Role of Kupffer cells in pathogenesis of sepsis-induced
drug metabolizing dysfunction
Tae-Hoon Kim*, Sang-Ho Lee* and Sun-Mee Lee
School of Pharmacy, Sungkyunkwan University, Suwon, South Korea
Introduction
Sepsis, severe sepsis and septic shock are worldwide
problems and continue to be the most common causes
of death in surgical intensive care units [1]. The patho-
genesis of sepsis has often been viewed to involve
excessive immune inflammation that can lead to lethal
multiple organ failure, suggesting that the downregula-
tion of immunity could be beneficial [2]. As a result of
its major implications in essential metabolic functions
and host defense, the liver plays an important role in
the development of multiple organ failure [3].
Patients who are diagnosed with sepsis receive vari-
ous therapeutic agents because of its complex patho-
physiology and varied symptoms; the main clinical
concern has been that patients on a stable drug regi-
men would have increased exposure to an incidence of
adverse drug events. The cytochrome P450 (CYP)
enzyme system constitutes one of the major aspects of
hepatocyte function and contributes to the metabolism
and elimination of exogenous and endogenous sub-
stances [4]. In various models and in clinical reports,
Keywords
CYP450; HMGB1; Kupffer cells; sepsis;
Toll-like receptor
Correspondence
S M. Lee, School of Pharmacy,
Sungkyunkwan University, 300 Cheoncheon-

CYP1A1 ⁄ 1A2 markedly increased 24 h after CLP, which was prevented by
GdCl
3
. The increased serum level of high mobility group box 1, hepatic
level of Toll-like receptors 2 and 4, and inducible nitric oxide synthase pro-
tein expression were prevented by GdCl
3
. In addition, elevated serum con-
centrations of tumor necrosis factor-a and interleukin-6, and increased
hepatic mRNA levels of tumor necrosis factor-a and interleukin-6 were
decreased by depletion of KCs. Our findings suggest that ablation of KCs
protects against hepatic drug-metabolizing dysfunction by modulation of
the inflammatory response.
Abbreviations
ALT, alanine aminotrasferase; AST, aspartate aminotrasferase; CLP, cecal ligation and puncture; CYP, cytochrome P450; GdCl
3
, gadolinium
chloride; GSH, glutathione; GSSG, glutathione disulfide; HMGB1, high mobility group box 1; IL, interleukin; iNOS, inducible nitric oxide
synthase; KCs, Kupffer cells; LPS, lipopolysaccharide; MDA, malondialdehyde; NO, nitric oxide; PAP, p-aminophenol;
RIPA, radioimmunoprecipitation assay; ROS, reactive oxygen species; TLR, Toll-like receptor; TNF, tumor necrosis factor.
FEBS Journal 278 (2011) 2307–2317 ª 2011 The Authors Journal compilation ª 2011 FEBS 2307
inflammation or infection is associated with a decrease
in hepatic expression and ⁄ or activities of CYPs [5].
A previous study showed that hepatic CYP-mediated
drug metabolism is suppressed during polymicrobial
sepsis, particularly in the late phase [6].
The complex Toll-like receptor (TLR) and associ-
ated downstream regulators of immune cells play a
crucial role in the innate system as a first line of
defense against pathogens [7]. TLR2 and TLR4 expres-

) alone
significantly decreased the hepatic mRNA level of
CD163 compared to that of sham group. Twenty-four
hours after cecal ligation and puncture (CLP), the
hepatic mRNA level of CD163 was similar with that
of the sham group, which markedly decreased to
approximately 7.0% of that of the CLP group (Fig. 1).
Serum aminotransferase activities and lipid
peroxidation
The serum level of alanine aminotransferase (ALT) in
sham-operated rats was 22.6 ± 1.6 UÆL
)1
at 24 h after
CLP. The serum ALT level in rats who underwent
CLP was 1.8-fold that of sham-operated rats at 24 h
after CLP, which was significantly attenuated by
GdCl
3
. Similar to the ALT level, the serum aspartate
aminotransferase (AST) level increased significantly at
24 h after CLP and this increase was attenuated by
GdCl
3
. The malondialdehyde (MDA) level in CLP rats
was 1.8-fold that of sham-operated rats. The increase
in the MDA level at 24 h after CLP was significantly
prevented by GdCl
3
(Table 1).
Hepatic glutathione (GSH)

lipid peroxidation after CLP. Each value is presented as the
mean ± SEM for eight to ten rats per group.
Groups ALT (UÆL
)1
) AST (UÆL
)1
)
MDA
(nmolÆmg
)1
protein)
Sham 22.6 ± 1.6 68.1 ± 3.4 0.98 ± 0.05
GdCl
3
20.0 ± 1.9 82.6 ± 4.1 1.19 ± 0.07
CLP 41.4 ± 1.7** 143.6 ± 8.9** 1.72 ± 0.10**
GdCl
3
+ CLP 24.7 ± 2.1
++
119.9 ± 5.0
+
1.18 ± 0.09
++
**P < 0.01, significantly different from sham.
+
P < 0.05,
++
P < 0.01, significantly different from CLP.
KCs in drug-metabolizing dysfunction during sepsis T H. Kim et al.

Hepatic microsomal CYP isozyme activities
The results for the CYP isozyme activities are summa-
rized in Table 4. At 24 h after CLP, CYP1A1, 1A2
and 2E1 activities were reduced to levels approximately
46.2%, 45.8% and 34.3% of that observed in micro-
somes in sham-operated rats, respectively. These
decreases were attenuated by GdCl
3
pretreatment.
CYP2B1 activity remained unchanged across all
experimental groups.
CYP isozyme protein expression
The amount of CYP1A1 and 1A2 protein expression in
the microsome showed a significant decrease at 24 h
after CLP. GdCl
3
pretreatment raised CYP1A1 and
1A2 protein expression levels without statistical signifi-
cance. No significant differences in CYP2B1 protein
expression level were observed among any experimental
groups. The amount of 2E1 protein expression showed
a significant decrease at 24 h after CLP. This decrease
was prevented by GdCl
3
pretreatment (Fig. 2).
CYP1A1
⁄
1A2 phosphorylation
The phosphorylation of CYP1A1 ⁄ 1A2 significantly
increased 24 h after CLP, which was attenuated by

Serum tumor necrosis factor (TNF)-a and
interleukin (IL)-6 levels
Compared to sham-operated rats, serum TNF-a and
IL-6 levels showed a significant increase at 24 h
Table 2. Effect of gadolinium chloride on concentrations of GSH,
GSSG and GSH ⁄ GSSG ratio after CLP. Each value is presented as
the mean ± SEM for eight to ten rats per group.
Groups
GSH
(nmolÆmg
)1
liver)
GSSG
(nmolÆmg
)1
liver)
GSH ⁄ GSSG
ratio
Sham 4.02 ± 0.25 0.24 ± 0.02 17.61 ± 2.37
GdCl
3
3.84 ± 0.14 0.24 ± 0.02 16.52 ± 1.86
CLP 3.01 ± 0.17** 0.33 ± 0.03* 9.32 ± 0.86**
GdCl
3
+ CLP 3.66 ± 0.17
+
0.30 ± 0.01 12.41 ± 0.51
+
*P < 0.05, **P < 0.01, significantly different from sham.

+
P < 0.05,
++
P < 0.01, significantly different from CLP.
T H. Kim et al. KCs in drug-metabolizing dysfunction during sepsis
FEBS Journal 278 (2011) 2307–2317 ª 2011 The Authors Journal compilation ª 2011 FEBS 2309
after CLP (450.8 ± 22.6 pgÆmL
)1
and 255.1 ±
40.8 pgÆmL
)1
, respectively). GdCl
3
pretreatment atten-
uated these increases (Fig. 7).
Hepatic TNF-a and IL-6 mNRA expression
As shown in Fig. 8, the hepatic level of TNF-a and
IL-6 mRNA expression showed a significant increase
at 24 h after CLP, and this increase was attenuated by
GdCl
3
.
Discussion
Several studies have shown that interactions between
KCs and endotoxin comprise the initiating event lead-
ing to hepatotoxicity in liver injury, including endotox-
emia and ischemia ⁄ reperfusion injury [14]. In our
studies, we employed GdCl
3
to inactivate KCs based

3
or saline alone 48 and 24 h before CLP. The values are represented as the mean ± SEM for eight to ten rats per
group. *P < 0.05, **P < 0.01, significantly different from sham. +P < 0.05, significantly different from CLP.
Phospho-CYP1A1
CYP1A2
CYP1A1
IP : CYP1A1/1A2
Blot : phosphoserine/
threonine
Phospho-CYP1A2
GdCl
3
– + – +
––++CLP
Fig. 3. Effects of KCs on the phosphorylation of CYP1A1 ⁄ 1A2 24 h
after CLP. Rats were pretreated intravenously with 7.5 mgÆkg
)1
GdCl
3
or saline alone 48 and 24 h before CLP. The livers samples
were subjected to immunoprecipitation (IP) using anti-CYP1A1 ⁄ 1A2
serum. Immunoprecipitates were subjected to immunoblot analysis
using anti-phosphoserine ⁄ threonine serum. The values are repre-
sented as the mean ± SEM for eight to ten rats per group.
*P < 0.05, **P < 0.01, significantly different from sham.
+
P < 0.05,
significantly different from CLP.
Table 4. Effect of gadolinium chloride on the hepatic microsomal cytochrome P450 isozyme activities after CLP. Each value is represented
as the mean ± SEM for eight to ten rats per group.

(ROS) and proinflammatory cytokines from KCs [16].
Our recent studies have shown that GdCl
3
attenuated
the imbalanced vascular stress gene expression induced
by sepsis [17]. In the present study, the depletion of
KCs was confirmed by dramatically reduced expression
of the KC marker gene CD163.
In humans and animals, infections or inflammatory
stimuli cause changes in the activities and expression
levels of various forms of CYP in the liver. In most
cases, CYPs and their activities are suppressed; how-
ever, some are unaffected or induced under these con-
ditions [18]. Our previous study reported on
abnormalities in microsomal drug-metabolizing func-
tion during the late phase of sepsis [6]. However, the
underlying mechanisms involved in hepatic dysfunction
during sepsis remain elusive.
Among various CYP isoforms, CYP1A1, 1A2, 2B1
and 2E1 are both present in hepatic microsome of
human and normal rats. The function and regulation
of these isozymes are highly conserved among mam-
malian species [19].
CYP1A1 is not expressed in normal adult tissues
but can be induced several fold by polycyclic or halo-
genated hydrocarbons [20]. CYP1A2, which is consti-
tutively expressed in the liver, is primarily involved in
the oxidative metabolism of xenobiotics and is capable
of the metabolic activation of numerous procarcino-
gens, including aflatoxin B1 [21]. In the present study,

)1
GdCl
3
or saline alone 48 and 24 h before CLP. The values are represented as the mean ± SEM for eight to ten rats per
group. **P < 0.01, significantly different from sham.
+
P < 0.05,
++
P < 0.01, significantly different from CLP.
Sham GdCl
3
CLP GdCl
3
+ CLP
β-actin
TLR4
TLR2
Fig. 5. Effects of KCs on hepatic TLR2 and TLR4 protein expres-
sion levels 24 h after CLP. Rats were pretreated intravenously with
7.5 mgÆkg
)1
GdCl
3
or saline alone 48 and 24 h before CLP. The val-
ues are represented as the mean ± SEM for eight to ten rats per
group. **P < 0.01, significantly different from sham.
++
P < 0.01,
significantly different from CLP.
T H. Kim et al. KCs in drug-metabolizing dysfunction during sepsis

Depression of CYP-dependent hepatic drug metabo-
lism in inflammatory reactions and infectious diseases
has been attributed to the inflammatory events. TLRs
play a critical role in the immune system by providing
an early recognition of pathogen invasion and a facili-
tation of the body’s subsequent immune responses [27].
The stimulation of these receptors activates inflamma-
tory responses characterized by the release of a wide
range of proinflammatory cytokines, including IL-6
Sham GdCl
3
CLP GdCl
3
+ CLP
HMGB1
iNOS
β-actin
Fig. 6. Effects of KCs on serum HMGB1 and hepatic iNOS protein expression levels 24 h after CLP. Rats were pretreated intravenously
with 7.5 mgÆkg
)1
GdCl
3
or saline alone 48 and 24 h before CLP. The values are represented as the mean ± SEM for eight to ten rats per
group. **P < 0.01, significantly different from sham.
++
P < 0.01, significantly different from CLP.
Fig. 7. Effects of KCs on serum TNF-a and IL-6 levels 24 h after
CLP. Rats were pretreated intravenously with 7.5 mgÆkg
)1
GdCl

KCs in drug-metabolizing dysfunction during sepsis T H. Kim et al.
2312 FEBS Journal 278 (2011) 2307–2317 ª 2011 The Authors Journal compilation ª 2011 FEBS
and TNF-a. KCs strongly express all TLRs, except
TLR5 [28]. TLR4 and TLR2 in hepatic and splenic
macrophages were significantly upregulated in mice
with experimental peritonitis induced by CLP [29]. It
has been reported that the regulation of hepatic CYP
gene expression elicited by chemically-induced inflam-
matory bowel disease was entirely dependent on TLR4
[30]. However, hepatic inflammation induced by Cit-
robacter rodentium infection was mainly TLR4-inde-
pendent because hepatic CYPs mRNA expression was
similarly downregulated and cytokine mRNAs were
similarly induced in both wild-type and TLR4-mutant
mice [31]. Recently, the TLR2 ligand, lipoteichoic acid,
altered the expression of hepatic genes associated with
drug metabolism and transport [10]. The results of the
present study show that inactivation of KCs by GdCl
3
pretreatment attenuates any increases in hepatic TLR4
and TLR2 protein expression levels at 24 h after CLP.
KCs mediated the specific downregulation of CYP2B1
via the release of TNF-a in a KCs ⁄ hepatocyte cocul-
ture system [32]. Moreover, proinflammatory cytokines
released from KC, although not the direct effects of
LPS, play an important role in downregulating hepatic
CYP1A2 expression in sepsis [33]. In the present study,
increased serum levels of TNF-a and IL-6 and the pro-
tein expression of iNOS were markedly suppressed by
GdCl

be elucidated; however, our results show that KCs dif-
ferentially regulate the expression of each form of
CYP among the various CYP subfamilies. These dif-
ferential regulations were attributed to the ability of
KCs to develop exaggerated inflammatory responses
through TLR overexpression, the release of HMGB1
and the upregulation of proinflammatory cytokines.
Materials and methods
Animals
Male Sprague-Dawley rats, weighing 280–320 g, were sup-
plied by the Jeil Animal Breeding Company (Deajeon,
Korea). The animals were housed in cages located in tem-
perature controlled rooms under a 12 : 12 h light ⁄ dark
photocycle, and received water and food ad libitum for at
least 1 week. All animal procedures were approved by the
Sungkyunkwan University Animal Care Committee and
were performed in accordance with the guidelines of the
National Institutes of Health.
Treatment with GdCl
3
and experimental groups
For depletion of KCs in vivo, 7.5 mgÆkg
)1
of GdCl
3
was
injected via the tail vein at 48 and 24 h before the perfor-
mance of CLP or sham operation, based on the findings of
other investigators who showed the destruction of KCs
after the intravenous administration of GdCl

abdominal cavity, and the abdominal incision was closed in
two layers. Sham-operated animals underwent the same
surgical procedure, except that the cecum was neither
ligated, nor punctured. All animals received normal saline
(3 mLÆ100 g
)1
body weight) subcutaneously immediately
after surgery (i.e. fluid resuscitation). At 24 h (i.e. late
phase of sepsis) after CLP, blood was obtained from the
T H. Kim et al. KCs in drug-metabolizing dysfunction during sepsis
FEBS Journal 278 (2011) 2307–2317 ª 2011 The Authors Journal compilation ª 2011 FEBS 2313
abdominal aorta. The left and median lobes of the liver
were isolated immediately, and stored at )75 °C until
assayed.
Isolation of hepatic microsomal fraction
The excised liver was minced and then homogenized in four
volumes of ice-cold 1.15% KCl for 1 g of liver, and centri-
fuged at 9000 g for 60 min. The supernatant was collected
and centrifuged at 105 000 g for 60 min, and the precipi-
tates (microsomal fractions) were resuspended with four
volumes of 0.1 m phosphate buffer at pH 7.4, for 1 g of
liver microsome and stored at )75 °C until assayed.
Analytical procedures
Serum ALT and AST activities were determined by stan-
dard spectrophotometric procedures using a diagnostic kit
(Sigma Chemical Co., St Louis, MO, USA). Lipid peroxide
was assayed by the method of Buege and Aust [38], and
1,1,3,3-tetraethoxypropane (MDA tetraethyl acetal) was
used as the standard. Total GSH was determined in liver
homogenates after precipitation with 1% picric acid, using

(Abcam, Cambridge, MA, USA). Binding of all of the anti-
bodies was detected using an ECL detection system (iNtRON
Biotechnology, Seoul, Korea) in accordance with the manu-
facturer’s instructions. The intensity of the immunoreactive
bands was determined using densitometric analysis software
(image gauge, version 3.12; Fujifilm, Tokyo, Japan).
Immunoprecipitation
Liver tissues were homogenized with ice-cold radioimmuno-
precipitation assay (RIPA) buffer (150 mm NaCl, 50 mm
Tris, 1% Triton X-100, 1% deoxycholic acid, 0.1% SDS,
pH 7.4) containing protease and phosphatase inhibitor
cocktail set (Calbiochem, La Jolla, CA, USA). Aliquots of
500 lg of protein diluted to 1 mgÆmL
)1
in RIPA buffer were
precleared with protein A ⁄ G agarose beads (Santa Cruz
Biotechnology) for 30 min and then incubated overnight at
4 °C with anti-CYP1A1 ⁄ 1A2 serum (Abcam) with a con-
stant rotation of the samples. Protein A ⁄ G agarose beads
was then added, and the samples were incubated for a fur-
ther 3 h at 4 °C with constant rotation. The immune com-
plexes were washed three times in a RIPA buffer for 30 s.
After the third wash, the immunoprecipitants were resus-
pended in Laemmli sample buffer. The samples were then
analyzed by western blotting using the polyclonal anti-phos-
phoserine ⁄ threonine serum (Abcam) as the primary anti-
body. Binding of all of the antibodies was detected using an
ECL detection system (iNtRON Biotechnology) in accor-
dance with the manufacturer’s instructions. The intensity of
the immunoreactive bands was determined using densito-

After RT-PCR, 10 lL samples of the amplified products
were resolved by electrophoresis in 1.5% agarose gel, and
stained with ethidium bromide. The intensity of each PCR
product was semi-quantitatively evaluated using a digital
camera (DC120; Eastman Kodak, New Haven, CT, USA)
and densitometric scanning analysis software (1d main;
Advanced American Biotechnology, Fullerton, CA, USA).
Statistical analysis
All results are presented as the mean ± SEM. Overall sig-
nificance was tested by one-way analysis of variance.
P < 0.05 was considered statistically significant for differ-
ences between groups at specific time points, with the
appropriate Bonferroni correction being made for multiple
comparisons.
Acknowledgements
This research was supported by Basic Science Research
Program through the National Research Foundation
of Korea (NRF) funded by the Ministry of Education,
Science and Technology (2010-0028646).
References
1 Jacob A, Zhou M, Wu R & Wang P (2009) The role of
hepatic cytochrome P-450 in sepsis. Int J Clin Exp Med
2, 203–211.
2 Llewelyn M & Cohen J (2002) Superantigens: microbial
agents that corrupt immunity. Lancet Infect Dis 2, 156–
162.
3 Aninat C, Seguin P, Descheemaeker PN, Morel F, Mal-
ledant Y & Guillouzo A (2008) Catecholamines induce
an inflammatory response in human hepatocytes. Crit
Care Med 36, 848–854.

12 Ojaniemi M, Liljeroos M, Harju K, Sormunen R, Vuol-
teenaho R & Hallman M (2006) TLR-2 is upregulated
Table 5. PCR primers used in the present study.
Gene
(accession
number)
Primer sequences
(5¢-to3¢)
Product
length (bp)
CD163
(XM_053094.2)
Sense:
AGCTGGGCTGTGCAGACAACG
Antisense:
TGAATGACCCCCGAGGATTTCAGC
736
CYP1A1
(X00469)
Sense:
CTGGTTCTGGATACCCAGCTG
Antisense:
CCTAGGGTTGGTTACCAGG
331
CYP1A2
(X01031)
Sense:
CAGTCACAACAGCCATCTTC
Antisense:
CCACTGCTTCTCATCATGGT

(BC063166)
Sense:
TTGTAACCAACTGGGACGATATGG
Antisense:
GATCTTGATCTTCATGGTGCTAG
764
T H. Kim et al. KCs in drug-metabolizing dysfunction during sepsis
FEBS Journal 278 (2011) 2307–2317 ª 2011 The Authors Journal compilation ª 2011 FEBS 2315
and mobilized to the hepatocyte plasma membrane in
the space of Disse and to the Kupffer cells TLR-4
dependently during acute endotoxemia in mice. Immunol
Lett 102, 158–168.
13 Zhao ZF, Han DW, Liu MS, Zhang GY, Zhang Y,
Yang H & Yang LX (2007) [Expression of HMGB-1
and its extracellular release of cultured primary hepatic
parenchymal cells and Kupffer cells induced by LPS].
Zhonghua Gan Zang Bing Za Zhi 15, 676–680.
14 Gregory SH & Wing EJ (2002) Neutrophil-Kupffer cell
interaction: a critical component of host defenses to sys-
temic bacterial infections. J Leukoc Biol 72, 239–248.
15 Hardonk MJ, Dijkhuis FW, Hulstaert CE & Koudstaal
J (1992) Heterogeneity of rat liver and spleen macro-
phages in gadolinium chloride-induced elimination and
repopulation. J Leukoc Biol 52, 296–302.
16 Xu DX, Wei W, Sun MF, Wu CY, Wang JP, Wei LZ
& Zhou CF (2004) Kupffer cells and reactive oxygen
species partially mediate lipopolysaccharide-induced
downregulation of nuclear receptor pregnane X receptor
and its target gene CYP3a in mouse liver. Free Radic
Biol Med 37, 10–22.

Mechanisms of down-regulation of CYP2E1 expression
by inflammatory cytokines in rat hepatoma cells. J
Pharmacol Exp Ther 304, 1048–1054.
27 Kumar H, Kawai T & Akira S (2009) Toll-like recep-
tors and innate immunity. Biochem Biophys Res Com-
mun 388, 621–625.
28 Wu J, Meng Z, Jiang M, Zhang E, Trippler M, Broer-
ing R, Bucchi A, Krux F, Dittmer U, Yang D et al.
(2010) Toll-like receptor-induced innate immune
responses in non-parenchymal liver cells are cell type-
specific. Immunology 129, 363–374.
29 Tsujimoto H, Ono S, Majima T, Kawarabayashi N, Ta-
kayama E, Kinoshita M, Seki S, Hiraide H, Moldawer
LL & Mochizuki H (2005) Neutrophil elastase, MIP-2,
and TLR-4 expression during human and experimental
sepsis. Shock 23, 39–44.
30 Chaluvadi MR, Nyagode BA, Kinloch RD & Morgan
ET (2009) TLR4-dependent and -independent regula-
tion of hepatic cytochrome P450 in mice with chemi-
cally induced inflammatory bowel disease. Biochem
Pharmacol 77, 464–471.
31 Richardson TA, Sherman M, Antonovic L, Kardar SS,
Strobel HW, Kalman D & Morgan ET (2006) Hepatic
and renal cytochrome p450 gene regulation during cit-
robacter rodentium infection in wild-type and toll-like
receptor 4 mutant mice. Drug Metab Dispos 34,
354–360.
32 Milosevic N, Schawalder H & Maier P (1999) Kupffer
cell-mediated differential down-regulation of cyto-
chrome P450 metabolism in rat hepatocytes. Eur J

purification, and properties. J Biol Chem 239, 2379–
2385.
41 Vermilion JL & Coon MJ (1978) Identification of the
high and low potential flavins of liver microsomal
NADPH-cytochrome P-450 reductase. J Biol Chem 253,
8812–8819.
42 Burke MD, Thompson S, Elcombe CR, Halpert J,
Haaparanta T & Mayer RT (1985) Ethoxy-, pentoxy-
and benzyloxyphenoxazones and homologues: a series
of substrates to distinguish between different induced
cytochromes P-450. Biochem Pharmacol 34, 3337–3345.
43 Schenkman JB, Remmer H & Estabrook RW (1967)
Spectral studies of drug interaction with hepatic micro-
somal cytochrome. Mol Pharmacol 3, 113–123.
44 Chomczynski P & Sacchi N (1987) Single-step method
of RNA isolation by acid guanidinium thiocyanate-
phenol-chloroform extraction. Anal Biochem 162,
156–159.
T H. Kim et al. KCs in drug-metabolizing dysfunction during sepsis
FEBS Journal 278 (2011) 2307–2317 ª 2011 The Authors Journal compilation ª 2011 FEBS 2317