Animal Reservoirs of Y. enterocolitica and
Y. pseudotuberculosis
Animals have long been suspected of being reser-voirs for pathogenic Y. enterocolitica and Y. pseudo-tuberculosis and, hence, sources of human infections.
These bacteria have been recovered from diverse ani-mal sources, ranging from farm aniani-mals, domestic pets, and experimental animals to wild and captive animals (Table 4). Pathogenic Y. enterocolitica bacte-ria have frequently been isolated only from slaugh-tered pigs. However, the seroprevalence of anti-Yop antibodies has been shown to be high in cattle (76%) and goats (70%) in Germany (Tomaso et al., 2006).
The highest prevalence of pathogenic Y. enterocolit-ica has been observed with pig tonsils, with serotype O:3 being the most common (Fredriksson-Ahomaa and Korkeala, 2003a). This serotype has a worldwide distribution in the pig population and also has been shown to be the most common serotype in the
United States (Bhaduri and Wesley, 2006). In the United Kingdom, serotypes O:5,27 and O:9 are com-mon in pigs, and serotype O:5,27 in cattle and sheep (McNally et al., 2004). Y. enterocolitica O:9 has fre-quently been isolated from stools of cattle and goats in France (Gourdon et al., 1999). Serotypes O:3 and O:8 have sporadically been isolated from captive monkeys (Iwata et al., 2005; Fredriksson-Ahomaa et al., 2007b), and serotypes O:3 and O:5,27 from dogs and cats (Fukushima et al., 1985; Fredriksson-Ahomaa et al., 2001c). In Japan, serotypes O:3, O:8, and O:9 have been isolated from wild rodents, espe-cially from fi eld mice (Hayashidani et al., 1995).
Several studies have been conducted to isolate Y. pseudotuberculosis from farm animals, pets, and wild animals (Table 5). The host range is broad, but the principal reservoir hosts are believed to be rodents, wild birds, and domestic animals (especially pigs and ruminants). Most animals are healthy carriers, but they may become ill and excrete the bacteria under stress, such as cold and humid weather or starvation.
Jerret et al. (1990) reported that Y. pseudotuberculo-sis is one of the most common infectious causes of death among farmed deer in Australia. Several out-breaks of Y. pseudotuberculosis infection among captive monkeys have been reported in Japan (Kageyama et al., 2002).
Some studies have been conducted to identify the primary reservoir of Y. enterocolitica and Y. pseudo-tuberculosis at the farm level for pigs (Gürtler et al., 2005; Bowman et al., 2007; Nesbakken et al., 2007;
Niskanen et al., 2008). In all studies, there was a trend of increasing prevalence as pigs mature. The highest prevalence was observed among fattening pigs. In contrast, the lowest prevalence was found mostly among suckling pigs, weaned pigs, and sows;
Figure 2. Distribution of different virulence determinants among high-, low-, and nonpathogenic Yersinia enterocolitica strains belonging to the most common bioserotypes.
however, Bowman et al. (2007) found 9% of the gestating sows positive but none of the farrowing sows. Nesbakken et al. (2007) have shown that pig-lets under 2 months of age are Y. enterocolitica nega-tive in both the tonsils and feces, and no antibodies against Y. enterocolitica can be measured. The young pigs became carriers in the tonsils and feces when they were about 60 to 80 days old and seropositive shortly thereafter. The tonsils remained positive for Y. enterocolitica during the last weeks of the lives of fattening pigs, which explains the high prevalence of this pathogen in the tonsils of slaughter pigs.
PCR has been applied in only a few studies inves-tigating the prevalence of pathogenic Y. enterocolitica in pigs (Fredriksson-Ahomaa et al., 2000a, 2007a;
Boyapalle et al., 2001; Korte et al., 2004; Bhaduri et al., 2005). In these studies, the detection rate was shown to be clearly higher with PCR than with cul-ture methods. Most (88%) of the pig tonsils were shown to be positive with PCR when pathogenic Y. enterocolitica could be isolated from only 35% of the tonsils (Fredriksson-Ahomaa et al., 2007a). Bhaduri et al. (2005) detected pathogenic Y. enterocolitica in 12% of fecal samples in the United States using PCR
compared to 4% using a culture method. Boyapalle et al. (2001) found 40% of the mesenteric lymph nodes of pigs Y. enterocolitica positive using PCR but none with culturing.
The Environment Contaminated with Y. enterocolitica and Y. pseudotuberculosis Y. enterocolitica strains belonging to pathogenic bioserotypes have rarely been isolated from the envi-ronment. However, strains of bioserotype 4/O:3 have occasionally been isolated from the environment in slaughterhouses and butcher shops (Nesbakken, 1988;
Fredriksson-Ahomaa et al., 2000b, 2004). Pathogenic Y. enterocolitica has been detected by PCR on a vari-ety of environmental sources in a Finnish slaughter-house (Fredriksson-Ahomaa et al., 2000b). This pathogen was detected on the brisket saw, the hook from which the pluck set (tongue, esophagus, trachea, lungs, heart, diaphragm, liver, and kidneys) hangs, evisceration knife, aprons used by trimming workers, the computer keyboard used in the meat inspection area, and the handle of the coffee maker used by slaughterhouse workers. Additionally, pathogenic Y. enterocolitica was isolated from the air in the
Table 4. Animal reservoirs for Yersinia enterocolitica and Yersinia pseudotuberculosis
Reservoir type Animal source Y. enterocolitica Y. pseudotuberculosis
Farm animals Pig O:3, O:5,27, O:8, O:9 O:1–O:4
Cattle, sheep, goat O:5,27, O:9 O:1–O:3
Pets Dog, cat O:3, O:5,27 O:1–O:5
Chinchilla O:1,2,3 O:3
Wild animals Bird O:1, O:2
Deer O:1–O:3
Hare, rabbit O:2,3, O:8 O.1–O:5
Rodent O:3, O:8, O:9 O:1–O:5
Zoo animals Monkey O:3, O:8 O:1–O:3
Table 5. Prevalence of Yersinia pseudotuberculosis in animal sources
Source No. of samples No. of positive samples % Positive Country Reference
Pig 1,200 52 4 Japan Fukushima et al. (1989)
210 8 4 Finland Niskanen et al. (2002)
Cattle 2,639 185 7 Australia Slee et al. (1988)
Sheep 449 21 5 Australia Slee and Skilbeck (1992)
Dog 252 16 6 Japan Fukushima et al. (1985)
Cat 318 4 1 Japan Fukushima et al. (1985)
Bird 259 2 1 Japan Fukushima and Gomyoda (1991)
468 3 1 Sweden Niskanen et al. (2003)
Deer 153 29 18 Australia Jerret et al. (1990)
215 8 4 Japan Fukushima and Gomyoda (1991)
Hare 139 2 1 Japan Fukushima and Gomyoda (1991)
Rabbit 148 4 3 China Zheng et al. (1995)
Mouse 107 9 8 China Zheng et al. (1995)
Rat 148 4 3 China Zheng et al. (1995)
237 4 2 Japan Kageyama et al. (2002)
bleeding area. PCR-positive samples were also obtained from the fl oor in the eviscerating and weighing areas and from the table in the meat-cutting area. Pathogenic Y. enterocolitica 4/O:3 has also been isolated occasionally from water (Thompson and Gravel, 1986). In Brazil, natural water and sewage have shown to be sporadically contaminated with pathogenic Y. enterocolitica belonging to serotype O:5,27 (Falcao et al., 2004). In Japan, Y. enterocolit-ica O:8 strains have been isolated from stream water (Iwata et al., 2005). Sandery et al. (1996) have shown with PCR that pathogenic strains of Y. enterocolitica can frequently be detected in environmental waters. In a case-control study, untreated drinking water has been reported to be a risk factor for sporadic Y. entero-colitica infections in Norway (Ostroff et al., 1994).
Y. pseudotuberculosis is widely spread in the environment (soil, water, vegetables, etc.), where it can survive for a long time. The environment can be contaminated by the feces of infected animals, mainly wild animals, like deer, rodents, and birds (Fukushima et al., 1998). Y. pseudotuberculosis has been isolated from fresh water, such as river, well, and mountain stream water, at a considerably high rate (Tsubokura et al., 1989). In a large point source outbreak caused by raw carrots contaminated by Y. pseudotuberculo-sis, the epidemic strain was also found on the soil and on the production line (washing and peeling equip-ment) at the farm of origin (Jalava et al., 2006).
Foods Contaminated with Y. enterocolitica and Y. pseudotuberculosis
Food has often been suggested to be the main source of Y. enterocolitica and Y. pseudotuberculosis, although pathogenic isolates have seldom been recov-ered from food samples. In epidemiological studies, Y. enterocolitica O:3 infections have been associated with eating raw or undercooked pork within 2 weeks before onset (Tauxe et al., 1987; Ostroff et al., 1994).
In the United States, Y. enterocolitica O:3 infections have been associated with the household preparation of chitterlings (intestines of pigs, which are a tradi-tional dish in the southern United States) (Lee et al., 1990; Jones et al., 2003). Raw pork products have been investigated widely due to the association between Y. enterocolitica and pigs. The only patho-genic bioserotype found in northern Europe and Germany is 4/O:3 (Johannessen et al., 2000;
Fredriksson-Ahomaa et al., 2001a; Fredriksson- Ahomaa and Korkeala, 2003b; Thisted Lambertz and Danielsson-Tham, 2005). Bioserotypes 2/O:5,27 and 2/O:9 have sporadically been found in the United Kingdom, and 3/O:3 and 3/O:5,27 in Japan in pork (Logue et al., 1996; Fukushima et al., 1997).
Y. pseudotuberculosis has very rarely been isolated from foods. This pathogen has sporadically been iso-lated from pork in Japan (Tsubokura et al., 1989;
Fukushima et al., 1997). Y. pseudotuberculosis has been isolated from iceberg lettuce and raw carrots implicated in food-borne outbreaks in Finland (Nuorti et al., 2004; Jalava et al., 2006). The detection rate of pathogenic Y. enterocolitica in foods has been shown to be clearly higher by PCR than by culturing (Fredriksson-Ahomaa and Korkeala, 2003a). The highest detection rate has been obtained from pig offal, including pig tongues, livers, hearts, and kid-neys, and from chitterlings (Table 6). Lambertz et al.
(2007) reported recently a relatively high prevalence (11%) of pathogenic Y. enterocolitica in fermented pork sausages by use of PCR. Vishnubhatla et al.
(2001) detected a high prevalence of yst-positive Y. enterocolitica in ground beef; however, pathogenic strains have only sporadically been isolated from beef. No pathogenic Y. enterocolitica has been detected in fi sh and chicken, but this pathogen has been detected in 3% of lettuce samples with PCR (Fredriksson-Ahomaa and Korkeala, 2003b). In Korea, Lee et al. (2004) isolated one ail-positive Y. enterocolitica strain of bioserotype 3/O:3 from 673 ready-to-eat vegetables.
Transmission of Y. enterocolitica and Y. pseudotuberculosis to Humans
The primary transmission route of human yer-siniosis is proposed to be fecal-oral via contaminated food (Figure 3). In particular, pork and pork prod-ucts have been implicated as the major source of human Y. enterocolitica infection, with some epide-miological studies linking consumption of uncooked or undercooked pork (Tauxe et al., 1987; Ostroff et al., 1994; Satterthwaite et al., 1999; Fredriksson-Ahomaa et al., 2001b, 2006b). Consumption of raw pork may play an important role in countries like Belgium, Germany, and The Netherlands, where raw minced pork with pepper and onion is a delicacy that can be purchased in ready-to-eat form from butcher shops. Transmission is likely to occur more often via cross-contamination of cooked pork or foods not normally harboring Y. enterocolitica.
Another transmission route may be from person to person, which can be direct or indirect. Direct person-to-person contact has not been demonstrated, but Lee et al. (1990) reported Y. enterocolitica O:3 infections in infants who were probably exposed to infection by their caretakers. This may happen when basic hygiene and hand-washing habits are inade-quate. Indirect person-to-person transmission has apparently occurred in several instances by transfusion
Y. pseudotuberculosis is also a food-borne patho-gen, but this pathogen has rarely been isolated from foods implicated in the illness. In the reported out-breaks, fresh produce and untreated surface water have been possible infection sources. Some recent epi-demiologic investigations in Finland have linked outbreaks of Y. pseudotuberculosis to domestically grown iceberg lettuce and carrots (Nuorti et al., 2004;
Jalava et al., 2006). Y. pseudotuberculosis occurs in water and in the environment and has been isolated from various animals; however, the transmission routes are unclear. In the outbreak linked to carrots, washing and peeling equipment was shown to be contaminated with the outbreak strain; however, the of contaminated blood products (Bottone, 1999). In
these cases, the most likely source of Yersinia has been blood donors with subclinical bacteremia. Direct contact with pigs, a common risk for pig farmers and slaughterhouse workers, may also be a transmission route. Elevated serum antibody concentrations have been found among people involved in swine breeding or pork production (Seuri and Granfors, 1992). Pet animals have also been suspected as being sources of human yersiniosis through close contact with humans, especially young children. Pathogenic Y. enterocolit-ica may be transmitted to humans indirectly from pork and offal via dogs and cats (Fredriksson-Ahomaa et al., 2001c).
Table 6. Prevalence of pathogenic Yersinia enterocolitica in foods with PCR and culture methods
Sample No. of samples No. of culture-positive samples (%)a
No. of PCR-positive
samples (%) Reference
Pig tongues 15 7 (47) 10 (67) Vishnubhatla et al. (2001)
99 79 (80) 82 (83) Fredriksson-Ahomaa and Korkeala (2003b)
Pig offalb 110 38 (35) 77 (70) Fredriksson-Ahomaa and Korkeala (2003b)
Chitterling 350 8 (2) 278 (79) Boyapalle et al. (2001)
Ground pork 350 0 133 (38) Boyapalle et al. (2001)
100 32 (32) 47 (47) Vishnubhatla et al. (2001)
Ground beef 100 23 (23) 31 (31) Vishnubhatla et al. (2001)
Minced pork 255 4 (2) 63 (25) Fredriksson-Ahomaa and Korkeala (2003b)
100 5 (5) 35 (35) Lambertz et al. (2007)
Porkc 300 6 (2) 50 (17) Johannessen et al. (2000)
91 6 (7) 9 (10) Thisted Lambertz and Danielsson-Tham (2005)
150 0 9 (6) Fredriksson-Ahomaa et al. (2007a)
97 0 11 (11) Lambertz et al. (2007)
Chicken 43 0 0 Fredriksson-Ahomaa and Korkeala (2003b)
Fish 200 0 0 Fredriksson-Ahomaa and Korkeala (2003b)
Lettuce 101 0 3 (3) Fredriksson-Ahomaa and Korkeala (2003b)
Tofu 50 0 6 (12) Vishnubhatla et al. (2001)
a Pathogenicity of isolates confi rmed.
b Liver, heart, and kidney.
c Except pig tongues and offal.
Figure 3. Transmission of Yersinia enterocolitica and Yersinia pseudotuberculosis to humans.
exact mechanism of contamination of carrots on the farm could not be clarifi ed (Jalava et al., 2006).
A combination of direct contact with wildlife feces during storage and cross-contamination of the equip-ment was the most likely contributing factor. In another outbreak, iceberg lettuce was probably con-taminated with irrigation water. Y. pseudotuberculo-sis could have been transmitted from infected animals to the lettuce and irrigation water by feces at the farm before distribution (Nuorti et al., 2004). Unchlori-nated drinking water from wells, springs, and streams contaminated with feces of wild animals is consid-ered an important transmission route in mountainous areas in Japan (Fukushima et al., 1998). In Korea, untreated mountain spring water could be linked to Y. pseudotuberculosis infection (Han et al., 2003).
An epidemiological study has shown evidence that Y. pseudotuberculosis was transmitted from infected pets to humans in a familial outbreak in Japan (Fukushima et al., 1994).
INTRINSIC AND EXTRINSIC FACTORS THAT AFFECT SURVIVAL AND GROWTH IN FOOD
PRODUCTS AND CONTRIBUTE TO OUTBREAKS
Intrinsic Factors
The most important intrinsic factors are nutri-tion, pH, and water activity. Nutritionally, Yersinia is not a fastidious organism. This bacterium is also able to multiply over a wide pH range from approximately pH 4 to 10, with an optimum pH of around 7.6.
However, as only few foods have an alkaline pH, the tolerance of a high pH is relatively unimportant.
Yersinia is generally considered to be sensitive to low pH conditions. Citric acid is less inhibitory than ace-tic acid. With citric acid, growth was detected even at a pH of 3.81 obtained with a concentration of 0.31%
(vol/vol), while acetic acid inhibited the growth at pH 5.58 obtained with a concentration greater than 0.16% (vol/vol) (Karapinar and Gonul, 1992).
Salt tolerance of Yersinia is moderate and strongly dependent on storage temperature. While 7% or more of NaCl is bacteriostatic (Robins-Browne, 1997) for Yersinia, the bacterium is able to grow in 5% NaCl. Thus, at the levels most commonly present in foods, salt alone will not prevent growth, and other preservative hurdles are required.
Y. enterocolitica can tolerate both sodium nitrate and nitrite of up to 20 mg/ml for 48 h in vitro (De Giusti and De Vito, 1992). However, a nitrite concentration of only 80 mg/kg has been reported to inhibit the growth of Y. enterocolitica in fermented sausages (Asplund et al., 1993).
Extrinsic Factors
The most important extrinsic factors include temperature and gas atmosphere. Yersinia is unusual among pathogenic enterobacteria in being psy-chrotrophic, thus having the ability to grow at refrig-erator temperatures. The doubling time at the optimum growth temperature (approximately 28 to 30°C) is around 34 min (Schiemann, 1989). Although Yersinia can grow at temperatures as low as 0°C, the organism grows much more slowly as temperatures drop below 5°C (Goverde et al., 1994; Harrison et al., 2000). It has been shown that the number of Y. enterocolitica on pork can reach 109 CFU per cm2 after 5 days at 10°C (Nissen et al., 2001). In some Finnish Y. pseudotuberculosis outbreaks detected in spring, the vehicles have been vegetables of the crop stored at chilled temperature from a previous year.
Yersinia can withstand freezing and can survive in frozen foods for extended periods, even after repeated freezing and thawing (Toora et al., 1992).
At the cellular level, the acute cold shock response, induced by any temperature downshift of 10°C or more, and longer-term adaptation to low temperature involve different mechanisms (Bresolin et al., 2006). As in most bacteria, the cold shock response in yersiniae is featured by the expression of cold shock proteins (Csp). Csp bind single-stranded nucleic acids and might thus play a role as RNA chaperones and transcription antiterminators (Jones and Inouye, 1994; Bae et al., 2000). Csp also bind mRNA and regulate ribosomal translation, mRNA decay, and termination of transcription (Ermolenko and Makhatadze, 2002; Chattopadhyay, 2006). At the acclimation phase, the production of Csp declines.
The early and mid-exponential growth of Y. entero-colitica at low temperature involves the activation of genes encoding environmental sensors and regula-tors related to signal transduction (Bresolin et al., 2006), catabolic events, utilization of endogenous resources, and defense against oxidative stress. Also, fl agellar synthesis and chemotaxis (Bresolin et al., 2006) are induced, a well-known phenomenon in many pathogens exposed to temperatures outside of the mammalian host. The late exponential and early stationary phase of adaptation of Y. enterocolitica to low temperature is dominated by mechanisms involved in biodegradative metabolism (Bresolin et al., 2006). Membrane rigidity and fl uidity are maintained by changing the membrane composition.
At low temperature, unsaturated fatty acids and phospholipids predominate, while saturated and cyclic fatty acid contents become a majority upon a rising temperature (Nagamachi et al., 1991; Goverde et al., 1998).
the widespread occurrence of pathogenic Yersinia in herds renders the control of this bacterium at farm level diffi cult. Furthermore, it is impossible to reject asymptomatic pigs contaminated with pathogenic Yersinia at postmortem meat inspection. There are several meat inspection procedures, like evisceration and incision of the submaxillary lymph nodes, in which contamination of carcass and offal can occur.
During evisceration, the spread of pathogenic Yer-sinia from tonsils to pluck set and other parts of the carcass by knives and hands is unavoidable because the tonsils are usually removed together with the pluck set (tongue, esophagus, trachea, lungs, heart, diaphragm, liver, and kidneys) (Fredriksson-Ahomaa et al., 2000b, 2001a). By removing the head together with the tonsils and tongue prior to evisceration, the contamination of pathogenic Yersinia could probably be reduced. Additionally, the use of a mechanized bung cutter in connection with enclosing the anus and rectum in a plastic bag to minimize fecal contamina-tion has been shown to reduce feces contaminacontamina-tion of carcasses (Nesbakken et al., 1994). The inspection of head, tonsils, and tongue should occur in a separate room with separate equipment. Furthermore, the head meat and tongue should be pasteurized before delivery. However, strict slaughter hygiene remains important in reducing contamination in slaughter-houses. Yersinia contamination in the later stages of the food chain has not been studied suffi ciently.
Fresh produce may become contaminated with pathogenic Yersinia, especially with Y. pseudotuber-culosis, during irrigation, harvesting, packing, ship-ping, and processing. Preventing the access of wild animals to irrigation water and fi elds could reduce the risk of contamination of surface water and soil.
Furthermore, the access by rodents and birds to stor-age facilities should be prevented, and the cleaning of processing equipment should be adequate (Nuorti et al., 2004; Jalava et al., 2006).
DISCRIMINATIVE DETECTION METHODS FOR CONFIRMATION AND TRACE-BACK OF
CONTAMINATED PRODUCTS
Several diffi culties have been associated with isolating pathogenic Yersinia from food and environ-mental samples. Conventional culture-dependent methods have several limitations, such as long incuba-tion steps taking up to 4 weeks, lack of identifi caincuba-tion between species, and lack of discrimination between pathogenic and nonpathogenic strains (Fredriksson-Ahomaa and Korkeala, 2003a). In addition, the ability of bacteria, including Y. enterocolitica, to per-sist in samples in a viable but noncultivable state can Bacterial virulence functions are often
tem-perature-regulated due to the high energy cost of unnecessary expression of virulence genes (Hurme and Rhen, 1998). Thus, many of the Yersinia virulence factors are expressed at 26°C but not at 37°C (Table 7).
The maximum growth temperature of Yersinia is between 42 and 44°C. Yersinia is relatively sensitive to heat, with thermal treatments employed in food processing destroying it readily. It is also destroyed by pasteurization at 72°C for 15 to 20 s. As a faculta-tively anaerobic bacterium, Yersinia can grow in anaerobic conditions. It can also grow well in modi-fi ed atmospheres (Harrison et al., 2000), but with higher levels of CO2 the length of the lag phase will increase and growth will be slower (Pin et al., 2000).
Y. enterocolitica has been shown to grow well on meat
Y. enterocolitica has been shown to grow well on meat