FEMS Microbiology Reviews 11 (1993) 145-152 145

0 1993 Federation of European Microbiological Societies 016%6445/93/$1.5.00

Published by Elsevier

FEMSRE 310

Microbial retention of mercury from waste streams

in a laboratory column
containing vzerA gene bacteria
M. Brunke, W.-D. Deckwer, A. Frischmuth, J.M. Horn, H. Liinsdorf, M. Rhode, M. RChricht,
K.N. Timmis and P. Weppen

GBF National Research Institute for Biotechnology, Braunschweig, FRG

Abstract: The microorganisms used for the mercury retention experiments were natural isolates and genetically engineered
bacteria. All mercury-resistant strains contained the merA gene. Column experiments with these strains were carried out by
immobilizing them on different support materials. To obtain kinetic data of the reductase activity for whole cells and the crude
extract, batch experiments were carried out under different conditions.

Key words: Mercury resistance; merA gene; Immobilized microorganisms

Introduction (erosion, volcanic eruptions) and various anthro-

Elimination of heavy metals from dilute waste
streams continues to require great efforts as a
result of their impact on the environment. Micro-
biology and biotechnology offer different ap-
proaches to eliminate heavy metals from dilute
aqueous media which are depicted elsewhere [1,2].
Removal of highly toxic elements like mercury is
expected to provide sufficient net profit to sup-
port even the development of unconventional
biotechnological processes using active microbial
metabolism. A selection of microbial interactions
with metals is given in Table 1. Mercury is dis-
tributed into the environment by natural events
Correspondence to: M. Brunke, GBF National Research Insti-
tute for Biotechnology, Mascheroderweg 1, D-W-3300 Braun-
schweig, FRG.
pogenic activities. For instance, off-gas from waste
incineration plants and power plants leads to the
emission of huge amounts of mercury which is
partly removed by gas washing facilities and is,
hence, accompanied by the production of mer-
cury-contaminated aqueous residues. Gold min-
ing causes serious problems because of mercury
contaminations in several developing countries.
Mercuric and organomercurial detoxifying en-
zymes are widespread among microorganisms and
are quite well-characterized mechanistically [3-51
and genetically [6]. Microbial communities are
subjected to adaptation stress when mercuric
compounds are present in the environment [7]. In
this study, the applicability of these detoxifying
mechanisms was investigated in order to retain
mercuric compounds from highly diluted efflu-
ents. The study aims to develop an alternative to
chemical precipitation and to evaluate the feasi-
146
Table 1
Interactions of microorganisms and metals (active processes,
energy-dependent)
• Complexation by metabolic products (Cu and Fe ions
by oxalic acid, Endothia parasitica)
• Precipitation at cell surface (H2S by Desulfol,ibrio
desulfuricans, carbonates, uranium phosphates by
Citrobacter with glycerol phosphate)
• Production of biopolymers which bind metals (proteins,
polysaccharides, (Pseudomonas sp. strains)
• Effiux mechanisms by ATP-driven ionic pump (Ni-
resistant Ak'aligenes eutrophus strains)
• Biomethylation of metals Wig, Sn, Pb) in presence of
methyl donors (anaerobic methane bacteria, methyl-
cobalamin)
• Reduction of metal ions to less toxic valencies (Hg(II),
Cr(IV))
bility of a biotechnological clean-up procedure at
low costs and with a high performance. Earlier
findings revealed a markable retention of mer-
cury by mercury-resistant bacteria in a continuous
column with immobilized microorganisms [8]. This
study focuses on kinetic information and the fea-
sibilities of an improvement of that technology. In
particular, the applicability and advantages of
genetically engineered organisms are considered.
Materials and Methods
Saccharomyces cereuisiae was immobilized in
alginate by the standard technique described by
Vorlop and Klein [9]. The alginate beads with
entrapped yeast cells were used in short columns
and successfully exposed to a continuous flow of
a mercuric chloride solution in order to enrich
natural consortia of bacteria containing the merA
gene. The immobilized yeast supported bacterial
growth and served as substrate. Isolates were
identified by DSM (Deutsche Sammlung yon
Mikroorganismen, Braunschweig) and were culti-
vated on synthetic solid and liquid media. Among
the various isolates, Aeromonas hydrophila (HGS
2) was selected for further investigations. This
strain was grown on Na-acetate 7.5 g l - l ; yeast
extract (Difco), 1.0 g l - l ; ( N H 4 ) 2 S O 4 , 4.5 g 1-1;
MgSO4" 7H20, 0.2 g 1-1; KH2PO4, 0.465 g l - t ;
N a 2 H P O 4. 2H20, 0.93 g 1 l; C a C I 2 . 2 H 2 0 ' 20
mg 1 ~; traces and F e - E D T A solution 1 ml 1 t
each [10] and Tris(hydroxymethyl)-aminomethane
12.1, g 1 f; pH was adjusted to 7.0 with HCI.
Column experiments were performed with the
same medium at 1/100 strength except for Na-
acetate, 50 mg 1-1; Piperazin, 1,4-bis-2(eth-
ane)sulfonic acid, 0.5 g 1-I and CaCI 2, 2 H 2 0 ,
1.47 g l-I (in order to stabilize alginate beads).
Mercuric chloride was added to the final concen-
tration from a filter-sterilized stock solution.
Pseudomonas putida strains were maintained and
investigated on M121 media [11].
Mercury was determined by a flow injection
system (FIAS 200, Perkin Elmer) which was linked
to an atomic absorption spectrometer (AAS 2100,
Perkin Elmer). All samples were digested by a
wet ashing procedure using a mixture of HC1 and
H N O 3. Samples were diluted to a concentration
of 2 0 - 1 0 0 / x g Hg when necessary.
Resistance genes were identified by colony and
Southern blotting using subfragments of Tn 501 of
1.8-kb and 6 kb as gene probes, respectively.
Genomic DNA was digested with AcaI and
Eco RI, respectively.
The transposon Tn501 encoding mercury resis-
tance by the rnerA gene was introduced stably
into the chromosome of Ps. putida KT2442 by
using a suicide vector. This was done by filtermat-
ing technique [6]. The structure of Tn501 is shown
in Fig. 1.
Screening of mercuric ion reductase activity
Schematic diagram of Tn501 mer-Operon
Operator-
Promoter
merR i merT merP merA merD
E l
merR affects the promoter when binding Hg 'm
merT and merP encode for transport proteins
merA encodes for mercuric ion reductase
merB (missing here) yields a lyase (broad substrate spectra)
Fig. 1. Schematic diagram of Tn501 mer operon.
 147

served to select the most promising strains out of
30 derivatives of Ps. putida KT2442. Reaction
kinetics of the mercury reductase were studied in
whole cells and crude extracts. Mercury reductase
was assayed spectrophotometrically [12]. Reduc-
tase activity of whole cells was assayed by an
atomic absorption spectrometer, which was linked
to an automatically operated minireactor [13].
Data allowed the calculation of specific activity
and reaction velocity according to the Michaelis-
Menten kinetics or the Haldane formalism by
conventional non-linear regression analysis.
Aeromonas hydrophila HGS2 and Is. putida
KT2442::Tn501-132 were immobilized on the
surface of Bi 54t (aluminum oxide-based ceram-
ics), Siran (porous glass), Biopor (silicone-based
ceramics) or by entrapment in alginate beads.
Microbial retention of mercury was studied in
fixed bed columns (10 cm length and 1.6 cm
internal diameter) in the long term. The columns
were operated at a hydraulic load of 2-4 bed
volumes per hour. The residence time of the
liquid phase ranged from 6 to 12 rain.
Scanning electron microscopy, EDAX analysis
and transmission electron microscopy were used
to localize entrapped mercury.

Hybridization of a m e r Probe to A v a I-Digested

Total Genomic DNA Isolated from Three Identified
Reactor Isolates

pJH2 : :Tn .~/- I Amromonal Xanthomonae

Tn 5 0 1 m e r Operon

, , ,,, , , , , , , m
I I I I i I
EcoRl Aval EcoRI Aval Aval EcoRl

probe (6 kb)
Fig. 2. Southern blot analysis of natural isolates using a 6-kb gene probe and AuaI-digested genomic DNA of merA isolates.
148

Colony Hybridization of 16 Reactor Isolates

O Q O

O i O i
P put/da K I 2 4 4 2 InS,O f

• P ~t~* KT2442 pUT Hq (+)

• gc,~h S M I O ~kp/t ( p U T - t t q )

"A •
P~udo~3 puhda K I 2442'
( )

pI.Jl Hq rner [)peron

~ l;>Z4>, ^ l:>Bt>
i 1 i i
! !
/co Ol [co Ol

probe(1 ~ kh)

Fig. 3. Colony blot of merA isolates using a 1.6-kb gene probe shows that the promoter gene is always present in isolated strains
and Ps. putida KT2242 :: Tn501 derivatives.

Results Mercury in the e f f t u e n t , Aerocnonashydrophi}a

Investigation of d i f f e r e n t porous support media
Enrichment and characterization of rnerA gene 200-
strains " solid support moteriol
~Biopor oSi,on oBi 5~t
The enrichment of natural consortia of rnerA
;~: . f :o* through
gene bacteria yielded more than 20 strains which
~. 120" i mg/'L 2 Mg/L it ag/~ ,5 mg/L ,I ~g,/~ 0 ~g/L 2 mg/L 5 mq/L 10 ag/L l0 mg/L
belonged to diverse genera of bacteria. Southern 8
2 B','/'h 2 [3V/h ~4 BV,'b 2 8V/h ,2 BV/h 1213V/hi2 BV/h 4 [~V/h 14 BV/h 14t~V/h
I I i I

blots of total genomic D N A from bacterial col- 5 8o: ....... !

. !' =
umn isolates revealed that Xanthomonas mal- ! i i i L i i
= 1 j ' I 1 , ' ' I
tophilia, Aerornonas hydrophila, Alcaligenes eutro- ,.3 40- , l~ , °
phus and other strains contained the rnerA gene
which encodes mercuric ion reductase (Fig. 2). 0 I' ,,,I,,'10 . . .20
. . . . . . . 30
.... 40 501. . . . 601. . . . . .710
. . . . . . . 80
. . . . . . 90 1(~ ~.10
Colony blots unambiguously showed hybridiza- time, d

tion of genomic D N A fragments which were ob- Fig. 4. L o n g - t e r m b e h a v i o r o f i m m o b i l i z e d Aeromonas hy-

tained from EcoR1 digests with a gene probe. drophila in a fixed bed column. Retention of mercury re-
mained unaffected by hydraulic load and concentration of
The colony blots were positive for 16 natural
HgCI 2, respectively. Three types of inorganic support media
isolates from enrichment experiments (Fig. 3). showed comparable results, but Biopor has the longest service
The phenotypic resistance to mercury was not life time.
 149

equally established in these strains, but the most
resistant natural isolates were capable of growth
at 50 mg Hg(II) l-1 (Table 2), when cultivated on
solid media. Related type strains obtained from
DSM only showed tolerance to mercury in some
cases. The strain Ps. putida KT2442 was sensitive
to Hg 2+ and acquired resistance only by transpo-
son mutagenesis.
Improvement of column performance
In earlier column experiments [14], the iso-
lated consortia retained 82-88% of the HgC12
input in a fixed bed column of 10 cm height. To
improve the technical applicability, the natural
consortia of resistant strains were replaced by
immobilized pure strains (merA +) or genetically
engineered strains. Growth and metabolic activity
were maintained by dilute synthetic culture me-
dia which was fed to the fixed bed column. Im-
mobilized pure strains retained 95-99% of the
mercury. Overall Hg balances yielded recoveries
of roughly 100%. The effluent concentration
never exceeded 50 /~g Hg 1-1 during several
long-term (up to 100 days) experiments. The mi-
crobial mercury trap resisted shifts of the mer-
cury concentration (1-10 mg Hg 2+ 1 1) as well as
shifts of the hydraulic load rate which ranged
from two to four bed volumes per hour (Fig. 4).
The hydraulic residence time of the bulk liquid
phase ranged from 6 to 12 min. The performance
depended only slightly on the carriers which were
used to immobilize the microorganisms. Reten-
tion of mercury was very effective in fixed-bed
columns which contained A. hydrophila immobi-

Table 2

Tolerance of natural and constructed merA s t r a i n s to m e r c u r i c c h l o r i d e ( p p m H g ( I I ) )

Strain Id-No. 1 2.5 5 7.5 10 25 50 75

Xanthomonas maltophilia HGS1 + + + + + + + -

Aeromonas hydrophila HGS2 + + + + + + + -
Alcaligenes eutrophus HGS4 + + + + + + + -
Pseudomonas paucimobilis HGS6 + + + + + + + -
(not identified) HGS9 + + + + + + + -
(not identified) HGS11 + + + + + + + -
(not identified) HGS13 + + + + + + + -
(not identified) HGS16 + + + + + + + -
(not identified) HGS20 + + + + + + + -
(not identified) HGS21 + + + + + + + -
(not identified) HGS22 + + + + + + - -
(not identified) HGS23 + + + + + + - -
Clavibacter sp. HGS10 + + + + + - - -
(not identified) HGS15 + + + + + - - -
Microbacterium sp. HGS7 + + + + . . . .
(not identified) HGS17 + + + + . . . .
Microbacterium sp. HGS8 + + + . . . . .
Bacillus sp. HGS12 + + + . . . . .
Bacillus cereus HGS14 + + + . . . . .
Microbacterium sp. HGS3 + + . . . . . .
(not identified) HGS19 + + . . . . . .

Pseudomonas putida KT2442 + + . . . . . .

Pseudoraonas putida K T 2 4 4 2 :: T n 501 + + + + + + + -

Xanthomonas maltophilia DSM 50170 + + + + + - - -

Xanthomonas maltophilia DSM 50173 + + + + + + - -
A. hydrophila ssp. hydrophila DSM 30187 + + . . . . . . .
A. hydrophila ssp. anaerogenes DSM 30188 + + + + + - - -
Alcaligenes eutrophus DSM 531 + . . . . . . .
Pseudomonas paucimobilis DSM 1098 + + + + + - - -
150

lized on Bi 54t (aluminium oxide based ceramics).
Other inorganic carrier materials, e.g. Siran (por-
ous glass), or Biopor (silicone based ceramics)
were also applicable. Alginate was the most rec-
o m m e n d e d matrix with regard to analytical moni-
toring of the bioprocess but it was ruled out for
technical application due to its poor chemical
stability.
Ps. putida K T 2 4 4 2 : : T n 5 0 1 demonstrated a
high degree of Hg tolerance and at least as much
Hg retention in columns as other isolates tested
in the same configuration. The strain showed
remarkable resistance to HgCI 2 when grown in
the presence of subtoxic concentrations of mer-
cury. In contrast, Ps. putida KT2442 showed a
typical breakthrough curve of mercury corre-
sponding with the saturation of its biosorptive
capacity.
When operated in a stirred tank, the rner r
bacteria released elemental mercury into the at-
mosphere. In the fixed-bed column [14] the 'ele-
mental' mercury formed particles of about 1-5
~ m diameter which accumulated outside the cells
and were associated with the mechanical struc-
ture of the matrix material ( T E M and SEM-stud-
ies). These particles were efficiently retained in
the column (Fig. 5). The spherical particles were
analysed with the aid of an X-ray microprobe
( E D A X ) which revealed the predominance of
mercury as a constituent of the particles. A sharp
gradient of mercurous deposits was formed at the
column inlet, and effluent data remained unaf-

Fig. 5. Scanning electron micrograph showing a spherical particle which represents the typical mode of mercury retention in the
column experiments.
 151

0
be described by a Michaelis-Menten formalism.
© ©o
o Figure 6 shows a plot of specific reductase activ-
(,l
o ~ ity vs. the available concentration of Hg(II). The
o
experimental design allowed the determination of
© o
o ~ i k~ = 170 ug Hg/1 SD=J2~ the activity without systematic errors since the
] '¢c0t : ?07 n f ~ o ~
amount of biomass was very low (approx. 1 mg
l - l ) and substrate turnover was much smaller
than the available amount.
o 5 to 15 20
i/[s] Improuement of strains
.... lobo . . . . 2~ .... ~bo . . . . ~bo . . . . Natural isolates needed cultivation at subtoxic
Hg(II), ug/l concentrations of Hg(II) to maintain resistance
Aeromonas hydrophila reoction kinetics of HgCI2 reduction and to induce expression of the detoxification
Fig. 6. In vivo r e d u c t a s e activity of Aeromonas hydrophila. mechanism. The localization of the merA genes is
unknown. Strain improvement by genetic engi-
neering yielded four families of strains. The Ps.
fected by increased hydraulic load or increased putida KT2442::Tn501-132 strain caused effi-
concentrations of mercuric chloride in the feed cient retention of mercury within the bioreactor
solution. The findings of the column experiments (column) even when it was exposed to a 6-fold
indicate that merA gene organisms show a high higher mercury concentration than the original
affinity to Hg(lI) and a high velocity of the detox- consortia isolates. A broad-range Hg resistance
ifying reactions in vivo. Effluent concentrations operon located within a mini T n 5 construct was
were usually below 50 /zg H g t o t I - t and only a randomly inserted into the Ps. putida KT242
negligible portion was identified as Hg(II). chromosome. This operon, in addition to specify-
Experimental data obtained from whole ceils ing the mercury reductase also encoded an
in suspension showed a high affinity to HgCI 2 organomercurial lyase which enzymatically cleaves
( K m) and a high specific activity (V~t), which can Hg from many organic compounds. After screen-

Table 3
P r o p e r t i e s of g e n e t i c a l l y o p t i m i z e d strains and a typical n a t u r a l isolate

Aeromonas Ps. putida Ps. putida Ps. putida

hydrophila ~ KT2442::Tn501-132 KT2442 :: p U T - H g - 7 3 FI:: p U T - H g - I - 1
Mercury resistance Inducible Inducible Constitutive Constitutive
L o c a l i s a t i o n of mer r n.t. Chromosome Chromosome Chromosome
Substrate spectra n.t. Narrow Broad Broad
D e g r a d a t i o n of P M A
organic r e s i d u e Not e x p e c t e d Not e x p e c t e d Probable
Antibiotics resistance n.t. Rif + Rif ÷ Ap +
M T C on solid m e d i a :
HgCI z 5 0 / x g ml - 1 4 0 / z g ml - I 45/~g ml- t 75 p.g ml i
PMA n.t. n.t. 80 ,u.g ml - 1 7 5 / z g ml - t
Specific activity
o f r e d u c t a s e a'b 1 0 7 n m o l s -1 mg -1 2 0 0 n m o l s - I mg t 760nmols-I rag.-1 680nmols-I mg - i
Affinity c o n s t a n t ,,b 860 nmol 1- I n.t. n.t. n.t.

n.t., not t e s t e d or u n k n o w n ;
MTC, m a x i m u m c o n c e n t r a t i o n t h a t allows significant g r o w t h of colonies.
D a t a on A. hydrophila must not be directly c o m p a r e d with others, b e c a u s e a d i f f e r e n t e x p e r i m e n t a l m e d i u m was applied.
h ln-vivo r e d u c t i o n of H g ( l I ) by cell s u s p e n s i o n s ( A A S - m e t h o d ) , d a t a w e r e c a l c u l a t e d by n o n - l i n e a r r e g r e s s i o n analysis of raw
analytical data, using e i t h e r a M i c h a e l i s M e n t e n f o r m a l i s m (A. hydrophila) or a H a l d a n e f o r m a l i s m (Ps. putida).
152
ing for microorganisms with'a high mercury resis-
tance, a strain was obtained which showed a
10-fold higher mercury resistance and a 5-fold
higher mercury reductase activity compared to
the Tn501 derivative described above. Table 3
summarizes the kinetic data of the different
strains with regard to their Hg(Il) transformation
properties.
Conclusions
The experimental results from genetic and en-
zymatic studies indicate that reduction of Hg(lI)
to elemental Hg could be the predominating
mechanism accounting for Hg retention by fixed
bed columns containing immobilized rnerA bacte-
ria. The mechanism of Hg ~ retention in the
columns is not yet clarified but did not depend
significantly on the immobilization matrix. There
were a number of indications for the formation of
particulate elemental mercury droplets ( E D A X -
microprobe).
The performance of Hg retention increased
with the replacement of natural bacterial consor-
tia by pure strains. Genetic instabilities could be
surmounted by insertion of resistance genes into
Ps. putida KT2442 by transposon mutagcnesis.
Strains showing constitutive expression of the
mercury reductase resulted from the insertion of
the merA genes downstream a strong host pro-
moter into the bacterial chromosome. These
strains showed the highest activity of the detoxify-
ing enzymes in vivo. The broad spectra strains
were capable of degrading and reducing
organomercurials (PMA) as well as Hg(II).
References
1 Macaskie, L.E. and Dean, A.C.R. (1989) Microbial
metabolism, desolubilization and deposition of heavy met-
als. In: Biological Waste T r e a t m e n t CA. Mizrahi, Ed.), Vol.
12, pp. 160--194. Alan, R. Liss, New York, NY.
2 R6hrichl, M., Weppen, P. and Deckwer W.-D. (1990)
A b t r e n n u n g wm Schwermetallen aus Abwasscrstr6men--
Biosorption im Vergleich zu herk6rnmlichen Verfahren.
Chem. Ing. Tech. 62, 582 583.
3 Miller, S.M., Ballou, D.P., Massey, V., Williams, C.tl. Jr.
and Walsh, C.T. (1986)Two-electron reduced mercuric
reductase binds tlg(ll) to the active site dithiol but does
not catalyze Hg(ll) reduction. J. Biol. Chem. 261. 6081
6(184.
4 Schottel, J.L. (1978) The mercuric and organomercurial
detoxifying enzymes from a plasmid-bearing strain of Es-
chcrichia coil, J. Biol. Chem. 253, 4341 4349.
5 Philippidis, G.P., Schottcl, J.L. and Hu, W.-S. (1991) A
model for mercuric ion reduction in recombinant Es
cherichia coll. Biotech. Bioeng. 37, 47 54.
6 llerrero, M., de Lorenzo, V. and Timmis, K.N. (1990)
Transposon vectors containing non-antibiotic resistance
selection markers for cloning and stable chromosomal
insertion of foreign genes in gram-negative bacteria. J.
Bacteriol. 172, 6557-6567.
7 Barkay, T. and Olson, B.|t. (1986) Phenotypic and geno-
typic adaptation of aerobic sediment bacterial communi-
ties to mercury stress. Appl. Environ. Microbiol. 52, 403-
4116.
8 Deckwer, W.-D., Frischmuth, A. and Weppen, P. (1990)
Patent DE 39111 217.4 GBF.
9 Vorlop, K.D. and Klein, J. (1983) New developments in
the field of cell immobilization---fl~rmation of biocatalysts
by ionotropic gelation. Enzyme Technology Berlin. (R,M.
Lafferty, Ed.), Springer Verlag, Berlin, pp. 21%235.
1(I Rippka, R., Deruelles, J.. Waterbups. J.B., tlerdmann, M.
and Stanicr, R.Y (1979) Genetic assignment, strain histo-
ries and properties of pure cultures of cyanobacteria. J.
Gen. Microbiol. 111. I 61.
11 Kreuzer. K., Pratt, C. and Torriani. A. (1975) Genetic
analysis of regulatory mutants of alcaline phosphatase
from E. coll. Genetics 81,459-468.
12 Fox, B. and Walsh, C.T. (1982) Mercuric r e d u c l a s e -
purification and characterization of a lransposon-encodcd
flavoprotein containing an oxidation-reduction-active
disulfide. J. Biol. Chem. 257, 2498 2503.
13 R6hricht, M., (1992) Dr.-lng. thesis, Universitfit Essen.
14 Frischmuth, A., Weppen, P. and Deckwer, W.-D. (1991)
Quecksilberentfernung aus w:,isscrigen Medien durch ak-
live mikrobielle Prozesse. BioEngineering 7, 38 48.