Mercuric reductase in environmental Gram-positive bacteria sensitive to mercury

FEMS Microbiology Letters 97 (1992) 95-100 O 1992 Federation of European Microbiological Societies 03'78-1097 /92/505.00 Published by Elsevier

FEMSLE 05071

Mercuric reductase in environmental Gram-positive bacteria sensitive to mercury Elena S. Bogdanova ", Sofia Z. Mindlin o, Eva Pakrov5 b, M. Kocur b and Duncan A. Rouch

'

t' " Institute of Molecular Genetics, Russian Academy of Sciences, Moscow, Russia, Czechoslot;ak CoLlection ol Microorganisms, Masaryk UniL;ersity, Brno, CzechosloL;akia, and ' School of Biological Sciences, Birminghant Unirersity, Birmingham, UK

Received 1 July 1992 Accepted 3 July 1992

Key words: Mercuric reductase; Mercury resistance; Cryptic mer operor, Environmental bacteria collections 1. SUMMARY

2. INTRODUCTION

According to existing data, mercury resistance operons (mer operons) are in general thought to be rare in bacteria, other than those from mer-

A common mechanism for bacterial resistance to non-organic mercury is mer operon-encoded

cury-contaminated sites. We have found that a high proportion of strains in environmental isolates of Gram-positive bacteria express mercuric reductase (MerA protein): the majority of these strains are apparently sensitive to mercury. The expression of MerA was also inducible in all cases. These results imply the presence of phenotypically cryptic zer resistance operons, with both the merA (mercuric reductase) and merR (regulatory) genes still present, but the possible absence of the transport function required to complete the resistance mechanism. This indicates

that mer operons or parts thereof are more widely in nature than is suggested by the frequency of mercury-resistant bacteria.

spread

to: E.S. Bogdanova, Institute of Molecular Genetics, Russian Academy of Sciences, Kurchatov Sq. 46, Correspondence

Moscow 123182, Russia.

enzymatic detoxification. mer operons from different sources appear to encode a number of common functions, which together confer resistance, viz. a regulatory protein (MerR), a mercury

transport system (MerT) and/or other proteins) and the enzyme mercuric reductase (MerA) [1]. In the presence of mercuric ions, the regulatory protein promotes expression of the structural proteins, whereupon mercuric ion is taken into the cell by the transport system and subsequently reduced by MerA to non-toxic metallic mercury, which then volatilizes from the cell. Among bacteria isolated from different environments mercury-resistant strains are rare (0.001 - 1 - 47a) t2- 51. In studies of environmentally isolated strains of Gram-positive bacteria from culture collections we have occasionally found the presence of inducible mercuric reductase in strains that are apparently sensitive to mercury. This suggested that mer determinants, or parts of these, are more widely spread in nature than can be in-

96

ferred from the frequency or mercury-resistant bacteria. To test the validity of this notion a number of strains of Gram-positive bacteria were examined for both the presence of MerA and resistance to mercury.

and after 1-1.5 h of further incubation the cells were harvested by centrifugation. 3.1. Determination of MerA in cell extracts

MerA activity was dctermined in extracts by spectrophotometry at 340 nm of the mercury-de-

3. MATERIALS AND METHODS 3.1. Bacteriol strains

The organisms uscd wcre te n Micrococcus strains (M. lttteus B110 and B1045, M. t'arians 825, 8485 and 81232, M. roseus 81236, Micrococcus sp. 8490, 81108, 81233 and 81042), two strains of Rhodococctts ( R. lentifragmentus Acl161 and R. ery-thropolis Ac11,50), two strains Arthrobacter (A. globiformis Ac1109 and A. oxydans Acll14) and one Mycobacterium strait (Mb. cyaneum Acll55) from All-Union Collection of Microorganisms (VKM); one strain of B re t' ib a c t e rium fla ttm B-42 (derive d ATCC 1 406 7) from A1l-Union Collection of Industrial Microorganisms (VKIM) and 54 strains of seven Micro-

pendent oxidation of NADPH [7]. One unit of McrA catalysed HgClr-dependent oxidation of 1.0 prmol of NADPH for 30 min at 30'C. Electrophoresis of total bacterial cxtract proteins followed by immunoblotting with Typc III serum (obtained against the mycobacterial MerA) was carried out as in [8]. In all the experiments, proteins of both induced and not induced cultures were analyscd simultaneously.

of

3.5. Mercury susceptibility, testing

coccus species from the Czechoslovak Collection

3.5.1. The disk diffu.sion method. PY-agar 1 containing I pg/ml HgCl, (25 ml per plate) was overlaid with 5 ml of soft PY-agar (0.5o/o agar) containing 0.5 ml of induced culture with ,4onn 2.4 (a I cm wide cuvette). 6-mm diameter filter paper disks loaded with 10 pg of HgCl, were placed on the plates. The growth inhibition zoncs were measured after incubation at 30"C for 20 h. 3.5.2. Determination of minimal inhibition concentration L,alues (MIC). A. Bacterial suspensions induced as in section 3.3. were diluted to 10,,,,: 0.15; 50 irl of diluted suspensions were spread

L'

of Microorganisms (CCM) (Table 1). The previously described mercury-resistant strain of Arthrobacter sp. IMG TC28-1 and its mercury-sensitive derivative were used as controls t6l.

Stock culturcs were maintained at 4-8'C on LB-agar [6]. 3.2. Media

Thc following nutrient media werc used: Luria

broth (LB) [6]; AP-broth [6]; PY-broth of thc following composition: Bactopeptone (Difco), 1%, yeast extract (Dif'co), 0.5%, NaCl, 1'lo; PY-broth sotidificd with 1.57o agar Difco (PY-agar t); PYbroth solidificd with l.8c/o agar Ferak (PY-agar 2).

over half of the plate with PY agar 2 supplemented with HgCl, at concentrations over the range ol 2-50 pg/ml. B. 5 ir,l of induced bacterial cells with Atoo: 0.1 were placed in drops on plates with PY-agar I supplemented with I-10 pg/ml of HgClr. All the assays in section 3.5. were repeated 3-7 times for each strain, with mean values determined.

3.6. Antihiotic st,tt.sitit ity tt'stirtg

3.3. MerA inductiort

Overnight cultures of bacteria grown at 30"C vcllumc of

in LB were supplemented with 1/7

AP-broth, 0.1% glucose and 0.15-0.3 pC/ml HgCl,. After 30 min of incubation,0.35-0.7 pC/ml HgCl, was addcd to bacterial suspensions

Bacteria to be tested wcrc grown for 36 h on PY-agar I and individual colonies wcre rcplicated to the same medium supplcmcntcd with antibiotics at the final cclncentrations (in pg/ml): penicillin, 50; ampicillin, 10; streptomycin, 5 and 20; chloramphenicol, 5; tetracyclin, 5; kanamycin, 5; erythromycin, 5 and 50.

Table

1

Mercuric reductase activity, mercury and antibiotic resistance of bacterial strains Bacterial

Mercuric reductase

strains

Activity

Presence

Inhibition

MIC Ab

(U

of MerA

zone

(ps/ml)

(immuno

(mm) "

/mg

prote in

)

HgCI, resistance

Antibiotic

MICBh

resi st ance

1pg/ml).

blot) tvt. tyt0e

RWHI l SLI66 CCM 2694 CCM 2696 TW226

CCM 2693

KHYI2

l.lti

+

5

l5

S

0.tt-5

+

12'l

2

S

0.59 . 0.04

+

S

+

l3 t7

NT

l3

3

S

2

S

2

S

0.

20" 20"

0.00 0.00 0.00

21

'

NT

Pen, Amp

M. luteu.t

CCM 144 CCM 1048 CCM 169

0.23

+

t6

4

S

0.31

+

16

NT

S

0.52 0.66

+

r'7

4

S

'

CCM 622

NT

+

16

4

CCM 559

0.02 0.06 0.03

+

12

2

12 l-2

CCM 410 CCM 33I

0.15

+

24

2

l2

0.09

+

l5

2

S

CCM 2494

0.15

+

20

3

S

CCM 1335

0.00

'

1g "

NT

S S

S

Pen

M. t'arians

CCM 825 CCM2492

0.09 0.18 0.02 0.03

CCM 2431

0.03

CCM 552 CCM 2490 CCM 547

0.03 0.00 0.00 0.00

10

' 11

' 12',|

CCM 2671

13 "

5

2

S

17"

2

1.,2

Pen

NT

1.-2

Pen

/-)

M. nishinomiyaensis CCM 2670 0.80 JL79 0.70 CCM 2672 0.(x)

CCM 2I4O

2

'

0.00 0.00

+

3

8

5

Pen

+

4

6

6

S

l3 19"

NT

2

t6

NT

I

Pen Pen Pen

3

12

M. kristinue

CCM 2690

0.20

l

S

0.21

8

NT NT

5

JM8

4

Pen

SM237 SM332

0.22 0.25 0.00 0.00

18

3

2

S

1B

4

1-2

Str, Erm

2

Pen, Amp Pen Pen, Amp

M. sedentarius CCM 2691

CCM 314 CCM 2699

'

1.09

5

0.61

6

0.00

20"

NT NT NT

2 1

98

Table 1 (continucd) Bacteria strai

Activity

Presencc

Inhibition

MICAb

(U/mg

of MerA

zone

(

protein)

(

immu no-

Antibiotic

I-lgCl, resistance

Mcrcrrric reductase

I

n s

p"g/ml)

MICBb pg/ml)

re si st

a

nce

(

(mm) "

blot) Rfut d ot

occ

u.s I e n t iJiu gme r tt tts

VKM Acl167 B rc t'

i but'

I

c riurn

0.79 0.26 J

lu

NT

l0

'

t urtt NT

0.49 0.27

B-42

0.139

'

Arthrobacter sp.

IMG TC28-1 ilg-r IMG TCl2l3-l Hg-s

+

0.91

0.00

2

40

>10

l0

5

2

NT NT

Abbreviations: Pen, penicillin: Amp, ampicillin; Str, strcptomycin; Erm, erythromycin; NT' not tested. " Inhibition zone diameter minus diameter of the disk. I' See MerlnreLs nNo MnrsoDS 3.5.2. A and B. respectively. ' Results of 2 3 experiments are shown. d In 50% of experiments bacteria formed a lawn only when double turbid suspension was used (1600:4.8). ' There were two growth inhibition zones around the disks: clear and turbid. The sum total of the two zone diameters is given. The presence of MerA ancl antibiotic resistance were tested in an additional 20 strains of Micrococcus from CCM. MerA has been found in immunoblot rn M. tytae RM32,l (S), M. roseus CCM 679 (S), CCM 570 (S), CCM 560 (S), CCM 837 (S), CCM 618 (S), CCM 1679 (S), M. kristinae CCM 2691 (S), MK322 (Pen). MerA has not been found in M. lylae MK312 (S), M. tarians CCM 884 (S), CCM 2139 (S), M. roseus CCM 839 (S), CCM 691 (S), CCM 1145 (S), M. nLshinomiyaensis CCM 2699 (Pen), PM297 (Pen), M. sedentarius CCM 2698 (Pen), CCM 3947 (Pen), CCM 4055 (Pen, Str): antibiotic resistance data are shown in parentheses.

4. RESULTS AND DISCUSSION

The prevalence of MerA in randomly chosen Gram-positive bacteria was first revealed in in-

vestigation of Micrococcus, Rhodococcus, Arthrobacter, Mycobocterium and Breuibacterium strains from VKM and VKIM (see naargntels AND METHODs). MerA was observed in six out of 16 strains under study, M. t:arians B25 and 8485, M. luteus 81045, M.sp.8490, R. lentifragmentus AC1167 and B. flaL:trm B-42 (data not shown; Table 1). This indicated the presence in these bacteria

of the merA

gene. Furthermore, the

expression of MerA was inducible by mercuric ion, implying the additional presence of the regulator protein gene, merR [1]. To examine further the frequency of mer operons in environmental

Gram-positive bacteria, the presence of MerA and mercury resistance was studied in 54 strains

of seven M. species from CCM.

Table 1 shows that MerA can be found in all Micrococcus species studied with high frequency of occurrence: from 297o for M. nishinomiyaensis to 897o for M. luteus. The values could possibly be underestimates, if MerA was not detected in some cases due to non-optimal induction conditions and the MerA was different enough not to be detected by the heterologous serum (against mycobacterial MerA). However, the proportion

of undetected MerA-encoding mer

operons

should not be significant as MerA was clearly seen on immunoblots in all cases in which MerA activity could be determined in extracts (Fig. 1 and Table 1). A high frequency of mercury resistance duc to mer operon-encoded enzymatic detoxification has been described only for microorganisms isolated from mercury-polluted areas 12-4,91 and clinical specimens [10-12]. Mercury is unlikely to occur in high concentrations in the habitats of micro-

99

KDA

VZ.

e

uteus

sldpls_-_JJL __1e4a_{4g _t2lgg__4/o , uru 57\ * 56- ffi **a .q --

t2- .rde

i-+l*+l-+r-+t-+lm. eqtae

std RtqHfl sL.t65 2696

+t 8.ffavum

26s4 sld *ret

:a

&

---_-

q

d.

@

r-+l-+l-+l-f1l-+l

resistant

to penicillin, show no correlation

bc-

tween the presence of the nrcr operon and antibi-

otic resistance. Thc majority

of

Mic:roc:occLts, BreL'ibacteriurn

and possibly Rhodocr,tcc:tts studied are apparently scnsitivc to HgCl, (Table l), in accordance with the usual criteria [2]. Only M. lylae RWHI I and Arthrobactet' sp. IMG TC2tt-I Hg-r, as positivc control, could be classcd as HgCl ,-rcsistant. Thc mercury-sensitivc phcnotypc of thcse strains suggests the possible absence of a functional transport systcm, which is needed in addition to MerA

to confer

resistance [15]. Consequently, thesc bacteria appear to contain phenotypically cryptic

m. zlarians

s/d. ztgt a4g2

62l- riz

Y-l--r

a4go

---

;-" l- +l- +l- +l- +l- + Fig. 1. Immunoblot of mercuric reductases of strains deI

scribed in Table 1. Strain designation is reported in the upper

part. ( ) and (+) denote the non-induced and induced cultures. Std lanes show the molecular masses of mercuric

reductases of strains: Arthrobacter sp. IMG TC43-4, Mycobacterium sp. IMG CHM22-8, IMG CHM19-3 and IMG CHM21-1 [6] used as molecular mass standards. Extract protein quantities in each lane are. M. luteus CCM 559, 40 pg; CCM 331, 30 pg: CCM 1048, 7 pg; CCM 169, I pg; CCM 2494, 15 pg; CCM 410. 60 pg; M. Iylae RWHll, 2.5 pe;

SL166, 0.5 pg: CCM 2696, 30 pg; CCM 2691, 60 pg; B. llaaun B-42, 20 pe; M. r'arran-r CCM 2431, 7 pg; CCM 2492, 13 rrg; CCM 825, 6 pg; CCM 552, '10 pg; CCM 2490,40 pg.

cocci, since the normal environments of these bacteria are soil, water, and human and animal

skin [13]. Micrococci are also unlikely to have

been exposed to thc same kind of selective pressures as clinical bacteria. For these, the

widespread occurrence of mercury resistance seems to be maintained by selection for genetically linked antibiotic resistance genes. Thus, as a rule these mercury-resistant clinical isolates are simultaneously resistant to at least one and commonly a number of antibiotics [10,11,14]. This is not so for Micrococcas in this study: Table 1 shows the members of M. lylae, M. luteus, M. uarians and M. roseus to be mainly sensitive to antibiotics: the strains of the M. nishinomiyaensis, M. kristinae and M. sedentorius species, though

mer operon derivatives, with thc rcgion that encodes the transport component possibly being inactive or deleted to some dcgree. The value for the frequency of cryptic ffc',' ()perons suggested by this study could be an underestimate bccause cryptic mer operons that failed to express MerA would not have been detected. In contrast with the proposed loss of transport function in most examples of cryptic mer operons seen in this study, in several cases a different or additional mer operon defect may occur. In M. lylae CCM 2694, M. luteus CCM 559 and M. uarians CCld 2492 and 2437 approximately nor-

mal quantities of MerA protein werc observed (Fig. 1), the corresponding enzymatic activities were inducible but were cxtremely low (Table 1). These low activities were found not to result from the presence of MerA inhibitors, tested by examining the effect of the extracts on MerA activity of M. kristinae JM8 (data not shown). A possible explanation for the data is a defect in the MerA

of these strains. Thc phenotypically cryptic nc,' opcrons of Micrococcus, BreL'ibacteritrm and possibly Rhodococcus may be analogous to othcr cryptic genes and

for example, the determinants for utilization of B-glucosides of E,nterobacteriaceae. Arbutin-spccific phospho-B-glucosidase activity

operons,

has been found in many wild-type Escherichia coli strains. but these strains cannot use arbutin bccause they do not express the transport and phosphorylation genes (arbT or bgl operon) nccessary

for arbutin utilization. -lhe arbT gcne or

bgl

operon can be activatcd by a variety of genetic

1(X)

events. wlrich then allows growth of E. coli on arbutin [16 l9]. This suggested the possibility that thc cryptic r/r(,r' opcrons coulcl be reacl.ivated

by one or two rrutational cvcnts. On thc othcr hand. these cr1'ptic rircr opcrons ntay havc clcvolvcd t