Free access
Research Article
14 April 2014

Multilocus Sequence Typing Identifies Epidemic Clones of Flavobacterium psychrophilum in Nordic Countries


Flavobacterium psychrophilum is the causative agent of bacterial cold water disease (BCWD), which affects a variety of freshwater-reared salmonid species. A large-scale study was performed to investigate the genetic diversity of F. psychrophilum in the four Nordic countries: Denmark, Finland, Norway, and Sweden. Multilocus sequence typing of 560 geographically and temporally disparate F. psychrophilum isolates collected from various sources between 1983 and 2012 revealed 81 different sequence types (STs) belonging to 12 clonal complexes (CCs) and 30 singleton STs. The largest CC, CC-ST10, which represented almost exclusively isolates from rainbow trout and included the most predominant genotype, ST2, comprised 65% of all isolates examined. In Norway, with a shorter history (<10 years) of BCWD in rainbow trout, ST2 was the only isolated CC-ST10 genotype, suggesting a recent introduction of an epidemic clone. The study identified five additional CCs shared between countries and five country-specific CCs, some with apparent host specificity. Almost 80% of the singleton STs were isolated from non-rainbow trout species or the environment. The present study reveals a simultaneous presence of genetically distinct CCs in the Nordic countries and points out specific F. psychrophilum STs posing a threat to the salmonid production. The study provides a significant contribution toward mapping the genetic diversity of F. psychrophilum globally and support for the existence of an epidemic population structure where recombination is a significant driver in F. psychrophilum evolution. Evidence indicating dissemination of a putatively virulent clonal complex (CC-ST10) with commercial movement of fish or fish products is strengthened.


Flavobacterium psychrophilum, a bacterium belonging to the family Flavobacteriaceae, is considered one of the most important bacterial pathogens in freshwater salmonid aquaculture worldwide. It is recognized as the causative agent of bacterial cold water disease (BCWD), a condition that has been variously referred to as rainbow trout fry syndrome, fry mortality syndrome, low-temperature disease, and peduncle disease (1, 2). Since no commercial vaccines are available today against BCWD, its control relies mainly on management hygiene and antibacterial treatment. Clinical symptoms associated with F. psychrophilum infections vary between bacterial strain and host fish species. If BCWD is left untreated, high mortalities due to sepsis generally occur in species such as rainbow trout (Oncorhynchus mykiss) and coho salmon (O. kisutch), which are considered particularly susceptible to the disease (3). The bacterium is also known to cause considerable losses in farmed and natural populations of ayu (Plecoglossus altivelis altivelis) (4). In less susceptible salmonid species, e.g., Atlantic salmon (Salmo salar) and brown trout (S. trutta), the disease often manifests as fin/tail rot or ulceration, although outbreaks of systemic infections in these species have also been reported (5). F. psychrophilum has also been identified in association with disease in several nonsalmonid species (1, 6, 7).
F. psychrophilum can be transmitted horizontally from fish to fish, and vertical transmission through eggs and sexual fluids is also suspected (810), although the occurrence of the bacterium inside eggs is debated (11, 12). However, it is still unclear whether virulent strains of the bacterium are endemic in the environment or whether pathogenic strains are spread to new areas with the trade of fish and eggs. Initially, BCWD seemed limited to North America (3, 13), until the mid-1980s, when the pathogen was described on the European continent (1419) and subsequently in the Pacific region (4, 2022). In the Nordic countries, where F. psychrophilum currently constitutes the main threat to rainbow trout production, BCWD was first reported in Denmark in 1985 (23, 24) and in Sweden and Finland in 1986 and 1993, respectively (25). In Norway, F. psychrophilum was considered a minor problem in brown trout and salmon hatcheries until 2008, when an epidemic of systemic disease caused high mortalities in several rainbow trout farms (26). In Iceland, the number of outbreaks in rainbow trout is low but appears to be increasing (Gísli Jónsson, personal communication).
Due to the fairly recent emergence (in relation to the original import of rainbow trout to the countries in question) of BCWD in the Nordic countries, the present study, utilizing a standardized multilocus sequence typing (MLST) scheme (27), was initiated to examine the population genetics of F. psychrophilum in this region and further investigate the hypothesis (26) of recent import of an epidemic clone. MLST of a variety of clinical and environmental isolates covering the period between initial recognition of BCWD in this part of Europe until present time represents the largest study of its type to date.


Bacterial isolates.

The 560 Flavobacterium psychrophilum isolates used in the present study (see Table S1 in the supplemental material) were collected between 1983 and 2012 from a wide variety of sources in Denmark (n = 209), Finland (n = 195), Norway (n = 117), and Sweden (n = 36) and included the reference strains NCIMB 1947T, THC02/90, and JIP02/86 originally isolated outside the Nordic region. The isolates were cultured on Anacker and Ordals medium (28) or TYES agar (29) and identified by using routine diagnostic procedures, i.e., biochemical characterization (30, 31), sequencing of 16S rRNA gene or species-specific PCR (32, 33). The collection comprised isolates from 10 different fish species: rainbow trout (n = 448), Atlantic salmon (n = 52), brown trout (n = 26), perch (Perca fluviatilis) (n = 11), three-spined stickleback ( Gasterosteus aculeatus) (n = 3), coho salmon (n = 2), arctic char (Salvelinus alpinus) (n = 1), brook trout (Salvelinus fontinalis) (n = 1), flounder (Platichthys flesus) (n = 1), and char hybrid (Salvelinus fontinalis × Salvelinus namaycush) (n = 1). F. psychrophilum from rainbow trout included isolates from eggs (n = 4), ovarian fluid (n = 10), and milt (n = 12). Environmental isolates from water samples (n = 11) from fish farm environments, as well as three rough laboratory mutant strains prepared from isolated smooth colony types by repeated passages in TYES broth (34), were included in the collection.


Genomic DNA was extracted using various commercially available kits according to the manufacturer's instructions. Partial sequences from atpA, dnaK, fumC, gyrB, murG, trpB, and tuf genes were amplified and sequenced according to the methods of Siekoula-Nguedia et al. (35) (Danish, Finnish, and Norwegian isolates) and Nicolas et al. (27) (Swedish isolates), respectively. Both strands of the amplified sequences were sequenced (Sanger), and all chromatograms were manually verified to ensure high quality before assignation of allele types (ATs) and sequence types (STs) using an “in-house” Perl script. Alleles, STs, and background information on the isolates were submitted to the F. psychrophilum MLST database ( Reference strains were included as internal controls in all amplification and sequencing analyses.

MLST data analysis.

After assignation of a seven-numbered allelic profile (i.e., ST), clonal complexes (CCs) were identified using eBURST v3 ( The predicted ancestral genotype (founder) of a CC containing three or more STs was defined as the ST with the highest number of single-locus variants. Major complexes included three or more STs and were named according to the ST of the predicted founder. Minor CCs containing only two STs were named after the most represented ST or by default by the ST with the lower numbering. Singletons were defined as an ST not belonging to any CC. Clonal relationships were examined using both relaxed and stringent default settings, i.e., five or six shared alleles, respectively. Bootstrapping (n = 1,000) was performed to evaluate model robustness.

Population genetic and phylogenetic analyses.

Gene diversity (H) and the level of linkage disequilibrium (nonrandom association) between the ATs found at the seven MLST loci, as measured by the standardized index of association (ISA) (36), were investigated using LIAN 3.6. Three data sets were tested: the whole population, single representatives of each ST (81 STs), and founder STs of each CC, together with singleton STs (42 STs). A network representation of the evolutionary relationships between STs was constructed on concatenated sequences using the neighbor-net algorithm (with default settings) available in SplitsTree 4 software ( Recombination within the concatenated sequences of each ST was tested using the pairwise homoplasy index (PHI) test also implemented in SplitsTree 4.

Nucleotide sequence accession numbers.

Nucleotide sequences have been deposited in the GenBank database under accession numbers KJ366498 to KJ370396.


MLST data analysis.

MLST of 560 F. psychrophilum isolates included in the study revealed 81 different STs, 68 of which were novel ( The greatest national variation of genotypes was found in the Finnish collection, with 36 different STs compared to 28, 20, and 17 found in Denmark, Norway, and Sweden, respectively. The three smooth-to-rough converted Finnish isolates did not change ST after conversion. eBURST clustering (Fig. 1) divided the 81 STs into seven major CCs, five minor CCs, and 30 singleton STs (Table 1). The major CCs were designated CC-ST10 (14 STs), CC-ST123 (6 STs), CC-ST124 (6 STs), CC-ST125 (3 STs), CC-ST138 (3 STs), CC-ST191 (6 STs), and CC-ST236 (3 STs).
FIG 1 Population snapshot of 560 F. psychrophilum isolates representing 81 STs divided into 12 CCs and 30 singleton STs by the eBURST algorithm. Each dot represents an ST, and the frequency of the ST correlates with the size of the dot. STs that differ by a single locus are linked together into a CC (circled).
TABLE 1 Clonal complexes and singletons with default and relaxed eBURST group definitionsa
Clonal complex or singletonbCountrySource(s)cYr(s) of isolationnSTAllelic profile
Clonal complexes      
    CC-ST10DenmarkRbT (84); RbTM (3); RbTOF (3); BrT (3); water (3)1990–201296ST22, 2, 2, 2, 2, 2, 2
    CC-ST10FinlandRbT20116ST22, 2, 2, 2, 2, 2, 2
    CC-ST10NorwayRbT2004–201254ST22, 2, 2, 2, 2, 2, 2
    CC-ST10SwedenRbT20021ST22, 2, 2, 2, 2, 2, 2
    CC-ST10DenmarkRbT1990–201125ST102, 8, 2, 2, 2, 2, 2
    CC-ST10FinlandWater19991ST102, 8, 2, 2, 2, 2, 2
    CC-ST10SwedenRbT1994–20032ST102, 8, 2, 2, 2, 2, 2
    CC-ST10DenmarkRbT1992–19954ST122, 8, 2, 2, 2, 7, 2
    CC-ST10FinlandRbT (11); RbTOF (1)1996–200312ST122, 8, 2, 2, 2, 7, 2
    CC-ST10SwedenRbT1988–20118ST122, 8, 2, 2, 2, 7, 2
    CC-ST10FranceRbT (reference strain JIP02/86)19861ST208, 8, 2, 2, 2, 2, 2
    CC-ST10DenmarkRbTOF19992ST262, 15, 2, 2, 2, 2, 2
    CC-ST10DenmarkRbT1988–199525ST792, 8, 8, 2, 2, 2, 2
    CC-ST10FinlandRbT1993–200630ST792, 8, 8, 2, 2, 2, 2
    CC-ST10SwedenRbT19913ST792, 8, 8, 2, 2, 2, 2
    CC-ST10FinlandRbT20093ST912, 2, 2, 2, 2, 41, 2
    CC-ST10DenmarkRbT2006–20117ST923, 2, 2, 2, 2, 41, 2
    CC-ST10FinlandRbT (56); RbTE (3)2008–201159ST923, 2, 2, 2, 2, 41, 2
    CC-ST10SwedenRbT20115ST923, 2, 2, 2, 2, 41, 2
    CC-ST10DenmarkRbT (3); RbTOF (2); RbTM (1); water (1)19997ST982, 2, 2, 2, 2, 48, 2
    CC-ST10SwedenRbT19901ST1282, 8, 2, 2, 2, 48, 2
    CC-ST10FinlandRbT2003–20063ST1282, 8, 2, 2, 2, 48, 2
    CC-ST10FinlandRbT20031ST1308, 8, 8, 2, 2, 2, 2
    CC-ST10SwedenRbT1992–19943ST2352, 8, 8, 2, 2, 3, 2
    CC-ST10SwedenRbT19941ST2382, 8, 8, 2, 2, 49, 2
    CC-ST10SwedenRbT1995–19963ST2402, 8, 8, 2, 2, 7, 2
    CC-ST124DenmarkRbT19901ST701, 13, 15, 14, 1, 7, 1
    CC-ST124FinlandAtS19991ST701, 13, 15, 14, 1, 7, 1
    CC-ST124NorwayAtS20099ST701, 13, 15, 14, 1, 7, 1
    CC-ST124FinlandRbT19961ST738, 13, 15, 14, 1, 7, 1
    CC-ST124FinlandRbT201118ST1242, 13, 15, 14, 1, 7, 1
    CC-ST124FinlandRbT20115ST1262, 13, 15, 16, 1, 7, 1
    CC-ST124SwedenRbT20112ST1262, 13, 15, 16, 1, 7, 1
    CC-ST124FinlandRbT20081ST1292, 13, 2, 14, 1, 7, 1
    CC-ST124NorwayAtS2008–20125ST1879, 13, 15, 14, 1, 7, 1
    CC-ST191FinlandRbT19971ST1404, 3, 22, 3, 3, 7, 3
    CC-ST191DenmarkRbT (2); RbTM (4)19996ST1764, 3, 22, 2, 3, 3, 3
    CC-ST191DenmarkRbT2001–20064ST1814, 3, 22, 3, 3, 3, 2
    CC-ST191NorwayRbT20093ST1814, 3, 22, 3, 3, 3, 2
    CC-ST191DenmarkRbT20061ST1834, 3, 22, 3, 32, 3, 3
    CC-ST191DenmarkBrT20001ST1914, 3, 22, 3, 3, 3, 3
    CC-ST191DenmarkBrT (1); RbT (4)1995–20005ST1924, 3, 22, 3, 20, 3, 3
    CC-ST123NorwayAtS19971ST1234, 49, 22, 11, 6, 3, 3
    CC-ST123NorwayAtS2006–20127ST1684, 49, 8, 11, 6, 3, 3
    CC-ST123NorwayAtS20101ST1729, 49, 8, 11, 6, 3, 3
    CC-ST123NorwayAtS20101ST1734, 52, 22, 11, 6, 3, 3
    CC-ST123NorwayAtS19961ST1844, 49, 22, 11, 6, 50, 3
    CC-ST123NorwayBrT19971ST1854, 49, 15, 11, 6, 3, 3
    CC-ST125FinlandRbT (8); RbTOF (1); water (1)1997–201210ST1251, 13, 15, 2, 2, 7, 2
    CC-ST125FinlandRbT20006ST1311, 1, 15, 2, 2, 7, 2
    CC-ST125FinlandRbTOF20001ST1351, 13, 15, 2, 2, 3, 2
    CC-ST138DenmarkRbT19955ST13811, 29, 8, 3, 32, 41, 31
    CC-ST138FinlandRbt20054ST13811, 29, 8, 3, 32, 41, 31
    CC-ST138DenmarkRbTM19992ST17815, 29, 8, 3, 32, 41, 31
    CC-ST138DenmarkRbTM19991ST17911, 29, 8, 3, 32, 53, 31
    CC-ST236 [CC-ST191]FinlandRbT19951ST1398, 50, 22, 3, 3, 3, 3
    CC-ST236 [CC-ST191]SwedenRbT19911ST2348, 8, 22, 3, 3, 2, 3
    CC-ST236 [CC-ST191]SwedenRbT (1); BrT (1)19922ST2368, 50, 22, 3, 3, 2, 3
    CC-ST121FinlandBrT19931ST12132, 23, 8, 10, 10, 22, 42
    CC-ST121NorwayAtS (1); BrT (2)1998–20093ST12132, 23, 8, 10, 10, 22, 42
    CC-ST121NorwayAtS20092ST17032, 23, 8, 10, 10, 52, 42
    CC-ST34FinlandPer20065ST341, 20, 15, 3, 16, 6, 18
    CC-ST34FinlandPer20061ST1961, 55, 15, 3, 16, 6, 18
    CC-ST193DenmarkRbT; StB (2)20003ST1934, 18, 22, 17, 2, 3, 47
    CC-ST193DenmarkBrT (1); StB (1)20002ST1944, 49, 22, 17, 2, 3, 47
    CC-ST35FinlandBrT19961ST3512, 21, 8, 7, 17, 19, 19
    CC-ST35FinlandRbT20081ST19712, 21, 8, 7, 17, 7, 19
    CC-ST13USACoS (reference strain THC02/90)19901ST94, 7, 6, 5, 6, 5, 4
    CC-ST13FinlandBrT19961ST134, 7, 6, 5, 6, 8, 4
    CC-ST13USACoS (type strain NCIMB 1947T)Unknown1ST134, 7, 6, 5, 6, 8, 4
    SingletonFinlandBrT (1); ACh (1); AtS (1); water (3)1996–20006ST239, 13, 11, 7, 9, 13, 12
    SingletonNorwayAtS (5); BrT (5); AtSOF (2)1997–201212ST239, 13, 11, 7, 9, 13, 12
    SingletonDenmarkRbT19951ST313, 19, 13, 9, 12, 16, 15
    SingletonFinlandRbT (1); AtS (1); BrT (1)1994, 20113ST3613, 22, 8, 10, 6, 20, 20
    SingletonFinlandBkT19961ST3713, 21, 16, 7, 17, 21, 21
    SingletonNorwayBrT19971ST3713, 21, 16, 7, 17, 21, 21
    SingletonDenmarkRbT19911ST12031, 46, 9, 10, 35, 50, 14
    SingletonFinlandRbT19831ST12210, 48, 12, 10, 10, 47, 14
    SingletonNorwayAtS1999–20084ST12210, 48, 12, 10, 10, 47, 14
    SingletonFinlandPer20062ST1278, 37, 8, 2, 36, 47, 18
    Singleton [CC-ST132]FinlandPer (2); water (1)20003ST13211, 49, 2, 11, 3, 3, 3
    SingletonFinlandWater20001ST13333, 37, 8, 5, 2, 26, 2
    Singleton [CC-ST134]FinlandPer20001ST13434, 19, 8, 5, 37, 24, 20
    SingletonFinlandRbTE20101ST1374, 21, 26, 5, 14, 49, 20
    SingletonFinlandAtS19991ST14121, 23, 8, 7, 38, 51, 42
    SingletonNorwayAtS20091ST1698, 3, 15, 1, 1, 1, 1
    SingletonNorwayAtS (6); BrT (1)1998–20127ST1717, 51, 8, 17, 33, 17, 43
    Singleton [CC-ST191]DenmarkRbT19991ST1754, 43, 2, 3, 22, 3, 3
    SingletonDenmarkRbTM19992ST1774, 29, 16, 5, 2, 49, 20
    SingletonDenmarkRbT20062ST18035, 21, 8, 7, 28, 19, 44
    Singleton [CC-ST182]DenmarkRbT20061ST18215, 53, 8, 7, 30, 54, 45
    SingletonNorwayBrT19972ST18613, 22, 8, 10, 17, 22, 42
    Singleton [CC-ST134]NorwayBrT19971ST18834, 19, 26, 5, 39, 24, 20
    SingletonDenmarkRbT20061ST1902, 37, 8, 2, 16, 47, 46
    SingletonDenmarkFlo20001ST19518, 54, 22, 3, 40, 3, 3
    SingletonFinlandAtS19991ST1982, 28, 8, 10, 17, 19, 42
    Singleton [CC-ST191]DenmarkRbT20051ST2074, 19, 2, 3, 3, 3, 3
    SingletonDenmarkRbT19951ST23138, 21, 28, 2, 18, 20, 53
    SingletonNorwayAtS20121ST23228, 62, 15, 2, 25, 1, 54
    Singleton [CC-ST182]SwedenRbT19911ST23339, 53, 8, 7, 30, 54, 2
    Singleton [CC-ST132]SwedenBrT19931ST23711, 49, 2, 11, 44, 2, 3
    SingletonSwedenChH19941ST2398, 63, 31, 3, 2, 3, 3
    SingletonSwedenRbT20011ST2412, 13, 15, 14, 1, 2, 2
The table shows the country of origin and the source of F. psychrophilum, the year of isolation, the number of isolates (n), the sequence type (ST), and the allelic profile (trpB, gyrB, dnaK, fumC, murG, tuf, and atpA). Sources: ACh (Salvelinus alpinus), AtS (Salmo salar), BkT (Salvelinus fontinalis), BrT (Salmo trutta), ChH (Salvelinus fontinalis x Salvelinus namaycush), CoS ( Oncorhynchus kisutch), Flo (Platichthys flesus), Per (Perca fluviatilis), RbT (Oncorhynchus mykiss), StB ( Gasterosteus aculeatus).
Clonal complexes (CC) and singletons with default and relaxed (indicated in brackets) eBURST group definitions are shown.
Superscripts: OF, ovarian fluids; M, milt; E, eggs. Where several sources are indicated, the numbers of isolates from each source are indicated within parentheses.
The largest CC, CC-ST10, represented 65% of the total number and 79, 59, 46, and 56% of the Danish, Finnish, Norwegian, and Swedish isolates, respectively. CC-ST10 included almost exclusively isolates from rainbow trout (Fig. 1, Table 1). Brown trout isolates belonging to this complex originated from gills in fish caught downstream from rainbow trout farms infected by F. psychrophilum. Isolates from rainbow trout belonged almost every year mainly to CC-ST10 (Fig. 2), which included the reference strain F. psychrophilum JIP02/86, a virulent strain belonging to ST20 isolated in France (37). CC-ST10 isolates were recovered both from internal and external organs, but also from milt, eggs, and ovarian fluid (see Table S1 in the supplemental material). CC-ST10 contained 14 different STs of which in this collection ST10 was predicted to be the primary founder (Fig. 1). Most ST10 isolates originated from Denmark and were collected recurrently through the years. ST2, subgroup founder of CC-ST10, was the most predominant genotype, being represented by 157 isolates, i.e., 28% of the total data set. Although predominantly isolated in Norway and Denmark, ST2 was also identified in Finland and Sweden. Interestingly, ST2 was the only genotype belonging to CC-ST10 found in Norway. The other predicted subgroup founder in CC-ST10, ST79 (n = 60), was mostly found in Denmark and Finland. In addition to causing problems in fish farms in these countries from 1988 to 2006, ST79 was isolated from a feral rainbow trout in Finland in 1993. ST12 (n = 25) and ST92 (n = 71) belonging to CC-ST10, were recurrently isolated in Denmark, Finland, and Sweden. Although ST12 was identified already in 1988, outbreaks of ST92 (n = 71) had been occurring more recently (Fig. 3).
FIG 2 Annual percentages (from 1988 to 2012) of 446 rainbow trout-associated F. psychrophilum isolates of CC-ST10, other CCs, and singleton STs.
FIG 3 Timeline of emergence of different CC-ST10 STs in Nordic countries. Colors indicate the country of origin (red, Denmark; blue, Finland, green, Norway; yellow, Sweden).
eBURST revealed five additional CCs shared between the Nordic countries. CC-ST124 (6 STs) found in each country shared only few common alleles with CC-ST10 and was dominated by Finnish rainbow trout isolates but also included Atlantic salmon isolates from Finland and Norway. CC-ST191 (6 STs) included isolates from all countries except Sweden and comprised mainly of Danish rainbow and brown trout isolates (Table 1). CC-ST121 (2 STs), CC-ST138 (3 STs), and CC-ST236 (3 STs) were all identified in Finland and included isolates from Norway, Denmark, and Sweden, respectively. Some CCs showed limited distribution. CC-ST123 containing 12 isolates from Atlantic salmon or brown trout was exclusively found in Norway. CC-ST125 with 17 rainbow trout-associated isolates was limited to Finland. CC-ST138 was limited to rainbow trout and originated from Finland and Denmark. The remainder of the isolates was distributed between four minor CCs and 30 singletons (Table 1), which were identified from a diverse array of fish hosts, including Atlantic salmon, brown trout, and perch. Interestingly, the type strain NCIMB 1947T isolated from Coho salmon in the United States seems closely related to a Finnish sea trout isolate (ST13) in CC-ST13.
MLST revealed that individual disease outbreaks in Atlantic salmon, perch, and rainbow trout sometimes involved more than one ST or CC. Coinfections of the same individual fish by distinct STs were, however, only observed in rainbow trout.

Population genetic and phylogenetic analyses.

Genetic diversity (H) in the total data set was estimated to be 0.6127 ± 0.0420. Limiting the analysis to rainbow trout isolates reduced H to 0.4589 ± 0.0622, while non-rainbow trout isolates alone resulted in an H of 0.8733 ± 0.0108. The rainbow trout isolates from Denmark (H = 0.3949 ± 0.0501), Finland (H = 0.5401 ± 0.0626), and Sweden (H = 0.4594 ± 0.0714) showed a similar range of diversity, while corresponding Norwegian isolates showed a much lower diversity (H = 0.0870 ± 0.0145). The ISA value for the whole population was estimated to be 0.5419, which differed significantly from zero (P < 0.05) and thus rejected the null hypothesis of linkage equilibrium. Avoiding overrepresentation of particular genotypes by restricting the analysis to single representatives of the 81 STs identified in the study, decreased the ISA value to 0.2353 (P < 0.05). By further limiting the analysis to one representative (the founder) for each CC (12 STs) and singleton (30 STs), the ISA value fell to 0.1253 (P < 0.05), closer to linkage equilibrium. A PHI test conducted for the concatenated sequences of each ST showed statistically significant evidence of recombination (P = 0.0). A reticulated structure consistent with the presence of recombination events in the history of the studied genotypes can be seen in the phylogenetic network reconstructed with SplitsTree (Fig. 4).
FIG 4 The reticulated phylogenetic structure of the neighbor-net analysis is indicative of extensive recombination of loci providing additional support for the high estimates of recombination supported by the PHI test (P = 0.0). Major CCs determined in eBURST are indicated. See the text for details.


As aquaculture production continues to expand and intensify, more efficient pathogen control strategies are required to ensure animal health. Infections with F. psychrophilum pose a significant threat to salmonid farming around the world and preventative measures are urgently needed. To be able to monitor the spread of this important pathogen on a global scale, quick and reproducible typing methods are required. MLST using sequences of (usually seven) housekeeping genes has become the gold standard for bacterial typing and population analysis due to its discriminatory power and portability between laboratories, making it amenable to international collaboration.
MLST data can be used to investigate the genetic relatedness among bacterial species. However, the true evolutionary relationship between STs can be distorted by the high level of recombination in F. psychrophilum reported in previous studies (27, 38). In our analyses, recombination is clearly visible from the results of the PHI test and from the level of reticulation of the phylogenetic network (Fig. 4). Despite the high levels of recombination, the ATs at the 7 MLST loci were found to be in linkage disequilibrium. This result should be interpreted as a direct consequence of the high level of clonality which is also in keeping with the output of eBURST analysis that delineated large CCs in our data set (Fig. 1). This interpretation is corroborated by the observed reduction in ISA when linkage analysis was limited to single ST representatives (ISA = 0.2353) and to one ST of each CC (founder) and singleton (ISA = 0.1253). The observed clonality in the whole population could be explained by the dominance of clonal lineages relatively unbroken by horizontal gene transmission either as a result of niche specialization (i.e., host adaptation), physical barriers, or recent introduction. Of note, the main part of the isolates were collected from disease outbreaks which certainly contribute to an overrepresentation of the virulent clones in our sample and thus to the observed level of clonality. Nevertheless, the present study also included isolates from sources where the bacterium was suspected to have a more opportunistic or accidental role.
Taking the large number of isolates into account, our results strongly support the proposed epidemic population structure of F. psychrophilum (27, 35, 39), where highly successful epidemic clones have arisen from a generally recombinant background population. The eBURST population snapshot (Fig. 1) displaying the presence of CCs against a background of diverse singletons confirms the epidemic structure (40).
The present study revealed a higher overall genetic diversity of F. psychrophilum in the Nordic countries (H = 0.6127 ± 0.0420) compared to the previous global study (H = 0.5333) by Nicolas et al. (27). However, the diversity in rainbow trout isolates in Denmark, Finland, and Sweden was within a similar range to that reported in France (H = 0.4313) (35), while the low diversity in corresponding Norwegian isolates reflected the dominance of one genotype (ST2) in that country. The higher diversity observed among non-rainbow trout isolates and the whole population indicates that BCWD in rainbow trout is caused by a genetically less diverse group of genotypes. Our study shows that since first recognition of BCWD in the Nordic countries, most outbreaks in rainbow trout have been caused by a group of closely related strains belonging to CC-ST10 (Table 1). Unfortunately, the previous work regarding MLST of Norwegian and Chilean F. psychrophilum isolates by Apablaza et al. (41) involved sequencing of shorter loci and did not allow assigning of comparable STs. However, CC-ST10 in the present investigation includes ST2, the predicted founder in French (35), Japanese (39), Swiss (42), and global (27) studies. The predicted founder of CC-ST10 in the present study has also been detected in North America, Scotland, Switzerland, Chile, and Japan ( The timeline of emergence of CC-ST10 genotypes (Fig. 3) does not provide additional support for identification of ST2 or ST10 as the founder of CC-ST10 as STs belonging to this CC were isolated earlier. The sole dominance of ST2 within CC-ST10 in Norway, post-2004, strongly supports the hypothesis of recent introduction of this ST to that country. The wide geographical distribution of CC-ST10 genotypes and their association with the rainbow trout host indicates a strong niche adaptation and a high transmissibility. The isolation of CC-ST10 genotypes from juvenile and adult fish, eggs, and sexual fluids indicates that trade not only in live fish but also in eggs and gametes provides the potential for spread of this epidemic clone between countries or regions.
With the exception of ST70, the five other shared CCs (CC-ST121, CC-ST124, CC-ST138, CC-ST191, and CC-ST236) contained unique STs not identified outside the study area (, indicating that these genotypes may be restricted to and possibly spread between the Nordic countries by trade of rainbow trout and associated products. On the contrary, CC-ST90, which was identified as the second dominant CC in France (35) and Switzerland (42), was absent in the Nordic countries. The presence of temporally stable country-specific CCs in Finland (CC-ST125) and Norway (CC-ST123) indicate that in addition to the shared CCs, endemic F. psychrophilum genotypes have remained in circulation causing recurrent infections.
Almost 80% of the isolates appearing as singleton STs were of non-rainbow trout origin. Some, e.g., ST23, were recurrently found in different host species. These singletons have been suggested to represent less virulent endemic isolates or environmental strains, acting more like opportunistic pathogens compared to other more host-specific strains (35). It is likely, therefore, that these endemic isolates appearing as singletons and minor CCs do not pose a direct threat to farmed rainbow trout.
Our study also revealed mixed infections in rainbow trout involving isolates of distinct ST and CC, a phenomenon previously reported in ayu in Japan (39). The ubiquity of F. psychrophilum creates an opportunity for coinfection and subsequent recombination in the environment and possibly within the host. These observations show that, for the purpose of detecting and monitoring emergent F. psychrophilum genotypes, it is important to examine more than a single colony from an infected fish. To aid in rapid differentiation of clinically relevant STs from coinfections with distinct genotypes, an applicable examination strategy based on phenotypically distinguishable characteristics, such as colony morphology and antimicrobial susceptibility (32), should be set up.
Thus far, attempts to link specific genotypes of F. psychrophilum with clinical information and high mortalities in fish farms have been limited. The predominance of ST2 in the Nordic countries, France (35), and Switzerland (42) and its association with severe BCWD outbreaks in rainbow trout make it a genotype of particular clinical significance. In Norway, where ST2 was responsible for mortalities of up to 90% during an epizootic in 2008 (26), its apparent host specificity was demonstrated during an outbreak in which only rainbow trout fry were affected on a site also producing Atlantic salmon juveniles. Since then, ST2 has also been responsible for recurrent outbreaks in brackish-water-farmed rainbow trout epidemiologically linked to the epizootics in 2008. These ST2 isolates were shown to have a decreased sensitivity toward quinolones (43), which could provide them with a selective advantage over other strains found in fish farm environments. However, ST2 being the only genotype of CC-ST10 found in Norway where BCWD has not been a serious problem until recent years, leads to the speculation that this newly emerged genotype of F. psychrophilum has not yet diversified in this country and may in the future give rise to new epidemic clones which may threaten Norwegian rainbow trout production. Against the current background, monitoring F. psychrophilum genotypes in Norway would provide an excellent opportunity for studying the evolutionary processes in F. psychrophilum from introduction of an epidemic clone to diversification.
As opposed to the situation in Norway, F. psychrophilum isolates from Finland showed high genotypic diversity. Although more than half of the 196 Finnish isolates belonged to CC-ST10, only six of them were of ST2. These ST2 isolates were collected from outbreaks in two different farms in 2011, suggesting that ST2 has until recently not been able to establish a foothold in Finnish fish farms. Historically, since the first reports of BCWD in Finland (and Denmark), ST79 was responsible for repeated outbreaks in rainbow trout farms but, more recently, disease outbreaks have been associated with ST92 (Fig. 3). The Finnish CC-ST125, which clustered together with CC-ST10 in the phylogenetic network reconstructed with SplitsTree (Fig. 4), is also causing recurrent outbreaks and contains genotypes associated with high virulence in rainbow trout (44). The small genetic distance between these two CCs suggest that STs may exist (or have existed) which would join these CCs together in an eBURST analysis. In Denmark, almost 80% of the isolates belonged to CC-ST10 of which ST2 and ST10 have been associated with severe disease outbreaks in rainbow trout since 1990 until present time (Fig. 3). Most Swedish isolates were collected from clinically affected farmed rainbow trout since 1988 and belonged to CC-ST10.
CC-ST191 and CC-ST236 are closely related CCs consisting mainly of rainbow trout isolates, although brown trout isolates are also represented. These CCs cluster together in the phylogenetic network (Fig. 4) and belong in a relaxed eBURST analysis to the same CC (Table 1), which is only distantly related to CC-ST10. This complex includes the three ST181 isolates from rainbow trout in Norway, all of which originated from a single outbreak in a site which has imported rainbow trout during recent years, supporting host transmission of pioneering strains to new geographical areas. Interestingly, this ST was also isolated in Denmark in the same period, but only from the surface of fish and not in connection with disease outbreaks.
In the present study, isolates originating from Atlantic salmon were mainly associated with ST70 and the Norwegian CC-ST123. ST70 was isolated several times from Norwegian Atlantic salmon associated with systemic infections with degradation of skeletal muscle and external lesions (5). ST70 has also been identified in the same host species in Chile and North America ( The association between infections in Atlantic salmon and specific F. psychrophilum genotypes was demonstrated by the country-specific CC-ST123, which contained almost exclusively Norwegian Atlantic salmon isolates both from commercial and restocking sites. Whether this represents a threat to the salmon industry remains to be seen. Phylogenetic analyses show that the rainbow trout-specific CC-ST10 is evolutionarily distinct from major CCs containing genotypes infecting Atlantic salmon (Fig. 4) and those infecting ayu in Japan (39), suggesting a mechanism for host specificity or niche adaptation in these genotypes.
We describe here the MLST of a large collection of F. psychrophilum isolates collected from diverse origins and years, from aquaculture in a whole region, including different countries. This represents a successful interlaboratory comparison of results between national laboratories and adds support to earlier hypotheses that virulent variants have spread to new areas within the studied region. Implementation of a reliable typing tool enabling global F. psychrophilum surveillance is of paramount importance to prevent the transmission of genotypes associated with severe disease outbreaks—in particular to countries with growing aquaculture production. Since the population of F. psychrophilum seems to be epidemic, identification of highly virulent bacterial variants is crucial in preventing the spread of epidemic clones. The simultaneous presence of genetically divergent clones (Fig. 2) poses an additional challenge for disease prevention and control. The identification of CCs enables a rational selection of representative genotypes for research aiming to determine the genomic diversity of F. psychrophilum. Further work aimed at the identification of virulence-related loci within particular STs or CCs is important for risk assessment and for developing tools for diagnostics and control. This could be accomplished by whole-genome sequencing of genotypes selected on the basis of knowledge gained in the present and related studies.


This study was funded by the EU EMIDA ERA-NET project “Control Flavobacteriaceae Infections in European Fish Farms” (RCN 202834/E40 for H.N. and D.J.C., 2114/311/210 for K.S. and T.W., ANR 2010-EMID-006-01 for E.D. and P.N., and 3405-10-0182 for I.D. and L.M.). Swedish investigations were supported through a climate research-directed governmental funding.
We thank all partners of the consortium. We thank Anne Berit Olsen for critically reading the manuscript, Åse Garseth and Kristine Gimservik for collection of isolates from mitigation sites, and Kirsten Bottolfsen, Lene Gertman, Maija Liisa Hoffrén, Sofia Lindström, Nora Martinussen Tandstad, and Fanny Örn for technical assistance.

Supplemental Material

File (zam999105297so1.pdf)
ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.


Barnes ME and Brown ML. 2011. A review of Flavobacterium psychrophilum biology, clinical signs, and bacterial cold water disease prevention and treatment. Open Fish Sci. J. 4:1–9.
Nematollahi A, Decostere A, Pasmans F, and Haesebrouck F. 2003. Flavobacterium psychrophilum infections in salmonid fish. J. Fish Dis. 26:563–574.
Borg AF. 1948. Studies on myxobacteria associated with diseases in salmonid fishes. Ph.D. thesis. University of Washington. Seattle, WA.
Wakabayashi H, Horinouchi M, Bunya T, and Hoshiai G. 1991. Outbreaks of cold-water disease in coho salmon in Japan. Fish Pathol. 26:211–212.
Nilsen H, Johansen R, Colquhoun DJ, Kaada I, Bottolfsen K, Vågnes Ø, and Olsen AB. 2011. Flavobacterium psychrophilum associated with septicemia and necrotic myositis in Atlantic salmon Salmo salar: a case report. Dis. Aquat. Org. 97:37–46.
Austin B and Austin DA. 2013. Bacterial fish pathogens: disease of farmed and wild fish, 5th ed. Springer, Dordrecht, Netherlands.
Cipriano RC and Holt RA. 2005. Flavobacterium psychrophilum, cause of bacterial cold-water disease and rainbow trout fry syndrome. Fish disease leaflet no. 86. U.S. Geological Survey, Leetown Science Center, National Fish Health Research Laboratory, Kearneysville, WV.
Brown LL, Cox WT, and Levine RP. 1997. Evidence that the causal agent of bacterial cold-water disease Flavobacterium psychrophilum is transmitted within salmonid eggs. Dis. Aquat. Org. 29:213–218.
Madsen L, Møller JD, and Dalsgaard I. 2005. Flavobacterium psychrophilum in rainbow trout, Oncorhynchus mykiss (Walbaum), hatcheries: studies on broodstock, eggs, fry, and environment. J. Fish Dis. 28:39–47.
Taylor PW. 2004. Detection of Flavobacterium psychrophilum in eggs and sexual fluids of pacific salmonids by a polymerase chain reaction assay: implications for vertical transmission of bacterial coldwater disease. J. Aquat. Anim. Health 16:104–108.
Kumagai A, Nakayasu C, and Oseko N. 2004. No evidence for the presence of Flavobacterium psychrophilum within ayu eggs. Fish Pathol. 39:183–187.
Madsen L and Dalsgaard I. 2008. Water recirculation and good management: potential methods to avoid disease outbreaks with Flavobacterium psychrophilum. J. Fish Dis. 31:799–810.
Borg AF. 1960. Studies on myxobacteria associated with diseases in salmonid fishes. Wildlife diseases, no. 8. American Association for the Advancement of Science, Washington, DC.
Austin B and Stobie M. 1991. Recovery of yellow-pigmented bacteria from dead and moribund fish during outbreaks of rainbow trout, Oncorhynchus mykiss (Walbaum), fry syndrome in England. J. Fish Dis. 14:677–682.
Balta F. 1997. Kültürü yapılan alabalıklarda (Oncorhynchus mykiss) görülen Flexibacter psychrophila enfeksiyonu. IX. Ulusal Su Ürünleri Sempozyumu. Eğirdir/Isparta 1997:641–648. (In Turkish.)
Bernardet JF, Baudin-Laurencin F, and Tixerant G. 1988. First identification of Cytophaga psychrophila in France. Bull. Eur. Assoc. Fish Pathol. 8:104–105.
Sarti BM, Giorgettii G, and Manfrin A. 1992. Method for the rapid diagnosis of visceral myxobacteriosis in reared trout in Italy. Bull. Eur. Assoc. Fish Pathol. 12:53.
Toranzo AE and Barja JL. 1993. Fry mortality syndrome (FMS) in Spain. Isolation of the causative bacterium Flexibacter psychrophilus. Bull. Eur. Assoc. Fish Pathol. 13:30–32.
Weis J. 1987. A cold water disease in rainbow trout. Über das Vorkommen einer Kaltwasserkrankheit bei Regenbogenforellen, Salmo gairdneri. Tierarztl. Umsch. 42:575–578.
Bustos PA, Calbuyahue J, Montaña J, Opazo B, Entrala P, and Solervisenc R. 1995. First isolation of Flexibacter psychrophilus, as causative agent of rainbow trout fry syndrome (RTFS), producing rainbow trout mortality in Chile. Bull. Eur. Assoc. Fish Pathol. 15:162–164.
Lee KB and Heo GJ. 1998. First isolation and identification of Cytophaga psychrophila from cultured ayu in Korea. Fish Pathol. 33:37–38.
Schmidtke LM and Carson J. 1995. Characteristics of Flexibacter psychrophilus isolated from Atlantic salmon in Australia. Dis. Aquat. Org. 21:157–161.
Dalsgaard I and Hørlyck V. 1990. Koldtvandssyge eller vintersår hos ørreder (cold water diseases or winter ulcers in trout). Ferskvandsfiskeribladet 88:118–120. (In Danish.)
Lorenzen E, Dalsgaard I, From J, Hansen EM, Hørlyck V, Korsholm H, Mellergaard S, and Olesen NJ. 1991. Preliminary investigation of fry mortality syndrome in rainbow trout. Bull. Eur. Assoc. Fish Pathol. 11:77–79.
Wiklund T, Kaas K, Lönnström L, and Dalsgaard I. 1994. Isolation of Cytophaga psychrophila (Flexibacter psychrophilus) from wild and farmed rainbow trout (Oncorhynchus mykiss) in Finland. Bull. Eur. Assoc. Fish Pathol. 14:44–46.
Nilsen H, Olsen AB, Vaagnes Ø, Helleberg H, Bottolfsen K, Skjelstad H, and Colquhoun DJ. 2011. Systemic Flavobacterium psychrophilum infection in rainbow trout, Oncorhynchus mykiss (Walbaum), farmed in fresh and brackish water in Norway. J. Fish Dis. 34:403–408.
Nicolas P, Mondot S, Achaz G, Bouchenot C, Bernardet JF, and Duchaud E. 2008. Population structure of the fish-pathogenic bacterium Flavobacterium psychrophilum. Appl. Environ. Microbiol. 74:3702–3709.
Anacker RL and Ordal EJ. 1959. Studies on the myxobacterium Chondrococcus columnaris. I. Serological typing. J. Bacteriol. 78:25–32.
Holt RA, Rohovec JS, and Fryer JL. 1993. Bacterial coldwater disease, p 3–22. In Inglis V, Roberts RJ, and Bromage NR (ed), Bacterial diseases of fish. Blackwell Scientific Publications, Oxford, United Kingdom.
Bernardet JF, Nakagawa Y, and Holmes B. 2002. Proposed minimal standards for describing new taxa of the family Flavobacteriaceae and emended description of the family. Int. J. Syst. Evol. Microbiol. 52:1049–1070.
Dalsgaard I and Madsen L. 2000. Bacterial pathogens in rainbow trout, Oncorhynchus mykiss (Walbaum), reared at Danish freshwater farms. J. Fish Dis. 23:199–209.
Sundell K, Heinikainen S, and Wiklund T. 2013. Structure of Flavobacterium psychrophilum populations infecting farmed rainbow trout (Oncorhynchus mykiss). Dis. Aquat. Org. 103:111–119.
Wiklund T, Madsen L, Bruun MS, and Dalsgaard I. 2000. Detection of Flavobacterium psychrophilum from fish tissue and water samples by PCR amplification. J. Appl. Microbiol. 88:299–307.
Högfors-Rönnholm E and Wiklund T. 2010. Phase variation in Flavobacterium psychrophilum: characterization of two distinct colony phenotypes. Dis. Aquat. Org. 90:43–53.
Siekoula-Nguedia C, Blanc G, Duchaud E, and Calvez S. 2012. Genetic diversity of Flavobacterium psychrophilum isolated from rainbow trout in France: predominance of a clonal complex. Vet. Microbiol. 161:169–178.
Haubold H and Hudson RR. 2000. LIAN 3.0: detecting linkage disequilibrium in multilocus data. Bioinformatics 16:847–848.
Duchaud E, Boussaha M, Loux V, Bernardet JF, Michel C, Kerouault B, Mondot S, Nicolas P, Bossy R, Caron C, Bessieres P, Gibrat JF, Claverol S, Dumetz F, Le Henaff M, and Benmansour A. 2007. Complete genome sequence of the fish pathogen Flavobacterium psychrophilum. Nat. Biotechnol. 25:763–769.
Vos M and Didelot X. 2009. A comparison of homologous recombination rates in bacteria and archaea. ISME J. 3:199–208.
Fujiwara-Nagata E, Chantry-Darmon C, Bernardet JF, Eguchi M, Duchaud E, and Nicolas P. 2013. Population structure of the fish pathogen Flavobacterium psychrophilum at whole-country and model river levels in Japan. Vet. Res. 44:34.
Smith JM, Feil EJ, and Smith NH. 2000. Population structure and evolutionary dynamics of pathogenic bacteria. Bioessays 22:1115–1122.
Apablaza P, Løland AD, Brevik OJ, Ilardi P, Battaglia J, and Nylund A. 2013. Genetic variation among Flavobacterium psychrophilum isolates from wild and farmed salmonids in Norway and Chile. J. Appl. Microbiol. 114:934–946.
Strepparava N, Nicolas P, Wahli T, Segner H, and Petrini O. 2013. Molecular epidemiology of Flavobacterium psychrophilum from Swiss fish farms. Dis. Aquat. Org. 105:203–210.
Shah SQA, Nilsen H, Bottolfsen K, Colquhoun DJ, and Sørum H. 2012. DNA gyrase and topoisomerase IV mutations in quinolone-resistant Flavobacterium psychrophilum isolated from diseased salmonids in Norway. Microb. Drug Resist. 18:207–214.
Madetoja J, Dalsgaard I, and Wiklund T. 2002. Occurrence of Flavobacterium psychrophilum in fish-farming environments. Dis. Aquat. Org. 52:109–118.

Information & Contributors


Published In

cover image Applied and Environmental Microbiology
Applied and Environmental Microbiology
Volume 80Number 91 May 2014
Pages: 2728 - 2736
Editor: C. A. Elkins
PubMed: 24561585


Received: 20 December 2013
Accepted: 13 February 2014
Published online: 14 April 2014


Request permissions for this article.



Hanne Nilsen
Norwegian Veterinary Institute, Bergen, Norway
Krister Sundell
Åbo Akademi University, Turku, Finland
Eric Duchaud
INRA, Virologie et Immunologie Moléculaires, UR892, Jouy-en-Josas, France
Pierre Nicolas
INRA, Mathématique Informatique et Génome, UR1077, Jouy-en-Josas, France
Inger Dalsgaard
National Veterinary Institute, Technical University of Denmark, Frederiksberg, Denmark
Lone Madsen
National Veterinary Institute, Technical University of Denmark, Frederiksberg, Denmark
Anna Aspán
National Veterinary Institute, Section of Fish, Uppsala, Sweden
Eva Jansson
National Veterinary Institute, Section of Fish, Uppsala, Sweden
Duncan J. Colquhoun
Norwegian Veterinary Institute, Bergen, Norway
Tom Wiklund
Åbo Akademi University, Turku, Finland


C. A. Elkins


Address correspondence to Hanne Nilsen, [email protected], or Krister Sundell, [email protected].
H.N. and K.S. contributed equally to this article.

Metrics & Citations


Note: There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.

Citation counts come from the Crossref Cited by service.


If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.

View Options

Figures and Media






Share the article link

Share with email

Email a colleague

Share on social media

American Society for Microbiology ("ASM") is committed to maintaining your confidence and trust with respect to the information we collect from you on websites owned and operated by ASM ("ASM Web Sites") and other sources. This Privacy Policy sets forth the information we collect about you, how we use this information and the choices you have about how we use such information.
FIND OUT MORE about the privacy policy