Salmonella Enteritidis Phage Types 1 and 4: Pheno- and Genotypic Epidemiology of Recent Outbreaks in Finland

Abstract

In the 1990s, Salmonella enterica subsp. enterica serovar Enteritidis has caused 15 outbreaks in Finland; 12 of them were caused by phage type 1 (PT1) and PT4. Thus far, there has been no clear evidence as to the source of these Salmonella Enteritidis PT1 and PT4 strains, so it was necessary to try to characterize them further. Salmonella Enteritidis PT1 (n = 57) and PT4 (n = 43) isolates from different sources were analyzed by genomic pulsed-field gel electrophoresis (PFGE), plasmid profiling, and antimicrobial resistance testing to investigate the distribution of their subtypes in Finland. It was also hoped that this investigation would help in identifying the sources of the infections, especially the sources of the outbreaks caused by PT1 and PT4 in the 1990s. The results showed that both PFGE and plasmid profiling, but not antimicrobial susceptibility testing, were capable of differentiating isolates of Salmonella Enteritidis PT1 and PT4. By genotypic methods, it was possible to divide both PT1 and PT4 isolates into 12 subtypes. It could also be shown that all PT1 outbreak isolates were identical and, at least with this collection of isolates, that the outbreaks did not originate from the Baltic countries or from Russia, where this phage type predominates. It was also established that the outbreaks caused by PT4 all had different origins. Valuable information for future investigations was gained on the distribution of molecular subtypes of strains that originated from the tourist resorts that are popular among Finns and of strains that were isolated from livestock.

Salmonella enterica subsp. enterica serovar Enteritidis is the most common serovar that causes salmonellosis in humans. Since the 1980s, there has been a dramatic increase in the number of reported findings of this serovar in Europe and worldwide (8, 13, 19). The most frequent reservoir for Salmonella Enteritidis has been poultry, especially eggs (19). The occurrence of the phage types of Salmonella Enteritidis in different geographical areas has varied: phage type 1 (PT1) has been common in the Baltic countries and Russia (7), whereas PT4 most often has been seen in Western European countries (9, 20). PT8 apparently has been a frequent finding in the United States (8).

The proportion of Salmonella Enteritidis causing indigenously acquired salmonellosis has greatly increased in Finland in the past few years. The strains that have been the cause of this increase have originated from abroad, whereas the proportion of domestic strains has remained fairly constant (Fig. 1). Salmonella Enteritidis has caused 15 domestic outbreaks during the 1990s; 9, 3, and 3 were caused by Salmonella Enteritidis PT1, Salmonella Enteritidis PT4, and other phage types, respectively (21a).

FIG. 1.

Number of all salmonella infections, number of domestic and foreign Salmonella (S.) Enteritidis infections in Finland between 1965 and 1997 (21a), and number of Finnish passengers on charter flights (statistics of the Civil Aviation Administration and Statistics Finland).

When one is tracing the origin of a salmonella infection, it is important to differentiate between strains in great detail. Traditional epidemiological methods include biotyping, serotyping, and phage typing of isolates, as well as antimicrobial susceptibility testing, although these methods do not always give enough information for epidemiological purposes. Also, phage typing is not available in all laboratories, and certain strains of Salmonella Enteritidis can change phage types (1, 16, 17, 27, 28). Some phage types may predominate in a geographical area, a fact which may limit the use of phage typing in investigating local outbreaks. However, more information can be gathered by analyzing plasmids from the isolates (2, 29) or chromosomal DNA by different molecular techniques, such as ribotyping (6), IS200 typing (23), or pulsed-field gel electrophoresis (PFGE) (15).

In Finland, data on the epidemiologic characteristics and potential reservoirs of Salmonella Enteritidis PT1 and PT4 have been insufficient. To gain more detailed information, the highly discriminatory technique PFGE was chosen for subtyping of the strains of these phage types (11, 14, 18). Plasmid analysis and antimicrobial susceptibility testing were used to complement the results of PFGE.

MATERIALS AND METHODS

Salmonella isolates.

All strains had previously been phage typed as Salmonella Enteritidis PT1 and PT4 according to the scheme of Ward et al. (31) at the Laboratory of Enteric Pathogens and stored at −70°C in sterilized skim milk or at room temperature in nutrient agar tubes. Along with a set of PT1 and PT4 outbreak isolates, representative sets of nonoutbreak isolates from humans, as well as all available PT1 and PT4 isolates from livestock and foodstuffs, were included in the study.

Salmonella Enteritidis PT1.

Fifty-seven strains of Salmonella Enteritidis PT1 were studied (Table 1); 26 of the 51 human isolates were of domestic origin, and 6 of them represented the PT1 outbreaks in a 5-year period from 1991 to 1995 (Fig. 2 and Table 2). Two isolates (chicken liver and eggshell) were obtained from the poultry farm that was the source of the outbreaks in Turku, Finland (10). Twenty-five strains were of foreign origin; 4 of them were obtained from the Baltic countries (kindly sent by Unna Jöks, Central Laboratory of Microbiology, Health Protection Inspectorate, Tallinn, Estonia), and 21 were isolated from Finnish tourists who had been abroad preceding their salmonella infections. These nonoutbreak strains were selected randomly to represent strains from various countries. Three nonoutbreak strains were from animals, and one was from foodstuffs.

TABLE 1.

Salmonella Enteritidis PT1 isolates

Strain	Geographic origin/source	Yr of isolation	PFGE subtype	Plasmid profile	Antimicrobial resistancea	Additional information
IH 68475	Finland, Jyväskylä	1996	1A	a		Human isolate
IH 68476	Finland, Rovaniemi	1996	1A	a		Human isolate
IH 68477	Finland, Kuopio	1996	1A	a		Human isolate
IH 104209	Finland, Helsinki	1996	1A	a		Human isolate
IH 68473	Finland, Kuopio	1996	1A	d		Human isolate
IH 69249	Finland, Haukipudas	1997	1A	a		Human isolate
IH 69250	Finland, Heinola	1997	1A	a		Human isolate
IH 69257	Finland, Kajaani	1997	1A	a		Human isolate
IH 69263	Finland, Helsinki	1997	1A	a		Human isolate
IH 68250	Baltic countries	1996	1A	a		Human isolateb
IH 68251	Baltic countries	1996	1A	a		Human isolateb
IH 68480	Estonia	1996	1A	a		Human isolatec
IH 68485	Norway	1996	1A	a		Human isolatec
IH 104498	Portugal	1996	1A	a		Human isolatec
IH 68484	Germany	1996	1A	c	Sul, Tet, Str	Human isolatec
IH 68481	Latvia	1996	1A	d		Human isolatec
IH 69286	Estonia	1997	1A	a		Human isolatec
IH 69441	Estonia	1998	1A	a		Human isolatec
IH 69043	Baltic countries	1998	1A	a		Human isolateb
IH 69045	Baltic countries	1998	1A	a		Human isolateb
IH 68478	Finland, Helsinki	1996	1B	a		Human isolate
IH 104682	Finland, Vihti	1996	1B	a		Human isolate
IH 68479	Finland, Helsinki	1996	1B	a		Human isolate
IH 69264	Finland, Seinäjoki	1997	1B	a	Nal	Human isolate
IH 68979	Finland, Ikaalinen	1998	1B	a		Human isolate
IH 69026	Finland, Helsinki	1998	1B	a		Human isolate
IH 68875	Finland, Nummi-Pusula	1998	1B	a		Human isolate
IH 69434	Finland, Hämeenlinna	1998	1B	e		Human isolate
IH 68489	Spain	1993	1B	a		Human isolatec
IH 68490	France	1994	1B	a		Human isolatec
IH 67741	Portugal	1995	1B	a		Human isolatec
IH 68482	Poland	1996	1B	a		Human isolatec
IH 68483	Spain	1996	1B	a		Human isolatec
IH 68486	Spain, Canary Islands	1996	1B	a		Human isolatec
IH 69241	Portugal	1997	1B	a		Human isolatec
IH 69246	Germany	1997	1B	f		Human isolatec
IH 69442	Spain, Canary Islands	1998	1B	a		Human isolatec
IH 69440	Spain, Canary Islands	1998	1B	a	Nal	Human isolatec
IH 69443	Spain	1998	1B	a	Nal	Human isolatec
IH 69435	Finland/strip of leek	1995	1B	a		Foodstuff
IH 68487	Finland/cow	1995	1B	a		Animal isolate, sweep sample
IH 68488	Finland/cow	1995	1B	a		Animal isolate, sweep sample
IH 68474	Finland, Vihti	1996	1C	a		Human isolate
IH 104210	Finland, Jyväskylä	1996	1D	g		Human isolate
IH 67791	Finland/cow	1994	1E	g		Animal isolate, feces
IH 59089	Finland, Vieremä	1991	1F	a		Outbreak strain, human isolate
IH 59439	Finland, Vieremä	1991	1F	a		Outbreak strain, human isolate
IH 59958	Finland, Turku	1992	1F	a		Outbreak strain, human isolate
IH 59943	Finland, Turku	1993	1F	a		Outbreak strain, human isolate
IH 66842	Finland, Turku	1994	1F	a		Outbreak strain, human isolate
IH 67745	Finland, Turku	1995	1F	a		Outbreak strain, human isolate
IH 68964	Finland, Tornio	1998	1F	a		Human isolate
IH 67832	Finland/chicken liver	1995	1F	a		Outbreak strain, animal isolated
IH 67920	Finland/eggshell	1995	1F	a		Outbreak strain, foodstuffd
IH 59430	Spain, Canary Islands	1992	1F	b		Human isolatec
IH 69277	Hungary	1997	1F	b		Human isolatec
IH 69282	Russia	1997	1G	d		Human isolatec

^aThe strain is sensitive to ampicillin (Amp), chloramphenicol (Chl), ceftriaxone (Cef), imipenem (Imi), mecillinam (Mec), nalidixic acid (Nal), neomycin (Neo), sulfonamide (Sul), tetracycline (Tet), trimethoprim (Tmp), streptomycin (Str), and ciprofloxacin (Cip) unless resistance is otherwise shown. 

^bThe strain was provided by the Central Laboratory of Microbiology, Health Protection Inspectorate, Tallinn, Estonia. 

^cThe strain was from a Finnish tourist who had visited the country listed before the salmonella finding. 

^dThe strain was from an outbreak in a Finnish commercial layer flock (10). 

FIG. 2.

Geographic occurrence of Salmonella Enteritidis PT1 (●) and PT4 (■) outbreaks in Finland in the 1990s. See Table 2 for the number of infections in each outbreak.

TABLE 2.

Number of infections in outbreaks

Phage type	Outbreak area	No. of infections during the indicated yr of the outbreak
		1991	1992	1993	1994	1995
PT1	Turkua	66	50	38	10	10
	Turkub	10		99		180
	Vieremä	20
PT4	Leppävirta			17
	Rauma					85
	Kotka					12

^aFirst outbreak. 

^bSecond outbreak. 

Salmonella Enteritidis PT4.

Forty-three Salmonella Enteritidis PT4 strains were studied (Table 3); 14 of the 32 human isolates were of domestic origin, and 6 of them represented three PT4 outbreaks between 1993 and 1995. Domestic nonoutbreak isolates from 1993 to 1995 were not available, but eight isolates from 1997 to 1998 were included in this study. Eighteen nonoutbreak strains of foreign origin, isolated in Finland from Finnish tourists who had been abroad preceding their salmonella infections, were selected randomly to represent strains from various countries. Eleven strains were from animals.

TABLE 3.

Salmonella Enteritidis PT4 isolates

Strain	Geographic origin/source	Yr of isolation	PFGE subtype	Plasmid profile	Antimicrobial resistancea	Additional information
IH 59741	Finland, Leppävirta	1993	4A	a		Outbreak strain, human isolate
IH 59742	Finland, Leppävirta	1993	4A	a		Outbreak strain, human isolate
IH 69444	Finland, Kuopio	1997	4A	b		Human isolate
IH 68274	Finland, Tampere	1997	4A	b		Human isolate
IH 69252	Finland, Helsinki	1997	4A	b		Human isolate
IH 69266	Finland, Tampere	1997	4A	b		Human isolate
IH 68336	Finland, Uusi Kaarlepyy	1997	4A	d		Human isolate
IH 69445	Finland, Juva	1998	4A	b		Human isolate
IH 69447	Finland, Palokka	1998	4A	b		Human isolate
IH 68178	Belgium/pork	1993	4A	b		Animal isolate, imported from Belgium
IH 59883	Finland/cow	1993	4A	b		Animal isolate, feces
IH 68501	Finland/cow	1995	4A	b		Animal isolate, feces
IH 68177	Finland/cow	1995	4A	b		Animal isolate, feces
IH 68232	England/chicken steak	1996	4A	g		Animal isolate, imported from England
IH 69455	Finland/cow	1997	4A	g		Animal isolate, feces
IH 69261	Finland/cow	1997	4A	g		Animal isolate, feces
IH 68651	Finland/turkey	1997	4A	g		Animal isolate, feces
IH 69265	Finland/turkey	1997	4A	g		Animal isolate, feces
IH 68502	France, Belgium	1993	4A	b		Human isolateb
IH 68544	Turkey	1993	4A	b		Human isolateb
IH 68503	Hungary	1993	4A	b		Human isolateb
IH 68509	Belgium	1995	4A	b		Human isolateb
IH 68506	Lithuania	1995	4A	b		Human isolateb
IH 68505	Cyprus	1995	4A	c	Sul, Str	Human isolateb
IH 68508	Germany	1995	4A	d		Human isolateb
IH 69225	Belgium	1997	4A	b		Human isolateb
IH 69228	England	1997	4A	b		Human isolateb
IH 69245	Germany	1997	4A	b		Human isolateb
IH 69279	Hungary	1997	4A	e	Mec	Human isolateb
IH 69000	Spain, Canary Islands	1998	4A	b		Human isolateb
IH 69453	Thailand	1998	4A	b		Human isolateb
IH 68952	Dominican Republic	1998	4A	b		Human isolateb
IH 69454	Spain, Canary Islands	1998	4A	c	Sul, Str	Human isolateb
IH 68189	Finland, Kotka	1995	4B	g		Outbreak strain, human isolate
IH 68190	Finland, Kotka	1995	4B	g		Outbreak strain, human isolate
IH 69456	Finland/cow	1997	4B	g		Animal isolate, feces
IH 68507	Spain	1995	4B	g		Human isolateb
IH 68545	Italy	1993	4C	b		Human isolateb
IH 68504	Turkey	1995	4D	b		Human isolateb
IH 67650	Finland, Rauma	1995	4E	b		Outbreak strain, human isolate
IH 67651	Finland, Rauma	1995	4E	b		Outbreak strain, human isolate
IH 69446	Finland, Mikkeli	1998	4F	c	Sul, Tet	Human isolate
IH 68656	France/turkey	1997	4G	f		Animal isolate, imported from France

^aThe strain is sensitive to ampicillin (Amp), chloramphenicol (Chl), ceftriaxone (Cef), imipenem (Imi), mecillinam (Mec), nalidixic acid (Nal), neomycin (Neo), sulfonamide (Sul), tetracycline (Tet), trimethoprim (Tmp), streptomycin (Str), and ciprofloxacin (Cip) unless resistance is otherwise shown. 

^bThe strain was from a Finnish tourist who had visited the country listed before the salmonella finding. 

PFGE.

Salmonella strains were grown overnight on Drigalski-Conradi agar plates at 37°C. Bacterial cells were suspended in 1,200 μl of TEN (0.1 M Tris-HCl, 0.15 M NaCl, 0.1 M EDTA [pH 7.5]) to an optical density at 600 nm of 0.150 to 0.200. This cell suspension was mixed in equal parts with molten 2% low-melting-point agarose (SeaPlaque agarose; FMC BioProducts, Rockland, Maine), and the mixture was pipetted into plug molds. The plugs were incubated overnight at 55 to 57°C in ES buffer (0.5 M EDTA, 1% N-(auroylsarcosine) with 0.15 mg of proteinase K per ml. The plugs were washed with TE buffer (10 mM Tris-HCl, 1 mM EDTA [pH 8.0]) for 30 min, with TE buffer–4 mM phenylmethylsulfonyl fluoride for 30 min to inactivate proteinase K, and with TE buffer three times for 30 min each time. The conditions for restriction endonuclease digestion and PFGE were essentially those described by Thong et al. (26). Chromosomal DNA was digested with 10 U of restriction enzymes XbaI, SpeI, and NotI (Boehringer Mannheim GmbH, Mannheim, Germany) overnight at 37°C. Electrophoresis was performed at 210 V on 1.0% SeaKem ME agarose gels (FMC BioProducts) with a Gene Navigator system (Pharmacia, Uppsala, Sweden) or with CHEF Mapper systems (Bio-Rad Laboratories, Richmond, Calif.). Running conditions were as follows: XbaI digests, 5 to 70 s for 24 h; SpeI digests, 5 to 40 s for 24 h; and NotI digests, 1 to 20 s for 20 h. A bacteriophage lambda ladder consisting of concatemers in increments of 48.5 kbp (New England BioLabs Inc., Beverly, Mass.) was used as a molecular weight standard. Any difference between two profiles was considered sufficient to distinguish two different PFGE profiles. XbaI-digested PFGE profiles were named with uppercase letters starting at A; the numbers 1 and 4 indicate whether the strain belonged to PT1 or PT4, respectively.

Plasmid profiling.

Plasmid DNA was isolated by the alkaline lysis method as described earlier (5). Samples were analyzed by electrophoresis at 120 V for 70 min on horizontal 0.9% SeaKem ME agarose gels with a CIBCO BRL Horizon 20.25 system (Life Technologies Inc., Gathersburg, Md.). The gels were stained for 30 min with ethidium bromide (0.5 μg/ml) and photographed under UV illumination. Plasmid-containing strains RH 4240 (39R861 [30]) and RH 4241 (V517 [12]) were used as controls. Plasmid profiles were named with lowercase letters starting at a.

Antimicrobial susceptibility testing.

The antimicrobial susceptibilities of the strains were determined by the agar diffusion method on semisynthetic Iso-Sensitest medium with the following antimicrobial agents (4): ampicillin, chloramphenicol, ceftriaxone, imipenem, mecillinam, nalidixic acid, neomycin, sulfonamide, tetracycline, trimethoprim, streptomycin, and ciprofloxacin (Neo-Sensitabs; A/S Rosco, Taastrup, Denmark). Inhibition zones were interpreted according to the recommendations of the Swedish Reference Group for Antibiotics (24).

RESULTS

Salmonella Enteritidis PT1.

Seven banding patterns (subtypes 1A to 1G) were observed among the 57 PT1 strains studied by PFGE (Table 1 and Fig. 3) when chromosomal DNAs of the strains were digested with restriction enzyme XbaI. Different test conditions showed that subtypes 1B, 1E, and 1F had fragments at 676 and 661 kb, although only one of these was visible in the fragment pattern (Fig. 3). All eight outbreak strains from Turku and Vieremä, in addition to three nonoutbreak strains belonged to PFGE subtype 1F (Table 1). Among the nonoutbreak strains, the most common PFGE subtypes were 1A (20 strains) and 1B (22 strains), whereas subtypes 1C, 1D, and 1E each included only one strain.

FIG. 3.

PFGE banding patterns of different subtypes (1A to 1G) of Salmonella Enteritidis PT1 isolates obtained when chromosomal DNA was digested with restriction enzyme XbaI. Lanes St, lambda ladder used as a molecular size marker. Arrows indicate the positions of the marker DNA fragments. Lane F, outbreak strains from Turku and Vieremä.

Six strains—one each from Turku and Vieremä, one from chicken liver, two from the Baltic countries, and one from Finnish cattle—were also digested with restriction enzymes SpeI and NotI. All six SpeI-digested strains had indistinguishable PFGE profiles. When digested with NotI, strains from Turku, Vieremä, and chicken liver had similar PFGE profiles, but they were different from those of the two strains from the Baltic countries and the single strain from Finnish cattle (data not shown).

Seven plasmid profiles (a to g) were identified by plasmid analysis (Table 1 and Fig. 4). Profile a had one plasmid; b had four plasmids; c had three; d, e, and f each had two; and g was plasmid free (Fig. 4). Forty-seven of the 57 strains studied, including all 8 outbreak strains, had plasmid profile a, regardless of the origin (human or nonhuman, domestic or foreign) (Table 1).

FIG. 4.

Plasmid profiles (a to g) of Salmonella Enteritidis PT1 isolates. Lane St1, Escherichia coli RH 4240 (39R861) (control). Lane St2, E. coli RH 4241 (V517) (control). The arrow indicates the position of the chromosomal DNA fragment.

When the PFGE and plasmid types were compiled, all outbreak strains belonged to subtype 1Fa. Most of the strains associated with Eastern Europe belonged to subtype 1Aa, and those associated with Western Europe belonged to subtype 1Ba. Any other clear categories based on the origin being either domestic or foreign could not be made. Two strains isolated from animals belonged to subtype 1Ba, and one belonged to subtype 1Eg (Table 1).

Twelve subtypes (Aa, Ac, Ad, Ba, Be, Bf, Ca, Dg, Eg, Fa, Fb, and Gd) were observed among the 57 PT1 strains when the results of PFGE and plasmid analyses were compiled. Four nonoutbreak strains were resistant to some antibiotics, but all other strains were sensitive to all antimicrobial agents tested (Table 1).

Salmonella Enteritidis PT4.

Seven banding patterns (subtypes 4A to 4G) were represented among the 43 PT4 strains studied by PFGE after XbaI restriction of genomic DNAs (Table 3 and Fig. 5). With different test conditions, it could be seen that subtypes 4A, 4B, 4C, 4F, and 4G had fragments at 646 and 661 kb, although only one of these was visible in the fragment pattern (Fig. 5). Of the 43 strains studied, 33, including the outbreak strains from Leppävirta, belonged to PFGE subtype 4A (Table 3).

FIG. 5.

PFGE banding patterns of different subtypes (4A to 4G) of Salmonella Enteritidis PT4 isolates obtained when chromosomal DNA was digested with restriction enzyme XbaI. Lanes St, lambda ladder used as a molecular size marker. Arrows indicate the positions of the marker DNA fragments. Lane A, outbreak strains from Leppävirta; lane B, outbreak strain from Kotka; lane E, outbreak strain from Rauma.

The outbreak strains (from Leppävirta, Rauma, and Kotka) and three other strains were digested with restriction enzymes SpeI and NotI (data not shown). When digested with SpeI, only the outbreak strain from Kotka differed from the other strains in its PFGE profile. Similarly, the PFGE profile of the outbreak strain from Rauma was the only one that differed from those of the other strains when digested with NotI.

Seven plasmid profiles (a to g) were identified by plasmid analysis (Table 3 and Fig. 6). Profile a had four plasmids; b had one; c, d, and e each had two; f had three; and g was plasmid free (Fig. 6). The most common profiles were b and g; 25 of the 43 strains had the former profile, and 9 had the latter profile (Table 3).

FIG. 6.

Plasmid profiles (a to g) of Salmonella Enteritidis PT4 isolates. Lane St1, E. coli RH 4240 (39R861) (control). Lane St2, E. coli RH 4241 (V517) (control). The arrow indicates the position of the chromosomal DNA fragment.

The six outbreak strains were divided into three subtypes: 4Aa (Leppävirta in 1993), 4Bg (Kotka in 1995), and 4Eb (Rauma in 1995) (Table 3). The subtype distribution among the nonoutbreak strains differed from these. However, subtype 4Bg, which was the cause of the outbreak in Kotka in 1995, was isolated from a patient who had returned from Spain in 1995; this subtype was also isolated from a Finnish cow in 1997 (Table 3).

When the results of PFGE and plasmid analyses were compiled, 12 subtypes (Aa, Ab, Ac, Ad, Ae, Ag, Bg, Cb, Db, Eb, Fc, and Gf) were observed among the 43 PT4 strains. Four nonoutbreak strains were resistant to some antibiotics, whereas all other strains were sensitive to all antimicrobial agents tested (Table 3).

DISCUSSION

A large set of Salmonella Enteritidis PT1 and PT4 isolates from different sources was analyzed by PFGE, plasmid profiling, and antimicrobial resistance testing to gain knowledge of the distribution of their potential subtypes in Finland. It was especially hoped that these analyses would help in determining the sources of the several outbreaks caused by PT1 and PT4 isolates in the 1990s.

Finland has traditionally had very strict salmonella control for livestock, a fact which has made it easier to keep indigenously acquired salmonella infections rare, unlike in many other European countries. The European Union allowed Finland to continue with this control system even after it joined the Union. Most of the salmonella infections diagnosed in Finland have been related to a trip abroad (21, 22). In the mid-1980s, infections caused by Salmonella Enteritidis began to increase in parallel with the increasing number of Finns vacationing abroad (Fig. 1). In the 1990s, Salmonella Enteritidis caused 15 outbreaks in Finland; 12 of them were caused by PT1 and PT4. However, it remained unclear from where these outbreaks originated; it was possible that the infections were connected to imported foodstuffs, that they were secondary infections caused by strains of foreign origin, or that there were some hidden reservoirs in Finnish production animals. Thus, it was necessary to try to characterize the PT1 and PT4 isolates further.

Salmonella Enteritidis PT1 was the cause of eight outbreaks in the area around Turku over a period of 5 years. In 1995, the infection source was finally found to be a poultry farm near Turku. This was the first time that Salmonella Enteritidis PT1 was detected in a Finnish commercial layer flock (10). As a result, the poultry farm was closed down, but the original source of the PT1 strain was never discovered. There were some suspicions that this strain was somehow imported from the Baltic countries, where Salmonella Enteritidis PT1 strains have been the most common Salmonella strains found in eggs (7).

The results of this study confirmed the earlier findings (10) that the Salmonella Enteritidis PT1 1Fa strain that had caused the outbreaks in Turku indeed had originated from the nearby poultry farm. It was interesting that the strains isolated in Estonia and those isolated from the Finnish tourists who had visited the Baltic countries were of type 1Aa, thus differing from the 1Fa outbreak strain and suggesting that the original source was not the Baltic countries. The 1Fa strain was also found to be the cause of the outbreak in Vieremä, which is located about 500 km northeast of Turku. Unfortunately, no connection between the outbreaks caused by identical strains in Vieremä and Turku could be found either at the time of the outbreaks or later on. Subtype 1Fa seemed to be very rare in Finland: only 1 of the 48 nonoutbreak strains belonged to this subtype, which differed from the other nonoutbreak strain subtypes by four to nine chromosomal fragments.

Among the nonoutbreak PT1 strains, subtype 1Aa seemed to be associated with Eastern Europe, and subtype 1Ba seemed to be associated with Western Europe. Type 1Ba also included two isolates from Finnish cattle. This was an important finding because there has been no identified reservoir for Salmonella Enteritidis PT1 in Finland since the commercial layer flock outbreaks in 1991 to 1995.

Salmonella Enteritidis PT4 has caused three outbreaks in different parts of Finland in the 1990s: the first was in Leppävirta in April 1993, the second was in Rauma in March 1995, and the third was in Kotka in December 1995. No connections were found between these three outbreaks at the time, and their sources were never established.

Salmonella Enteritidis PT4 is the most common phage type that contaminates eggs in Central Europe (20) and in England (9). In Finland, however, Salmonella Enteritidis PT4 has been found very rarely in cattle, less than 10 times in turkeys, and never in eggs. Imported meat has occasionally tested positive for Salmonella Enteritidis PT4.

Most of the PT4 strains belonged to PFGE subtype 4A. However, plasmid profiling was able to distinguish six profiles within this PFGE subtype. Among the 33 subtype 4A strains, only the outbreak strain in Leppävirta belonged to subtype 4Aa. Subtype 4Eb, which caused the outbreak in Rauma, differed from subtype 4Aa by three PFGE and plasmid fragments and was not found in other strains investigated here. The outbreak strain isolated in Kotka belonged to subtype 4Bg, which differed from subtype 4Ab only by one plasmid (about 45 to 60 kb), which was about the same size as the fragment that distinguished subtypes 4B and 4A in PFGE. Thus, strains of subtypes 4Bg and 4Ab could be of the same origin if these strains had lost or acquired a plasmid of this size. This finding suggests that subtypes 4Bg and 4Ab isolated from Finnish cows potentially have a reservoir in Finnish cattle and that this reservoir might have been the source of the outbreak in Kotka if these strains had lost or acquired a plasmid. These suggestions would also explain the 1997 and 1998 human isolates that belonged to subtype 4Ab. In addition, 4Bg and 4Eb differed by four PFGE fragments, whereas 4Bg and 4Aa differed by only one PFGE fragment. This one fragment was also smaller than 125 kb, a result which could indicate interference from an unstable fragment of plasmid DNA (14). However, these two subtypes, 4Bg and 4Aa, differed by four plasmids. Considering all of these results, it seems that the outbreaks in Leppävirta, Rauma, and Kotka had different origins.

This study showed that both PFGE and plasmid profiling were capable of differentiating between Salmonella Enteritidis PT1 and PT4 isolates. Some of the isolates with different PFGE subtypes had a difference of only one or two fragments in their PFGE banding patterns when digested with XbaI. Recent point mutations within a subtype might account for these minor differences (25), although the study of Murase et al. (13) concluded that the PFGE profiles of strains digested with XbaI were not affected by point mutations. Salmonellae can spontaneously lose or acquire plasmids (3), a fact which limits the use of plasmid analysis for epidemiological investigations. In this study, however, if there was a difference of several fragments in the plasmid profiles of two strains, it was considered to support the results of PFGE showing fragment patterns that differed by only a few fragments. Additional tests were done with two other restriction enzymes, SpeI and NotI. The discriminatory power of these two enzymes was, however, lower than that of XbaI.

It is also worth mentioning that without phage typing, PFGE would not have been effective in distinguishing between PT1 and PT4, since some strains showed almost identical PFGE banding patterns (1F and 4G; 1B and 4A), regardless of their phage type. In addition, antimicrobial susceptibility testing was not useful because most of the strains were sensitive to all antimicrobial agents tested.

With this collection of isolates, the genotypic methods used here showed that all PT1 isolates associated with the outbreaks were identical to each other and differed from those associated with the Baltic countries and Russia, where this phage type predominates (7). It was also established that the outbreaks caused by PT4 all had different origins. Furthermore, valuable information for future investigations was gained on the distributions of the molecular subtypes of strains that originated from the tourist resorts that are popular among Finns and of the subtypes of strains isolated from domestic production animals.