RESEARCH ARTICLE

Persistence of bacterial proteolytic enzymes in lake ecosystems

Bartosz Kiersztyn, Waldemar Siuda & Ryszard J. Chrost
Department of Microbial Ecology, Institute of Botany, University of Warsaw, Warsaw, Poland

Correspondence: Ryszard J. Chrost,

Microbial Ecology Department, Institute of
Botany, University of Warsaw, ul.
Miecznikowa 1, 02-096 Warsaw, Poland.
Tel./fax: +4822 554 1413; e-mail:
chrost@biol.uw.edu.pl
Received 26 May 2011; revised 19 November
2011; accepted 30 November 2011.
Final version published online 9 January
2012.
DOI: 10.1111/j.1574-6941.2011.01276.x
Editor: Riks Laanbroek

MICROBIOLOGY ECOLOGY

Keywords
extracellular enzymes; enzyme stability;
protein degradation; lakes.

Abstract
This study analyzes proteolytic enzyme persistence and the role of dead (or
metabolically inactive) aquatic bacteria in organic matter cycling. Samples from
four lakes of different trophic status were used. Irrespective of the trophic status of the examined lakes, bacterial aminopeptidases remained active even 72 h
after the death of the bacteria that produced them. The total pool of proteolytic enzymes in natural lake water samples was also stable. We found that the
rates of amino acid enzymatic release from proteinaceous matter added to preserved lake water sample were constant for at least 96 h (r2 = 0.99, n = 17,
P
0.0001, Vmax = 84.6 nM h1). We also observed that proteases built into
bacterial cell debris fragments remained active for a long time, even after the
total destruction of cells. Moreover, during 24 h of incubation time, about
20% of these enzymatically active fragments adsorbed onto natural seston particles, becoming a part of the attached enzymes system that is regarded as the
hot-spot of protein degradation in aquatic ecosystems.

Introduction
It is well documented that a majority of bacteria present
in natural aquatic ecosystems might be metabolically
inactive (dormant cells) or dead (Stevenson, 1978; Zweifel
& Hagstrom, 1995; Ouverney & Fuhrman, 1999; Luna
et al., 2002). Various methods for the separation of metabolically active bacteria from metabolically inactive ones
have been developed (Table 1). Results obtained using
these methods showed that the share of dead (or inactive)
bacterial cells in both marine and freshwater environments ranged from a few to several dozen percent (Berman et al., 2001; Adamczewski et al., 2010). In contrast
to the relatively high level of information concerning
many aspects of metabolic activity of living bacteria, and
their role in the microbial loop (Grossart et al., 2001;
Herndl et al., 2008), there is much less research focused
on the influence of dead bacteria debris on the active bacteria community. In particular, the stability of aquatic
bacterial enzymes and their possible role in the cycling of
organic matter is poorly elucidated. One of the few
exceptions is the early work of Vives-Rego et al. (1985),
who pointed to the possibility of the presence of catalytically active dead bacterial cell fragments in natural water.
2011 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved

Later studies focused on the measurement of enzyme

activity in preserved natural water samples suggested that
enzymes produced by bacteria remained active even after
their death (Haemejko & Chrost, 1986). However, up to
now, there have been no detailed analyses of enzymes
from dead bacteria treated as an important, active element of the organic matter transformation system in
lakes.
In this study, we attempt to answer the following
questions: (1) how long do hydrolytic enzymes attached
to bacterial debris remain active after the death of the
bacteria, which have produced them? (2) Is the integrity of the bacterial envelope necessary for the functioning of this group of enzymes? (3) Are active enzymes
built into fragments of bacterial walls and membranes
able to adsorb on natural seston surfaces? (4) What
kind of role may these types of enzymes play in aquatic
ecosystems? We chose extracellular leucine-aminopeptidase (LAP) as a model bacterial proteolytic enzyme.
There were three main reasons for this choice. First,
LAP is one of the best known and most studied
aquatic bacterial extracellular enzymes (Somville & Billen, 1983; Fontigny et al., 1987; Chrost, 1991; Hoppe
et al., 1993; Martinez et al., 1996); second, it is proFEMS Microbiol Ecol 80 (2012) 124134

125

Persistence of extracellular enzymes in lakes

Table 1. The most commonly used methods to distinguish metabolically active bacteria from metabolically inactive (or dead) bacteria
Target

Basis of measurement

Reference

Electron transfer system

Intracellular reduction of CTC

(5-cyano-2,3-ditolyl-tetrazolium chloride)
Anionic or cationic membrane-specific
dye staining (Sytox GreenR)
LIVE&DEAD kit (Invitrogen, Molecular Probes)
mRNA detection, rRNA detection
Microautoradiography

Rodriguez et al. (1992)

[3H]methyl-thymidine incorporation rates

DNA-DAPI staining followed by propanol washing

Fuhrman & Azam (1982)

Zweifel & Hagstrom (1995)

Membrane potential
Membrane integrity
Translation, transcription
Active membrane transport
(substrate uptake)
Bacterial DNA synthesis
Presence of nucleoid

duced by almost all species of aquatic bacteria (Chrost,

1991); and third, in aquatic ecosystems, extracellular
aminopeptidase is produced mainly by heterotrophic
bacteria (Martinez et al., 1996).

Suller & Lloyd (1999)

Luna et al. (2002)
Karner & Fuhrman (1997)
Ouverney & Fuhrman (1999)

3.9 mL of water samples, 0.1 mL of appropriate substrate

L-leucine-4-methyl-cumarinylamide hydrochloride (LeuMCA) solutions in ethanol was added, yielding final LeuMCA concentrations of 0.25, 0.5, 1.0, 2.5, 5.0, 7.5, 10.0,
12.5, and 15.0 lM. Separate stock solutions for each LeuMCA concentrations were prepared. The same final concentrations of ethanol were used for the preparation of
calibration standard solutions for fluorescence product
measurement (MCA). Fluorescence of the product
7-amino-4-methylcoumarin (AMC) was determined spectrofluorometrically (380 nm Ex. and 460 nm Em.) in a
Shimadzu RF 1501 spectrofluorometer at zero time and
after 0.51.0 h of sample incubation (temp. 20 C). The
tested enzyme-substrate system followed first-order
MichaelisMenten kinetics. The plot of the reaction
velocity (v) against substrate concentration [S] displayed
a rectangular hyperbola relationship, described by the
equation v = Vmax 9 [S]/(Km + [S]). Nonlinear regression analysis was applied to calculate the kinetic parameters of enzymatic reactions by means of PC software
Origin 6.1 (OriginLab Corporation, Northampton). Total

Materials and methods

Study area and sampling

Samples were collected from lakes of different trophic status presented in Table 2. Subsamples (3 9 2 L) of pelagic
water were collected from the surface (0.5 m) and from
different depths (4, 8, 10 and 16 m) of the studied lakes.
The three subsamples, taken from each depth of the studied lakes, were mixed together (v/v) and transported in
the polyethylene containers to the laboratory in < 3 h.
Leucine-aminopeptidase activity

Maximal potential leucine-aminopeptidase activity (Vmax

LAP) was measured fluorometrically (Chrost, 1990). To
Table 2. Basic characteristics of the studied lakes
Lake GPS*
position
Kuc
5349N
2124E
Tatowisko
5352N
2134E
Mikoajskie
5347N
2135E
Zegrzynskie
5247N
2106E

Depth (m)
Area (km2)

Average

Maximum

Trophic Status

TSI

Secchi disc (m)

Chlorophyll a (lg L1)

Ptot(lmol P-PO43 L1)

0.99

8.0

28.8

Mesotrophic

38 2

4.5 0.2

3.3 1.4

0.3 0.1

3.27

14.0

39.5

Hypereutrophic

70 4

0.5 0.1

87.3 7.2

2.8 0.4

4.98

11.2

25.9

Eutrophic

55 5

1.4 0.6

33.1 3.6

1.9 0.4

33.00

3.0

15.0

Hypereutrophic

72 2

0.3 0.1

59.7 1.6

3.12 0.6

*Global Positioning System.

Trophic State Index according to Carlson (1977).

Total phosphorus.

FEMS Microbiol Ecol 80 (2012) 124134

2011 Federation of European Microbiological Societies

Published by Blackwell Publishing Ltd. All rights reserved

B. Kiersztyn et al.

126

Vmax LAP was measured in nonfiltered samples (total

fraction). Free dissolved Vmax LAP activity (fraction
< 0.2 lm) was measured after filtration through 0.2-lm
polycarbonate (Nuclepore) filters. Vmax LAP of the freeliving bacteria fraction (0.21 lm) was calculated as the
difference between LAP activity measured after filtration
through 1.0-lm filters (polycarbonate, Nuclepore) and
LAP activity in the 0.2-lm filtrate. Vmax LAP of attached
bacteria and enzymes trapped in biofilms surrounding
detritus (fraction > 1 lm) was calculated as the difference
between total Vmax LAP activity and Vmax LAP activity in
the fraction < 1 lm.
Number of bacteria and bacterial production

The number of bacteria (BN) was determined by direct

counting of cells on 0.2-lm, black polycarbonate membrane filters (Millipore) under epifluorescence microscope
according to Porter & Feig (1980). DAPI (4,6-diamidino2-phenylindole) in final concentration 1 lg L1 was used
for bacteria staining (10 min, temp. 24 C). For bacterial
counting, a computer image analyzing system composed
of a Nikon epifluorescence E450 microscope, Nikon Digital Camera DXM 1200F and LUCIA 4.8 software (Laboratory Imaging Ltd, Czech Republic) was used. The bacteria
were counted from digital images of 1030 random fields
of each membrane filters (from 50 to 100 bacteria per
field, 10003000 bacteria cells per each membrane filter,
picture area: 5510 lm2, UV2A Nikon fluorescence filter
Ex. 380420 nm, DM. 430 nm, Em. 435485 nm). The
total number of bacteria was determined after mild sonication of the sample with addition of sodium pyrophosphate (56 s impulses, 300 W; sonication wavelength
amplitude: 22 lm; sodium pyrophosphate concentration,
50 mM) according to Griebler et al. (2001). The number
of free-living bacteria was determined in the sample fraction < 1 lm (filtered through a 1-lm polycarbonate filter). The number of attached bacteria was calculated as
the difference between total number of bacteria and number of free-living bacteria.
Bacterial production was measured by means of tritiated thymidine (TdR, 9097.5 Ci nmol1; NEN Du
Pont) incorporation according to Chrost & Rai (1993).
Metabolically active bacteria

To determine the share of bacteria with intact cell membranes in total number of bacteria, LIVE/DEAD BacLightTM Bacterial Viability Kit, Invitrogen Molecular
Probes, was used according to Invitrogen, Molecular
Probes standard procedure (Luna et al., 2002). The share
of bacteria with active oxidation chains was estimated by
CTC (5-cyano-2,3-ditolyl-tetrazolium chloride) reduction
2011 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved

test (Creach et al., 2003). The bacteria were stained in

unpreserved samples, two hours after the samples were
taken. After staining (both CTC test and LIVE/DEAD
test), the bacteria were collected on 0.2-lm black polycarbonate membrane filters (Millipore), dried, and the filters
were mounted in BacLight mounting oil on microscopic
slides. CTC+ bacteria and LIVE/DEAD bacteria were
stained on separate membrane filters. The bacteria were
counted immediately (CTC test) or after a 1-week storage
in 25 C (LIVE/DEAD test). We used the same counting methodology as for the DAPI-stained samples
described above. For counting CTC+ bacteria, we used
B3A Nikon filter (Ex. 430490 nm, DM. 505 nm, Em.
> 520 nm) and for LIVE/DEAD bacteria, B2A Nikon filter (Ex. 450490 nm, DM. 500 nm, Em. > 515 nm). We
calculated the participation of CTC+ and MEM+ bacteria
in the total DAPI visible bacteria number in the vertical
profile of the eutrophic Lake Mikoajskie, which we treat
as a model system of changes in the proportion of active
bacteria.
Concentration of amino acids and
oligopeptides

The summarized concentrations of free amino acids

(FAA) and short oligopeptides (SOP) were measured by
the o-phthaldialdehyde method (Roth, 1971), modified
by Siuda et al. (2007) as follows. Triplicate water samples
(3.9 mL) were supplemented with 0.05 mL of borate buffer (0.4 M, pH = 9.0) and 32 lL of 2-mercaptoethanol,
mixed and, after the addition of 0.05 mL of o-phthaldialdehyde (OPA) solution (1.25 mg mL1 in ethanol),
immediately mixed again. The final concentrations of the
reagent in each replicate were 1.0 lM, 114.4 mM, and
116.5 lM. The fluorescence of the samples was measured
in a Shimadzu RF 1501 spectrofluorometer (Ex. 330 nm,
Em. 455 nm) within 15 min. A calibration curve was
obtained by measurement of L-leucine standard concentrations (0.1, 0.5, 1.0, 1.5, 2.0, 2.5 lM). We did not
observe any significant influence of NH
4 ions on the
method used.
Analysis of enzyme stability

For the analysis of enzyme stability in preserved samples,

sodium azide (Formally, sodium azide is rather an inhibitor that a total preservative. Nevertheless, sodium azide is
frequently used as a preservative in many reagents and
stock solutions that are utilized in science laboratories
and healthcare facilities. Although we are aware that the
term preservative is not fully precise, we decided to use
it for simplicity.) in final concentration 0.3% was used.
The following criteria motivated us to choose sodium
FEMS Microbiol Ecol 80 (2012) 124134

127

Persistence of extracellular enzymes in lakes

azide as a fixing agent: (1) even in low (0.3%) concentrations, it completely inhibits the metabolic activity of
aerobic microorganisms (Touminen et al., 1994; Arunmozi et al., 1997); (2) sodium azide does not damage cell
walls or cell membranes and therefore prevents the rapid
leakage of cellular content into the external environment
(Maselli et al., 2002); (3) sodium azide does not directly
affect the activity of proteolytic enzymes (Guellil et al.,
2001).
Analysis of LAP sorption and stability

Sorption of proteolytic enzymes bound to the fragments of

bacterial envelopes was conducted using a natural water
sample (2 L) taken, under nonsterile conditions, from the
surface layer (1 m depth) of highly eutrophied Lake
Zegrzynskie (Table 2). The sample was divided into two
subsamples (for the scheme of sample preparation see
Fig. 1). One of them was autoclaved twice (134 C,
20 min) to completely inactivate bacterial enzymes and
then filtered through a 1-lm polycarbonate filter (Nuclepore). Subsequently, thermally inactivated seston collected
on the 1-lm filter was resuspended in the autoclaved
(enzymatically inactive), seston-free (filtered through
0.2 lm) lake water (AW lake water filtered through
0.2 m membrane filter, autoclaved and filtered again
through another 0.2 m membrane filter). By this procedure, a suspension of natural seston particles larger than
1 lm deprived of proteolytic activity was obtained (S
enzymatically inactive seston particles larger than 1 m
suspended in AW water). The second subsample was sonicated in ice (10 impulses, 30 s each, 0.8 kw, 40 lm) to
completely disintegrate all microbial cells. After sonication,
two portions of the subsample were filtered through
0.2-lm and 1-lm polycarbonate filters to obtain filtrates
containing enzymatically active fragments of cells smaller
than 0.2 lm (E0.2 free proteolytic enzymes and enzymes
in cell debris smaller than 0.2 m) and smaller than 1 lm
(E1.0 free proteolytic enzymes and enzymes in cell debris
smaller than 1 m), respectively. Finally, six samples (variants of the experiment) in three repetitions were prepared:
(I) lake water + AW (mixed v/v): control of enzymes activity in natural lake water; (II) E0.2 + AW (mixed v/v): testing the activity of < 0.2 lm in size fraction of cell
fragments containing active proteolytic enzymes; (III)
E1.0 + AW (mixed v/v): testing the activity of smaller than
1.0 lm size fraction of cell fragments containing active
proteolytic enzymes; (IV) E0.2 + S (mixed v/v): testing the
sorption efficiency of < 0.2 lm in size enzymatically active
fragments of bacteria cells on larger inactive seston particles; (V) E1.0 + S (mixed v/v): testing the sorption efficiency of smaller than 1.0 lm enzymatically active
fragments of bacteria cells on larger inactive seston partiFEMS Microbiol Ecol 80 (2012) 124134

cles; and (VI) S + AW (mixed v/v): control of activity of

enzymatically inactive seston particles. All samples were
preserved with sodium azide (final conc. 0.3%) and incubated for 48 h at 20 C. After incubation, Vmax of LAP was
measured in samples I, II and III. Samples IV, V, VI were
filtered through 1.0-lm filters. The seston collected on the
filters was resuspended in AW. In the suspension, Vmax of
LAP bound to cell debris size fragments smaller than
0.2 lm (variant IV), and fragments < 1 lm in size (variant
V), and adsorbed on the seston particles larger than 1 lm
(initially lacking LAP activity) was measured.
Analyses of water samples in vertical profiles

The share of metabolically active bacteria in the whole

bacterial community was evaluated in samples taken from
the vertical profile (depth 0.5, 4, 8, 10, 16 m) of the
eutrophic Lake Mikoajskie during the summer stratification period (July).
Analyses of LAP stability in lakes of different
trophic status

LAP stability was evaluated in the samples of surface

water taken in July from lakes of different trophic status
(Lake Kuc, Lake Mikoajskie, Lake Tatowisko; Table 2).
Vmax of LAP was measured repeatedly in the samples
fixed with 0.3% sodium azide over a period of 72 h.
Leucine-aminopeptidase persistence in
bacterial size fractions

To compare LAP persistence in the fraction of free-living

bacteria (smaller than 1 lm) and the fraction predominated by attached bacteria colonizing seston particles larger than 1 lm, samples of surface water from highly
eutrophied Lake Zegrzynskie were collected. Samples were
fixed with 0.3% sodium azide and then incubated for
8 days at 20 C. After fixation, maximal potential LAP
activity in the fraction of free-living and attached bacteria
was measured several times during the period of incubation (8 days). As controls, we used unfixed samples treated in the same way.
Stability of proteases in lake water

To determine the stability of various (not only LAP) proteases present in the lake water, a methodology developed
by the authors and described in detail by Siuda et al.
(2007) was used. Three 2-liter (in three repetitions) samples taken nonsterilely from the surface water (0.5 m
depth) of the eutrophic Lake Mikoajskie were preserved
with sodium azide (0.3%), enriched with High Powder
2011 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved

B. Kiersztyn et al.

128

Fig. 1. Scheme of the samples preparation for analysis of LAP sorption and stability.

Azure (HPA, Sigma) to the final concentration of

0.23 mg HPA mL1 and incubated for 91 h at 20 C.
HPA (Hide Powder Azure, also known as Remazol Brilliant Blue RHide) is a hide powder (mainly scleroproteins) covalently linked to Remazol Brilliant Blue dye (for
spectrophotometry). It is a general substrate for proteases,
including trypsin, commonly used as a substrate for a
wide spectrum of bacterial extracellular exo- and endoproteases. We have used HPA from Sigma, catalog number 89435. The samples were supplemented with HPA to
avoid the reduction of enzymatic hydrolysis rate that
would have been caused by a decrease in natural protein
concentration during the incubation period. In subsamples taken during incubation, the increase in the concentration of amino acids and SOP enzymatically liberated
2011 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved

from HPA was measured by OPA method modified by

Siuda et al. (2007). Autoclaved samples supplemented
with HPA and treated the same way as the experimental
samples were used as a control for no enzymatic release
of amino acids.
All measurements and determinations were done with
triplicate subsamples.

Results
Metabolically active bacteria in the water
column

The analysis of the samples from the vertical profile of

the eutrophic Lake Mikoajskie showed that only about
FEMS Microbiol Ecol 80 (2012) 124134

Persistence of extracellular enzymes in lakes

12% (12.7% LIVE/DEAD test, 12.1% CTC test) of

the bacteria present in the surface water and from 3% to
4% (4.01% LIVE/DEAD test, 3.05% CTC test) of the bacteria present in the hypolimnion were metabolically active
(Fig. 2). Vertical changes in the number of potentially
active bacteria obtained by the two independent methods
(CTC and LIVE/DEAD test) were strongly and positively
correlated (r2 = 0.97, P
0.005). The total number of
bacteria was relatively constant (depth: 0.5 m
6.6 0.3*106 mL1, 4 m 6.8 0.24*106 mL1, 6 m
7.0 0.2*106 mL1, 10 m 7.2 0.3*106 mL1, 16 m
7.2 0.3*106 mL1). The maximal potential aminopeptidase activity was relatively constant throughout the whole
vertical profile of the lake (from 199 39.5 at the depth
of 0.5 m to 158.53 29.8 nM h1 at the depth of 16 m)
and did not follow the rapid decrease in the bacteria CTC
+ and MEM+ (intact cell membrane) abundance (Fig. 2).
This suggested that aminopeptidase activity might not be
related only to living, metabolically active bacterial cells
but also to dead and/or inactive bacteria.
Persistence of aminopeptidase activity in lake
water

This interpretation was confirmed by the model experiments. We found that, irrespective of the trophic status

Fig. 2. Percentage of contribution of bacterial cells with active

oxidation chain (CTC+) and cells with intact cell membranes (MEM+)
in the total bacteria number (bars) and maximal potential
aminopeptidase activity Vmax LAP (curve with dots) in the vertical
profile of eutrophic Lake Mikoajskie.

FEMS Microbiol Ecol 80 (2012) 124134

129

of the examined lakes, aminopeptidase was active even

72 h after the death of the bacteria that produced it (It is
widely accepted that the major pool of extracellular proteases able to act in outer cell aquatic conditions is produced by bacteria (Chrost, 1991; Martinez et al., 1996).
Microscopic observation of the samples (data not shown)
revealed that there were no other potential sources of
extracellular proteases, like fungi or numerous cyanobacteria cells, in the studied lakes.). The largest decrease of
LAP Vmax was observed in the samples from the eutrophied Lake Mikoajskie and the smallest (to 74.5% of initial activity) in the samples from the mesotrophic Lake
Kuc (Fig. 3). We assumed that the greater decrease in
potential proteolytic activity in the highly eutrophied
lakes in comparison with the mesotrophic lake might
have been a consequence of a decrease in the activity of
enzymes associated with seston particles and attached
bacteria. We tested the latter hypothesis by comparing
aminopeptidase persistence in a fraction of free-living
bacteria and in a fraction containing bacteria and
enzymes attached to seston particles larger than 1 lm.
We found that LAP was more stable in the fraction
< 1 lm than in the fraction larger than 1 lm. After
8 days of incubation of the samples preserved with
sodium azide (0.3% final conc.), the maximal potential
LAP activity decreased from 350 to about 50 nM h1 in
the fraction of attached bacteria. In the fraction of freeliving bacteria, Vmax of LAP remained almost unchanged
during the whole time of incubation after death of the
microorganisms (Fig. 4). We did not observe any activity
of free, dissolved LAP (in the fraction < 0.2 lm), in
water samples from the examined lakes. To ascertain that
persistence is not only a specific attribute of aminopeptidase, we tested the stability of the whole pool of proteolytic enzymes in natural lake water. We found that the
rates of amino acid enzymatic release from proteinaceous
matter (HPA model substrate) added to the preserved
lake water sample was constant for at least 96 h
(r2 = 0.99, n = 17, P
0.0001, Vmax = 84.6 nM h1),
suggesting the substantial stability of this kind of
enzymes.
To confirm that the integrity of bacterial envelopes is
not a prerequisite for the activity of the enzymes, we
applied sonication prior to measuring potential aminopeptidase activity. The preparation of enzymes and sample is described in detail in the methodology chapter
(Analysis of LAP sorption and stability) and on Fig. 1.
We observed no changes in Vmax LAP in the fraction
smaller than 1 lm (E1.0, see Fig.1) and an about 75%
decrease in LAP activity in the fraction smaller than
0.2 lm of sonicated samples (E0.2, see Fig. 1) compared
to the nonsonicated samples (Fig. 4a). After 24-h incubation of proteolytically active fragments of bacterial cells in
2011 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved

B. Kiersztyn et al.

130

Fig. 3. Time course changes of maximal potential aminopeptidase

activity (Vmax LAP) in lake water samples fixed with 0.3% sodium
azide in the studied lakes: Kuc, Mikoajskie, and Tatowisko.

Fig. 4. Time course changes of maximal potential aminopeptidase

activity (Vmax LAP) in the fraction of free-living bacteria (< 1 m) and
fraction of attached bacteria (> 1 m) in lake water samples fixed
with 0.3% sodium azide sample from highly eutrophic Lake
Zegrzynskie.

the presence of natural seston particles larger than 1 lm

(artificially deprived of proteolytic activity, see Fig. 1), we
found that 17% and 20% of the LAP activity in the sonicated fraction smaller than 0.2 (variant IV) and 1 lm
(variant V), respectively, was transferred onto the fraction
of particles larger than 1 lm (Fig. 5b). This was because
of the adsorption of active LAP bound to small bacterial
cell debris on the surface of natural seston particles.
In our analyses of water samples taken from the surface
layer of three Mazurian lakes (Lake Mikoajskie, Kuc, and
Tatowisko) during different vegetation periods (Table 2),
we did not find any correlation between the number of
bacteria and aminopeptidase activity (LAP Vmax), either
in the fraction of free-living (r2 = 0.19, n = 12, P = 0.14)
or that of attached bacteria (r2 = 0.22, n = 12, P = 0.12).
Neither did we observe any correlation between the total
2011 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved

Fig. 5. Comparison of Vmax LAP in the natural, non-sonicated,

samples of surface lake water from Lake Zegrzynskie (I) and Vmax LAP
in size fragments of cell debris < 0.2 m (II) and smaller than 1.0 m
(III) after destruction of microorganisms by sonication (a), and (b)
comparison of Vmax LAP of size fragments of bacterial cell debris
< 0.2 m (variant IV) and smaller than 1.0 m (variant V) attached
and unattached to seston particles larger than 1.0 m in size after
24 h of incubation. The percentages refer to share of initial LAP
activity transferred to the surface of seston larger than 1.0 m after
24 h incubation time. Variant VI represents the control of seston
enzymatic activity without cell debris addition.

number of bacteria and bacterial production (r2 = 0.23,

n = 9, P = 0.19). However, we found a strong positive
linear relationship between FAA + SOP concentration
and LAP activity in the fraction of attached bacteria
(r2 = 0.72, n = 12, P
0.008).

Discussion
The results of the experiments described in this article
suggest that the extracellular proteolytic enzymes present
in natural lake water are relatively stable and a large pool
of them remain active for a long time. They performed
their function for several days after the death of the bacteria that produced them. This was not very surprising
because, as opposed to the intracellular enzymes which
do not act on substrate effectively in outer cell space
(Hoffman & Decho, 2000), extracellular enzymes are
FEMS Microbiol Ecol 80 (2012) 124134

Persistence of extracellular enzymes in lakes

adapted to outer membrane conditions such as pH,

salinity, redox potential and the presence of various reaction activators and inhibitors dissolved in water. From a
biochemical point of view, all proteolytic enzymes are
proteins. Therefore, the lifetime of each protease in aquatic environments depends mainly on the rate of its enzymatic degradation by other proteolytic enzymes, which in
turn depends on the probability of contact between the
digesting and digested enzymes. It is well documented
that attached bacteria usually express very high potential
maximal aminopeptidase activity, much higher than freeliving bacteria (Becquevort et al., 1998; Unanue et al.,
1998). Because of the higher probability of enzymeenzyme contact, at least theoretically, seston-attached
enzymes should be degraded faster than the enzymes of
free-living bacteria. This assumption was confirmed by
the results of our experiments. We found that proteases
attached to the seston particles colonized by bacteria were
less stable than those produced by free-living microorganisms (Fig. 4). Moreover, enzyme stability in the highly
eutrophic lake (where the number of attached bacteria is
high) was usually much lower than in less eutrophic lakes
where free-living bacteria predominated (Fig. 3).
The idea of cross-hydrolysis as the main cause of falling protease activity was also supported by the result of
the HPA degradation experiment. The addition of HPA
to the sample containing nonliving bacteria led to an
increase in the stability of proteolytic enzymes (the rate
of enzymatic release of the amino acids from HPA was
constant for at least 91 h). We concluded that the high
HPA concentration created conditions for proteolytic
enzyme saturation, reducing the probability of physical
contact between two protease molecules and in that way
prevented their mutual destruction. It is worth noting
that the particulate substrate (HPA) was hydrolyzed without the active colonization of HPA by living bacteria.
This suggests that passive adsorption of dead bacteria
with still active enzymes is sufficient for the effective
degradation of HPA. The findings of Haemejko & Chrost
(1986), who demonstrated the enzymatic degradation of
particulate substrate in preserved water samples, indirectly
point to the same conclusions. We also found that the
integrity of bacterial envelopes was not a necessary condition for the functioning of proteases. Proteases bound to
small fragments of bacterial cell debris remained active.
These kinds of enzymes may play an important ecological
role during protein degradation after, for example, phage
lysis of bacteria. Lysis, leading to the fragmentation of
bacterial envelopes (Shibata et al., 1997), is assumed to
be one of the main reasons of bacterial mortality in aquatic ecosystems (Wommack & Colwell, 2000).
The tendency of small particles to aggregate in larger
structures (McCave, 1984; Armstrong & Barlocher, 1989)
FEMS Microbiol Ecol 80 (2012) 124134

131

may have important consequences in the context of the

enzymatic activity of dead bacterial cells. Considering that
the average size of an aquatic bacterial cell is about
0.6 lm (length) 9 0.3 lm (width) and bacterial concentration is about 109 cells L1, their total surface area
exceeds 1000 mm2 L1. Assuming that (i) the percentage
of dead bacteria in lake water is usually quite high and
(ii) proteolytic enzymes are relatively stable, dead and
metabolically inactive bacteria may constitute a huge
enzymatically active platform for protein degradation
processes in lake water.
Owing to the methodological difficulties associated
with distinguishing between enzymes attached to the surface of alive bacteria and enzymes attached to dead bacteria, it is hard to show directly which part of the total
proteolytic activity in natural lake water is associated with
dead bacterial cells or their fragments. Nevertheless, there
are reasons to assume the widespread presence of such an
activity in aquatic environments. These reasons are
related to the fact that enzymes attached to dead bacteria
and their fragments hydrolyze proteins and polypeptides
without the coupled assimilation of amino acids (end
products of catalysis).
The large share of such enzymes in natural lakes can
easily explain the following observations: (1) the lack of a
strong correlation between bacterial numbers and proteolytic activity; (2) the presence of free dissolved amino
acids, even in bacteria-abundant environments, where at
least theoretically, potential bacterial demand for carbon
and nitrogen should be high (Kiersztyn, 2005); (3) the
lack of a strong correlation between bacterial production
and protease activity; and (4) the release of FAA from
organic particles colonized by bacteria observed by Hoppe
(1991) and Smith et al. (1992). It is the latter observation, confirmed by the results of our 2-year investigations,
which we consider especially interesting. We found a
strong positive correlation (r2 = 0.70, P
0.008)
between amino acid concentration and aminopeptidase
activity in the seston particles fraction. This result suggests that amino acids are liberated from particulate matter into the surrounding water, as was proposed by
Hoppe (1991). Such a loss of FAA is disadvantageous for
attached bacteria producing proteolytic enzymes, because
they spend energy on enzyme synthesis but do not gain
the full benefits of monomer possession, as the latter are
consumed only partially. This paradoxical loss of monomers after hydrolysis is inconsistent with the idea of a
hydrolysis-uptake coupling system (Fuhrman, 1987),
which assumes that all monomers enzymatically liberated
by bacteria are immediately assimilated by them. If we
suppose that the biofilm surrounding the particles functions as a trap for dead but still enzymatically active
bacteria and bacterial cell debris, the release of amino
2011 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved

B. Kiersztyn et al.

132

acids from the seston particles into the surrounding environment can be easily explained.
The results of our experiments lead us to propose a
division of the dead bacterial proteolytic enzyme pool
into two fractions. The first pool slow but stable proteases includes the enzymes of free-living bacteria and
those connected with small (<1.0 lm) bacterial cell debris
suspended in the environment. Proteases in this fraction
usually are characterized by low potential maximal proteolytic activity (not exceeding 100 nM h1) but are relatively stable and almost constantly active even for many
days after the death of their producer. In our opinion,
this type of proteases is probably responsible for the slow
but constant release of amino acids from proteins. The
second pool fast but unstable proteases fraction includes

attached enzymes, characterized by high activity, usually

many times greater than that associated with the slow
but stable proteases pool. The activity of attached proteases rapidly decreases after the death of their producers
but may be responsible for short bursts of monomers
instantly released from proteins. It is worth noting that
the potential maximal activity of enzymes within the fast
but unstable fraction decreased to the level of the freeliving bacteria fraction. We suppose that this low level of
Vmax of LAP (a few dozen nM h1) corresponds to such
an enzyme concentration, in which the probability of
enzyme-enzyme contact becomes sufficiently low to inhibit mutual hydrolysis of proteolytic enzyme molecules. In
such a situation, the enzymes can remain stable and
active for a long time.

Fig. 6. The hypothetical role of dead bacteria in the hydrolysis of proteins in lake ecosystems.

2011 Federation of European Microbiological Societies

Published by Blackwell Publishing Ltd. All rights reserved

FEMS Microbiol Ecol 80 (2012) 124134

133

Persistence of extracellular enzymes in lakes

We would like to stress that proteolytic enzymes

attached to dead bacteria and their fragments seem to be
an important element of the ecosystem effectively enriching aquatic environments with easily utilizable FAA by a
variety of aquatic microorganisms, not only heterotrophic
bacteria. Amino acids liberated by these enzymes may not
only stimulate bacterial growth but also, after their extracellular deamination, serve as an ammonium nitrogen
source for autotrophic phytoplankton (Berman & Bronk,
2003). We propose the conceptual model of the hypothetical role of dead bacteria in the hydrolysis of proteins in
lake ecosystems (Fig. 6). Processes such as autolysis and
viral lysis, sloppy feeding of zooplankton, lethal mutation,
solubilization of fecal material or active protein excretion
enrich the aquatic environment with labile particulate
combined amino acids (PCAA) and dissolved combined
amino acids (DCAA). PCAA and DCAA aggregate or
adsorb on the surface of large particles, becoming a component of an effective particle-associated protein degradation system. Consistent with the coupling system idea,
live, active bacteria assimilate all the amino acids released
through DCAA or PCAA hydrolysis. However, according
to the results of our research, enzymes associated with
the surfaces of dead bacteria can also hydrolyze particulate organic matter (POM) or their fragments trapped in
the biofilm matrix surrounding organic matter particles.
Such enzymes (together with the trapped free extracellular enzymes) hydrolyze proteins without the parallel
assimilation of amino acids. These amino acids, after
their partial release from the particles, stimulate the
growth of free-living bacteria community. A similar
mechanism, although with less intensity, may apply to
free-living bacteria proteases, which were also found to be
relatively stable. As a conclusion, we propose to treat
dead bacteria not only as a source of nutrients but also as
an active part of the microbial food web, enriching the
aquatic environment in the monomeric, labile fraction of
dissolved organic matter. Thus, dead bacteria may be
treated, not only as a source of detritus but also as an
active catalytic component of the biogenic transformation
system, a component which can hydrolyze polymers without parallel assimilation of monomeric products.

Conclusions
Proteolytic enzymes attached to seston (dead bacteria
and their fragments) seem to be an important element
of the ecosystem, which very effectively catalyze hydrolysis of proteinaceous matter in lake water. Persistent
enzyme activity associated with dead bacteria might be
also considered a form of seemingly altruistic behavior.
Dead, but still long-lasting enzymatically active bacteria
may provide an easily utilizable pool of monomers, proFEMS Microbiol Ecol 80 (2012) 124134

viding an easy start for the next bacterial generations

(Fig. 6).

Acknowledgements
We would like to thank professor Aleksander Swiatecki
and dr. Dorota Gorniak from the University of Warmia
and Mazury for helping with the CTC experiments. The
critical comments of professor Zdzisaw Markiewicz that
helped to improve our manuscript are greatly appreciated. The authors would also like to thank the anonymous reviewers for their valuable comments. This study
was supported by the Polish Ministry of Science and
Higher Education projects: N304 023237 awarded to B.K.
and Nr N304 017540 awarded to R.J.C.

References
Adamczewski T, Chrost RJ, Kalinowska K & Skowronska A
(2010) Relationships between bacteria and heterotrophic
nanoflagellates in lake water examined by different techniques
controlling grazing pressure. Aquat Microb Ecol 60: 203213.
Armstrong S & Barlocher F (1989) Adsorption and release of
amino acids from epilitic biofilms in stream. Freshw Biol 22:
153159.
Arunmozi G, Jayavel R & Subramanian C (1997) Experimental
determination of metastable zone width, induction period
and interfacial energy of LAP family crystals. J Crystal
Growth 178: 387392.
Becquevort S, Rousseau V & Lancelot C (1998) Major and
comparable roles for free-living and attached bacteria in the
degradation of Phaeocystis-derived organic matter in Belgian
coastal waters of the North Sea. Aquat Microb Ecol 14: 3948.
Berman T & Bronk A (2003) Dissolved organic nitrogen: a
dynamic participant in aquatic ecosystems. Aquat Microb
Ecol 31: 279305.
Berman T, Kaplan B, Chava S, Viner Y, Sherr BF & Sherr EB
(2001) Metabolically active bacteria in Lake Kinneret. Aquat
Microb Ecol 23: 213224.
Carlson RE (1977) A trophic state index for lakes. Limnol
Oceanogr 22: 361369.
Chrost RJ (1990) Microbial ectoenzymes in aquatic
environments. Aquatic Microbial Ecology: Biochemical and
Molecular Approaches (Overbeck J & Chrost RJ, eds), pp.
4778. Springer-Verlag, New York.
Chrost RJ (1991) Environmental control of synthesis and
activity of aquatic microbial ectoenzymes. Microbial Enzymes
In Aquatic Environments (Chrost RJ, ed), pp. 5670.
Springer-Verlag, New York.
Chrost RJ & Rai H (1993) Bacterial secondary production.
Microbial Ecology of Lake Plubsee (Overbeck J & Chrost RJ,
eds), pp. 92117. Springer-Verlag, New York.
Creach V, Baudoux A, Bertru G & Le Rouzic B (2003) Direct
estimate of active bacteria: CTC use and limitations. J
Microbiol Meth 52: 1928.

2011 Federation of European Microbiological Societies

Published by Blackwell Publishing Ltd. All rights reserved

134

Fontigny A, Billen G & Vives-Rego J (1987) Some kinetic

characteristics of exoproteolytic activity in coastal seawater.
Estuar Coast Shelf Sci 25: 127133.
Fuhrman J (1987) Close coupling between release and uptake
of dissolved free amino acids in seawater studied by an
isotope dilution approach. Mar Ecol Prog Ser 37: 4552.
Fuhrman J & Azam F (1982) Thymidine incorporation as a
measure of heterotrophic bacterioplankton production in
marine surface waters: evaluation and field results. Mar Biol
66: 109120.
Griebler C, Mindle B & Slezak D (2001) Combining DAPI and
SYBR Green II for the enumeration of total bacteria numbers
in aquatic sediments. Int Rev Hydrobiol 86: 453465.
Grossart HP, Riemann L & Azam F (2001) Bacterial motility
in the sea and its ecological Implications. Aquat Microb Ecol
12: 247258.
Guellil A, Boualam M, Quiquampoix H, Ginestet P, Audic JM
& Block JC (2001) Hydrolysis of wastewater colloidal
organic matter by extracellular enzymes extracted from
activated sludge flocs. Water Sci Technol 43: 3340.
Haemejko GZ & Chrost RJ (1986) Enzymatic hydrolysis of
proteinaceous particulate and dissolved material in
eutrophic lake. Arch Hydrobiol 107: 121.
Herndl GJ, Agogue H, Baltar F Reinthaler ES& Varela MM
(2008) Regulation of aquatic microbial processes: the
microbial loop of the sunlit surface waters and the dark
ocean dissected. Aquat Microb Ecol 58: 5968.
Hoffman M & Decho AW (2000) Proteolityc enzymes in
marine bacterium Pseudoalteromonas atlantica: postsecretional activation and effects of environmental
condition. Aquat Microb Ecol 23: 2939.
Hoppe HG (1991) Microbial extracellular enzyme activity.
Microbial Enzymes In Aquatic Environments (Chrost RJ &
Overbeck J, eds), pp. 6083. Springer-Verlag, New York.
Hoppe HG, Ducklow H & Karrasch B (1993) Evidence for
dependency of bacterial growth on enzymatic hydrolysis of
particulate organic matter in the mesopelagic ocean. Mar
Ecol Prog Ser 93: 277283.
Karner M & Fuhrman JA (1997) Determination of active
marine bacterioplankton: a comparison of universal 16s
rRNA probes, autoradiography and nucleoid staining. Appl
Environ Microbiol 63: 12081213.
Kiersztyn B (2005) Enzymatyczny rozkad substancji
biakowych w jeziorach o roznym stopniu eutrofizacji
proba opisu, implikacje ekologiczne. PhD Thesis, University
of Warsaw, Warsaw.
Luna GM, Mannini E & Donovaro R (2002) Large fraction of
dead and inactive bacteria in coastal marine sediments:
comparison of protocols for determination and ecological
significance. Appl Environ Microbiol 68: 35093513.
Martinez J, Smith DC, Steward GF & Azam F (1996)
Variability in ectohydrolitic enzyme activities of pelagic

2011 Federation of European Microbiological Societies

Published by Blackwell Publishing Ltd. All rights reserved

B. Kiersztyn et al.

marine bacteria and its significance for substrate processing

in the sea. Aquat Microb Ecol 10: 223352.
Maselli A, Laevsky G & Knecht D (2002) Kinetics of binding,
uptake and degradation of live fluorescent (DsRed) bacteria
by Dictyostelium discoideum. Microbiology 148: 413420.
McCave IN (1984) Size spectra and aggregation of suspended
particles in deep ocean. Deep Sea Res 31: 329352.
Ouverney CC & Fuhrman JA (1999) Combined
microautoradiography 16s rRNA probe techniques for
determination of radioisotope uptake by specific microbial
cell types in situ. Appl Environ Microbiol 65: 17461752.
Porter KG & Feig YS (1980) Use of DAPI for identifying and
counting aquatic microflora. Limnol Oceanogr 25: 943948.
Rodriguez GG, Phipps D, Ishiguro K & Ridgway HF (1992) Use
of fluorescent redox probe for direct visualization of actively
respiring bacteria. Appl Environ Microbiol 58: 18011808.
Roth M (1971) Fluorescence reaction for amino acids. Anal
Chem 43: 880882.
Shibata A, Kogure K, Koike I & Ohwada K (1997) Formation
of submicron colloidal particles from marine bacteria by
viral infection. Mar Ecol Prog Ser 155: 303307.
Siuda W, Kiersztyn B & Chrost RJ (2007) The Dynamics of
protein decomposition in lakes of different trophic status
reflections on the assessment of the real proteolytic activity
in situ. J Microbiol Biotechnol 17: 897904.
Smith DC,Simon M, Alldredge AL &Azam F (1992) Intense
hydrolytic enzyme activity on marineaggregates and its
implications for rapid particle dissolution. Nature 359: 139142.
Somville M & Billen G (1983) A method for determining
exoproteolytic activity in natural waters. Limnol Oceanogr
28: 190193.
Stevenson LH (1978) A case for bacterial dormancy in aquatic
systems. Microb Ecol 4: 127133.
Suller MTE & Lloyd D (1999) Fluorescence monitoring of
antibiotic-induced bacterial damage using flow cytometry.
Cytometry 35: 235241.
Touminen E, Kairesalo T & Hartikainen H (1994) Comparison
of methods for inhibiting bacterial activity in sediment. Appl
Environ Microbiol 60: 34543457.
Unanue M, Azua I, Arrieta JM, Labirua-Iturburu A, Egea L &
Irriberri J (1998) Bacterial colonization and ectoenzymatic
activity in phytoplankton-derived model particles. Cleavage
of peptides and uptake of amino acids. Microbial and
epidemiological analysis. Microbiology 143: 30713084.
Vives-Rego J, Billen G, Fontigny A & Somville M (1985) Free
and attached proteolytic activity in water environments.
Mar Ecol Prog Ser 21: 245249.
Wommack KE & Colwell RR (2000) Virioplankton: viruses in
aquatic ecosystems. Microbiol Mol Biol Rev 64: 69114.
Zweifel UL & Hagstrom A (1995) Total counts of marine
bacteria include a large fraction of non nucleoid containing
bacteria (ghosts). Appl Environ Microbiol 61: 21802185.

FEMS Microbiol Ecol 80 (2012) 124134