Heavy Metal Resistance Patterns of Frankia Strains

The sensitivity of 12 Frankia strains to heavy metals was determinedby a growth inhibition assay. In general, all of the strainswere sensitive to low concentrations (<0.5 mM) of Ag¹⁺, AsO₂^1-,Cd²⁺, SbO₂^1-, and Ni²⁺, but most of the strains were less sensitiveto Pb²⁺ (6 to 8 mM), CrO₄^2- (1.0 to 1.75 mM), AsO₄^3- (>50mM), and SeO₂^2- (1.5 to 3.5 mM). While most strains were sensitiveto 0.1 mM Cu²⁺, four strains were resistant to elevated levelsof Cu²⁺ (2 to 5 mM and concentrations as high as 20 mM). Themechanism of SeO₂^2- resistance seems to involve reduction ofthe selenite oxyanion to insoluble elemental selenium, whereasPb²⁺ resistance and Cu²⁺ resistance may involve sequestrationor binding mechanisms. Indications of the resistance mechanismsfor the other heavy metals were not as clear.

INTRODUCTION

Top

Introduction

Frankia, a member of the order Actinomycetales, forms a symbioticnitrogen-fixing association with a variety of woody dicotyledonousplants (for reviews see references 3 and 28). The members ofthis bacterial genus are known to be associated with over 200species of plants representing eight plant families. These bacteriafix N₂ from the atmosphere and produce a significant amountof the fixed nitrogen on the planet. Actinorhizal plants areecologically important as pioneer community plants and haveeconomic value in land reclamation, reforestation, and soilstabilization.

The lack of a well-established genetic system is a major obstaclein the elucidation of the mechanism of actinorhizal nitrogenfixation and plant-microbe interactions (for reviews see references22, 23, and 25). There is a paucity of genetic markers for Frankia.Some of the most useful genetic markers include resistance toantibiotics, resistance to antimetabolites, and resistance toheavy metals. These directly selectable traits provide a mechanismfor positive selection in genetic studies and are also usefulin the development of cloning vectors. For example, metal resistancehas been useful in the development of cloning vectors for Rhodococcus,another member of the Actinomycetales (8). Previously, we developeda growth inhibition assay that was used to screen several Frankiastrains for resistance to antibiotics (27). Although severalimportant antibiotic resistance markers were described in thatstudy, we were interested in identifying other selectable geneticmarkers. Since actinorhizal plants have been used for land reclamationin strip-mined areas, we reasoned that it is possible that thebacteria are resistant to heavy metals. The purpose of thisstudy was to extend the use of our growth inhibition assay toidentify new selectable markers, resistance to heavy metals.

Frankia strains ACN1^AG (14), CcI3 (29), Cc1.17 (18), CN3 (20),CpI1 succinate variant (CpI1-S) (5, 26), CpI1 propionate variant(CpI1-P) (5, 26), DC12 (1), EI5c (17), EAN1pec (16), EuI1c (2),EUN1f (14), and QA3 (12) were grown and maintained in basalgrowth medium with NH₄Cl as the nitrogen source, as describedpreviously (26, 27). For the heavy metal sensitivity assays,the basal growth medium contained 20 mM glucose and 20 mM succinateas the carbon and energy sources (glucose-succinate medium).For strains EUN1f, Cc1.17, CpI1-P, and QA3, the growth mediumcontained 5 mM propionate instead of glucose and succinate.For comparative purposes, Escherichia coli W3110, Bacillus subtilis168, Micromonospora echinospora ATCC 15836, and Streptomycesviridochromogenes NRRL B-1511 were also used in this study.

The sensitivities of Frankia strains to heavy metals were determinedby a growth inhibition plate assay that was developed initiallyto test for antibiotic sensitivity (27). For Cu²⁺, AsO₂^1-, AsO₄^3-,and Pb²⁺, the standard growth medium was replaced with a low-phosphategrowth medium that contained 1 mM K₂HPO₄. Metal ions, includingAgNO₃, Na₂HAsO₄, NaAsO₂, CdCl₂, CoCl₂, K₂CrO₄, CuCl₂, NiCl₂,Pb(NO₃)₂, K(SbO)C₄H₄O₆, and Na₂SeO₂, were added to the growthmedia at concentrations of 0.005, 0.01, 0.05, 0.1, 0.25, 0.5,0.75, 1.0, 2.0, 4.0, 5.0, 6.0, 8.0, 10.0, and 20.0 mM; AsO₄^3-was also tested at concentrations of 25 and 50 mM. For theseexperiments, 14- to 21-day-old cultures were used as the inocula,and the hyphae were fragmented by using a tissue homogenizer.Each inoculum (0.5 ml) was added to a 10-ml sterile Falcon tube,and 3.0 ml of 0.8% Bacto Agar (Difco) at 45°C was also addedto the tube. The contents of the tube were mixed by agitationwith a vortex mixer and were poured onto the surface of a plate.Most plates were incubated at 30°C; the exceptions werethe plates that contained strains DC12 and EAN1pec, which wereincubated at 25°C, and the plates that contained strainCc1.17, which were incubated at 28°C.

Growth was scored after 2 weeks and was monitored for an additional6 weeks. When this assay was used, growth could be easily scoredvisually. Growth was scored relative to the growth of the controlby using a scale ranging from + to ++++. With some Frankia strains,growth on the 10⁰ control plates resulted in a thick bacteriallawn. After dilution and plating, the total number of hyphalfragments was also estimated. An average of about 10⁷ fragmentswas plated for each assay. Plates that had the same amount ofgrowth as the undiluted control were scored as ++++, while platesthat showed a 2-log decrease were scored as +++. Thus, the +++,++, and + growth scores were equivalent to the growth on the10^-2, 10^-4, and 10^-6 control plates, respectively. A negativeresult indicated that there was no growth. For E. coli and B.subtilis, cells were incubated at 37°C and growth was scoredon days 1 and 2. For S. viridochromogenes and M. echinospora,cells were incubated at 28°C and growth was scored on days7 and 14.

Growth was quantified relative to the growth of a control containingno metals as described above. To evaluate the levels of resistance,the following two parameters were used: MIC and maximum tolerableconcentration (MTC). The MTC is the highest concentration ofmetal which does not affect the viable count and is used withselective media designed to promote the growth of a resistantbacterium by providing efficient counterselection for undesiredbacteria (19). The values were determined by inspection afterthe relative growth was plotted as a function of the log ofthe metal concentration. The MIC was determined by determiningthe intersection of the survival curve with the horizontal axis(Fig. 1). In each experiment each metal was tested in duplicate,and each metal was tested in three to seven replicate experiments.In all cases, similar results were obtained in the experimentsperformed.

View larger version (16K):
[in this window]
[in a new window]
FIG. 1. Frankia growth as a function of heavy metal (chromate) concentration. Growth was scored on a scale from 0 to ++++ as described in the text.

Heavy metal sensitivity.

Top

Heavy metal sensitivity.

Strains that showed resistance to metals as determined by theplate assay (Tables 1 and 2) were also resistant to the samemetals in broth cultures (data not shown). All of the Frankiastrains were sensitive to Ag¹⁺, Ni²⁺, Cd²⁺, SbO₂^1-, and AsO₃^1-.Compared to the control bacteria, a few Frankia strains hadhigh MICs ( $>=$ 0.4 mM) for Co²⁺ and were considered cobalt resistant.Most Frankia strains were resistant to AsO₄^3-, Pb²⁺, SeO₂^2-,and CrO₄^2-. Several of these metal resistance properties andthe effect of Cu²⁺ are described below.

View this table:
[in this window]
[in a new window]
TABLE 1. MICs of 10 heavy metals for 12 Frankia strains and other organisms

View this table:
[in this window]
[in a new window]
TABLE 2. MTCs of nine heavy metals for 12 Frankia strains and other organisms

CrO₄^2- resistance.

Top

Cro42- resistance.

Most of the isolates were resistant to elevated levels of CrO₄^2-(Table 1). The MICs for two isolates, CcI3 and EAN1pec, were<0.5 mM, and these isolates were considered chromate sensitive.For the chromate-resistant isolates the CrO₄^2- MICs ranged from1.0 to 1.75 mM. The control bacteria E. coli and B. subtiliswere chromate sensitive, while the actinomycetes M. echinosporaand S. viridochromogenes were chromate resistant. When observedby phase-contrast microscopy, these bacteria exhibited no obviousmorphological or structural changes (data not shown). As a mechanismof resistance to CrO₄^2- some organisms reduce the highly solubleoxyanion Cr(VI) to the less toxic cationic form Cr(III), whichreadily precipitates (6, 19, 24). No precipitate was observedwith resistant strains that were grown in medium containingchromate. The amount of Cr(VI) in the spent growth medium wasdetermined by measuring the absorbance at 380 nm. Cr(VI) absorbslight at 380 nm, while Cr(III) does not. The measurements showedthat there was no change in the amount of Cr(VI) during growth.These results suggest that CrO₄^2- resistance does not occurvia a chromate reduction mechanism.

SeO₂^2- resistance.

Top

Seo22- resistance.

One-half of the Frankia isolates tested were resistant to elevatedlevels of SeO₂^2- (Table 2). Isolates for which this metalloidhad an MIC of <0.5 mM were considered selenite sensitive,while the SeO₂^2- MICs for selenite-resistant isolates rangedfrom 1.5 to 3.5 mM. The control bacteria E. coli and S. viridochromogeneswere selenite sensitive, while B. subtilis and M. echinosporawere selenite resistant. Frankia isolates that were resistantto elevated levels of selenite formed red colonies on the surfaceof the growth medium. The red form is elemental selenium, aninsoluble, generally less toxic state of the element. In liquidcultures, a red precipitate was found to be associated withcells (data not shown). Frankia strains resistant to SeO₂^2-formed globular structures associated with their hyphae. Theseresults suggest that the selenium oxyanion is reduced to Se⁰,which remains associated with the cells. Selenite resistanceis probably due to reduction of colorless soluble SeO₂^2- toan insoluble red form of Se⁰, which is much less toxic thanselenite. This is a common mechanism of selenite resistance.In B. subtilis, an inducible detoxification system reduces seleniteto elemental selenium (9). Rhodobacter sphaeroides and Rhodospirillumrubrum exhibit intracellular oxyanion reduction. R. sphaeroidesdeposits the metal in the cytoplasmic membrane (21), while R.rubrum expels the elemental selenium across the cytoplasmicmembrane and the cell wall (13). From our results, it is notclear whether selenite reduction occurs intracellularly, extracellularly,or at the Frankia membrane.

Cu²⁺ resistance.

Top

Cu2+ resistance.

An unusual but consistent pattern of resistance was observedwith Cu²⁺ (Fig 2). With strains DC12, EuI1c, and CN3, growthoccurred in the presence of elevated levels of Cu²⁺ (2 to 5mM and even up to 20 mM with strain CN3), but no growth wasobserved in the presence of moderate levels of Cu²⁺ (0.5 to1.0 mM for EuI1c and CN3; 1.0 mM for DC12). None of the otherFrankia strains tested grew in the presence of >0.1 mM CuCl₂(data not shown). For B. subtilis, S. viridochromogenes, andM. echinospora the MIC of Cu²⁺ was 0.2 mM, while for E. colithe MIC was 0.3 mM. Similar patterns of sensitivity and resistancewere observed with CuSO₄ (data not shown). Blue colonies formedon the surface of medium containing Cu²⁺ (Fig. 2). In some cases,the growth medium surrounding the colonies appeared to be lighterblue or white. This observation suggests that colonies boundor absorbed the Cu²⁺ in the medium, possibly by producing adiffusible binding compound. Several bacteria have this typeof Cu²⁺ resistance mechanism, and the genes encoding resistanceare plasmid borne or located on the chromosome (4, 7, 10, 24).In broth medium, the Cu²⁺ was also found to be associated withFrankia cells. As determined by phase-contrast microscopy, Cu²⁺-resistantFrankia strains formed unusual globular structures that wereassociated with their hyphae (data not shown). All of the Cu²⁺-resistantstrains either formed ineffective nodules (EuI1c) or were membersof the atypical Frankia group (DC12 and CN3) containing organismswhich are unable to reinfect host plants.

View larger version (91K):
[in this window]
[in a new window]
FIG. 2. Frankia strain CN3 colonies grown on succinate-glucose-NH₄Cl low-phosphate medium containing CuCl₂. Plates containing CuCl₂ at the following concentrations were incubated for 35 days at 30°C: no CuCl₂ (control) (A), 0.5 mM CuCl₂ (B), 2.0 mM CuCl₂ (C), 5.0 mM CuCl₂ (D), and 10.0 mM CuCl₂ (E).

Pb²⁺ resistance.

Top

Pb2+ resistance.

Several Frankia isolates were resistant to elevated levels ofPb(NO₃)₂, and the range of MICs for these strains was modest(Table 1). Compared to the four other bacteria tested, Frankiastrains were more resistant to Pb²⁺. The MIC of Pb(NO₃)₂ forthe most sensitive Frankia isolates was 5 mM, while for themost resistant isolates 8 mM Pb(NO₃)₂ was required to inhibitgrowth. Most of the Frankia lead-resistant strains had MTC valuesof $>=$ 4 mM, while two strains (ACN1^AG and QA3) had a Pb²⁺ MTC of6 mM. In broth media, Pb(NO₃)₂ forms a cloudy solution. Growthof the resistant isolates cleared the media. When Pb²⁺ was presentin the growth medium, resistant Frankia isolates produced anamorphous matrix that was associated with the hyphae.

Our results suggest that Pb²⁺ resistance and Cu²⁺ resistancemay result from a sequestration or binding mechanism. Cu²⁺ andPb²⁺ appear to bind to material on the cell surface. Polysaccharidesor proteins are potential binding sites. Pb²⁺-resistant strainsof Staphylococcus aureus and Citrobacter freundii accumulatelead as Pb²⁺ phosphate (15, 16). Cells of Streptomyces sp. alsoaccumulate Pb²⁺ (11). Cu²⁺ is accumulated in the periplasm ofPseudomonas by periplasmic CopA and CopB proteins, which alsocauses the bacterial colonies to turn blue (7, 24). Our resultsalso suggest that Cu²⁺ is bound to the cells. Cu¹⁺ rapidly oxidizesto Cu²⁺ under aerobic conditions. With some systems, Cu¹⁺ resistanceis linked with Ag¹⁺ resistance (7, 24), but Frankia isolateswere very sensitive to Ag¹⁺.

Most of the Frankia strains were resistant to elevated levelsof several heavy metals. In genetic studies, these heavy metalresistance traits should be extremely valuable as positive selectionmarkers. Frankia strains grow slowly and require long incubationtimes. Compared to antibiotics, heavy metals are more stableand do not degrade over long incubation times. The high MTCsof Pb²⁺, CrO₄^2-, and SeO₂^2- (Table 2) suggest that these metalscould be used as selective agents for some Frankia strains.We are currently testing the use of heavy metal resistance asa counterselection marker in conjugation experiments.

ACKNOWLEDGMENTS

This investigation was supported in part by Hatch grant 377,by a grant from The University of New Hampshire Vice Presidentfor Research Discretionary Funds, by a faculty summer fellowshipfrom the Graduate School of The University of New Hampshire,and by the College of Life Science and Agriculture, The Universityof New Hampshire-Durham. A summer undergraduate research fellowshipfrom the University of New Hampshire-Durham supported M.S.C.

We thank Robert Mooney for his help with the photography andJennifer Pinard and Julia Burrows for their contributions duringthe initial stages of this project.

FOOTNOTES

* Corresponding author. Mailing address: Department of Microbiology, University of New Hampshire, 46 College Rd., Durham, NH 03824-2617. Phone: (603) 862-2442. Fax: (603) 862-2621. E-mail: LST{at}hypatia.unh.edu.

This is scientific contribution number 2076 from the New HampshireAgricultural Experiment Station.

Present address: Proteome Inc., Beverly, MA 01915.

Present address: Department of Microbiology, Michigan StateUniversity, East Lansing, MI 48824.

REFERENCES

Top