr concentrations used. No Cr was detected in the filtrate from untreated sediments. In control experiments, mixtures of
VOL. 46, 1983
Cr(III) and Cr(VI) in ddH2O were assayed for Cr(VI), oxidized, and assayed again for Cr(VI). In ddH2O, Cr(III) did not interfere with the Cr(VI) assay, and it was quantitatively oxidized by permanganate and then detected as Cr(VI). Cr(VI) added to PYE was not reduced or lost through volatization after 48 h. Comparison of the five different acid digestion procedures is shown in Table 2. The three standard methods procedures yielded approximately 50% recovery of Cr from Ottawa River sediments. The modified nitric acid digestion technique gave a higher recovery but was subject to a high standard deviation. We found the nitric to hydrochloric acid (aqua regia) procedure to give the best recovery with a low standard deviation. When internal standards were used, all procedures recovered approximately 60% of the total Cr, with the exception of nitricsulfuric acid digestion, which recovered only about 40%. When the acid digests were assayed for total and hexavalent Cr by the colorimetric assay, no difference was found between the values for total and hexavalent Cr. This indicates that the Cr recovered from acid-digested sediments was essentially all hexavalent. Extraction of the compounds known to interfere with the colorimetric assay (Fe, Hg, Mo, Na, and MnO4) did not increase the amount of Cr detected nor did it influence the ratio of hexavalent to total Cr. Of these compounds, Mo, Hg, and Va are not present in the sediments at high enough concentrations to represent significant interferences. The permanganate was eliminated by the addition of azide during oxidation of the samples for Cr(VI) determinations (1). Thus, only Fe represented a major interfering agent. Fe was present in high concentrations (Table 1) and may not have been completely extracted. Viable plate counts. Although the number of bacteria recovered from both sites on PYE containing 0 to 100 ,g of Cr(VI) ml-' decreased with increasing Cr concentration, viable plate counts for Cr-resistant bacteria showed no apparent difference between the two sites for any concentration of Cr tested (Table 3). From both sites, control plates [0 pug of Cr(VI) ml-'] contained 105 to 106 CFU g (dry weight)-1, whereas only 100 to 102 CFU g (dry weight)-' were recovered on PYE medium containing 100 ,ug of Cr(VI) ml-' (Table 3). Continuous flow enrichment. After the inoculated culture vessels were incubated as batch cultures for 24 h, the continuous flow of PYE medium containing 100 ,ug of Cr(VI) ml-' was begun. The concentration of Cr in the vessels rapidly increased and was stable at an average value of 87 ,ug ml-' by the third residence time. Because all results at both dilution rates were
CHROMIUM-RESISTANT BACTERIA c
c
0
(A (A =
0)0
=) =) -D 00
CD
0E
0)CL
C)
(D 0
0) 0
(A
3I
CD --=S
r) 0 j"
o
x
3
0
1+ 1+
0)
\0
1+ 1+
_5.
zco
0
O)C)
00
3 _. 0 0 I+ 3
0. 0) 0. 0)0. CA
cn C) 0)
000O
L'
I+ I+
0
BFf
aS
P0
5.-_l.
1+ 1+
_l.
G M.
00
C) 00k
00
0'
\0 0
ON
-.
1+ 1+ 00o
r)
C) CD
co
0
0 0
1+ 1+
_ On On
I+ 00
1+ 1+
.~-I
0
B C)
a -0
O.
.
_.
N 03
0 cn
CL 0.
0.
0.
o v~
0. 0.
0
00
0
~ 00
0. 3-C
00
00 -v
0. oo1+ 0 t'
0
00
*--v
00.v 1+ 1+
1+ 1+
.~>A 00
S
849
850
LULI ET AL.
APPL. ENVIRON. MICROBIOL.
TABLE 2. Recovery of Cr from Ottawa River sediments by five different acid digestion procedures Digestion procedure
Recovery with internal standarda
% Recoverya
Reference
50.0 ± 20.8 Nitric acid 1 84.2 ± 24.9 Modified nitric This paper This paper 76.1 ± 10.9 Nitric-hydrochloric 1 42.0 ± 7.4 Nitric-sulfuric 52.6 ± 13.4 1 Nitric-perchloric a Mean of three samples ± standard deviation. Recovery based on NAA data. b Internal standard was by addition and recovery of 50 ,ug of Cr(VI) ml-'.
64.5 61.5 67.5 39.9 59.7
± ± ± ± ±
6.4 16.3 12.7 5.4 18.3
recovered on 75 and 100 ,ug of Cr(VI) decreased slightly after the third residence time (Fig. 1). These trends held for all samples and both dilution rates. Chromium resistance potential. A total of 89 bacterial isolates (from both sites) from plates containing no Cr were tested for their level of resistance to Cr(III) and Cr(VI) (Fig. 2). Of the isolates tested, 47% were resistant to 100 jig of Cr(VI) ml-', and 88% were resistant to 100 jig of Cr(III) ml-'. However, at 250 jig of Cr ml-l, 29% were resistant to Cr(VI), whereas only 2% were resistant to Cr(III). To determine whether the Cl- ion was the toxic agent in CrC13, the organisms were plated onto media containing NaCl at a final Cl- concentration equal to that present in 250 ,ug of CrCl3 ml-; no Cl- toxicity was observed. Growth curves. Isolates which were known from plating experiments to be resistant to 100 ,ug of Cr(VI) ml-' were not visually affected in terms of growth rate or final turbidity by 100 jig of Cr(VI) ml-1 (Fig. 3). However, growth of Crsensitive cultures was inhibited by the addition of 50 jig of Cr(VI) ml-'. To test whether Cr was bacteriocidal or bacteriostatic, the two Cr-sensitive cultures were harvested, washed, and resus-
essentially identical, Fig. 1 represents the averaged values of six experiments (three at each dilution rate). The turbidity in each vessel remained constant after an initial decrease due to sediment particulate matter from the inoculum which was eventually washed out. At each residence time, samples were plated onto PYE agar containing the same Cr concentrations as used in direct plating procedures. Two- to three-log higher CFU were recovered over the values obtained by direct plating of diluted sediments. However, the ratio of different colony morphologies at each Cr(VI) concentration was comparable to direct plating (data not shown). The number of bacteria recovered from the continuous flow vessels decreased with increased Cr concentrations in the plates. Colony counts were in the range of 109 for 0 and 10 jig of Cr(VI) ml-, 108 for 25 and 50 ,ug of Cr(VI) ml-', and 107 for 75 and 100 ,ug of Cr(VI) ml-. The number of colonies recovered on plates containing 0 and 10 ,ug of Cr(VI) ml-' did not significantly change during the first six residence times. However, the population of bacteria capable of growth when plated onto 25 and 50 ,ug of Cr(VI) ml-' initially increased and then stabilized after the third residence time. The CFU
TABLE 3. Viable plate counts from direct plating of diluted sediments onto PYE agar containing various concentrations of Cr(VI) Date and Viable plate counts at the following Cr(VI) concn (>g ml-') x 104 sample sitea
0
10
25
50
75
100
9.00 1.14
0.05 0.006
0.003 0.0003
0.0009 0.0002
1.08 0.14
0.009 0.007
0.008 0.008
0.002 0.006
0.25 0.13
0.13 0.08
0.07 0.08
May 1981 1 2
June 1981 1 2
10.20 33.00
b
4.10 9.60
1.26 4.50
July 1981 1 2
252.00 2.52 1.38 78.00 1.14 0.75 a Site 1 is the contaminated site, and site 2 is downstream. b , Not done.
VOL. 46, 1983
CHROMIUM-RESISTANT BACTERIA
851
9 S z
In
07
..-------t0
..
ILIn
a 30 E
o
a4 g 20
0
1
2
0
o
3 4 RESIDENCE TIMES
FIG. 1. Average recovery of Cr(VI)-resistant bacteria from six continuous culture enrichments (three each at D = 0.02 and 0.10) plated onto PYE medium containing concentrations of Cr(VI), in micrograms per milliliter: 0 (0), 10 (0), 25 (O), 50 (E), 75 (A), and 100 (A). X, Total Cr concentration in each vessel.
pended in fresh PYE broth without Cr(VI). These cultures returned to normal growth rates and final turbidities. However, one Bacillus sp. was shown to completely lyse in the presence of 100 ,ug of Cr(VI) ml-. Antibiotics and other metals. To further differentiate strains and because heavy metal resistance has been linked to antibiotic resistance on plasmids (2), nine isolates which were resistant to Cr(VI) at 100 ,g ml-' and two which were sensitive to 10 ,ug ml-' were tested for their sensitivity to other heavy metals and several antibiotics (Tables 4 and 5, respectively). With the exception of one strain (P6), all of the isolates tested were generally resistant to other heavy metals at the concentrations used. Moreover, all of the Cr(VI)-resistant and -sensitive isolates demonstrated apparent similarities in antibiotic sensitivity profiles. Biochemical testing of isolates. Biochemical identification of the isolates showed that 64% belonged to the family Pseudomonadaceae (Table 6). These isolates were gram-negative, aerobic, respiratory rods which gave negative results for methyl red and Voges-Proskauer tests. They were motile, they reduced nitrate, and they produced yellow to green water-soluble pigments. Thirty-two percent of this group was resistant to 100 ,ug of Cr(VI) ml-'. Fifteen isolates (17%) were fermentative, gram-negative, methyl red positive, Voges-Proskauer negative, colorless, and identified as enterobacteria. Of this group, 47% were resistant to 100 ,ug of Cr(VI) ml-1. Nine gram-positive isolates were recovered: three Bacillus spp., which were resistant to 50 but not to 100 ,ug of Cr(VI) ml-', and two Streptomyces spp. and four corynebacteria, all
1to
1
so
CR CONCENTRATION (UJG/ML)
FIG. 2. Relative toxicities of Cr(VI) (U) and Cr(III) (0) to 89 isolates from Ottawa River sediments. Cr was incorporated into PYE agar, and resistance was positive if growth occurred within 48 h.
of which were resistant to 100 ,ug of Cr(VI)
ml'.
DISCUSSION Neutron activation analysis (NAA) proved to be a simpler method than AAS for simultaneously determining the levels of a broad range of elements in Ottawa River sediments, even though long periods of time were required for aquisition of the data. The advantages of NAA were apparent in that no acid digestion of sedi-
A
2.2
i
2 w
1.6
1l 0
tJ
1.21.4/
5
s
It
15
25
TIME (HR)
FIG. 3. Effect of 100 ,ug of Cr(VI) ml-' on the growth of pure cultures of Cr(VI)-resistant and sensitive isolates. Seven Cr(VI)-resistant isolates (0) showed no change in growth rate and final turbidity after the addition of Cr(VI) (long arrow). Growth of two Cr(VI)-sensitive isolates was inhibited (A) but returned to normal growth rates and final turbidities (U) after resuspension in fresh PYE medium containing no Cr(VI) (short arrow). One Bacillus sp. lysed in the presence of 100 ,ug of Cr(VI) ml-' (0).
852
LULI ET AL.
APPL. ENVIRON. MICROBIOL.
TABLE 4. Resistance profiles of several isolates to various metalsa Resistance
Strain
K2Cr2O7
CdC12
Co(NO3)2
Cu(NO3)2
Pb(NO3)2
NiC12
ZnSO4
+ + +
+ + +
+
+
+
+ +
+ +
+ + +
+ + + +
+ +
-
+ + + + +
Corynebacteria sp. strain C1 C2 C3 C4
+
+ +
Pseudomonad sp. strain + + + P1 + + + P2 + + + P3 + + P4 + + P5 P6 a Resistance was negative (-) if cleared zones were observed
ments was required. The high percent standard deviations for the NAA analysis of Cr, Sb, and Zn were thought to be due to incomplete homogenization of the dried sediment samples. Recovery of Cr from acid-digested sediments during Cr speciation experiments was subject to low efficiency and high standard deviation. Chromium speciation was limited by low recovery of Cr from acid-digested sediments. Of the five digestion procedures used, the nitrichydrochloric acid digestion gave the best recovery of Cr with a low standard deviation. Nitricsulfuric acid digestion gave the lowest recovery. It is not known whether the low recovery of Cr after acid digestion was due to matix interfer-
-
+ +
+ +
+ + + +
-
+ +
around impregnated disks after 48 h.
-
ences during AAS. Chlorides and perchlorates are preferred matrices, with nitrate being acceptable and sulfate being poor (1). The Cr which was recovered from digested sediments was essentially all hexavalent, even though previous work has determined the sediments to have a low redox state (J. A. Titus and R. M. Pfister, personal communication). This would indicate that the Cr in the sediments may be bound and not available to redox equilibria (20). The Cr could not be washed from sediments with water, and there was an apparent higher
affinity of the Ottawa River sediments for Cr(III) over Cr(VI). Because site 1 was located in the proximity of
TABLE 5. Resistance profiles of several isolates to various antibiotics'
Resistance' Strain
Ampicillin (10 U)
Baci-
tracin (10 U)
Erythromycin
Genta-
Neo-
Peni-
(15 U)
micin (10 U)
mycin (30 U)
cillin (10 U)
-
-
-
-
+
-
-
+
Polymyxin B (50 U)
Rifampin (50 U)
Streptomycin (10 U)
cycline (30 U)
-
-
-
Tetra-
Corynebacteria sp. strain C1 C2 C3 C4
-
+
-
+
-
+
-
-
-
-
-
-
+
-
-
-
-
+
-
-
-
+ + -
+ + +
-
-
-
-
-
-
-
-
+ +
+ +
+ +
-
-
+
+ + + + +
-
-
+ + +
-
-
+ + +
+ + _ _
+
+
+
Pseudomonad sp. strain P1 P2 P3 P4 P5 P6
-
-
-
a Resistance was negative (-) if cleared zones were observed around impregnated disks after 48 h. b Concentrations are in micrograms per milliliter except where noted as units (U).
-
+ _ _
VOL. 46, 1983
TABLE 6. Taxonomic identification of 89 isolates and the number of each group capable of growth on Cr(VI)
CHROMIUM-RESISTANT BACTERIA
853
ence may represent a population of cells which may not have been exposed to 87 jig ml-' in the continuous flow vessel. However, the bacterioNo. of isolates resistant at the static action of Cr(VI) was demonstrated in pure following Cr(VI) concn cultures of Cr-sensitive isolates (Fig. 3). The Group (p.gmMl1): isolates were able to recover normal growth 50 100 0 rates and final turbidity after being removed 18 from an inhibitory concentration of Cr and re57 39 Pseudomonads suspended in fresh medium without Cr. 0 8 0 Flavobacteria We have not determined whether the total 7 15 15 Enterobacteria 3 0 Cr(VI) added to the medium is actually "avail3 Bacillus spp. 2 2 2 Streptomyces spp. able" to the cells or whether there are toxicity 4 4 4 Corynebacteria differences between the solid versus liquid types of medium. We developed PYE medium with a low organic content to reduce the binding of Cr a plating industry, higher concentrations of Cr to medium components (19). Furthermore, it and other metals were expected and obtained. was determined that the Cr(VI) added to PYE Direct dilution and plating of sediments onto broth was not significantly reduced by comvarious concentrations of Cr(VI) showed a de- pounds present in the medium, nor was it volatilized. crease in the number of colonies recovered as the concentrations of Cr in the medium were The resistance to particular heavy metals has increased. Our data did not support the hypothe- been correlated to antibiotic and other heavy sis that a greater number of Cr-resistant bacteria metal resistance in a variety of organisms (2). would be recovered on plates containing high The Cr-resistant isolates tested in this study levels of Cr(VI) from the heavy metal-contam- were also resistant to a number of antibiotics inated site over the number recovered from and other metals. However, two Cr-sensitive sediments downstream. Chemical speciation of isolates were also resistant to the antibiotics and the Cr extractable from sediments showed that other metals, indicating that Cr resistance may the Cr present was predominantly hexavalent. not be directly correlated with generalized metal The Cr present was probably chemically bound and antibiotic resistance. We are currently atand not available to the bacteria (19). tempting to determine whether these phenoWhen sediments were incubated for 24 h types are correlated with the presence of plasbefore continuous culture enrichment was be- mid DNA, as has been suggested by other investigators (2, 23, 31). gun, a greater number of bacteria were recovThe isolates recovered on PYE medium were ered which were capable of growth or survival in the presence of 100 gxg of Cr(VI) ml-' over the predominantly pseudomonads. This group probnumber obtained by direct plating. As selection ably does not, however, represent the majority for Cr-resistant bacteria occurred in the continu- of active bacteria in the sediments at the contaminated site because of the highly reduced ous cultures, the number of colonies recovered on 75 and 100 ,ug of Cr(VI) ml-' declined after nature of the sediments (J. A. Titus and R. M. the third residence time. Because we have Pfister, personal communication). In addition, shown that the growth rates of Cr-resistant the medium and growth conditions used selected bacteria are not altered by Cr(VI), this decline for aerobic, heterotrophic bacteria. The purpose must be a reduction in the number of Cr-resist- of this study was not to determine the environant bacteria as a result of competition within the mental impact of Cr on the bacterial population mixed population of Cr-resistant bacteria in re- in Ottawa River sediments, but to determine sponse to the established experimental condi- whether there was a population of Cr-resistant bacteria present and to what level of Cr(VI) tions. We have shown that Cr(VI) is bacteriostatic to these bacteria were resistant. We have determined that high concentrations some sensitive strains by the difference in the number of colonies recovered on Cr(VI) at the of Cr are present in Ottawa River sediments at 50 ,ug ml-' level versus 100 ,ug ml-'. At the end an industry-contaminated site. Our data suggest of the continuous flow, the concentration of Cr that the Cr is predominantly hexavalent and is in the vessel was determined to be about 87 p.g chemically bound in the sediments. Because ml-'. Thus, all of the cells in the vessel would be there was no difference in the number and types expected to be resistant to that level of Cr(VI). of Cr-resistant bacteria isolated from contamOur data show a greater number of colonies inated sediments relative to downstream sedirecovered at low Cr concentrations at the end of ments, it is unlikely that the Cr present in these the continuous culture relative to the number sediments influences natural selection for Crrecovered on high Cr concentrations. The differ- resistant phenotypes.
854
LULI ET AL.
ACKNOWLEDGMENTS We are grateful for the assistance given by T. Adelman, D. Bucci, M. Burns, and M. Otstot in this project. We would especially like to thank J. Titus for his advice and contributions in the preparation of the manuscript. This work was partially funded by contract B-079-OHIO from the U.S. Department of the Interior, Office of Water Research and Technology.
LITERATURE CITED 1. American Public Health Association. 1971. Standard methods for the examination of water and wastewater, 13th ed. American Public Health Association, Inc., Washington, D.C. 2. Austin, D. A., and R. R. Colwell. 1977. Antibiotic resistance patterns of metal-tolerant bacteria isolated from an estuary. Antimicrob. Agents Chemother. 12:545-547. 3. Bartlett, R. J., and J. M. Kimble. 1976. Behavior of chromium in soils. I. Trivalent forms. J. Environ. Qual. 5:379-383. 4. Bartlett, R. J., and J. M. Kimble. 1976. Behavior of chromium in soils. II. Hexavalent forms. J. Environ. Qual. 5:383-386. 5. Bordner, R., and P. Scarpino (ed.). 1978. Microbiological methods for monitoring the environment-water and wastes. EPA-600/8-78-017. U.S. Environmental Protection Agency, Cincinnati, Ohio. 6. Brinton, H. P., E. S. Frazier, and A. L. Koven. 1952. Morbidity and mortality experience among chromate workers. Public Health Rep. 67:835-847. 7. Browning, E. 1969. Toxicity of industrial metals. Butterworth and Co., Ltd., London. 8. Colwell, R. R., and W. J. Wiebe. 1970. "Core" characteristics for use in classifying aerobic, heterotrophic bacteria by numerical taxonomy. Bull. Georgia Acad. Sci. 28:165185. 9. Duursma, E. K. 1972. Specific activity of radionuclides sorbed by marine sediments in relation to the stable element composition. IAEA-SM-158/4. International Atomic Energy Agency. 10. Dybczynski, R., and 0. Suschny. 1979. Final report on the intercomparison run SL-1 for the determination of trace elements in lake sediment samples. IAEA-RL-64. International Atomic Energy Agency, Vienna, Austria. 11. Dye, D. W. 1968. A taxonomic study of the genus Erwinia. I. The "Amylovora" group. N. Z. J. Sci. 11:590607. 12. Efstathiou, J. D., and L. L. McKay. 1977. Inorganic salts resistance associated with a lactose-fermenting plasmid in Staphylococcus lactis. J. Bacteriol. 130:257-265. 13. Elderfield, H. 1970. Chromium speciation in sea water. Earth Planet. Sci. Lett. 9:10-16. 14. Forstner, U., and G. T. W. Wittman. 1979. Metal pollution in the aquatic environment. Springer-Verlag, New York. 15. Gale, T. F. 1978. Embryonic effects of chromium trioxide in hamsters. Environ. Res. 16:101-109. 16. Huff, J. W., K. Sivarama Sastry, M. P. Gordon, and W. E. C. Wacker. 1964. The action of metal ions on tobacco mosaic virus ribonucleic acid. Biochemistry. 3:501-506. 17. Loeb, L. A., and R. A. Zakour. 1980. Metals and genetic miscoding, p. 115-144. In T. G. Spiro (ed.), Metals in biology, vol. 1. Wiley Interscience, New York.
APPL. ENVIRON. MICROBIOL. 18. Mertz, W. 1969. Chromium occurrence and function in biological systems. Physiol. Rev. 49:163-239. 19. Nakayama, E., T. Kuwamoto, S. Tsurubo, H. Tokoro, and T. Fujinaga. 1981. Chemical speciation of chromium in sea water. I. Effect of naturally occurring organic materials on the complex formation of chromium (III). Anal. Chim. Acta 130:289-294. 20. Nakayama, E., T. Kuwamoto, S. Tsurubo, H. Tokoro, and T. Fujinaga. 1981. Chemical speciation of chromium in sea water. II. Effects of manganese oxides and reducible organic materials on the redox processes of chromium. Anal. Chim. Acta 130:401-404. 21. Nakayama, E., T. Kuwamoto, S. Tsurubo, H. Tokoro, and T. Fujinaga. 1981. Chemical speciation of chromium in sea water. III. The determination of chromium species. Anal. Chim. Acta 131:247-254. 22. Nishioka, H. 1975. Mutagenic activities of metal compounds in bacteria. Mutat. Res. 31:185-189. 23. Novick, R. P., and C. Roth. 1968. Plasmid-linked resistance to inorganic salts in Staphylococcus aureus. J. Bacteriol. 95:1335-1342. 24. Petrilli, F. L., and S. DeFlora. 1977. Toxicity and mutagenicity of hexavalent chromium on Salmonella typhimurium. Appl. Environ. Microbiol. 33:805-809. 25. Petrilli, F. L., and S. DeFlora. 1978. Metabolic deactivation of hexavalent chromium mutagenicity. Mutat. Res. 54:139-147. 26. Petrilli, F. L., and S. DeFlora. 1978. Oxidation of inactive trivalent chromium to the mutagenic hexavalent form. Mutat. Res. 58:167-173. 27. Rades-Rohkohl, E., P. Hirsch, and 0. Franzle. 1979. Neutron activation analysis for the determination of amphibolite rock-weathering activity by a yeast. Appl. Environ. Microbiol. 38:1061-1068. 28. Roe, F. J. C., and R. L. Carter. 1969. Chromium carcinogenesis: calcium chromate as a potent carcinogen for the subcutaneous tissues of rats. Br. J. Cancer 23:172-179. 29. Schroeder, D. C., and G. F. Lee. 1975. Potential transformations of chromium in natural waters. Water Air Soil Pollut. 4:355-365. 30. Smillie, R. H., K. Hunter, and M. Loutit. 1981. Reduction of chromium (VI) by bacterially produced hydrogen sulfide in a marine environment. Water Res. 15:1351-1354. 31. Summers, A. O., and G. A. Jacoby. 1978. Plasmid-determined resistance to boron and chromium compounds in Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 13:637-640. 32. Summers, R. J., D. P. Boudreaux, and V. R. Srinivasan. 1979. Continuous cultivation for apparent optimization of defined media for Cellulomonas sp. and Bacillus cereus. Appl. Environ. Microbiol. 38:66-71. 33. Venitt, S., and L. S. Levy. 1974. Mutagenicity of chromates in bacteria and its relevance to chromate carcinogenesis. Nature (London) 250:493-494. 34. Waters, L. J., Jr., T. J. Wolery, and R. D. Myser. 1974. Occurrence of As, Cd, Co, Cr, Cu, Fe, Hg, Ni, Sb, and Zn in Lake Erie sediments, p. 219-234. In Proceedings of the 17th Conference on Great Lakes Research. International Association of Great Lakes Research. Braun-Brumfield, Inc., Ann Arbor, Mich. 35. Zubkoff, P. L., and W. E. Carey. 1970. Neutron activation analysis of sediments in western Lake Erie, p. 319325. In )Proceedings of the 13th Conference on Great Lakes Research. International Association of Great Lakes Research. Braun-Brumfield, Inc., Ann Arbor, Mich.
