Comparative analysis and mutation effects of
fpp2–fpp1 tandem genes encoding proteolytic
extracellular enzymes of Flavobacterium
David Pe ´rez-Pascual,1Esther Go ´mez,1Beatriz A´lvarez,2Jessica Me ´ndez,1
Pilar Reimundo,1Roberto Navais,1Eric Duchaud3and Jose ´ A. Guijarro1
Jose ´ A. Guijarro
firstname.lastname@example.org or jaga@uniovi.
Received 22 November 2010
Revised13 January 2011
Accepted 1 February 2011
1A´rea de Microbiologı ´a, Departamento de Biologı ´a Funcional, Facultad de Medicina, IUBA,
Universidad de Oviedo, 33006 Oviedo, Spain
2Karolinska Institutet, Institutionen fo ¨r Laboratoriemedicin, Karolinska Universitetssjukhuset, 14186
3Unite ´ de Virologie et Immunologie Mole ´culaires, Equipe Infection et Immunite ´ des Poissons,
INRA-Domaine de Vilvert, 78352 Jouy en Josas Ce ´dex, France
Flavobacterium psychrophilum is a very significant fish pathogen that secretes two biochemically
characterized extracellular proteolytic enzymes, Fpp1 and Fpp2. The genes encoding these
enzymes are organized as an fpp2–fpp1 tandem in the genome of strain F. psychrophilum
THC02/90. Analysis of the corresponding encoded proteins showed that they belong to two
different protease families. For gene function analysis, new genetic tools were developed in F.
psychrophilum by constructing stable isogenic fpp1 and fpp2 mutants via single-crossover
homologous recombination. RT-PCR analysis of wild-type and mutant strains suggested that both
genes are transcribed as a single mRNA from the promoter located upstream of the fpp2 gene.
Phenotypic characterization of the fpp2 mutant showed lack of caseinolytic activity and higher
colony spreading compared with the wild-type strain. Both characteristics were recovered in the
complemented strain. One objective of this work was to assess the contribution to virulence of
these proteolytic enzymes. LD50experiments using the wild-type strain and mutants showed no
significant differences in virulence in a rainbow trout challenge model, suggesting instead a
possible nutritional role. The gene disruption procedure developed in this work, together with the
knowledge of the complete genome sequence of F. psychrophilum, open new perspectives for
the study of gene function in this bacterium.
Flavobacterium psychrophilum is the aetiological agent of
‘cold water disease’ and ‘rainbow trout fry syndrome’ in
salmonid fish. The disease mainly affects fingerlings, and is
currently considered one of the most significant bacterial
infections in the salmonid industry, causing high mortality
and serious economic losses worldwide. Although F.
psychrophilum is considered a rather fastidious bacterium
because it is difficult to isolate and handle (Michel et al.,
1999), significant advances have been made over the last
few years in culture (Michel et al., 1999; Cepeda et al.,
2004; A´lvarez & Guijarro, 2007), typing (Izumi et al., 2003;
Soule et al., 2005; Arai et al., 2007; Nicolas et al., 2008),
molecular diagnosis (Urdaci et al., 1998; Cepeda & Santos,
2000; del Cerro et al., 2002; Fujiwara-Nagata & Eguchi,
2009) and genomics (Duchaud et al., 2007). Moreover, a
transformation and transposon mutagenesis has been
developed (A´lvarez et al., 2004). Nevertheless, the very
low frequency of conjugation and the very low number of
F. psychrophilum strains that can be genetically manipu-
lated (A´lvarez et al., 2004) have hampered the development
of mutagenesis techniques and tools. Furthermore, genetic
systems for targeted gene mutation have not been reported.
Several molecules secreted by F. psychrophilum and
involved in adherence (Møller et al., 2003), haemaggluti-
nation and haemolysis (Lorenzen & Olesen, 1997; Møller
et al., 2003) and iron acquisition (A´lvarez et al., 2008) have
been suggested to play a role in virulence. In addition,
different authors have proposed extracellular proteolytic
activities as putative virulence factors (Bertolini et al., 1994;
Madsen & Dalsgaard, 1998). Among them, the extracellular
proteases Fpp1 and Fpp2 have been purified and
enzymically characterized as metalloproteases, and the N-
terminal sequences of both proteins have been determined
(Secades et al., 2001, 2003). In addition, 13 genes, encoding
Microbiology (2011), 157, 1196–1204
1196046938G2011 SGMPrinted in Great Britain
putative extracellular proteolytic activities, including Fpp1
and Fpp2, have been identified in the genome of strains
JIP02/86 (Duchaud et al., 2007) and THC02/90 (E.
Duchaud, unpublished data). There is not conclusive
evidence, however, of their involvement in disease
development. Although extracellular proteolytic enzymes
have been considered as virulence factors in many bacterial
fish pathogens, comparative analyses of wild-type strains
and isogenic mutants devoid of proteolytic activity have
seldom been performed. In particular, proteases from
Vibrio anguillarum (Yang et al., 2007), Yersinia ruckeri
(Ferna ´ndez et al., 2003), Pseudomonas fluorescens (Zhang et
al., 2009) and Moritella viscosa (Bjornsdottir et al., 2009)
have been clearly proved to be involved in bacterial
virulence in fish.
The aim of this study was to develop a disruption system
for specific gene inactivation of F. psychrophilum. For this
task, the fpp1 and fpp2 genes, encoding extracellular
proteolytic activities, were chosen. The results showed that
insertional mutation of both genes was stable in vivo.
Additionally, fpp12and fpp22mutants were used to study
the roles of the corresponding proteases in the physiology
and virulence of the bacterium.
Bacterial strains, plasmids, growth conditions and proteolytic
activity. Escherichia coli strain S17-1lpir (Simon et al., 1983) was
grown at 37 uC in 26 TY medium (per litre: 10 g tryptone, 10 g yeast
extract, 5 g NaCl) with 20 g agar l21added for solid medium. This
strain was used to transfer DNA into F. psychrophilum. The wild-type
strain F. psychrophilum THC02/90 was grown at 12 or 18 uC in
nutrient broth (NB; Pronadisa) or nutrient broth containing 10 mM
CaCl2(NBF) (Secades et al., 2001). Nutrient agar (NA; NB containing
15 g agar l21) or nutrient agar charcoal [NAC; NA supplemented with
activated charcoal (Sigma)] were used for solid cultures, as previously
described (A´lvarez & Guijarro, 2007). For selective growth of E. coli
S17-1lpir, 50 mg streptomycin ml21was used and transformants were
selected with 100 mg ampicillin ml21. Selection of F. psychrophilum
transconjugants was carried out with 10 mg erythromycin ml21and
10 mg tetracycline ml21. The plasmids and primers used are listed in
Colony spreading and biomass production were analysed according
to Pe ´rez-Pascual et al. (2009). Briefly, plates containing one-sixth NA
were inoculated in the centre with 6 ml liquid culture in mid-
exponential growth phase. The plates were then incubated at 20 uC,
and spreading and biomass production were evaluated at 72, 96 and
120 h by measuring the colony diameter (cm) and biomass
production (OD525) of 1 ml aqueous suspensions of colonies.
Extracellular proteolytic activity on solid medium was visualized by
using NBF containing 15 g agar l21and supplemented with 0.75%
(w/v) gelatin or 1% (w/v) casein. In these media, proteolytic activity
was visualized as the formation of clear haloes around the colonies.
Azocasein assays using supernatants of liquid cultures were carried
out according to Secades et al. (2001). One unit of enzyme activity
(enzymic units; EU) was defined as the amount of enzyme which
yielded an increase in the A420of 0.01 in 2 h at 30 uC.
Generation of fpp1”and fpp2”mutants by single-crossover
homologous recombination and complementation of the fpp2”
mutant. Fpp1 and Fpp2 had previously been purified and
biochemically characterized (Secades et al., 2001, 2003). The amino
acid N-terminal sequences for Fpp1 (SSTGQLKTMRLAQS(C/
W)NGQYA; our unpublished results) and Fpp2 (Secades et al.,
2003) had previously been determined and their sequences used to
locate the fpp1–fpp2 tandem genes in the genome of the THC02/90
Internal fragments of 2481 and 1967 bp of the predicted fpp1 and fpp2
genes, respectively, were amplified by PCR with the following
primers: Fpp1-F (nucleotides 315–323) and Fpp1-R (nucleotides
2785–2764) for fpp1 (both with a SalI restriction site), and Fpp2-F
(nucleotides 452–470) and Fpp2-R (nucleotides 2426–2406) for fpp2
(with XbaI and PstI restriction sites) (Table 1). All PCRs were
performed in a Perkin-Elmer 9700 GeneAmp thermocycler with an
initial denaturation cycle at 94 uC for 5 min, followed by 25 cycles of
amplification (denaturation at 94 uC for 30 s, annealing at 50 uC for
30 s, and extension at 72 uC for 3 min), and a final 7 min elongation
step at 72 uC. The generated amplicons were digested with the
corresponding restriction enzymes and ligated into the pLYL03
suicide plasmid, previously digested with the same enzymes and
dephosphorylated when necessary. Each ligation mixture of target
genes was used to transform electrocompetent cells of E. coli S17-
1lpir. Clones carrying pLYL03 containing inserts (pLYL03Fpp1 and
pLY03Fpp2) were used to conjugate with F. psychrophilum THC02/
90, as previously described (A´lvarez et al., 2004), by selecting the
transconjugants using NAC plates supplemented with erythromycin.
Erythromycin-resistant colonies of F. psychrophilum appeared after 7–
10 days of incubation at 20 uC. Selected transconjugant colonies were
analysed by Southern blotting using a DIG DNA detection kit
(Roche) following the protocol described by the manufacturer. Total
DNA from transconjugants and the wild-type strain was isolated and
digested with HindIII and XbaI restriction enzymes. Then, DNA
fragments were subjected to 0.8% agarose gel electrophoresis at 100 V
for 2 h, transferred to a nylon membrane (Amersham Biosciences)
and fixed with UV irradiation. The labelled internal regions of fpp1
(2.5 kb) and fpp2 (1.9 kb) genes were used as probes to perform
hybridization for each mutant.
To complement the fpp22mutant strain, the fpp2 gene was amplified
from the wild-type strain by PCR using an Expand Long Template
PCR system (Roche) and primers Fpp2C-F (nucleotides 2384 to
2366) and Fpp2C-R (nucleotides 3128 to 3146), obtaining a
fragment of 3530 bp (Table 1). The KpnI and SphI restriction sites
were introduced into the sequences of Fpp2C-F and Fpp2C-R,
respectively, to clone the PCR product digested with KpnI and SphI
into pCP23. The resulting plasmid was designated pCP23-Fpp2
(Table 1). Transfer of pCP23-Fpp2 to F. psychrophilum was carried
out by conjugation, as described by A´lvarez et al. (2004), and pCP23-
Fpp2 was recovered from the transconjugants, digested with KpnI and
SphI, and analysed by agarose gel electrophoresis to confirm the
presence of the insert.
RT-PCR. Total RNA was obtained from 2 ml of early stationary phase
cultures of wild-type strain THC02/90 and mutant strains fpp12and
fpp22grown in NBF at 12 uC. RNA was isolated using a High Pure
RNA Isolation kit (Roche) and treated with DNase I (RNase-free)
(Ambion) to eliminate traces of DNA. Reverse transcription (RT-
PCR) was performed using Superscript One-Step with Platinum Taq
(Invitrogen Life Technologies). Ten nanograms of RNA were used in
each reaction. In order to determine whether RNA was free of
contaminant DNA, PCRs using Platinum Taq DNA polymerase
(Invitrogen Life Technologies) were performed and also reactions
omitting the reverse-transcription step were included in each run as
negative controls. The primers used for RT-PCR amplifications were
1F (nucleotides 2109–2032) and 1R (nucleotides 2526–2544) for the
intragenic region of fpp2, 2F (nucleotides 2674–2695) and 2R
Mutation of fpp2–fpp1 genes of F. psychrophilum
(nucleotides 95–75) for the fpp2–fpp1 intergenic region, and finally 3F
(nucleotides 1898–1920) and 3R (nucleotides 3013–3035) for the
intragenic region of fpp1 (Table 1).
Animal experiments. Animal experiments were performed in
accordance with the European legislation governing animal welfare,
Experimentation Ethics Committee of Universidad de Oviedo.
Rainbow trout (Oncorhynchus mykiss) fingerlings weighing between 5
and 7 g used for the animal experiments were obtained from a
commercial fish farm. Fish were kept in 60 l tanks at 14±1 uC in
continually flowing dechlorinated water.
To determine the stability of fpp1 and fpp2 mutations in vivo, two
groups of 10 rainbow trout were injected intramuscularly with 50 ml
of a cell suspension in PBS containing 106c.f.u. from an exponential-
phase culture of each mutant grown in NB. After 24, 48 and 72 h, fish
were euthanized, and muscle, spleen, liver and intestine from each
infected fish were homogenized together in NB. Aliquots of the
suspensions were plated onto NAC and incubated for 5 days at 20 uC.
Randomly taken F. psychrophilum colonies were used to inoculate
flasks containing NB, and cultures were incubated at 12 uC for 24 h.
Total DNA from each culture was obtained and analysed by Southern
blotting as described above. In total, eight different colonies (two per
time point) from each mutant were analysed. As a control for the
recovery of the mutants, erythromycin was used in all the steps,
except during the fish infections. Control fish injected with 106wild-
type strain cells were also analysed.
For LD50determinations, F. psychrophilum cultures were grown to
exponential phase, harvested by centrifugation and washed with PBS.
Cells were resuspended in PBS and serial dilutions were prepared.
Groups of 10 fish were challenged by intramuscular injection of 50 ml
of dilutions containing 103–109c.f.u., and LD50 was calculated
according to the method of Reed & Muench (1938). LD50
experiments were performed twice.
Information (NCBI) Basic Local Alignment Search Tool (BLAST) was
used to compare protein sequences and Simple Modular Architecture
Research Tools (SMART) to detect conserved domains. MotifScan
software from MyHits was used to identify the motifs present in each
sequence. The MEROPS peptidase database was used to classify each
protease. The ProtParam program (ExPASy) was used for molecular
mass computation and SignalP3.0 (Center for Biological Sequence
Analysis; CBS) to predict the location of a signal peptide cleavage site.
The alignment of the protein sequences of both Fpp1 and Fpp2 was
Table 1. Strains, plasmids and primers used in this study
Plasmid, strain or
Description or sequence*Source or reference
E. coli strain
ColE1 ori, (pCP1 ori), Apr(Tcr), E. coli–F. psychrophilum shuttle plasmid
fpp2 gene, derived from pCP23
ColE1; Apr(Emrb); suicide plasmid
fpp1 internal fragment, derived from pLYL03
fpp2 internal fragment, derived from pLYL03
Agarwal et al. (1997)
Li et al. (1995)
fpp1 mutant created by specific homology recombination
fpp2 mutant created by specific homology recombination
Mutant fpp22carrying pCP23-Fpp2, complemented strain
Chakroun et al. (1998)
lpir hsdR pro thi; RP4-2 Tc::Mu Km::Tn7Simon et al. (1983)
*Antibiotic-resistance phenotypes: Apr, ampicillin; Tcr, tetracycline; Emrb, erythromycin. Antibiotic-resistance phenotypes and other features listed
in parentheses are those expressed by F. psychrophilum but not by E. coli.
DRestriction sites for cloning are in bold type.
D. Pe ´rez-Pascual and others
performed with CLUSTAL_X software. fpp1 and fpp2 gene sequences
were obtained from the ongoing genome sequence of F. psychrophi-
lum THC02/90 (E. Duchaud, unpublished data) deposited in EMBL
under accession number FR667216.
fpp1 and fpp2 bioinformatic analysis
The genes encoding Fpp1 and Fpp2 proteases have been
definitely localized as an fpp2–fpp1 tandem in the F.
psychrophilum JIP02/86 genome (Duchaud et al., 2007) and
also in the ongoing genome of strain THC02/90 (Fig. 1a),
where these genes are separated by an intergenic space of
109 bp. Putative Flavobacterium 27 (TAnnTTTG) and
233 (TTG) promoter sequences, and a ribosome-binding
site (TAAAA) (Chen et al., 2007), were identified upstream
of the respective ORFs, and stem–loop palindromic
sequences corresponding to rho-independent terminators
were located at the 39 ends of both genes.
Additionally, Fpp2 and Fpp1 putative prosequences of 82
and 194 aa, respectively, could be deduced based on the
comparison between the encoding sequence and the N-
terminal amino acid sequence of both proteins.
The fpp2 gene encoded a predicted 942 aa protein with a
theoretical molecular mass of 100240 Da. The deduced
Fpp2 protein (Fig. 2) displays a modular structure
encompassing (i) an N-terminal signal peptide of 26 aa;
(ii) a region similar to other metalloproteases (MEROPS
M43, cytophagalysin family), with a typical HEXXH-
XXGXXH motif (amino acids 284–294); (iii) a ‘cleaved
adhesin domain’ (amino acids 526–686) of 160 aa charac-
teristic of a group of adhesins/peptidases represented by
the extracellular virulence complex RgpA, RgpB and Kgp of
Porphyromonas gingivalis (Pike et al., 1994; Bhogal et al.,
1997; Rangarajan et al., 1997), a species belonging to the
same phylum as F. psychrophilum; and (iv) a highly
conserved C-terminal region of 95 aa (amino acids 847–
942) (Duchaud et al., 2007). This module is likely involved
in cell-anchoring of proteins to the outer membrane
(Karlsson et al., 2004).
The fpp1 gene encoded a protein of 1138 aa with a
predicted molecular mass of 119587 Da. The putative
Fpp1 (Fig. 2) also displays a modular structure encom-
passing (i) an N-terminal signal peptide of 25 aa; (ii) a
region similar to other metalloproteases (MEROPS M12,
adamalysin family), with a typical HEXXHXXGXXH zinc
metalloprotease motif (amino acids 364–374); (iii) a
Drmip–Hesp domain, characteristic of a mitogen-activated
protein kinase involved in growth and development in
many organisms (Szeto et al., 2007) in the central region
(amino acids 600–675), and a regulatory P domain (amino
Fig. 1. Gene organization of the fpp2–fpp1 locus in the F. psychrophilum genome (a). The position of each gene and the
direction of transcription are shown by arrows. The fpp2–fpp1 promoter (P), the rho-independent transcriptional terminator
(hairpin loop), and the flanking genes ribF (riboflavin kinase; FP0233) and pth (peptidyl-tRNA hydrolase; FP0230) are indicated.
(b) RT-PCR analysis of the fpp2–fpp1 locus in F. psychrophilum. Each PCR was performed with 10 ng F. psychrophilum RNA.
As controls for DNA contamination, PCRs were also carried out with the primers used for RT-PCR analysis by omitting the
reverse-transcription step (lanes 2, 6 and 9). Lanes: 1, 100 bp molecular marker (Biotools); 6–8, RT-PCRs performed with the
RNA from the fpp2”mutant strain; 9–11, RT-PCRs performed with the RNA from the fpp1”mutant strain. The following primers
were used: 3R and 3F for lanes 5, 8 and 11; 2R and 2F for lane 4; 1F and 1R for lanes 3, 7 and 10. The localization of the
primers used for RT-PCR analysis and the lengths of the amplicons generated are also shown.
Mutation of fpp2–fpp1 genes of F. psychrophilum
acids 938–1043) containing an AGDL motif (amino acids
952–955) similar to the RGDL motif conserved in most
eukaryotic proprotein convertases belonging to the S8
family (Fig. 2); and (iv) a highly conserved C-terminal
region of 95 aa (amino acids 1043–1138) likely involved in
cell-anchoring (Karlsson et al., 2004).
Generation of fpp1”and fpp2”mutants by
single-crossover homologous recombination
Internal fragments from the fpp1 and fpp2 genes were used
to generate the suicide pLYL03Fpp1 and pLYL03Fpp2
plasmids, respectively. These vectors were introduced into
E. coli S17-1lpir and mobilized into F. psychrophilum
THC02/90 by conjugation, and transconjugants were
selected on NAC plates supplemented with erythromycin
as described in Methods. Southern blot analysis of the
selected transconjugants showed that the pLYL03Fpp1 and
pLYL03Fpp2 plasmids were integrated into the fpp1 and
fpp2 genes of the chromosome of F. psychrophilum THC02/
90, generating the corresponding fpp12
mutants (Fig. 3a, b). Two new XbaI fragments of 7.2 and
0.5 kb appeared from fpp12DNA digestion in addition to
the 4, 2 and 0.7 kb fragments originating from the parental
wild-type strain (Fig. 3a, b). In the same way, a new XbaI
fragment of 7.9 kb was found from the fpp22DNA
digestion in addition to the 2.7 kb fragment from the
parental wild-type strain (Fig. 3a, b). Similar results were
obtained when the HindIII enzyme was used for DNA
digestion (data not shown). Furthermore, and according to
the patterns of labelled DNA bands after Southern blot
hybridization, one can conclude that the integration of
pLYL03Fpp1 and pLYL03Fpp2 into the chromosome had
occurred by a single-crossover event between the fpp1 and
fpp2 internal DNA fragments harboured in the plasmids
and the respective homologous sequences from the
bacterial chromosome (Fig. 3). The resulting mutants were
stable during in vivo infections. Southern blot hybridiza-
tion showed the presence of the corresponding labelled
XbaI fragments resulting from fpp1 and fpp2 gene
disruption in the genome of all the colonies analysed that
were recovered from fish (Fig. 3a, b). Therefore, no
reversion was observed in the absence of selective antibiotic
pressure in vivo. Identical results were obtained when
erythromycin was used to recover the mutants from the
RT-PCR assays were carried out to elucidate the transcrip-
tional structure of the fpp1 and fpp2 genes. According to
sequence analysis, two putative promoter regions may be
present at the 59 ends of both the fpp2 gene and the fpp1
gene. When mRNA from the wild-type strain was used as
template for RT-PCR analysis, fpp2 and fpp1, as well as the
intergenic region, were amplified (Fig. 1b, lanes 3–5).
However, when the fpp1 gene was disrupted the fpp2
upstream gene was amplified (Fig. 1b, lane 10), whereas no
amplification was detected when the fpp2 gene was mutated
(Fig. 1b, lanes 7 and 8). These results indicate that under the
conditions used, there is a unique promoter activity located
upstream of the fpp2 gene that drives the transcription of
both genes, generating a single polycistronic mRNA.
Therefore, genes fpp2 and fpp1 constitute an operon. The
resulting transcriptional structure is presented in Fig. 1(a).
Phenotypic characterization and LD50
determination of fpp1”and fpp2”mutants
To further characterize the F. psychrophilum fpp12and
fpp22mutants, some of their phenotypic traits were
studied. The growth of fpp12(data not shown) and fpp22
mutants in NBF at 12 uC was similar to that of the wild-
type strain (Fig. 4a). The levels of proteolytic activity in the
culture supernatants were similar in the wild-type strain
and the fpp12mutant (data not shown). However, the
fpp22mutant exhibited a caseinolytic activity significantly
lower than that of the wild-type strain (Fig. 4a).
Fig. 2. Structural organization and domains of the Fpp1 and Fpp2 proteolytic enzymes from F. psychrophilum strain THC02/90.
Domain regions are labelled, motifs are shown and the initial and final residues of each domain are indicated above each protein.
D. Pe ´rez-Pascual and others
Analysis of colony spreading on diluted NA medium
showed that when fpp12and fpp22mutants were grown
on half-strength NA the diameters of the colonies were
similar to those of the wild-type strain (data not shown).
By contrast, when they were grown on one-sixth NA, the
colonies of the fpp22mutant presented a significantly
larger diameter than those of the wild-type strain (Fig. 4b,
c). This difference was correlated with an increase in
biomass production (Fig. 4b).
Plasmid pCP23-Fpp2 was constructed as described in
Methods to complement the fpp22mutant. Introduction
of this plasmid into the fpp22mutant restored the
extracellular caseinolytic activity, determined both qualita-
tively on NBF plates supplemented with casein (data not
shown), and quantitatively, showing an approximately
twofold higher azocasein activity of the complemented
strain compared with the wild-type after 120 h of
incubation (Fig. 4a). Additionally, colony spreading of the
complemented fpp22mutant was considerably decreased
In order to evaluate the effects of the fpp1 and fpp2
mutations on virulence, LD50experiments were carried out
on rainbow trout. The LD5010 days post-injection was
calculated to be 2.536106c.f.u. for the wild-type strain.
Under the same conditions, strains fpp12and fpp22
showed LD50 values of 2.126106and 1.736106c.f.u.,
respectively. These results indicate that the disruption of
the fpp1 and fpp2 genes had no significant effect on the
virulence of the bacteria in rainbow trout under the
conditions used. Additionally, visual examination of skin
lesions showed that the wild-type strain and the two
mutants caused the same tissue damage around the
injection site (data not shown).
The fpp1 and fpp2 genes have been unambiguously
identified in the genomes of F. psychrophilum strains
JIP02/86 (Duchaud et al., 2007) and THC02/90 (this
study). The deduced Fpp1 and Fpp2 proteins display
modular structures encompassing an unusual mixture of
The Fpp2 protein is composed of four domains, including
a typical ‘cleaved adhesin domain’ of 160 aa. This domain
is characteristic of the RgpA–Kgp protease–adhesin com-
plexes of P. gingivalis (Curtis et al., 2001; Pathirana et al.,
2006). Both Rgp and Kgp belong to the cysteine proteinase
family, but also display some characteristics of serine
proteases and metalloproteases (Kadowaki et al., 1994; Abe
et al., 1998). Nevertheless, the Fpp2 protein sequence
showed the highest similarity with the M43 family of
metalloproteases, in concordance with the biochemical and
enzymic properties of the purified Fpp2 enzyme (Secades
et al., 2003).
The Fpp1 protein is also composed of four domains and
contains a P domain (amino acids 938–1043). This domain
is found in subtilisin proprotein convertases, and seems to
be necessary to fold and maintain the endopeptidase
Fig. 3. Construction of fpp1”and fpp2”isogenic mutants by insertional mutagenesis. (a) XbaI restriction map of the area
surrounding the fpp2–fpp1 locus of the wild-type THC02/90 strain and isogenic fpp1”and fpp2”mutants. The labelled DNA
fragments of fpp1 and fpp2 used as probes and the size of the fragments revealed by Southern blot hybridization are indicated
in kb. X, XbaI; (i) THC02/90 strain, (ii) fpp1”mutant, (iii) fpp2”mutant. (b) Southern blot analysis of the wild-type F.
psychrophilum strain and of the fpp1”and fpp2”mutants. Genomic DNA from the wild-type F. psychrophilum strain (lanes 1
and 3) and the mutant strains fpp1”(lane 2) and fpp2”(lane 4) was digested with XbaI and hybridized with the 2.5 kb fpp1 and
1.9 kb fpp2 fragments. The sizes of a molecular mass ladder are indicated in the centre in kb.
Mutation of fpp2–fpp1 genes of F. psychrophilum
catalytic domain and to regulate the calcium- and acidic
pH-dependence of this type of protease. This P domain is
therefore likely responsible for the calcium-dependent
activity of the purified Fpp1 protein. Nevertheless, Fpp1
has metalloprotease characteristics, as previously deter-
mined (Secades et al., 2001), and belongs to the M12 family
Initially, the tandem organization of the fpp2 and fpp1
genes with putative regulatory sequences for each gene
suggested that they might be transcribed as two units.
However, RT-PCR analysis of the transcripts in the fpp12
and fpp22mutants showed that only one active promoter
was detected upstream of the fpp2 gene and that it was
responsible for the generation of an fpp2–fpp1 polycistronic
mRNA. The Fpp1 putative promoter seems not to be active
under the assayed conditions.
Proteolytic activity quantification during in vitro growth
indicated that fpp2 is likely the main gene driving
caseinolytic activity in the bacterium. Intriguingly, the
fpp2 mutation caused a conditional hypergliding pheno-
type and consequently a biomass increase when growing on
diluted NA. Therefore, colony spreading and Fpp2
extracellular proteolytic activity are negatively correlated
in this bacterium, suggesting that these two cellular
processes may be in some way related in F. psychrophilum.
It should be emphasized that a low nutrient concentration
induces colony spreading in F. psychrophilum (Pe ´rez-
Pascual et al., 2009). There was no clear decrease of growth
of the fpp12and fpp22mutants compared with the wild-
type strain in both standard and diluted broth cultures.
Growth conditions are considerably different between solid
and liquid media, especially regarding the factors that affect
motility, such as nutrient availability, cell density, surface
characteristics, and diffusion of extracellular factors related
to nutrient utilization. Therefore, the high-spreading
phenotype of the fpp22mutant may result from the
alteration of some of these characteristics, which are
reverted in the complemented mutant. For instance, the
absence of fpp2 may decrease nutrient availability,
stimulating colony spreading to colonize unexploited areas
of the agar plate. The alteration of bacterial motility by the
mutation of genes encoding proteolytic activity that has
been reported in different bacteria is usually related to
regulatory enzymes with a proteolytic function, such as the
ClpP protease from Bacillus subtilis (Gupta & Rao, 2009)
and the Lon protease from Proteus mirabilis (Clemmer &
Rather, 2008), rather than to extracellular proteases.
Alteration in motility phenotypes may also result from
changes in expression of virulence factors, including
Fig. 4. Growth, extracellular proteolytic activity and spreading of F.
psychrophilum THC02/90, the fpp2”mutant and the pCP23-
Fpp2-complemented fpp2”mutant. (a) Bacterial growth at 12 6C
in NBF was monitored by determining OD525. Solid lines, growth
($) and proteolytic activity (#) of strain THC02/90; dotted lines,
growth (m) and proteolytic activity (g) of the fpp2”mutant; dashed
and dotted lines, growth (X) and proteolytic activity (e) of the
pCP23-Fpp2-complemented fpp2”mutant. The growth curve and
proteolytic activity corresponding to the fpp1”mutant were similar
to those of strain THC02/90 (not shown). The extracellular
proteolytic activity in cell-free supernatants was determined by
azocasein assay (Secades et al., 2001); EU, enzymic units. (b)
Spreading of and relative biomass production by colonies of F.
psychrophilum THC02/90 and the fpp2”mutant. The histogram
represents the diameters of colonies of strain THC02/90 (white
bars) and the fpp2”mutant (grey bars) at different incubation
times. Lines represent the relative biomass production of strain
THC02/90 ($, solid line) and of the fpp2”mutant (m, dashed line)
at different incubation times. (c) Colony spreading of (i) F.
psychrophilum THC02/90, (ii) the fpp2”mutant and (iii) the
pCP23-Fpp2-complemented fpp2”mutant. The strains were
grown on one-sixth NA for 96 h at 20 6C according to Pe ´rez-
Pascual et al. (2009). Bar, 1 cm.
D. Pe ´rez-Pascual and others
proteolytic enzymes, as reported in Vibrio cholerae, in
which production is increased in hyperswarmer mutants
(Gardel & Mekalanos, 1996).
Extracellular proteolytic activities have been reported to be
involved in the virulence of different pathogens, including
a number of fish pathogenic bacteria (Ferna ´ndez et al.,
2003; Yang et al., 2007; Zhang et al., 2009). However, the
implication of proteases in pathogenesis is not a general
rule, as shown in this study in a rainbow trout model.
Although purified Fpp1 and Fpp2 proteases have been
shown to degrade tissue proteins in vitro (Secades et al.,
2001, 2003), the inactivation of the fpp1 and fpp2 loci did
not have any effect on bacterial virulence. This, together
with the fact that similar necrotic lesions appeared around
the injection site of the wild-type strain and the two
mutants in this study, suggests that these proteins are not
directly involved in virulence, at least in rainbow trout
challenged by intramuscular injection. Nevertheless, it
cannot be ruled out that they may play a role in virulence
in a natural infection process or in another fish host.
Because the genome of F. psychrophilum encodes 11 other
extracellular proteases, this genetic redundancy may
conceal the loss of Fpp1 and Fpp2.
This paper describes for the first time a method for specific
gene inactivation in F. psychrophilum by using homologous
recombination. The resulting mutants were stable, and no
reversion was observed in the absence of selective antibiotic
pressure during in vivo growth.
This technique, together with the knowledge of the
complete genome sequence, opens promising perspectives
for the study of this bacterium.
This work was supported in part by the Spanish Ministry of Sciences
and Education (MEC) (grant AGL2009-07003) and the Agence
Nationale de la Recherche, France (grant 07-GMGE). D.P.-P., E.G.,
P.R. and R.N. were the recipients of a grant from the Ministerio de
Ciencia e Innovacio ´n (MICINN), Spain. We also thank Miguel Sotelo
and BioMar S.A. for their support. Particular thanks go to J.-F.
Bernardet for critical reading of the manuscript.
Abe, N., Kadowaki, T., Okamoto, K., Nakayama, K., Ohishi, M. &
Yamamoto, K. (1998). Biochemical and functional properties of lysine-
specific cysteine proteinase (Lys-gingipain) as a virulence factor of
Porphyromonasgingivalis in periodontaldisease. J Biochem 123, 305–312.
Agarwal, S., Hunnicutt, D. W. & McBride, M. J. (1997). Cloning and
characterization of the Flavobacterium johnsoniae (Cytophaga johnso-
nae) gliding motility gene, gldA. Proc Natl Acad Sci U S A 94, 12139–
A´lvarez, B. & Guijarro, J. A. (2007). Recovery of Flavobacterium
psychrophilum viable cells using a charcoal-based solid medium. Lett
Appl Microbiol 44, 569–572.
A´lvarez, B., Secades, P., McBride, M. J. & Guijarro, J. A. (2004).
Development of genetic techniques for the psychrotrophic fish
pathogen Flavobacterium psychrophilum. Appl Environ Microbiol 70,
A´lvarez, B., A´lvarez, J., Mene ´ndez, A. & Guijarro, J. A. (2008). A
mutant in one of two exbD loci of a TonB system in Flavobacterium
psychrophilum shows attenuated virulence and confers protection
against cold water disease. Microbiology 154, 1144–1151.
Arai, H., Morita, Y., Izumi, S., Katagiri, T. & Kimura, H. (2007).
Molecular typing by pulsed-field gel electrophoresis of Flavobacterium
psychrophilum isolates derived from Japanese fish. J Fish Dis 30, 345–
Bertolini, J. M., Wakabayashi, H., Watral, V. G., Whipple, M. J. &
Rohovec, J. S. (1994). Electrophoretic detection of proteases from
selected strains of Flexibacter psychrophilum and assessment of their
variability. J Aquat Anim Health 6, 224–233.
Bhogal, P. S., Slakeski, N. & Reynolds, E. C. (1997). A cell-associated
protein complex of Porphyromonas gingivalis W50 composed of Arg-
and Lys-specific cysteine proteinases and adhesins. Microbiology 143,
Bjornsdottir, B., Fridjonsson, O. H., Magnusdottir, S., Andresdottir, V.,
Hreggvidsson, G. O. & Gudmundsdottir, B. K. (2009). Charac-
terisation of an extracellular vibriolysin of the fish pathogen Moritella
viscosa. Vet Microbiol 136, 326–334.
Cepeda, C. & Santos, Y. (2000). Rapid and low-level toxic PCR-based
method for routine identification of Flavobacterium psychrophilum.
Int Microbiol 3, 235–238.
Cepeda, C., Garcı ´a-Ma ´rquez, S. & Santos, Y. (2004). Improved
growth of Flavobacterium psychrophilum using a new culture medium.
Aquaculture 238, 75–82.
Chakroun, C., Grimont, F., Urdaci, M. C. & Bernardet, J.-F. (1998).
Fingerprinting of Flavobacterium psychrophilum isolates by ribotyping
and plasmid profiling. Dis Aquat Organ 33, 167–177.
Chen, S., Bagdasarian, M., Kaufman, M. G. & Walker, E. D. (2007).
Characterization of strong promoters from an environmental
Flavobacterium hibernum strain by using a green fluorescent protein-
based reporter system. Appl Environ Microbiol 73, 1089–1100.
Clemmer, K. M. & Rather, P. N. (2008). The Lon protease regulates
swarming motility and virulence gene expression in Proteus mirabilis.
J Med Microbiol 57, 931–937.
Curtis, M. A., Aduse-Opoku, J. & Rangarajan, M. (2001). Cysteine
proteases of Porphyromonas gingivalis. Crit Rev Oral Biol Med 12,
del Cerro, A., Mendoza, M. C. & Guijarro, J. A. (2002). Usefulness of a
TaqMan-based polymerase chain reaction assay for the detection of
the fish pathogen Flavobacterium psychrophilum. J Appl Microbiol 93,
Duchaud, E., Boussaha, M., Loux, V., Bernardet, J. F., Michel, C.,
Kerouault, B., Mondot, S., Nicolas, P., Bossy, R. & other authors
(2007). Complete genome sequence of the fish pathogen Flavobac-
terium psychrophilum. Nat Biotechnol 25, 763–769.
Ferna ´ndez, L., Lo ´pez, J. R., Secades, P., Mene ´ndez, A., Ma ´rquez, I. &
Guijarro, J. A. (2003). In vitro and in vivo studies of the Yrp1 protease
from Yersinia ruckeri and its role in protective immunity against
enteric red mouth disease of salmonids. Appl Environ Microbiol 69,
Fujiwara-Nagata, E. & Eguchi, M. (2009). Development and
evaluation of a loop-mediated isothermal amplification assay for
rapid and simple detection of Flavobacterium psychrophilum. J Fish
Dis 32, 873–881.
Gardel, C. L. & Mekalanos, J. J. (1996). Alterations in Vibrio cholerae
motility phenotypes correlate with changes in virulence factor
expression. Infect Immun 64, 2246–2255.
Mutation of fpp2–fpp1 genes of F. psychrophilum
Gupta, M. & Rao, K. K. (2009). Epr plays a key role in DegU-mediated
swarming motility of Bacillus subtilis. FEMS Microbiol Lett 295, 187–
Izumi, S., Aranishi, F. & Wakabayashi, H. (2003). Genotyping of
Flavobacterium psychrophilum using PCR-RFLP analysis. Dis Aquat
Organ 56, 207–214.
Kadowaki, T., Yoneda, M., Okamoto, K., Maeda, K. & Yamamoto, K.
(1994). Purification and characterization of a novel arginine-specific
cysteine proteinase (argingipain) involved in the pathogenesis of
periodontal disease from the culture supernatant of Porphyromonas
gingivalis. J Biol Chem 269, 21371–21378.
Karlsson, E. N., Hachem, M. A., Ramchuran, S., Costa, H., Holst, O.,
Svenningsen, S. F. & Hreggvidsson, G. O. (2004). The modular
xylanase Xyn10A from Rhodothermus marinus is cell-attached, and its
C-terminal domain has several putative homologues among cell-
attached proteins within the phylum Bacteroidetes. FEMS Microbiol
Lett 241, 233–242.
Li, L.-Y., Shoemaker, N. B. & Salyers, A. A. (1995). Location and
characteristics of the transfer region of a Bacteroides conjugative
transposon and regulation of transfer genes. J Bacteriol 177, 4992–
Lorenzen, E. & Olesen, N. J. (1997). Characterization of isolated
Flavobacterium psychrophilum associated with coldwater disease or
rainbow trout fry syndrome. II. Serological studies. Dis Aquat Organ
Madsen, L. & Dalsgaard, I. (1998). Characterization of Flavobac-
terium psychrophilum; comparison of proteolytic activity and
virulence of strains isolated from trout (Oncorhynchus mykiss). In
Methodology in Fish Disease Research, pp. 45–52. Edited by A.
C. Barnes, G. A. Davidson, M. P. Hiney & D. McIntosh. Aberdeen,
Scotland: Fisheries Research Service.
Michel, C., Antonio, D. & Hedrick, R. P. (1999). Production of viable
cultures of Flavobacterium psychrophilum: approach and control. Res
Microbiol 150, 351–358.
Møller, J. D., Larsen, J. L., Madsen, L. & Dalsgaard, I. (2003).
Involvement of a sialic acid-binding lectin with hemagglutination and
hydrophobicity of Flavobacterium psychrophilum. Appl Environ
Microbiol 69, 5275–5280.
Nicolas, P., Mondot, S., Achaz, G., Bouchenot, C., Bernardet, J.-F. &
Duchaud, E. (2008). Population structure of the fish-pathogenic
bacterium Flavobacterium psychrophilum. Appl Environ Microbiol 74,
Pathirana, R. D., O’Brien-Simpson, N. M., Veith, P. D., Riley, P. F. &
Reynolds, E. C. (2006). Characterization of proteinase–adhesin
complexes of Porphyromonas gingivalis. Microbiology 152, 2381–2394.
Pe ´rez-Pascual, D., Mene ´ndez, A., Ferna ´ndez, L., Me ´ndez, J.,
Reimundo, P., Navais, R. & Guijarro, J. A. (2009). Spreading versus
biomass production by colonies of the fish pathogen Flavobacterium
psychrophilum: role of the nutrient concentration. Int Microbiol 12,
Pike, R. N., McGraw, W., Potempa, J. & Travis, J. (1994). Lysine- and
Isolation, characterization, and evidence for the existence of com-
plexes with hemagglutinins. J Biol Chem 269, 406–411.
Rangarajan, M., Smith, S. J. M. & Curtis, M. A. (1997). Biochemical
characterisation of the arginine-specific proteases of Porphyromonas
gingivalis W50 suggests a common precursor. Biochem J 323, 701–709.
Reed, L. J. & Muench, H. (1938). A simple method of estimating fifty
percent endpoints. Am J Hyg 27, 493–497.
Secades, P., A´lvarez, B. & Guijarro, J. A. (2001). Purification and
characterization of a psychrophilic, calcium-induced, growth-phase-
dependent metalloprotease from the fish pathogen Flavobacterium
psychrophilum. Appl Environ Microbiol 67, 2436–2444.
Secades, P., A´lvarez, B. & Guijarro, J. A. (2003). Purification and
properties of a new psychrophilic metalloprotease (Fpp2) in the fish
pathogen Flavobacterium psychrophilum. FEMS Microbiol Lett 226,
Simon, R., Priefer, V. & Puhler, A. (1983). A broad host range
mobilization system for in vitro genetic engineering: transposon
mutagenesis in Gram negative bacteria. Biotechnology 1, 784–791.
Soule, M., LaFrentz, S., Cain, K., LaPatra, S. & Call, D. R. (2005).
Polymorphisms in 16S rRNA genes of Flavobacterium psychrophilum
correlate with elastin hydrolysis and tetracycline resistance. Dis Aquat
Organ 65, 209–216.
Szeto, C. Y. Y., Leung, G. S. & Kwan, H. S. (2007). Le.MAPK and its
interacting partner, Le.DRMIP, in fruiting body development in
Lentinula edodes. Gene 393, 87–93.
Urdaci, M. C., Chakroun, C., Faure, D. & Bernardet, J.-F. (1998).
Development of a polymerase chain reaction assay for identification
and detection of the fish pathogen Flavobacterium psychrophilum. Res
Microbiol 149, 519–530.
Yang, H., Chen, J., Yang, G., Zhang, X.-H. & Li, Y. (2007). Mutational
analysis of the zinc metalloprotease EmpA of Vibrio anguillarum.
FEMS Microbiol Lett 267, 56–63.
Zhang, W. W., Hu, Y. H., Wang, H. L. & Sun, L. (2009). Identification
and characterization of a virulence-associated protease from a
pathogenic Pseudomonas fluorescens strain. Vet Microbiol 139, 183–
Edited by: P. C. F. Oyston
D. Pe ´rez-Pascual and others