Comparative studies of endonuclease I from cold-adapted
Vibrio salmonicida and mesophilic Vibrio cholerae
Bjørn Altermark
1
, Laila Niiranen
2
, Nils P. Willassen
1,2
, Arne O. Smala
˚
s
1
and Elin Moe
1
1 Norwegian Structural Biology Centre, Faculty of Science, University of Tromsø, Norway
2 Department of Molecular Biotechnology, Faculty of Medicine, University of Tromsø, Norway
The marine and estuarine environment harbors a vast
diversity of bacteria. Some of the most extensively
studied marine or estuarine bacteria belong to the
genus Vibrio, with Vibrio cholerae being the most
notorious species as it is the cause of cholera in
humans. V. cholerae is found in tropical and temper-
ate areas, and can be classified as a mesophilic bac-
terium with growth optimum around 37 °C. It prefers
estuarine waters, is halotolerant, and does not require
NaCl for growth [1,2]. The bacterium with one of the
lowest growth optimum temperatures found in the
genus Vibrio is the fish pathogen Vibrio salmonicida.
It has an optimal growth temperature of % 15 °C and
requires NaCl for growth [3]. It can therefore be
classified as a psychrophilic and mildly halophilic

2006)
doi:10.1111/j.1742-4658.2006.05580.x
Endonuclease I is a periplasmic or extracellular enzyme present in many
different Proteobacteria. The endA gene encoding endonuclease I from the
psychrophilic and mildly halophilic bacterium Vibrio salmonicida and from
the mesophilic brackish water bacterium Vibrio cholerae have been cloned,
over-expressed in Escherichia coli, and purified. A comparison of the enzy-
matic properties shows large differences in NaCl requirements, optimum
pH, temperature stability and catalytic efficiency of the two proteins. The
V. salmonicida EndA shows typical cold-adapted features such as lower
unfolding temperature, lower temperature optimum for activity, and higher
specific activity than V. cholerae EndA. The thermodynamic activation
parameters confirm the psychrophilic nature of V. salmonicida EndA with
a much lower activation enthalpy. The optimal conditions for enzymatic
activity coincide well with the corresponding optimal requirements for
growth of the organisms, and the enzymes function predominantly as
DNases at physiological concentrations of NaCl. The periplasmic or extra-
cellular localization of the enzymes, which renders them constantly exposed
to the outer environment of the cell, may explain this fine-tuning of bio-
chemical properties.
Abbreviations
DSC, differential scanning calorimetry; VcEndA, recombinant Vibrio cholerae endonuclease I; VsEndA, recombinant Vibrio salmonicida
endonuclease I.
252 FEBS Journal 274 (2007) 252–263 ª 2006 The Authors Journal compilation ª 2006 FEBS
by a more flexible structure, and the increased flexi-
bility is thought to be the reason for the lower thermo-
stability of cold-adapted enzymes [4].
Endonuclease I is a periplasmic or extracellular
enzyme known to cleave both RNA and DNA at
unspecific internal (endo) sites. It also cleaves plasmids

expressed in V. vulnificus [5] and Erwinia chrysanthemi
[18].
Here we report the cloning, expression and purifica-
tion of the endonuclease I enzymes from the psychro-
phile V. salmonicida (VsEndA) and the mesophile
V. cholerae (VcEndA). The two orthologous enzymes
have been biochemically and biophysically character-
ized to reveal possible differences related to environ-
mental adaptation.
Results
Sequence similarity and composition
VsEndA and VcEndA show 71% identity and 80%
similarity (Blosum62) at the amino acid level, when the
active enzymes are compared without their N-terminal
signal peptide. A sequence alignment of VcEndA and
VsEndA is shown in Fig. 1. The first two amino acids
at the N-terminus (Thr and Met) are encoded by the
expression vector.
An analysis of the amino-acid composition shows
that VsEndA contains 13 more lysines and two fewer
arginines than Vc EndA, resulting in a very high R ⁄ K
ratio for the mesophilic enzyme (1.6 versus 0.6,
respectively). In addition VsEndA contains two less
D + E than VcEndA. However all the cysteines
involved in disulfide bridge formation in VcEndA are
also found in the sequences of VsEndA (Fig. 1). The
theoretical pI was 9.61 for VsEndA and 8.62 for
VcEndA.
Expression and purification
From 7 L Escherichia coli culture, a total of 24 and

as shown in Fig. 4. The pH optimum was unaffected
by the NaCl concentration in the buffers. When tested
in glycine buffer at pH 9.0, the enzymes showed very
low activity compared with that in diethanolamine and
Tris buffers at the same pH, indicating that glycine
inhibits the enzymes. VcEndA activity decreases stee-
ply below pH 6.5 (measured in Bis-Tris buffer, data
not shown).
The optimum temperature for activity was deter-
mined using a modified Kunitz assay. The results
showed optimum activity at % 45 °C for VsEndA and
50 °C for VcEndA, as shown in Fig. 5.
Kinetic constants for VsEndA and VcEndA were
measured by incubating the enzymes in the presence of
substrate with different concentrations and at different
temperatures. The kinetic constants for the two
enzymes at 5, 15, 25, 30 and 37 °C are shown in
200
116.3
97.4
66.3
55.4
36.5
31.0
21.5
14.4
6.0
3.5
Fig. 2. SDS ⁄ PAGE. Lane 1, Mark12 MW ladder; lane 2, % 5 lg
VcEndA; lane 3, % 5 lg VsEndA. The relative molecular masses of

pH optimum VcEndA
6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5
1.0
1.5
2.0
2.5
3.0
Tris
DEA
pH
Activity (Rfu/s)
Fig. 4. Optimum pH for activity. (A) VsEndA; (B) VcEndA. Buffers
used are 75 m
M Tris ⁄ HCl, pH 7–9, and 75 mM diethanolamine ⁄ HCl,
pH 8–10. DNaseAlert was used as a substrate in the assay. Each
replicate is plotted and the mean values are drawn.
Endonuclease I from V. salmonicida and V. cholerae B. Altermark et al.
254 FEBS Journal 274 (2007) 252–263 ª 2006 The Authors Journal compilation ª 2006 FEBS
Table 1. VsEndA possesses a higher k
cat
than VcEndA
at all temperatures, and the K
m
values of VcEndA are
slightly lower than for VsEndA at all temperatures.
The physiological efficiency is highest for VsEndA, but
the difference decreased with concomitant increase in
temperature.
As determined from Arrhenius plots, the energy
of activation (E

The rate of irreversible unfolding was analyzed by
incubating both enzymes at 70 °C. Samples were
removed after 10 min and incubated for 1 h on ice
0 102030405060
0
25
50
75
100
VsEndA
VcEndA
Temperature (°C)
Activity (%)
Fig. 5. Optimum temperature for activity. The enzymes were
assayed using the modified Kunitz assay. Each replicate is plotted
and the mean values are drawn.
Table 1. Kinetic constants for VsEndA and VcEndA at 5, 15, 25, 30 and 37 °C.
T(°C) VsEndA VcEndA VsEndA ⁄ VcEndA
K
m
(nM) 5 246 ± 15 118 ± 13 2.1
15 202 ± 9.6 131 ± 10 1.5
25 169 ± 20 156 ± 17 1.1
30 208 ± 14 161 ± 12 1.3
37 181 ± 10 174 ± 10 1.0
k
cat
(s
)1
) 5 9.41 1.03 9.1

)
p-m
D(DH
#
)
p-m
TD(DS
#
)
p-m
5 p 62.8 33.4 ) 29.4 ) 5.1 ) 40.6 ) 35.4
m 67.9 74.0 6.1
15 p 64.0 33.3 ) 30.7 ) 3.7 ) 40.6 ) 36.8
m 67.8 73.9 6.1
25 p 65.8 33.2 ) 32.6 ) 2.3 ) 40.6 ) 38.2
m 68.1 73.8 5.7
30 p 65.5 33.2 ) 32.3 ) 1.9 ) 40.6 ) 38.7
m 67.4 73.7 6.4
37 p 66.1 33.1 ) 32.9 ) 1.1 ) 40.6 ) 39.5
m 67.1 73.7 6.6
B. Altermark et al. Endonuclease I from V. salmonicida and V. cholerae
FEBS Journal 274 (2007) 252–263 ª 2006 The Authors Journal compilation ª 2006 FEBS 255
before being assayed. Figure 7 shows that the rate
of irreversible unfolding for VsEndA is higher than
for Vc EndA, with a half-life of % 13 and 33 min,
respectively.
Substrate specificity analysis
An analysis of the substrate specificity for DNA of the
enzymes shows that they both cleave plasmid DNA,
dsDNA and ssDNA as shown in Fig. 8.

enzymes may be involved in temperature adaptation;
however, two properties, which are not related to tem-
perature adaptation, clearly distinguish these enzymes.
The two enzymes respond notably differently to varia-
tions in both NaCl concentration and pH.
A notable increase in activity against the DNase-
Alert substrate was observed for the two enzymes
when NaCl was added to the assay buffer. The optimal
NaCl concentrations coincide with the salinities
encountered by the bacteria in their natural habitats.
Seawater at 3.5% salinity is composed of about
470 mm Na
+
ions and 540 mm Cl
–
ions [20]. The
30 40 50 60
0
10
20
30
40
50
60
VsEndA
VcEndA
Temperature (°C)
Cp (kJ mol
-1
K

buffer. The optima may be higher in a Tris buffer of
lower ionic strength, but this was not tested. Two
terrestrial orthologous endonucleases, one from the
plant pathogen Erwinia chrysanthemi and one from the
ruminal bacterium Fibrobacter succinogenes, are also
described in the literature [11,21]. The optimum
concentrations of NaCl for these enzymes are 0–
75 mm and 10 mm, respectively, with DNA as sub-
strate. It seems that the salt optima of the enzymes are
fine-tuned to match the salinity of their environment.
The outer membrane and cell wall of Gram-negative
bacteria do not restrict passage of ions, and the peri-
plasmic proteins are, like the extracellular proteins,
constantly exposed to the salinity of the surrounding
water. Knowledge on cold adaptation is in many
cases based on marine secreted enzymes. Detailed data
on salt adaptation of marine cold-adapted secreted
enzymes is lacking and may be a source of error in
the conclusions drawn [22]. For the endonuclease I
enzymes studied here, the effect of NaCl is very prom-
inent and underlines the need to dissect the different
adaptational strategies in future studies. The differ-
ences observed in the number of charged residues,
especially lysine, are probably related to adaptation to
both salinity and temperature. The K
m
of VsEndA is
higher than that of VcEndA; therefore, the more posit-
ive surface of VsEndA does not seem to significantly
increase the affinity for the negatively charged sub-

around the phosphates of DNA before
catalysis can take place. It seems clear that the salt-
adapted and cold-adapted properties of VsEndA are
intertwined.
The differences in optimum pH for activity were
% 0.5–1 unit between the two enzymes as shown in
Fig. 4, with the optimum for VsEndA being shifted to
VsEndA
A
0 125 250 375 500
0.0
0.3
0.6
0.9
1.2
1.5
1.8
RNase
DNase
[NaCl] (mM)
Rfu/s
B
VcEndA
0 50 100 150 200
0.0
0.3
0.6
0.9
1.2
1.5

the differences in catalytic efficiency (k
cat
⁄ K
m
) increase
with concomitant decrease in temperature. K
m
values
for VcEndA are lower than for VsEndA, indicating
that the former has slightly greater affinity for the sub-
strate. However, k
cat
is very different for the two
enzymes, especially at low temperatures, being 9 times
higher for VsEndA than for VcEndA. It is clear that
VsEndA adapts to lower temperatures by increasing
the k
cat
. The similar K
m
values of the two enzymes
may indicate that VsEndA is meant to function at high
substrate concentrations, at which the increase in k
cat
is more important for adaptation to low temperatures
[25]. The k
cat
values associated with both VsEndA
and VcEndA increase exponentially at temperatures
between 5 °C and 37 °C in accordance with the

(and flexible) VsEndA. However, the method of calcu-
lation, especially for DS
#
, must be carefully interpreted
as stated by Cornish-Bowden [26]. Enthalpy calcula-
tions based on the experimentally determined values of
E
a
give more precise information, and it is clear that
VsEndA has adapted to low temperatures by lowering
the enthalpy of activation.
DSC measurements show that VsEndA is less ther-
mostable than VcEndA with an unfolding temperature
that is 8 °C lower. This is in agreement with results
from stability analysis of other cold-adapted enzymes,
which show reduced temperature stability compared
with their mesophilic homologues [27,28]. The results
support the theory of increased structural flexibility
leading to lower thermostability in cold-adapted
enzymes. The NaCl concentrations in which the ther-
mal scans were performed mimic the physiological con-
ditions that each of the enzymes face in their natural
environments. Thermal scans of VcEndA at [NaCl]
optimal for VsEndA (425 mm) revealed a higher T
m
,
and a thermal scan performed on VsEndA at [NaCl]
optimal for VcEndA (175 mm) revealed a lower T
m
than those found in optimal buffers (data not shown).

played decreasing activity against the RNaseAlert sub-
strate with concomitant increase in [NaCl], as shown
in Fig. 9. At physiological NaCl concentration, the two
enzymes have extremely low RNase activity and may be
Endonuclease I from V. salmonicida and V. cholerae B. Altermark et al.
258 FEBS Journal 274 (2007) 252–263 ª 2006 The Authors Journal compilation ª 2006 FEBS
considered solely as DNases. The highest RNase activity
is in buffer without added NaCl, but it is still % 3.5 times
(VsEndA) and 14 times (VcEndA) lower than the
DNase activity in the same buffer. The opposite effect
that NaCl addition seems to have on the RNase activity
of the enzymes may be linked to an increase in Na
+
around the phosphate groups and the 2¢-OH, which
reduces the negative charge, and hence the affinity of the
enzyme decreases with increasing NaCl concentration.
However, it seems clear that both VsEndA and VcEndA
are intended to function purely as DNases in vivo.
Conclusion
Endonuclease I from the psychrophilic bacterium
V. salmonicida is an enzyme that shows cold-adapted
features, such as lower thermal stability, lower tem-
perature optimum, and higher catalytic efficiency,
when compared with the corresponding enzyme from
the related mesophilic bacterium V. cholerae. The peri-
plasmic or extracellular localization of these enzymes
means that they are constantly exposed to the external
environment of the bacterium. Their differences in
enzymatic properties, such as pH optimum, salt opti-
mum and catalytic efficiency, seem to be fine-tuned to

and
RNaseAlert
TM
QC System kit was purchased from Ambion
Inc. (Austin, TX, USA) and Integrated DNA Technologies
(Coralville, IA, USA).
Construction of the expression plasmids
The nucleotide sequences of VsEndA and VcEndA have the
GenBank accession nos. DQ263597 and DQ263605,
respectively.
To facilitate cloning of the VsEndA gene into the pBAD ⁄
gIII b vector, a restriction site for SalI was first removed by
point mutation using the overlap extension procedure [30].
PCR was conducted using primers 3 +4 and 1 +4 (Table 3),
with genomic DNA from V. salmonicida as a template. In a
0.2-mL PCR tube, a total of 50 lL reaction mix containing
37.5 lL water, 5 lL10· ThermoPol reaction buffer, 3 lL
25 mm MgCl
2
,1lL10mm dNTP, 1 lL each primer
(10 lm), 1 lL template, and 1 U Vent polymerase was sub-
jected to PCR using a DNA Engine (PTC-200) Peltier Ther-
mal Cycler from Bio-Rad (Hercules, CA, USA). Thermal
cycling conditions were 3 min at 94 °C followed by 30 cycles
of 30 s at 94 °C, 30 s at 50 °C and 90 s at 72 °C. The pro-
gram was ended by an extension step at 72 °C for 5 min, and
then cooled to 4 °C. This PCR yielded one 656-bp and a 254-
bp product when run on a 1% agarose gel. The 656-bp frag-
ment was used as a template in a second PCR conducted
under the same conditions as above, but with primers 3 +2.

extraction kit. Vector and insert were ligated overnight at
16 °C using T4 DNA ligase before transformation into
E. coli Top10 cells using the heat shock method. Positive col-
onies resistant to ampicillin were selected and used for
expression. The VcEndA was cloned using the same proce-
dure as for VsEndA, but no mutation was necessary. The clo-
ning primers for VcEndA are listed in Table 3. The plasmids
were thereafter sequenced using the PE Biosystems BigDye
Terminator Cycle Sequencing kit, ABI 377 Genetic Analyzer
and ABI Sequence Analysis software according to the proto-
col supplied by Applied Biosystems (Foster City, CA, USA).
Enzyme expression and purification
A Chemap CF 3000 fermentor (Chemap AG,
1
Volketswil,
Switzerland) was used for production of the recombinant
nucleases. First 7 L 2 · Luria–Bertani medium supplemen-
ted with 60 mL 20% glucose was inoculated with a 200-mL
overnight preculture and grown at 22 °C. The enzyme pro-
duction was induced by adding 50 mL 14.5% l-arabinose
when the glucose was depleted. The pH was held constant
at 7.4 by addition of 1 m NaOH or 2 m H
2
SO
4
. Oxygen
levels were automatically adjusted by increasing agitation
speed when the level went below 20% of maximum. The
cells were harvested 7 h after induction by centrifugation at
4225 g for 15 min at 4 °C. The cells were subjected to a

protein concentration was determined using Bio-Rad Pro-
tein Assay based on the method of Bradford [32] and
according to the microtiter plate protocol described by the
manufacturer using BSA as standard. N-Terminal signal
sequence cleavage sites were predicted using the SignalP
server [33]. Sequence alignment was performed using
BioEdit [34], and the alignment was visualized using the
ESPript server [35]. Theoretical isoelectric point, molecular
mass and sequence composition were calculated using the
protparam web-tool at ExPASy [36].
Enzyme assay
The DNaseAlert
TM
QC System kit was used in the deter-
mination of kinetic constants, pH optimum and optimum
NaCl concentration of the two enzymes. The DNase-
Alert
TM
substrate is a synthetic DNA oligonucleotide that
has a HEX
TM
reporter dye (hexachlorofluorescein) on one
end and a dark quencher on the other end. In all reactions,
except for the kinetic measurements, 200 nm substrate was
used. The reaction volumes were adjusted to 90 lL with
nuclease-free water. Reactions were started by pipetting
10 lL of the diluted enzyme solution into eight wells with a
multichannel pipette to a total reaction volume of 100 lL.
Non-binding 1.5-mL tubes from Eppendorf (Hamburg,
Germany) were used for enzyme dilution. New dilutions

2
; VcEndA, 175 mm
NaCl ⁄ 20 mm Tris ⁄ HCl (pH 8.0) ⁄ 5mm MgCl
2
], and the
buffer pH was adjusted at the respective assay tempera-
tures. The total reaction volume was 1 mL. Reaction mix-
tures were preincubated for 5–10 min at the respective
assay temperatures before the addition of enzyme. The
same amount of enzyme (VsEndA 1.5 ng, VcEndA 4.2 ng)
was used at each temperature. Reactions were allowed to
proceed for 20 min and then stopped by adding 0.5 mL ice-
cold 12% perchloric acid. For blank reactions, enzyme was
added after the addition of perchloric acid. Quenched assay
solutions were incubated on ice for 20 min, centrifuged for
10 min at 16 000 g, and the A
260
was measured for the
supernatants in triplicate.
Enzyme kinetic measurements
Fixed amounts of enzyme were incubated at seven different
substrate concentrations ranging from 23 to 1470 nm at 5,
15, 25, 30 and 37 °C in a total reaction volume of 100 lL.
The amounts of VsEndA enzyme used were 0.69, 0.44,
0.21, 0.15 and 0.069 ng at 5, 15, 25, 30 and 37 °C, respect-
ively. For VcEndA the amounts used at these temperatures
were 4.2, 0.21, 0.56, 0.31 and 0.14 ng, respectively. Assay
buffer was optimal for each enzyme [VsEndA, 425 mm
NaCl ⁄ 75 mm diethanolamine (pH 8.5) ⁄ 5mm MgCl
2

unitsÆs
)1
to nmolÆs
)1
. The calculated molecular masses
for VsEndA and VcEndA were 25005.41 gÆmol
)1
and
24731.72 gÆmol
)1
, respectively.
Thermodynamic activation parameters were calculated as
described by Lonhienne et al. [37]. Activation energy, E
a
,
was extracted from the slope of the linear regression curve
obtained from an Arrhenius plot of 1 ⁄ T versus lnk
cat
.
Stability measurements
DSC measurements were performed using the Nano-Differ-
ential Scanning Calorimeter III, model CSC6300 (Calori-
metry Sciences Corp., Lindon, UT, USA). The IUPAC
(International Union of Pure and Applied Chemistry)
recommendations for DSC measurements and analysis [38]
were used as a guideline. The scan rate was set to
1 °CÆmin
)1
, and the scans were performed from 25 to 85 °C
at a constant pressure of 304 kPa. All samples were dia-

tured by incubation at 98 °C in a PCR machine for 3 min,
and then kept on ice. Approximately 300 ng of the various
substrates was mixed with 30 ng enzyme in a total volume
of 20 lL containing 1 mm MgCl
2
and 75 mm diethanolam-
ine buffer with optimal [NaCl] and pH for each enzyme.
After 5 min of incubation at 23 °C, the reaction was
stopped by the addition of 5 lL 0.5 m EDTA. The samples
were analyzed on a 1% agarose gel for 1 h at 90 V and
visualized by ethidium bromide staining. The substrates
were also incubated without enzyme as a reference.
Activity towards RNA was measured using the RNase-
Alert QC System kit with the same instrumental set up as
for the DNaseAlert system mentioned above, except that
the wavelengths used for excitation ⁄ emission were
490 ⁄ 520 nm, respectively. Measurements were taken every
64 s for 20 min. The effect of [NaCl] on the RNase activity
was measured in 75 mm diethanolamine ⁄ HCl at pH 8.5 for
VsEndA and pH 8.0 for VcEndA with increasing concen-
trations of NaCl (0–425 mm for VsEndA, 0–175 mm for
VcEndA) including 5 mm MgCl
2
per 100 lL reaction mix-
ture. The maximum fluorescence obtained with 200 nm
RNaseAlert and DNaseAlert was measured by adding 5 lL
RNase A (0.01 UÆmL
)1
) to wells with RNaseAlert substrate
after the initial measurements and 2 lL undiluted VcEndA

Cloning and characterization of a periplasmic nuclease
of Vibrio vulnificus and its role in preventing uptake of
foreign DNA. Appl Environ Microbiol 67, 82–88.
6 Li CL, Hor LI, Chang ZF, Tsai LC, Yang WZ & Yuan
HS (2003) DNA binding and cleavage by the periplas-
mic nuclease Vvn: a novel structure with a known active
site. EMBO J 22 , 4014–4025.
7 Altermark B, Smalas AO, Willassen NP & Helland R
(2006) The structure of Vibrio cholerae extracellular
endonuclease I reveals the presence of a buried
chloride ion. Acta Crystallogr D Biol Crystallogr 62,
1387–1391.
8 Chang MC, Chang SY, Chen SL & Chuang SM (1992)
Cloning and expression in Escherichia coli of the gene
encoding an extracellular deoxyribonuclease (DNase)
from Aeromonas hydrophila. Gene 122, 175–180.
9 Focareta T & Manning PA (1987) Extracellular proteins
of Vibrio cholerae: molecular cloning, nucleotide
sequence and characterization of the deoxyribonuclease
(DNase) together with its periplasmic localization in
Escherichia coli K-12. Gene 53, 31–40.
10 Lehman IR, Roussos GG & Pratt EA (1962) The
deoxyribonucleases of Escherichia coli. II. Purification
and properties of a ribonucleic acid-inhibitable endonu-
clease. J Biol Chem 237, 819–828.
11 MacLellan SR & Forsberg CW (2001) Properties of the
major non-specific endonuclease from the strict anae-
robe Fibrobacter succinogenes and evidence for disulfide
bond formation in vivo. Microbiology 147, 315–323.
12 Nakasone K, Kato C & Horikoshi K (2002) Cloning

of low temperature enzyme activity. Biochim Biophys
Acta 1543, 417–433.
20 Turekian KK (1968) Oceans. Prentice Hall, Englewood
Cliffs, NJ.
21 Moulard M, Condemine G, Nasser W & Robert-Bau-
douy J (1995) Purification and characterization of the
nuclease NucM of Erwinia chrysanthemi. Biochim Bio-
phys Acta 1262, 133–138.
22 Siddiqui KS & Cavicchioli R (2006) Cold-adapted
enzymes. Annu Rev Biochem 75, 403–433.
23 Mevarech M, Frolow F & Gloss LM (2000) Halophilic
enzymes: proteins with a grain of salt. Biophys Chem
86, 155–164.
24 Oren A, Larimer F, Richardson P, Lapidus A &
Csonka LN (2005) How to be moderately halophilic
with broad salt tolerance: clues from the genome of
Chromohalobacter salexigens. Extremophiles 9, 275–279.
25 Feller G & Gerday C (1997) Psychrophilic enzymes:
molecular basis of cold adaptation. Cell Mol Life Sci
53, 830–841.
26 Cornish-Bowden A (2002) Enthalpy-entropy compensa-
tion: a phantom phenomenon. J Biosci 27, 121–126.
Endonuclease I from V. salmonicida and V. cholerae B. Altermark et al.
262 FEBS Journal 274 (2007) 252–263 ª 2006 The Authors Journal compilation ª 2006 FEBS
27 Bae E & Phillips GN Jr (2004) Structures and analysis
of highly homologous psychrophilic, mesophilic, and
thermophilic adenylate kinases. J Biol Chem 279,
28202–28208.
28 Georlette D, Damien B, Blaise V, Depiereux E, Uversky
VN, Gerday C & Feller G (2003) Structural and func-

Acids Res 31, 3784–3788.
37 Lonhienne T, Gerday C & Feller G (2000) Psychrophilic
enzymes: revisiting the thermodynamic parameters of
activation may explain local flexibility. Biochim Biophys
Acta 1543, 1–10.
38 Hinz HJ & Schwarz FP (2001) Measurement and analy-
sis of results obtained on biological substances with
differential scanning calorimetry. Pure Appl Chem 73,
745–759.
B. Altermark et al. Endonuclease I from V. salmonicida and V. cholerae
FEBS Journal 274 (2007) 252–263 ª 2006 The Authors Journal compilation ª 2006 FEBS 263