ARTICLE IN PRESS

WAT E R R E S E A R C H 42 (2008) 1867 – 1878

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/watres

Protein extraction from activated sludge:

An analytical approach

M. Rasa, E. Girbal-Neuhausera, E. Paulb, M. Spérandiob, D. Lefebvrea,

a
LBAE, Laboratoire de Biologie appliquée à l’Agro-alimentaire et à l’Environnement, Institut Universitaire de Technologie,
Université Toulouse III, 24 Rue d’Embaquès 32000 Auch, France
b
Laboratory of Biosystems and Process Engineering, UMR5504 CNRS/INSA & UMR792 INRA/INSA, 135 av. de Rangueil,
31077 Toulouse Cedex 4, France

art i cle info ab st rac t

Article history: To investigate the efficiency of different methods on exopolymeric substance (EPS)
Received 24 July 2007 extraction, mechanical and chemical treatments were applied on two activated sludges,
Received in revised form regarding the yield of protein extraction as well as their compatibility with usual
12 November 2007 quantification methods.
Accepted 13 November 2007 Mechanical disruption methods do not drastically affect protein measurements by both
Available online 19 November 2007 bicinchoninic acid (BCA) and modified Lowry methods. Chemical compounds such as
cationic exchange resin and triton show high interference with modified Lowry method
Keywords:
while the protein quantification by BCA method is not affected. In addition, inner sludge
Activated sludge
compounds were shown to interfere with both methods: BCA and modified Lowry
BCA method
measurement respectively overestimate and underestimate protein content. According to
Modified Lowry method
these data, BCA method was chosen in this study as the most appropriate protein
EPS
quantification method in sludge extracts.
Extraction methods
Comparison of various extraction protocols, combining mechanical and/or chemical
Protease
treatments, shows that efficiency can be increased by repeating the same method or by
applying a prior mechanical treatment. Proteins are preferably extracted by triton
treatments, indicating the importance of hydrophobic interactions linking proteins to the
EPS matrix. The amount of extracted proteins reaches 182 and 148 mg eq. BSA g1 VSS using
triton/triton and ultraturax/triton extractions, respectively. Protease activity/extracted
protein ratios vary widely depending on extraction protocols. Protease seemed to be
preferably extracted by ultrasound and triton treatments (150–220 U mg1 protein). This
study underlines that the choice of a relevant coupled quantification/extraction method is
of great importance for efficient EPS determination.
& 2007 Elsevier Ltd. All rights reserved.

1. Introduction interactions such as hydrophobic and cationic bonds. Good

Activated sludge and membrane bioreactor account for major
wastewater treatment processes. They involve particulate
materials, i.e., flocs resulting from microbial and exopoly-
meric substances (EPS) agglutination assembled by different
sludge settleability and dehydration capacity as well as low
membrane fouling are among the main objectives to improve
the mentioned treatment processes. Since the structural
properties of the flocs and of their polymeric components
undoubtedly have a great influence on the separation process

Corresponding author. Tel.: +33 5 62 61 63 05; fax: +33 5 62 61 63 01.

E-mail address: dominique.lefebvre@iut-tlse3.fr (D. Lefebvre).
0043-1354/$ - see front matter & 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.watres.2007.11.011
 ARTICLE IN PRESS
1868 WAT E R R E S E A R C H
efficiencies, they must therefore be carefully characterised
(Ehlers and Cloete, 1999; Esparza-Soto and Westerhoff, 2001;
Garnier et al., 2005; Ramesh et al., 2006).
Floc structures and sludge properties (surface charge,
hydrophobicity) are strongly influenced by the amount and
the nature of the biopolymers (Jorand et al., 1998; Müller et al.,
2005). These biopolymers are responsible for the cohesive
forces which keep these aggregates together. Bivalent cation
binding (Bruus et al., 1992; Morgan et al., 1990) and hydro-
phobic interactions (Urbain et al., 1993; Yu et al., 2004) are
shown as predominant bonds linking the matrix compounds
together. However, nonspecific interactions such as Van der
Waals interactions and polymer entanglement are also
important (Wilén et al., 2003). The EPS pool in activated
sludge originates from microbial secretion and lysis as well as
from the incoming wastewater (Urbain et al., 1993). Therefore,
a great variety of molecules have been reported such as
polysaccharides and proteins as major constituents (Jorand
et al., 1994; Wilén et al., 2003). Humic acids (Frohlund et al.,
1995) and nucleic acids (Palmgren and Nielsen, 1996) are also
reported but in smaller quantities. Several studies have
pointed out proteins as the dominant polymeric compound
in wastewater influents (Raunjker et al., 1994) and in activated
sludge (Comte et al., 2007; Dignac et al., 1998; Frohlund et al.,
1995). Moreover, proteins seem to have a great importance in
the aggregation process (Lazarova and Manem, 1995; Martinez
et al., 2004; Urbain et al., 1993). Wilén et al. (2003) showed a
positive correlation between protein content and flocculation
ability.
Protein quantification in activated sludge and sludge
extracts shows a great variability. Direct protein measure-
ments in municipal activated sludge showed between 224 and
462 mg protein g1 VSS of sludge (Frohlund et al., 1995; Wilén
et al., 2003), but these values can vary widely depending on
sludge origin and process treatments (Watson et al., 2004).
Using cation exchange resin (CER) which is the most
popular method for EPS extraction, protein content in soluble
extracts can vary between 41 and 215 mg protein g1 VSS
depending on sludge origin (Cadoret et al., 2002; Frohlund
et al., 1995; Liu and Fang, 2002; Wilén et al., 2003). Extraction
methods using chemicals such as formaldehyde, gluteralde-
hyde or EDTA, are generally less efficient in protein extraction
compared to mechanical methods. Between 2 and 4 times less
proteins are extracted (Azerdo et al., 1999; Comte et al., 2006).
Thus, besides sludge origin, variability in protein quantifica-
tion could also be due to the type of extraction method.
Negatively charged and hydrophobic amino acids play an
important role in EPS electrostatic and hydrophobic interac-
tions (Dignac et al., 1998). Therefore, ionic (i.e. CER) or
hydrophobic (i.e. Triton) extraction methods are most
likely to extract different types of proteins, respectively,
ionic and hydrophobic proteins. Comte et al. (2007) have
recently shown that extraction methods, particularly chemi-
cal methods (formaldehyde and gluteraldehyde) extract
different types of EPS compounds compared to mechanical
methods.
Another cause of variability in protein quantification is the
choice of the measurement method. Protein quantification
methods are numerous (Bradford, 1976; Lowry et al., 1951;
Smith et al., 1985) but rely on the same copper coloured
42 (2008) 1867– 1878
complex. Colorimetric protein measurement methods (Lowry,
Bicinchononic acid, Bradford) can be subjected to interfer-
ences with other organic compounds already found in the
sludge such as humic acids and polysaccharides. Lowry
method has been modified in order to correct interferences
due to humic substances (Frohlund et al., 1995) and since
then, this method has been predominantly used for protein
quantification in sludge.
In addition, measurement methods have been compared
regarding potential interfering chemicals used in extraction
methods (Brown and Lester, 1980; Davis, 1988; Krieg et al.,
2005). Triton and EDTA which are used for protein and
enzyme extraction (Gessesse et al., 2003; Krieg et al., 2005),
show stronger interference towards Lowry measurements
compared to bicinchoninic acid (BCA) method (Smith et al.,
1985). EDTA reduces protein response by 17% and Triton
precipitates proteins with Lowry method, while the response
is unchanged with BCA method. Protein measurements in
sludge extracts have been sometimes omitted due to inter-
ferences of EDTA and glutaraldehyde with modified Lowry
method (Comte et al., 2006).
Although numerous studies report various interfering
compounds for both Lowry and BCA colorimetric methods,
few data are available concerning the use of BCA method in
activated sludge. Moreover, accounting for the complexity of
the biological matrix, it appears essential to identify the
interfering factors in aggregates in order to determine the
reliability and the accuracy of each method.
Besides the structural role of proteins in bacterial aggre-
gates, exocellular proteins also offer a hydrolytic potential
(Dignac et al., 1998) which can contribute to a more efficient
water treatment and sludge reduction (Watson et al., 2004). As
previously described, the release of enzymes from the matrix
was observed performing extraction treatment on sludge
samples (Frohlund et al., 1995; Gessesse et al., 2003; Goel
et al., 1997), and protease activity seemed to be predominant
over other investigated enzymes (Jung et al., 2002). Moreover,
commercial enzymes have also been investigated as extrac-
tion methods (Dey et al., 2005; Watson et al., 2004) and
protease enzyme applications show maximal EPS release
from sludge flocs (Sessay et al., 2006). Consequently, protease
activity could be a key element to disrupt the matrix and
therefore enhance other enzyme activity by increasing
substrate accessibility.
Regarding the quantitative and functional importance of
proteins in wastewater treatment processes, strong interest
has been given to extract and characterise these proteins
(Ehlers and Cloete, 1999; Massé, 2004) and both the extraction
procedures and the protein quantification methods have to
be improved.
The aim of this study is to introduce a suitable and reliable
protein quantification method in EPS aggregates. Various
potential interfering agents were investigated including inner
compounds of the sludge and the reagents involved in the
extraction procedures described in this study.
Mechanical, ionic and hydrophobic methods were com-
pared and combined in order to obtain maximal extracted
proteins. In addition, the exoenzyme activity (particularly
protease) and protein content were both investigated in a
same extracted sludge sample.
 ARTICLE IN PRESS
WAT E R R E S E A R C H 42 (2008) 1867 – 1878 1869

2.2. EPS extraction

2. Material and methods
Before extraction, samples from sludge 1 and sludge 2 were
2.1. Sludge samples diluted to 1 g VSS L1 and centrifuged at 12 000g 15 min 4 1C.
This first supernatant (soluble fraction) was collected for EPS
Activated sludge samples were collected from two different analysis The pellet was washed twice with Tris–HCl buffer
pilot plants, pilot 1 and pilot 2, supplied by wastewater from (10 mmol L1, pH 8) before EPS extraction.
the WWTP of Toulouse, France. The main characteristics of The tested extraction protocols are summarised in Fig. 1.
the pilots are described in Table 1. Extraction methods were Two mechanical methods were investigated to homogenise
tested on these two types of sludge samples, called sludge 1 the re-suspended pellets (in Tris–HCl buffer): ultraturax (shear
and sludge 2 from pilot 1 and pilot 2, respectively. Sludge 2 stress), 2 min, 20 000 rpm, 4 1C or sonication for 20 s, 37 W, 4 1C.
was sampled twice at 2 solid residence times (SRT): SRTa of 10 The extraction potential was also tested by doubling ultra-
days and SRTb of 12 days. turax treatment and by increasing sonication time (5  2 min
with 5 min interval).
CER (Dowex, Fluka) was used according to Frohlund et al.
(1996) (70 g g1 VSS, 45 min, 500 rpm). Triton X-100 (0.5%)
Table 1 – Main characteristics of wastewater pilots and (Sigma) treatment (1 h, 500 rpm) prepared in Tris–HCl buffer
sludge samples used for EPS extraction was applied according to Gessesse et al. (2003). Both of these
extraction methods were tested alone and with a prior
Parameters Pilot 1 Pilot 2 physical disruption step (ultraturax and sonication). The
soluble fraction (first supernatant) and the extracted fractions
Feed 10% domestic 100%
were kept at 4 1C and analysed the same day.
sewage 90% domestic
synthetic effluenta sewage
TSS (g L1) 2.2 2.5 2.3. EPS analysis
VSS (g L1) 1.6 2
Organic load 3 0.2
2.3.1. Modified Lowry method
(kg DCO kg1 MES d1)
Protein and humic substances mutually interfere with colour
SVI (mL g1) 150 80
Hydraulic retention 24 15 development when CuSO4 is added. Frohlund et al. (1995)
time (h) modified Lowry procedure (Lowry et al., 1951) to correct this
bias. Total absorbance or Atotal (with CuSO4) and Blind
a
Synthetic effluent composition: tryptone, meat extract, sodium absorbance or Ablind (without CuSO4) measurements at
propionate, magnesium sulphate and Na2HPO4.

Sludge 1 diluted to 1g VSS / L Sludge 2 diluted to 1g VSS / L

Centrifugation 12 000g 15 min 4°C

Pellet re-suspended in Tris-HCl buffer

Centrifugation 12 000g 15 min 4°C

CER 70g/g
0.5% Triton X- Sonication 5 × 2 min,
VSS, 45
100 solution, Ultraturax 2 min,20000 rpm, 4°C (37W, 4°C) Sonication (20 sec, 37W, 4°C)
min, 500
1h 500 rpm
rpm

Centrifug ation 12 000g 15 min 4°C

CER 70g/g 0.5% Triton

CER 70g/g CER 70g/g VSS 45 min X-100, 1h 500
0.5% Triton X- Ultraturax 2min 0.5% Triton
VSS, 45 VSS, 45 500 rpm rpm
100, 1h, 500 20000 rpm, 4°C X-100, 1h,
min, 500 min, 500
rpm 500 rpm
rpm rpm

Centrifugation 12 000g 15 minutes 4°C Centrifugation 12 000g 15 min 4°C

Fig. 1 – Extraction procedures for sludge 1 and sludge 2 samples.

 ARTICLE IN PRESS
1870 WAT E R R E S E A R C H
750 nm, corrected protein absorbance (A) values by the
following formula:
Atotal ¼ Aprotein þ Ahumic ,
Ablind ¼ 0:2Aprotein þ Ahumic ,
Aprotein ¼ 1:25ðAtotal  Ablind Þ,
Ahumic ¼ Ablind  0:2Aprotein .
Bovin serum albumin (BSA) (Sigma) and commercial Humic
acid (Fluka) were used as standards for protein and humic
acid quantification, respectively.
2.3.2. Bicinchoninic acid (BCA) method
A second protein measurement method was performed with
the BCA reagent (Sigma), according to Smith et al. (1985). About
20 mL assay sample was added to 1 mL of BCA reagent. Samples
were incubated (60 1C) for 30 min. Standard and samples were
measured spectrophotometrically at 562 nm. BSA (Sigma) was
used as the standard for protein quantification.
For the two protein measurement methods, Ribonuclease,
Ovalbumine, Catalase and Thyroglobuline (Pharmacia, Gel
Calibration Kit) were compared as standard proteins.
2.3.3. Anthrone method
Polysaccharide content was measured by the anthrone
(Sigma) method using glucose as the standard. About 1 mL
assay samples were added to 2 mL of anthrone–H2SO4
reagent. After vortexing, samples were incubated at 100 1C
for 10 min and then cooled down in cold water. Standards and
samples were measured spectrophotometrically at 625 nm.
2.4. Enzyme assays
2.4.1. Glucose-6-phosphate-deshydrogenase (G6P-DH)
The intracellular G6P-DH enzyme was measured according to
Lessie and Vander Wyk (1972). About 0.8 mL of substrate and
0.2 mL of each extracts were incubated at room temperature.
Spectrophotometer kinetics were undertaken at 340 nm over
a 10 min period.
2.4.2. Protease
Protease activity was measured using 0.5% azocasein (Sigma)
in 20 mmol L1 Tris–HCl buffer pH 8. Assays (800 mL of
substrate with 200 mL of sample) were incubated at room
temperature for 1 h. Proteolysis was stopped by adding
trichloroacetic acid. After 30 min, the particulate phase was
eliminated by centrifugation (14000 rpm, 15 min.). Sulphani-
lamide colours were developed by mixing the supernatant
with NaOH 2 mol L1 (4/1 v/v). Absorbance was measured at
440 nm against blank reagent. One unit of protease activity
was defined as the amount of enzyme which results in an
absorbance increase of 0.01 after 1 h incubation at room
temperature (Gessesse et al., 2003).
2.4.3. Lipase
Lipase activity was measured with p-nitrophenol palmitate
(Sigma) as substrate 20 mmol L1 (prepared with isopropanol)
and diluted (1:20) in Tris–HCl buffer 20 mmol L1 pH 8
(containing 0.1% gum Arabic and 0.4% Triton X-100). Soluble
assay absorbance was measured continuously for 10 min at
42 (2008) 1867– 1878
410 nm (Gessesse et al., 2003). Particulate assays (nonex-
tracted samples) were incubated in a larger volume and
samples then centrifuged every 10 min. Resulting super-
natants were measured at 410 nm.
2.5. Laboratory instruments
Ultrasound treatments were undertaken with a Vibra Cell
sonication probe from Bioblock. Ultraturax disruption was
performed with Janke & Kunkel ultraturax 25 from IKA-
Labortechnik. Absorbance measurements were carried out on
a Miltron Roy spectrophotometer (Spectronic 1201).
2.6. Statistical analysis
Considering the colorimetric method response of BSA sam-
ples ranging from 0 to 1 g L1 versus their real concentration,
the slope (ao) and intercept (bo) values of the various linear
regressions were evaluated by the least squares method.
The linearity domains were defined using the Fisher test.
The confidence intervals for ao and bo values were evaluated
at 95% level with the Student distribution after determination
of their standard deviations (Sao and Sbo).
The detection limit (xDL) and quantification limit (xQL)
parameters were calculated as follows:
xDL ¼ ðao þ 3SaoÞ=bo;
xQL ¼ ðao þ 10SaoÞ=bo:
3. Results and discussion
3.1. Comparison of two protein measurement methods
In order to evaluate the reliability of BCA and modified Lowry,
both methods were tested to quantify different proteins.
The concentration of six pure proteins was determined
using BCA and modified Lowry methods and expressed in
equivalent BSA. Fig. 2 shows that the tested proteins do not
respond in the same manner as BSA. For both methods,
strong protein to protein variation is put forward since
proteins can either be overestimated (i.e. ribonuclease), or
underestimated (i.e. azocasein and catalase), with a variation
factor from 0.5 to 1.4. Over the five tested proteins,
ovalbumine and thyroglobulin respond correctly towards
BCA and modified Lowry measurement methods.
BSA response is intermediate for the five tested proteins.
Therefore, BSA can be considered as a correct reference for
both methods. These results are in accordance with Smith
et al. (1985), who showed intermediate response for BSA over
seven different proteins measured by BCA method.
However, sludge origin (Bura et al., 1998), treatment
systems (i.e. aerobic or anaerobic (Liu and Fang, 2002)) and
functional parameters (Sessay et al., 2006) have a great
influence on protein content. An increase in SRT has shown
to have a significant effect on EPS concentration and
particularly for protein content. Sessay et al. (2006) showed
an increase from 20 to 50 mg protein g1 VSS over a 4–20 day
period. Moreover, SRT has also an impact on the type of
 ARTICLE IN PRESS
WAT E R R E S E A R C H
1.8
Bicinchonic acid
1.6
Modified Lowry
1.4
g eq. BSA/g protein
1.2
1
0.8
0.6
0.4
0.2
0
A
ei
as
BS
as
le
oc
uc
Az
on
ib
R
Fig. 2 – Comparison of protein response towards bicinchoninic acid and modified Lowry methods for six pure protein
solutions. Six solutions of pure proteins were prepared at 1 g L1. The quantification of each protein was evaluated in
g eq. BSA g1 of protein.
proteins produced by micro-organisms (Liao et al., 2001).
Considering that different proteins do not respond identically
to BSA, it is thus difficult to compare absolute protein
concentrations from different treatment systems. On the
other hand, protein determination in sludge samples collected
from the same treatment system and the same functional
state can inform on relative protein concentrations.
The accuracy of both BCA and modified Lowry methods
towards solvents was investigated by preparing BSA solutions
in distilled water and in a classical biological buffer such as
Tris–HCl buffer. Results are presented in Fig. 3A. Between 0
and 1 g BSA L1 in distilled water, measurements show better
linearity and slope values for BCA method (R2 ¼ 1.00 and
slope ¼ 0.99) than for modified Lowry (R2 ¼ 0.96 and
slope ¼ 0.84). However, between 0 and 0.8 g BSA L1, protein
determination by both methods is considered accurate.
Reliability tests for both methods were applied on the
accurate concentration range (0–0.8 g L1 data set) in order to
distinguish the most appropriate measurement method
(Feindberg, 2001). For both methods, 95% confidence intervals
were calculated for slope and intercept estimation (Fig. 3B and
C). Between 0 and 0.8 g BSA L1 in distilled water, both
methods are reliable for protein quantification. Tris–HCl
buffer definitely underestimates protein measurements for
both methods. However, the observed underestimation is less
important towards BCA method (5%) compared to modified
Lowry method (30%).
Protein detection and quantification limits for both meth-
ods were estimated with the Student test (Table 2). The
calculated thresholds for protein detection in distilled water
are 5 mg L1 for BCA method and 43 mg L1 for modified
Lowry. Quantification thresholds are estimated at 38 mg L1
for BCA method and 185 mg L1 for modified Lowry. Tris–HCl
buffer increases these values to 95 and 225 mg L1 for BCA and
modified Lowry respectively (Table 2). These data underline
the better performances of the BCA method compared to the
modified Lowry method for quantification of pure proteins
42 (2008) 1867 – 1878 1871
in
in
as
um
ul
ob
al
lb
at
gl
va
C
ro
O
y
Th
solubilised in water or in Tris buffered solutions. Moreover,
coloured EPS extracts from sludge must be diluted in order to
reduce interferences. Therefore, BCA method with its low
quantification limit offers a better precision in EPS quantification.
3.2. Interference of sludge compounds on colorimetric
quantification methods
Simple media such as a buffer solution has an effect on
protein detection and quantification by both investigated
measurement methods. Activated sludge is a much more
complex media which offers a great diversity of compounds
which could interfere in these protein determination meth-
ods. Sugars are potential reducing agents which can respond
like proteins in BCA method (Raunjker et al., 1994; Smith
et al., 1985). Simple sugars such as glucose seem to have a
greater influence compared to slightly more complex sugars
such as saccharose (results not shown). Interfering simple
sugars are rapidly assimilated by bacteria and therefore
are less likely to persist in bacterial aggregates. Humic
substances are described as the predominant interfering
agent for protein measurements and Lowry method has
already been modified to that effect: modified Lowry method
(Frohlund et al., 1995).
The influence of humic substances on protein determina-
tion (0–1 g BSA L1) was investigated by both BCA and
modified Lowry (Fig. 4). Between 0 and 0.2 g humic sub-
stances L1, protein determination by modified Lowry is not
affected in comparison to protein measurement in buffer
alone. Above 0.2 g L1, humic substances significantly under-
estimate BSA. BCA method on the other hand globally
overestimates protein measurements in contact with humic
acids (Fig. 4B). More precisely, for concentration values
ranging from 0 to 0.8 g L1 range, the slopes are not affected
whereas the intercepts as well as the detection and quanti-
fication limit increase with the amount of humic acids in
solution (Table 2). Data from literature show that humic
 ARTICLE IN PRESS
1872 WAT E R R E S E A R C H 42 (2008) 1867– 1878

1.2
water / BCA : y = 0.99 x + 0.01 R2 = 1
water / Lowry : y = 0.84 x + 0.04 R2 = 0.96

1 buffer / BCA : y = 0.95 x + 0.03 R2 = 0.99

buffer / Lowry: y = 0.62 x + 0.05 R2 = 0.98

0.8
Measured [BSA] (g/L)

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
[BSA] in solution (g/L)

1.2 0.1
BCA method BCA method
modified Lowry method 0.08 modified Lowry method
1
0.06

0.8
0.04
Intercept
Slope

0.6 0.02
'
0
0.4
-0.02
Water : Water : Buffer :
0.2 0-1 g/L 0-0.8 g/L 0-0.8 g/L
-0.04

0 -0.06
Water : Water : Buffer :
0-1 g/L 0-0.8 g/L 0-0.8 g/L

Fig. 3 – Comparison of BCA and modified Lowry methods upon BSA solutions, respectively in water and tris–HCl buffer (A).
The 95% statistical confidence intervals obtained for slopes (B) and for intercept (C).

substance concentration varies between 0.5 and 1.6 g L1 in
different sludge samples (Nielsen et al., 1996; Wilén et al.,
2003). However, in order to enter sludge samples in the
calibration curve, protocols generally include dilutions factors
up to 10. Thus, taking this factor into consideration, humic
substances should not interfere in protein measurements
using either method.
Besides polysaccharides and humic acids, sludge contains a
mixture of other compounds such as lipids, uronic and
nucleic acids as well as some yet unknown (Azerdo et al.,
1999; Liu and Fang, 2002) which could have simple, additive
or synergistic effects on protein measurements (Krieg
et al., 2005). In order to evaluate the interferences of all
sludge compounds towards protein quantification, known
 ARTICLE IN PRESS
WAT E R R E S E A R C H 42 (2008) 1867 – 1878 1873

Table 2 – Statistical analysis of the effect of various conditions upon BCA and modified Lowry method

Colorimetric Conditions Slope Offset Detection limit Quantification

method (mg L1) limit (mg L1)

BCA Water 1.0470.01 0.0170.01 5 38

Modified Lowry 0.9970.04 0.0270.02 43 185

BCA Buffer 0.9570.02 0.0370.01 52 95

Modified Lowry 0.7070.04 0.0370.01 96 225

BCA Water US 30 min 1.0470.05 0.0070.02 49 169

Modified Lowry 0.9170.05 0.0070.02 59 195

BCA Water UT 2 min 1.3370.05 0.0370.02 13 104

Modified Lowry 0.9170.05 0.0170.02 53 191

BCA Triton 0.5% 1.0470.04 0.0470.01 81 179

Modified Lowry 0.4970.09 0.1170.03 436 910

BCA CER 45 min 1.0070.02 0.0170.01 14 62

Modified Lowry 0.9770.06 0.0170.03 75 263

BCA Humic acid 0.1 g L1 1.0770.05 0.0670.02 115 247

Modified Lowry 0.7370.04 0.0270.02 85 232

BCA Humic acid 0.5 g L1 0.9770.07 0.2470.03 337 542

Modified Lowry 0.6370.05 0.0470.02 50 309

Slope and offset are respectively the slope and the offset of measured BSA concentration versus BSA concentration (g L1) in the range
0–0.8 g L1.
US: sonication; UT: ultraturax; CER: cation exchange resin.

1.4 1.4
In buffer : y = 0.62 x + 0.05 In buffer : y = 0.95 x + 0.03
R2 = 0.98 R2 = 1
Measured [BSA] using modified Lowry

1.2 1.2
Measured [BSA] using bicinchoninic

HA 0.1 g/L : y = 0.64 x + 0.04 HA 0.1 g/L :

R2 = 0.97 y = 1.06 x + 0.05
1.0 HA 0.2 g/L : y = 0.66 x + 0.02 1.0 R2 = 0.99
R2 = 0.99
method (g/L)
method (g/L)

0.8 HA 0.5 g/L : y = 0.58 x - 0.03 0.8

R2 = 0.97
HA 1 g/L : y = 0.57 x - 0.09
0.6 R2 = 0.93 0.6 HA 0.2 g/L :
y = 0.91 x + 0.13
R2 = 1
0.4 0.4
HA 0.5 g/L : y = 0.95 x + 0.25
R2 = 0.99
0.2 0.2
HA 1 g/L : y = 0.82 x + 0.56
R2 = 0.96
0.0 0.0
0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2
[BSA] in solution (g/L) [BSA] insolution (g/L)

Fig. 4 – Comparison between modified Lowry (A) and bicinchoninic acid (B) methods upon BSA solutions in Tris–HCl buffer
containing humic acids (HA) in the range 0.1–1 g L1.

proportions of BSA (2, 4 and 8 g L1) were added in sludge (error bars between 0.5 and 0.8 g L1) and rarely covers the
samples and then measured by BCA and modified Lowry expected values compared to modified Lowry (error bars
method (Fig. 5). Values are corrected with protein content between 0.3 and 0.4 g L1). These results are in agreement
initially found in the sludge sample. Protein determination by with Fig. 4 data showing that BCA and modified Lowry
modified Lowry in sludge 2 at SRTb alone matches the method respectively overestimate and underestimate the
awaited values. BCA measurements show more variability protein concentration in presence of humic substances.
 ARTICLE IN PRESS
1874 WAT E R R E S E A R C H 42 (2008) 1867– 1878

12 12
Measured [BSA] using modified Lowry

Sludge 2 / SRTa

Measured [BSA] using bicinchoninic

Sludge 2 / SRTa
10 Sludge 2 / SRTb 10 Sludge 2 / SRTb
Sludge 1 Sludge 1

acid method (g/L)

8 Expected values 8 Expected values
method (g/L)

6 6

4 4

2 2

0 0
2 4 8 2 4 8
Added BSA (g/L) Added BSA (g/L)

Fig. 5 – Comparison of protein response towards modified Lowry (A) and BCA (B) methods after addition of known
proportions of BSA. Bovine serum albumin (BSA) was added to three different sludge samples: sludge 1 and sludge 2 collected
at 2 different residential times (SRTa and SRTb). Samples were incubated overnight at 4 1C, protein measurements (BCA and
modified Lowry) were undertaken the following day. Indicated values are corrected with protein content found initially in the
sludge.

Protein determination in all sludge samples is system-
atically overestimated with BCA method with a variation of
between +20% and +25% for three types of sludge. Modified
Lowry method on the other hand shows a greater variation
between measurements in different sludge samples (+4%
overestimation for sludge 2 SRTb and 40% underestimation
for sludge 1).
Thus interfering agents (type and concentration) found in
different types of sludge create greater variability in protein
determination by modified Lowry than for BCA measure-
ments.
by 50% (slope ¼ 0.49), the detection limit is increased
by a factor 10 (436 mg L1) and the quantification limit
increased by a factor 5 (910 mg L1). These results join those
presented by Smith et al. (1985), who showed a loss in
protein response with Triton 1% (precipitation) and Tris
0.1 mol L1 (25% underestimation) for Lowry method and no
influence for BCA method. CER does not affect modified
Lowry measurements.
3.4. Protein and enzyme extraction

In order to optimise the amount of extracted proteins, sludge

3.3. Extraction methods interferences for colometric
protein quantification in sludge
In addition to compounds already found in the sludge,
treatment methods used in the extraction protocols may also
affect the reliability of protein measurements. Mechanical
extraction methods were first of all investigated for their
influence on both BCA and modified Lowry methods. 30 min
sonication does not affect accuracy in protein determination
for both methods. However, in the 0–0.8 g L1 range, BCA
detection and quantification limits are increased to thresh-
olds comparable to modified Lowry method (Table 2). Ultra-
turax affects BCA method by reducing protein determination
accuracy (30% overestimation) and increasing the quantifica-
tion limit (104 mg L1).
The effects of chemicals used in extraction protocols on
protein determination methods were then investigated on
BSA solutions (Fig. 6). Protein determination accuracy
with BCA method is not affected by Triton and CER applica-
tion (slope ¼ 1.04 and 1.00 respectively in the range
0–0.8 g BSA L1) compared to control measurements in water.
Detection and quantification limits are slightly increased
particularly with Triton. Modified Lowry is greatly influenced
by Triton application. Measured proteins are underestimated
samples were submitted to various extraction treatments.
Since EPS compounds are bound together by low-energy
interactions, such as ionic and hydrophobic interactions,
different extraction methods acting upon these specific
linkages were investigated: mechanical desegregation, CER
and nonionic detergent Triton. The various extraction treat-
ments were arranged by doubling a same extraction method
or by combining mechanical disruption with a chemical
extraction method on the same sample (Fig. 1). BCA method
was chosen for protein determination for its better accuracy
and reliability towards the studied interferents compared to
modified Lowry.
After each extraction step, protein concentrations were
measured by BCA method. The protein extracts obtained
from sludge 1 samples are illustrated in Table 3. Results
show maximal protein release with Triton treatments. A
second Triton extraction releases 50% of the proteins
extracted with a first Triton treatment (respectively 60
and 122 mg eq. BSA g1 VSS). A prior ultraturax application
does not increase significantly Triton efficiency (113 mg eq.
BSA g1 VSS). Ultraturax, a soft physical breakdown method,
generally used for sludge or bacterial culture homogenisation
(Salhani and Uelker-Deffur, 1997) was also investigated as an
extraction method. However, results should be discussed with
 ARTICLE IN PRESS
WAT E R R E S E A R C H 42 (2008) 1867 – 1878 1875

1.2
1.2
In water : y = 0.99 x +0.00 In water : y = 0.99 x + 0.01

Measured [BSA] using bicinchoninic method

R2 = 0.99 R2 = 1
Measured [BSA] using modified Lowry

1.0 1.0 In buffer : y = 0.95 x + 0.03

In buffer : y = 0.62 x + 0.05 R2 = 1
R2 = 0.98
In buffer + triton 0.5% :
0.8 In buffer + triton 0.5% : 0.8
y = 0.99 x + 0.04
y = 0.48 x + 0.14
method (g/L)

R2 = 0.99
R2 = 0.89

(g/L)
0.6 0.6

CER 45 min
0.4 0.4 y = 0.97 x - 0.00
CER 45 min R2 = 1
y = 1.15 x - 0.03
R2 = 0.98 0.2 CER 145 min
0.2
y = 0.98 x - 0.00
CER 145 min : y = 1.18 x - 0.03 R2 = 1
R2 = 0.97
0.0 0.0
0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2
[BSA] in solution (g/L) [BSA] in solution (g/L)

Fig. 6 – Comparison between modified Lowry (A) and BCA (B) methods upon BSA solutions treated ‘‘chemically’’. (CER 45 min
and CER 145 min: BSA in tris–HCl buffer 10 mM pH 8 after respectively 45 and 145 min CER treatment).

care, for the effect of ultraturax on BCA measurement has (Jorand et al., 1994; Tiehm et al., 2001; Zhang et al., 2007).
been underlined previously. Besides the fact that CER and Only 20 s ultrasounds were performed on sludge samples in
ultraturax release less proteins than Triton treatments, these order to disrupt the matrix, increase chemical accessibility
methods, applied a second time, extract as much proteins as and assure intact cells. Low protein content in extracts from
the first time. A second CER extraction releases as much 20 s sonication was shown with sludge 2 samples (10 mg eq.
proteins as the first CER treatment (respectively 47 and BSA g1 VSS). By increasing sonication time, extracted pro-
39 mg eq. BSA g1 VSS) and a prior ultraturax application does teins reached 107 mg eq. BSA g1 VSS which are in comparable
not increase its efficiency (respectively 35 and 29 mg eq. proportions with combined sonication and triton extracts
BSA g1 VSS). This suggests either a limiting equilibrium (Fig. 7A). In all extracts, the measurement of intracellular G6P-
between the available Dowex cations and the divalent ions DH activity showed that no cell lysis occurred during the
linking the polymeric proteins or limited accessibility. extraction methods.
Total protein extracts from sludge 1 and sludge 2 are Polysaccharides are EPS compounds which are linked to the
illustrated in Fig. 7A. Both sludge samples differ from the EPS matrix as proteins, and are found in the analysed
soluble fraction (44 and 10 mg eq. BSA g1 VSS, for sludge 1 extracts. Strong P/S ratios indicate that the according method
and sludge 2, respectively). Doubled triton treatments allowed extracts preferentially proteins than polysaccharides. This
a total amount of extracted proteins of 182 mg eq. BSA g1 VSS ratio should be considered not only to reduce the potential
from sludge 1, i.e.; 226 mg g1 VSS in the total soluble fraction interference of sugars in the extracts, but also to optimise the
(Fig. 7A). Maximum protein extraction from sludge 2 is shown amount of proteins in the extracts. Triton shows the highest
with combined sonication and the nonionic detergent triton P/S ratios among the other treatments, whether alone
treatments (105 mg eq. BSA g1 VSS, i.e., 115 mg g1 VSS in the (P/S ¼ 14), doubled (P/S ¼ 22) or combined with ultraturax
total soluble fraction). In other words, a major part of proteins (P/S ¼ 17) (Table 3).
in sludge 1 and sludge 2 may be linked to the EPS matrix by Enzymes are part of the protein pool. Protease/protein
hydrophobic interactions. Dignac et al. (1998) pointed out the ratio informs on the specificity of the extraction method for
importance of hydrophobic amino acids in activated sludge, one type of protein. This ratio varies from 60 to
which represented 47% of the 16 investigated amino acids 220 U mg1 eq. BSA, showing that for the different investi-
extracted by CER. These data suggest that hydrophobic gated treatments, proteases are not extracted in the same
interactions are also disrupted by ionic treatments and proportions as other proteins (Fig. 7B). 20s sonication shows
that other more solid hydrophobic interactions can persist the highest ratio (220 U mg1 eq. BSA), but this result should
in the matrix. be considered with care, for the low protein content in the
Studies which proceed with sonication have shown the extract (close to the quantification limit) can create strong
dispersive and disruptive effect on sludge flocs (Jorand et al., error on the ratio value. Besides 20s sonication, Protease/eq.
1994). Convincing results shown by Urbain et al. (1993) explain BSA protein ratio shows maximum values for triton doubled
the wide utility of this method. However, extended ultra- or combined with a mechanical treatment (between 150
sounds treatments enhance cell disintegration, thereby con- and 180 U mg1 eq. BSA). Thus, triton treatment is favourable
taminating the liquid phase with intracellular material for proteolytic enzyme extraction compared to doubled
 ARTICLE IN PRESS
1876 WAT E R R E S E A R C H 42 (2008) 1867– 1878

ultraturax, doubled CER or 30 min sonication (between 60

Both extraction methods are undertaken on the same sample (sludge1). The same extraction method applied twice on the same sample is termed ‘‘doubled’’ and two different extraction methods
P/S

17
3
and 80 U mg1 eq. BSA). These results are in agreement
with Gessesse et al. (2003), who showed that Triton 0.5%
was the most efficient method for protease extraction
UT/Triton (4000 U g1 VSS). CER or triton combined with sonication

Protein (mg eq.

BSA g1 VSS)
shows higher values (140 or 170 U mg1 eq. BSA, respectively)

148712
11379
3573
than combined with ultraturax (80 or 150 U mg1 eq. BSA
respectively). This suggests either a difference in efficiency
in favour of ultrasound treatments compared to ultraturax, or
a difference in protease content between both sludge
samples. Indeed, proteolytic activity has been shown to differ
P/S

between extracts from different sludge samples (Whiteley

11
3

et al., 2002).
On the other hand, investigated extraction methods were
not efficient for lipase enzymes. Indeed, extracted super-
(mg eq. BSA g1 VSS)
UT/CER

natants showed no lipase activity, which is in agreement with

Protein

Gessesse et al. (2003). Solubility and functional tests on

3571
2972
6473

commercial lipase confirmed that extraction was not limited

by lipase solubility and also that the extraction methods did
Table 3 – Protein and protein/sugar (P/S) ratio content in extracts obtained by a first and second extraction method

not affect lipase activity (results not shown). Measurements

on re-suspended pellets revealed that lipase remained in the
particle fraction. As long as lipidic nutrients are available,
P/S

3
6

lipase is known to be anchored in the cell membrane (Alonso

et al., 2005; Wilhelm et al., 1999), which is in agreement with
our data and the idea that the matrix structure protects cells
(mg eq. BSA g1 VSS)

from nutrient impoverishment and prevents further enzyme

UT/UT

activity.
Protein

3575
3171
6676

4. Conclusion

This study highlights the following:

P/S

6
5

 BCA and modified Lowry are both accurate methods for

protein determination in water fractions in a restricted
range of 0–0.8 g L1.
(mg eq. BSA g1 VSS)
CER/CER

 BCA systematically overestimates proteins in sludge while

modified Lowry underestimates. However, BCA measure-
Protein

3972
4775
8677

ments in different types of sludge show less variability

than modified Lowry.
 The investigated chemicals added for extraction treat-
ments affect protein determination by both methods but
with a greater effect on modified Lowry than BCA method.
applied successively on the same sample ‘‘combined’’.

 Triton treatments extract the most proteins compared to

P/S

14
22

other extraction methods, whether alone or combined

with other treatments. Moreover, Triton is compatible with
Triton/Triton

BCA method.
UT: ultraturax; CER: cation exchange resin.
(mg eq. BSA g1 VSS)

 Triton also extracts a greater proportion of proteases

compared to other extraction methods.
Protein

12273

18274
6071

Here, Triton is put forward as the most efficient extraction

method and BCA as the most appropriate method towards
Triton. The extraction/quantification couple for protein
analysis is crucial to determine extraction efficiency in
complex mixtures such as activated sludge.
Second extract

Moreover, it is important to point out the specificity of

Total extract
First/Second

First extract

extraction methods towards certain types of proteins. Thus

extraction
method

an extraction strategy based on different methods can be of

an important relevance to have a representative pool of
proteins initially found in the mixture.
 ARTICLE IN PRESS
WAT E R R E S E A R C H 42 (2008) 1867 – 1878 1877

250 250
Extracted fraction

Protease / Protein ratio (U/mg eq. BSA)

Soluble fraction
[Proteins] (mg eq.BSA/g VS)

200 200

150 150

100 100

50 50

0 0

US 20 s + triton

US 20 s + triton
2 × Triton

2 × CER

2 × UT

UT + CER

UT + triton

US 20 s

US 20 s + CER

2 × Triton

2 × CER

2 × UT

UT + CER

UT + triton

US 30 min

US 20 s

US 20 s + CER
US 30 min

sludge 1 sludge 2 sludge 1 sludge 2

Fig. 7 – Protein release (A) and protease activity (B) in different extracts after doubled and combined extraction methods.
Triton: X-100 0.5%; CER: cation exchange resin 70 g g1 VSS 45 min; UT: ultraturax 2 min, 20000 rpm; US 20 s: ultrasound 20 s;
US 30 min: 5  2 min ultrasonic treatment with 5 min interval. Indicated protease activity values include the soluble fraction.

meric substances (EPS) and EPS complexion properties. Part I.

Acknowledgments Comparison of the efficiency of eight EPS extraction methods.
Enzyme Microb. Technol. 38 (1–2), 235–247.
The authors wish to thank Jean-Denis Fournier for his Comte, S., Guibaud, G., Baudu, M., 2007. Effect of extraction
assistance in statistical analysis and Isabelle Ribera for her method on EPS from activated sludge: an HPSEC investigation.
J. Hazard. Mater. 140 (1–2), 129–137.
technical participation in this study.
Davis, E.M., 1988. Protein assay: a review of common techniques.
Am. Biotechnol. Lab. 6 (July), 28–37.
R E F E R E N C E S
Dey, E., Szewczyk, E., Wawrznczyk, J., Norrlöw, O., 2005. A novel
approach for characterization of exopolymeric material in
sewage sludge. In: IWA-WISA Conference: The Management of
Alonso, F.O.M., Oliveira, E.B.L., Dellamora-Ortiz, G.M., Pereira- Residues Amanating from Water and Wastewater Treatment,
Meirelles, F.V., 2005. Improvement of lipase production at Johannesburg, South Africa.
different stirring speeds and oxygen levels. Braz. J. Chem. Eng. Dignac, M.F., Urbain, V., Rybacki, D., Bruchet, A., Snidaro, D.,
22 (1), 9–18. Scribe, P., 1998. Chemical description of extracellular poly-
Azerdo, J., Lazarova, V., Oliveira, R., 1999. Methods to extract the mers: implication on activated sludge floc structure. Water Sci.
exopolymeric matrix from biofilms: a comparative study. Technol. 38 (8–9), 45–53.
Water Sci. Technol. 39 (7), 243–250. Ehlers, M.M., Cloete, T.E., 1999. Comparing the protein profiles of
Bradford, M.M., 1976. A rapid and sensitive method for the 21 different activated sludge systems after SDS-PAGE. Water
quantification of micrograms quantities of protein utilising the Res. 33 (5), 1181–1186.
principle of protein-dye binding. Anal. Biochem. 72 (1–2), 248–254. Esparza-Soto, M., Westerhoff, P.K., 2001. Fluorescence spectro-
Brown, M.J., Lester, J.N., 1980. Comparison of bacterial extra- scopy and molecular weight distribution of extracellular
cellular polymer extraction methods. Appl. Environ. Microbiol. polymers from full-scale activated sludge biomass. Water Sci.
40 (2), 179–185. Technol. 43 (6), 87–95.
Bruus, J.H., Nielsen, P.H., Keiding, K., 1992. On the stability of Feindberg, M., 2001. International validation of analytical meth-
activated sludge flocs with implications to dewatering. Water ods (Validation interne des méthodes d’analyse). In: Techni-
Res. 26 (12), 1597–1604. ques de l’Ingénieur, Analyse et caractérisation,
Bura, R., Cheung, M., Liao, B., Finlayson, J., Lee, B.C., Droppo, I.G., pp. 224.1–224.23.
Leppard, G.G., Liss, S.N., 1998. Composition of extracellular Frohlund, B., Griebe, T., Nielsen, P.H., 1995. Enzyme activity in
polymeric substances in the activated sludge matrix. Water activated-sludge floc matrix. Appl. Microbiol. Biotechnol. 43
Sci. Technol. 37 (4–5), 325–333. (4), 755–761.
Cadoret, A., Conrad, A., Block, J.C., 2002. Availability of low and Frohlund, B., Palmgren, R., Keiding, K., Nielsen, P.H.,
high molecular weight substrates to extracellular enzymes in 1996. Extraction of extracellular polymers from activated
whole and dispersed activated sludges. Enzyme Microb. sludge using a cation exchange resin. Water Res. 30 (8),
Technol. 31 (1–2), 179–186. 1749–1758.
Comte, S., Guibaud, G., Baudu, M., 2006. Relations between Garnier, C., Görner, T., Lartiges, B.S., Abdelouhab, S., de Donato, P.,
extraction protocols for activated sludge extracellular poly- 2005. Characterization of activated sludge exopolymers from
 ARTICLE IN PRESS
1878 WAT E R R E S E A R C H
various origins: a combined size-exclusion chromatography
and infrared microscopy study. Water Res. 39 (13),
3044–3054.
Gessesse, A., Dueholm, T., Petersen, S.B., Nielsen, P.H., 2003.
Lipase and protease extraction from activated sludge. Water
Res. 37 (15), 3652–3657.
Goel, R., Mino, T., Satoh, H., Matsuo, T., 1997. Enzyme
activities under anaerobic and aerobic conditions in acti-
vated sludge sequencing batch reactor. Water Res. 32 (7),
2081–2088.
Jorand, F., Zartarian, F., Thomas, F., Block, J.C., Bottero, J.Y.,
Villemin, G., Urbain, V., Manem, J., 1994. Chemical and
structural (2D) linkage between bacteria within activated
sludge flocs. Water Res. 29 (7), 1639–1647.
Jorand, F., Boué-Bigne, F., Block, J.C., Urbain, V., 1998. Hydro-
phobic/hydrophilic properties of activated sludge exopoly-
meric substances. Water Sci. Technol. 37 (4–5), 307–315.
Jung, J., Xing, X.H., Matsumoto, K., 2002. Recoverability of protease
released from disrupted excess sludge and its potential
application to enhanced hydrolysis of proteins in wastewater.
Biochem. Eng. J. 10 (1), 67–72.
Krieg, R.C., Dong, Y., Schwamborn, K., Knuechel, R., 2005. Protein
quantification and its tolerence for different interfering
reagent using the BCA-method with regard to 2D SDS PAGE.
J. Biochem. Biophys. Methods 65 (1), 13–19.
Lazarova, V., Manem, J., 1995. Biofilm characterization and
activity analysis in water and wastewater treatment. Water
Res. 29 (10), 2227–2245.
Lessie, T.G., Vander Wyk, J.C., 1972. Multiple forms of Pseudomonas
multivarans glucose-6-phosphate and 6-phosphogluconate
deshydrogenase: differences in size, pyridine nucleotide
specificity and susceptibility to inhibition by adenosine
50 -triphosphate. J. Bacteriol. 110 (3), 1107–1117.
Liao, B.Q., Allen, D.G., Droppo, I.G., Leppard, G.G., Liss, S.N., 2001.
Surface properties of sludge and their role on bioflocculation
and settleability. Water Res. 35 (2), 339–350.
Liu, H., Fang, H.P., 2002. Extraction of extracellular polymeric
substances (EPS) of sludges. J. Biotechnol. 95 (3), 249–256.
Lowry, O.H., Rosebrough, N.J., Lewis Farr, A., Randall, R.J., 1951.
Protein measurement with Folin phenol reagent. J. Biol. Chem.
193 (1), 265–275.
Martinez, F.O., Lema, J., Mendéz, R., Cuervo-Lopez, F., Gomez, J.,
2004. Role of exopolymeric protein on the settleability of
nitrifying sludges. Bioresour. Technol. 94 (1), 43–48.
Massé, A., 2004. Urban wastewater treatment with immersed
membrane bioreactors: physico-chemical properties of the
biological media and biofouling (Bioréacteur à membranes
immergées pour le traitement des eaux résiduaires urbaines:
spécificités physico-chimiques du milieu biologique et col-
matage). Ph.D. Thesis, INSA, Toulouse.
Morgan, J.W., Forster, C.F., Evison, L., 1990. A comparative study of
the nature of biopolymers extracted from anaerobic and
activated sludges. Water Res. 24 (6), 743–750.
42 (2008) 1867– 1878
Müller, E., Lind, G., Lemmerb, H., Wilderera, P.A., 2005. Population
structure and chemical EPS analyses of activated sludge and
scum. Acta Hydrochim. Hydrobiol. 33 (3), 189–196.
Nielsen, P.H., Frolund, B., Keiding, K., 1996. Changes in the
composition of extracellular polymeric substances in acti-
vated sludge during anaerobic storage. Appl. Microbiol.
Biotechnol. 44 (6), 823–830.
Palmgren, R., Nielsen, P.H., 1996. Accumulation of DNA in the
exopolymeric matrix of activated sludge and bacterial cul-
tures. Water Sci. Technol. 34 (5–6), 233–240.
Ramesh, A., Lee, D.J., Hong, S.G., 2006. Soluble microbial products
(SMP) and soluble extracellular polymeric substances (EPS)
from wastewater sludge. Appl. Microbiol. Biotechnol. 73 (1),
219–225.
Raunjker, K., Hvittved-Jacobsen, T., Nielsen, P.H., 1994. Measure-
ment of pools of protein, carbohydrate and lipid in domestic
wastewater. Water Res. 28 (2), 251–262.
Salhani, N., Uelker-Deffur, A., 1997. Improved quantification of
aggregated bacteria by combined enzymatic and mechanical
treatment of flocs and biofilms from a rotating drum
bioreactor. Water Res. 32 (4), 1287–1295.
Sessay, L.M., Özcengiz, G., Sanin, F.D., 2006. Enzymatic extraction
of activated sludge extracellular polymers and implications on
biofloculation. Water Res. 40 (7), 1359–1366.
Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner,
F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J.,
Klenk, D.C., 1985. Measurement of protein using bicinchoninic
acid. Anal. Biochem. 150 (1), 76–85.
Tiehm, A., Nickel, K., Zellhorn, M., Neis, U., 2001. Ultrasonic waste
activated sludge disintegration for improving anaerobic
stabilization. Water Res. 35 (8), 2003–2009.
Urbain, V., Block, J.C., Manem, J., 1993. Biofloculation in activated
sludge: an analytical approach. Water Res. 27 (5), 829–838.
Watson, S.D., Akhurst, T., Whitley, C.G., Rose, P.D., Pletschke, B.I.,
2004. Primary floc degradation is accelerated under biosul-
phidogenic conditions: enzymological aspects. Enzyme
Microb. Technol. 34 (6), 595–602.
Whiteley, C.G., Heron, P., Pletschke, B., Rose, P.D., Tshivhunge, S.,
Van Jaarsveld, F.P., 2002. The enzymology of sludge solubilisa-
tion utilising sulphate reducing systems. Properties of proteases
and phosphatases. Enzyme Microb. Technol. 31 (4), 419–424.
Wilén, B.M., Jin, B., Lant, P., 2003. The influence of key chemical
constituents in activated sludge on surface and floculating
properties. Water Res. 37 (9), 2127–2139.
Wilhelm, S., Tommassen, J., Jaeger, K.E., 1999. A novel lipolytic
enzyme located in the outer membrane of Pseudomonas
aeruginosa. J. Bacteriol. 181 (22), 6977–6986.
Yu, L., Shu-Fang, Y., Joo-Hwa, T., Qi-Shan, L., Lei, Q., Yong, L., 2004.
Cell hydrophobicity is a triggered force of biogranulation.
Enzyme Microb. Technol. 34 (5), 371–379.
Zhang, P., Zhang, G., Wang, W., 2007. Ultrasonic treatment of
biological sludge: floc disintegration, cell lysis and inactiva-
tion. Bioresour. Technol. 98 (1), 207–210.