150 Journal o f Food Protection, Vol. 77, No. 1, 2014, Pages 150-170 doi: 10.4315/0362-028X.JFP-13-150 Copyright © , International Association for Food Protection
Review
Listeria monocytogenes Persistence in Food-Associated Environments: Epidemiology, Strain Characteristics, and Implications for Public Health V. FER REIRA ,12 M. WIEDMANN,2 P. TEIXEIRA,1 AND M. J. STASIEWICZ2* 1Centro de Biotecnologia e Quimica Fina, Laboratorio Associado, Escola Superior de Biotecnologia, Universidade Catolica Portuguesa Porto, Rua Dr. Antonio Bernardino Almeida, 4200-072 Porto, Portugal; and 2Department o f Food Science, Cornell University, Ithaca, New York 14853, USA MS 13-150: Received 10 April 2013/Accepted 11 September 2013
ABSTRACT Over the last 10 to 15 years, increasing evidence suggests that persistence of Listeria monocytogenes in food processing plants for years or even decades is an important factor in the transmission of this foodbome pathogen and the root cause of a number of human listeriosis outbreaks. L. monocytogenes persistence in other food-associated environments (e.g., farms and retail establishments) may also contribute to food contamination and transmission of the pathogen to humans. Although L. monocytogenes persistence is typically identified through isolation of a specific molecular subtype from samples collected in a given environment over time, formal (statistical) criteria for identification of persistence are undefined. Environmental factors (e.g., facilities and equipment that are difficult to clean) have been identified as key contributors to persistence; however, the mechanisms are less well understood. Although some researchers have reported that persistent strains possess specific characteristics that may facilitate persistence (e.g., biofilm formation and better adaptation to stress conditions), other researchers have not found significant differences between persistent and nonpersistent strains in the phenotypic characteristics that might facilitate persistence. This review includes a discussion of our current knowledge concerning some key issues associated with the persistence of L. monocytogenes, with special focus on (i) persistence in food processing plants and other food-associated environments, (ii) persistence in the general environment, (iii) phenotypic and genetic characteristics of persistent strains, (iv) niches, and (v) public health and economic implications of persistence. Although the available data clearly indicate that L. monocytogenes persistence at various stages of the food chain contributes to contamination of finished products, continued efforts to quantitatively integrate data on L. monocytogenes persistence (e.g., meta-analysis or quantitative microbial risk assessment) will be needed to advance our understanding of persistence of this pathogen and its economic and public health impacts.
Listeria monocytogenes is a gram-positive, facultative intracellular foodbome pathogen that can cause severe invasive illness (listeriosis) in humans and other animal species, including mammals (e.g., ruminants) and birds (155). Although invasive human listeriosis can occur in healthy individuals, the vast majority of cases occur in young, elderly, or immunocompromised individuals and manifest as septicemia, meningitis, or other infections of the central nervous system. In pregnant women, infections may lead to spontaneous abortion, still birth, or fetal death (177). Scallan et al. (154) estimated that this foodbome pathogen causes approximately 1,460 hospitalizations each year in the United States, resulting in 260 deaths. In developed countries worldwide, the incidence of listeriosis is 0.36 to 5 cases annually per million people (42, 43, 154); however, the number of reported cases can be very low in countries with limited surveillance for this disease. Al though some authors have suggested that the majority of * Author for correspondence. Tel: 607-255-1266; Fax: 607-254-4868; E-mail: [email protected].
human listeriosis cases are sporadic, findings from some studies that have included cluster analysis of human cases based on molecular subtyping data suggest that more human listeriosis cases than previously assumed may represent outbreaks (28, 147), supporting use of improved molecular subtyping methods for improved outbreak detection. In the first decade of the 21st century, coinciding with enhanced implementation of foodbome pathogen subtyping proce dures, a large number of human listeriosis outbreaks have been reported in various countries, including the United States (23, 58, 165), Canada (76, 136), Chile (77, 78), Germany and Austria (52), the United Kingdom (37), Sweden (20), Czech Republic (178), Japan (104), and Australia (128). Although L. monocytogenes is a facultative intracellular pathogen usually of homeothermic animals, it also can grow and survive outside a host at a wide pH range (4.7 to 9.2) (131, 132), at high salt concentrations (10%, wt/vol) (109), and most importantly at refrigeration temperatures (—0.5 to 9.3°C) (180). This growth range allows this pathogen to subsist in the food processing plant environment, survive
J. Food Prot., Vol. 77, No. 1
various food processing hurdles, and proliferate in food products. L. monocytogenes has been isolated from a variety of raw and processed food products, including milk and dairy products, meat products, egg products, seafood, vegetables, and other ready-to-eat (RTE) foods (44). In addition to food processing plant environments, this pathogen has been isolated from other environments, including natural, urban, and farm environments (66, 117, 149, 152), and types of samples, including soil, decaying vegetation, stream water, sewage, and human and animal feces (46, 125, 181, 182). Its frequent presence in different environments and its unusual ability to adapt to and survive under stressful conditions make control of L. monocyto genes in the food processing environment a considerable challenge. L. monocytogenes also has been isolated from environmental samples at retail establishments (148, 151) and in consumer homes (12, 31). However, our understand ing of L. monocytogenes transmission and ecology in food-associated environments other than those used for commercial food processing is limited. Results from two independent but similar risk assessments in the United States (41, 135) suggest that in >50% of human listeriosis cases associated with consumption of RTE deli meats the contamination may have occurred at the retail level. Cross contamination of retail products was recently found during a major listeriosis outbreak in Canada linked to contaminated cheese made from pasteurized milk (55). Although L. monocytogenes is inactivated by thermal treatments used for production of RTE foods, postproces sing cross-contamination from equipment and the environ ment represents a major concern (93, 95). Various approaches are used to control postprocessing contamina tion in food processing plants, including stringent imple mentation of good manufacturing practices, sanitation standard operating procedures, sterile packaging technolo gies (36), in-package postlethality treatments, and reformu lation of products with antimicrobial agents. Use of molecular subtyping approaches to better understand L. monocytogenes ecology and transmission in various foodassociated environments will continue to improve control of L. monocytogenes and prevention of cross-contamination. For example, molecular typing methods such as pulsed-field gel electrophoresis (PFGE), amplified fragment length polymorphism, and ribotyping, which have high discrimi natory power, have been essential in studies of L. monocytogenes sources and contamination routes in food processing plants (4, 56, 66,130). Studies including various molecular subtyping methods have revealed that specific L. monocytogenes subtypes can persist in food processing plants over long periods of time, whereas other subtypes are recovered only sporadically (96, 101, 113, 121). Subtypes that persisted were occasionally isolated from contaminated foods and were sometimes associated with listeriosis outbreaks. Although many field studies have revealed L. monocytogenes persistence in various food-associated environments and laboratory studies have revealed L. monocytogenes characteristics that facilitate persistence, our understanding of L. monocytogenes persistence and its impact on food safety and public health is still fragmented.
PERSISTENCE OF L. MONOCYTOGENES
151
This review includes a discussion of our current knowledge on some key issues regarding L. monocytogenes persistence: (i) persistence in food processing plants and other foodassociated environments (farm, retail, and consumers home environments), (ii) persistence in the general environment (natural and urban environments), (iii) phenotypic and genetic characteristics of persistent strains, (iv) environ mental niches, and (v) public health and economic implications of L. monocytogenes persistence. PERSISTENCE OF L. MONOCYTOGENES The term “ persistence” can acquire a variety of meanings within the context of a bacterial foodborne pathogen. Persistence can describe the long-term survival of a pathogen in a human host (a persistent infection). Although persistent human or animal infections have been described for various pathogens (e.g., persistent infection of the nostrils with Staphylococcus aureus), including foodborne bacterial pathogens (e.g., persistent infection of food handlers with Salmonella) (150, 167), no clear evidence of persistent infection of humans with L. monocytogenes has been published. Persistence also is used to describe long term survival (typically without growth) of a foodborne pathogen in a food or food matrix (often in laboratory incubation experiments), e.g., meat products (57, 75), cheese (94, 115), smoked salmon (60,142), and vegetables (11). Persistence can similarly be used to describe long-term survival of a pathogen in a simple, defined matrix (e.g., soil, water, or stainless steel surfaces) or a complex natural or human-made environment (e.g., a processing plant). In this review, we use the term persistence to describe the presence over time of L. monocytogenes in complex natural or human-made environments. This persistence likely requires both growth and survival in specific compartments of the complex environment. For example, L. monocytogenes could persist in a processing plant in a specific abiotic environment (e.g., a hollow roller in a piece of equipment or an uncleanable part of a slicer) or as a long-term host population inside a plant (e.g., in rodents in a plant lacking appropriate pest control). Differences in criteria used to define persistence complicate determination of persistence in a complex environment (e.g., a food processing plant or a farm). Typically, persistence is defined by repeated isolation on different dates of L. monocytogenes strains that are subsequently identified as identical subtypes (as determined by phenotypic or genotypic methods). However, some subtyping methods have limited discriminatory power, either for L. monocytogenes as a species or for specific subtypes (e.g., serotype 4b strains). Thus, a method with limited discriminatory power may not be useful for determining persistence of the same subtype in multiple sample collections. Some L. monocytogenes subtypes (e.g., specific PFGE types) are relatively common and widely distributed, whereas others appear rare. Fugett et al. (53) used the standard Centers for Disease Control and Prevention (CDC) two-enzyme PFGE protocol for L. monocytogenes and found one PFGE type that had caused
152
FERREIRA ET AL.
outbreaks in Switzerland in 1983 through 1985 and in the United States in 1985; this type also was isolated from multiple farms and urban environments in New York state in 2001 and 2002. Isolation of a very common subtype from multiple sample collections provides weaker evidence of persistence compared with isolation of a rare subtype from multiple sample collections. The relative frequency of a given subtype can be determined with a statistical testing approach developed by Malley et al. (105) that compares the observed frequency of the subtype to the historical frequency of that subtype. Of the five instances in which a subtype was observed more than once in a food processing plant, only three were identified as showing persistence; the other two cases of repeated isolation (two observations of the given subtype) were not significantly different from isolation expected by chance. Appropriate frequency information for subtypes may be very costly to acquire. The relative frequency of particular L. monocytogenes subtypes among human clinical isolates may be available, but the same information may not be available for other source populations (e.g., turkeys and turkey farms, small ruminants, and raw sheep milk cheeses), limiting frequency comparison within understud ied environments. Differentiation between repeated reintroduction of a specific L. monocytogenes subtype into a facility and true persistence in the facility is challenging based on subtyping information alone and almost always also involves infor mation on facility set up and design. To illustrate extremes, isolation of the same subtype over multiple sampling times in a facility that uses only pathogen-free ingredients and is separated from its surrounding environments (e.g., through premises design and constmction and/or implementation of good manufacturing practices) is likely to indicate true persistence. In contrast, isolation over multiple sampling times in a facility that has few measures to prevent introduction of L. monocytogenes from ingredients or surrounding environments (e.g., a farm or a retail facility) may indicate either reintroduction or persistence. In the following examples, criteria used to define persistence differ among studies and often include only isolation of the same subtype in a given facility over at least two or three sampling times. Although these repeat isolations may indicate persistence of the pathogen, we cannot exclude the possibility that in some instances, particularly when subtyping methods with low discriminatory power were used for isolate characterization, repeat isolation of a subtype may not necessarily indicate persistence. We recommend careful evaluation of data on strain persistence and hope to stimulate future efforts to develop and use statistical methods to quantify the likelihood that a set of subtyping data tmly indicates persistence. Persistence of L. monocytogenes in food processing plants. Harvey and Gilmour (61) were among the earliest authors to suggest colonization of food production and processing environments by persistent L. monocytogenes subtypes (Table 1). Multilocus enzyme electrophoresis (MEE) and restriction fragment length polymorphism
J. Food Prot., Vol. 77, No. 1
(RFLP) of L. monocytogenes isolates obtained from raw milk and nondairy food products from various producers and manufacturers revealed the recurrent isolation of L. monocytogenes of the same subtypes. Although lacking environmental samples, the authors suggested that specific subtypes of L. monocytogenes can persist in food processing and farm environments for long periods and recontaminate raw milk. Rprvik et al. (141) first reported persistence of L. monocytogenes subtypes within a fish plant: a salmon smokehouse in Norway (Table 2). MEE of the L. monocy togenes isolates revealed that those belonging to one electrophoretic type (ET-6) were present during the entire 8-month investigation in the environment, in the fish during processing, and in vacuum-packed smoked salmon. In the same year, Lawrence and Gilmour (95) (Table 3) reported characterization of L. monocytogenes isolates from a poultry meat processing plant by random amplification of polymor phic DNA (RAPD). Two of 18 RAPD types found (A and B) persisted throughout a 6- and 5-month period, respec tively, with one type more prevalent in raw poultry products and equipment and the other more prevalent in the cooked poultry processing environment. Samples were contaminat ed by the same RAPD types up to 1 year later. Subsequently, persistence of L. monocytogenes strains for months to several years has been reported in a variety of food processing plants, including those for meat, fish, dairy, and RTE products (Tables 1 through 4). In some studies, environmental sources have been identified as the main source of postprocessing L. monocytogenes contamination of food products within processing plants (124). Persistent L. monocytogenes strains are more often isolated from food processing environments (e.g., drains and equipment), including sites close to food contact surfaces (e.g., dicing machines), rather than from raw materials (99,113,156,179). The recovery of persistent strains from the environment and equipment after cleaning and disinfection emphasizes the risk of growth and establishment of L. monocytogenes, particularly in sites difficult to access, leading to ongoing food product contamination (4, 124, 186). Microbiological testing of the processing environment and equipment is therefore neces sary to detect particular niches of L. monocytogenes and validate the efficiency of sanitation procedures. Persistent strains also can be transferred between facilities through a contaminated environment. For example, a meat dicer harboring L. monocytogenes in an internal niche transferred identical PFGE types to two subsequent processing plants that purchased the contaminated equipment (99). Other authors have emphasized the importance of cross-contamination by raw materials. Berrang et al. (10) reported that three of the four resident L. monocytogenes strains in a chicken processing plant were recovered from the raw product at some point during the 1-year study, suggesting persistent reintroduction of this strain from raw product into the plant environment. Persistence of L. monocytogenes in other foodassociated environments. Although molecular subtyping
J. Food Prot., Vol. 77, No. 1
P E R S I S T E N C E O F L. MONOCYTOGENES
£> c -3 2 8 -a .23 B. 0 S£ £ 8.
°*0
X ON ON
ON ON ON
x 2 T3 C/3 s 1/3 O ’c5—4 /3 ’£ D > o cd T3 s -c c '3 2C /3
•S o 3 VJ 0 3e 2 ^ ■£ 1 ° C ^L 0—t c/3 © O 'c/3 ~ .2 3 c o £ '> 03 CL 3 3d> 3e O 3 , 3 X •S
T3 3cd
T A B L E 1. Selected studies of the persistence ofL. m o n o c y t o g e n e s in dairy processing plants
9O 00
E £cd DC
1>5 .aio ja £ % §
3
T3 O
cd
X ~ ao * “ .o cd X
g> xcd O -n
3
h§C /3 03
O 3 d) 3 t/3 00 ON . « •s .S d > 03 • CL ^ ’ .3
•
Vl ’ jDCL xX c y >,* • X ,cd 4 c— d) ■ £o 3CL
02
M 1"O J
-o 3
3
60 3 '33 03 yo di
) < L > oo ■0-3 dX 3 1 2 3 ■o « o CL 00 Cx0 2 s ,0 -w" 3 C 5. °o •o 1 £ £C/3 dJ s o '3 3 03 d) f.2
X
w
CL
O
E 7
E
GO 3 'a >>
G O
GO
c
a .
2s
vi
X
0 X £ X •&
0
(-H >V
d> 5— < 2
0 C0fl
m
PJ pft .
o Ph CL,
CL
GO 3 *CL •w* O X C4
3 1) _o 3 OGO cd N 'C Id 0
60 3 ‘
5-
>>
2
^ T> y3
CL £
Number of persistent isolates for each subtype is given in parentheses; multiple numbers in parentheses indicate multiple persistent subtypes in a given plant. MEE, multilocus enzyme electrophoresis; PFGE, pulsed-field gel electrophoresis. NA, data not available.
On O
K
ra w onge
> x
^4
15 3
FE R R E IR A E T A L .
J. Food Prot., V o l. 77 , No. 1
O
3
On On ON
C
a
£ .3 Si s X co
o U
Cl- tfc CD o O o •g (D JS o «
1
3
X
P --1 . j-H C/3
Q I CL —
> £ cr *s X 3 o
S.J
I s
1 | Qh ^ x £ d) c -3 0 ^ °
CL O
H
X X O O
£
§
CL o £ H
■g 3 w
E
3 ctf 3 —r r ~ \ T
XL o fc 3 CL £ W c =* & ?' o
.5 5
i
s CJ
oo-
“
3
O
X ^ C O
^
^
8
3 £
5 d> a >. 73
c | 8 o OX) *c
3 X
"
oo X
OX)
U2
cl a) g c p «* Xd> p^ o 5o p .
XI
d> »-
>>
X
D 3 "g .N
OX) L) OX) 3 X c - C - C N OX) • > X > i- c 3 ■ X C X ^ 3 .. u 3 o C CO TO 1/1 TO TO
8„ T3
~ x
CJ. ^
fll* O H
o
CL
5 X CL £
3 7?
o c CL d)
(N O o (N
3 dj CL CL
2 1/3 .52 p
*o _8 ^ 3
CL ~ ^ M
3 > 3 >
o C CL d)
s P CL S O C P OX)
is c
.
OX)
‘cl
o
d) 3
O 3 X O O u.
O
j*
O
co
c H -s
£ 3
CO oo i OX) £ ^ d> — • ^ _c X > OX) d> x 3 .s - i .£ £ « 2 CO Q, oo . 3 “ ^■ 5 E .- P d) 3 X o as 3 o g p d> u cd CL T3 3 c X c OX) P O d> 3 c X o, 3c ’■^ — X ’ n 7j r J 3 « OX) P ..5i 2 o X c .£ •p P o X c3 D , c3 « 3
5’. O — w
3 X
CL g
o o < N
OX)
d> X
c d)
OX) ^
^
§ i 3 X
w> 8
O o o (N
c *c 3 X
p tt> ft d> p
o 3
O O o (N
£
-d P
P
IO 3^
o
NO On
0O
i 0C u Oh
On
DC
£ 2
CO
m
o o
>n
o o
(N
o o
2
,—s o s—^ cni
I
s h* r*>
"g«n
in cn „ ON CN „ in 00 s—■ /
o it (N
eg
X3 S
*3 eg
Cl —
H
P
O cu
O
C eg M £ o fe DT3 & C *3 eg 3
O X OX) eg Ev: VO
r-~ O Z
z
t~-
T3
173 CA 3 CX 73 cd E NW N r_ • a P4 © "O c § E u
P-
*
j5
Eo
E
6 | Z ex
Zlj
&
H
C P TD c ? -o ) T3 W
Review
Listeria monocytogenes Persistence in Food-Associated Environments: Epidemiology, Strain Characteristics, and Implications for Public Health V. FER REIRA ,12 M. WIEDMANN,2 P. TEIXEIRA,1 AND M. J. STASIEWICZ2* 1Centro de Biotecnologia e Quimica Fina, Laboratorio Associado, Escola Superior de Biotecnologia, Universidade Catolica Portuguesa Porto, Rua Dr. Antonio Bernardino Almeida, 4200-072 Porto, Portugal; and 2Department o f Food Science, Cornell University, Ithaca, New York 14853, USA MS 13-150: Received 10 April 2013/Accepted 11 September 2013
ABSTRACT Over the last 10 to 15 years, increasing evidence suggests that persistence of Listeria monocytogenes in food processing plants for years or even decades is an important factor in the transmission of this foodbome pathogen and the root cause of a number of human listeriosis outbreaks. L. monocytogenes persistence in other food-associated environments (e.g., farms and retail establishments) may also contribute to food contamination and transmission of the pathogen to humans. Although L. monocytogenes persistence is typically identified through isolation of a specific molecular subtype from samples collected in a given environment over time, formal (statistical) criteria for identification of persistence are undefined. Environmental factors (e.g., facilities and equipment that are difficult to clean) have been identified as key contributors to persistence; however, the mechanisms are less well understood. Although some researchers have reported that persistent strains possess specific characteristics that may facilitate persistence (e.g., biofilm formation and better adaptation to stress conditions), other researchers have not found significant differences between persistent and nonpersistent strains in the phenotypic characteristics that might facilitate persistence. This review includes a discussion of our current knowledge concerning some key issues associated with the persistence of L. monocytogenes, with special focus on (i) persistence in food processing plants and other food-associated environments, (ii) persistence in the general environment, (iii) phenotypic and genetic characteristics of persistent strains, (iv) niches, and (v) public health and economic implications of persistence. Although the available data clearly indicate that L. monocytogenes persistence at various stages of the food chain contributes to contamination of finished products, continued efforts to quantitatively integrate data on L. monocytogenes persistence (e.g., meta-analysis or quantitative microbial risk assessment) will be needed to advance our understanding of persistence of this pathogen and its economic and public health impacts.
Listeria monocytogenes is a gram-positive, facultative intracellular foodbome pathogen that can cause severe invasive illness (listeriosis) in humans and other animal species, including mammals (e.g., ruminants) and birds (155). Although invasive human listeriosis can occur in healthy individuals, the vast majority of cases occur in young, elderly, or immunocompromised individuals and manifest as septicemia, meningitis, or other infections of the central nervous system. In pregnant women, infections may lead to spontaneous abortion, still birth, or fetal death (177). Scallan et al. (154) estimated that this foodbome pathogen causes approximately 1,460 hospitalizations each year in the United States, resulting in 260 deaths. In developed countries worldwide, the incidence of listeriosis is 0.36 to 5 cases annually per million people (42, 43, 154); however, the number of reported cases can be very low in countries with limited surveillance for this disease. Al though some authors have suggested that the majority of * Author for correspondence. Tel: 607-255-1266; Fax: 607-254-4868; E-mail: [email protected].
human listeriosis cases are sporadic, findings from some studies that have included cluster analysis of human cases based on molecular subtyping data suggest that more human listeriosis cases than previously assumed may represent outbreaks (28, 147), supporting use of improved molecular subtyping methods for improved outbreak detection. In the first decade of the 21st century, coinciding with enhanced implementation of foodbome pathogen subtyping proce dures, a large number of human listeriosis outbreaks have been reported in various countries, including the United States (23, 58, 165), Canada (76, 136), Chile (77, 78), Germany and Austria (52), the United Kingdom (37), Sweden (20), Czech Republic (178), Japan (104), and Australia (128). Although L. monocytogenes is a facultative intracellular pathogen usually of homeothermic animals, it also can grow and survive outside a host at a wide pH range (4.7 to 9.2) (131, 132), at high salt concentrations (10%, wt/vol) (109), and most importantly at refrigeration temperatures (—0.5 to 9.3°C) (180). This growth range allows this pathogen to subsist in the food processing plant environment, survive
J. Food Prot., Vol. 77, No. 1
various food processing hurdles, and proliferate in food products. L. monocytogenes has been isolated from a variety of raw and processed food products, including milk and dairy products, meat products, egg products, seafood, vegetables, and other ready-to-eat (RTE) foods (44). In addition to food processing plant environments, this pathogen has been isolated from other environments, including natural, urban, and farm environments (66, 117, 149, 152), and types of samples, including soil, decaying vegetation, stream water, sewage, and human and animal feces (46, 125, 181, 182). Its frequent presence in different environments and its unusual ability to adapt to and survive under stressful conditions make control of L. monocyto genes in the food processing environment a considerable challenge. L. monocytogenes also has been isolated from environmental samples at retail establishments (148, 151) and in consumer homes (12, 31). However, our understand ing of L. monocytogenes transmission and ecology in food-associated environments other than those used for commercial food processing is limited. Results from two independent but similar risk assessments in the United States (41, 135) suggest that in >50% of human listeriosis cases associated with consumption of RTE deli meats the contamination may have occurred at the retail level. Cross contamination of retail products was recently found during a major listeriosis outbreak in Canada linked to contaminated cheese made from pasteurized milk (55). Although L. monocytogenes is inactivated by thermal treatments used for production of RTE foods, postproces sing cross-contamination from equipment and the environ ment represents a major concern (93, 95). Various approaches are used to control postprocessing contamina tion in food processing plants, including stringent imple mentation of good manufacturing practices, sanitation standard operating procedures, sterile packaging technolo gies (36), in-package postlethality treatments, and reformu lation of products with antimicrobial agents. Use of molecular subtyping approaches to better understand L. monocytogenes ecology and transmission in various foodassociated environments will continue to improve control of L. monocytogenes and prevention of cross-contamination. For example, molecular typing methods such as pulsed-field gel electrophoresis (PFGE), amplified fragment length polymorphism, and ribotyping, which have high discrimi natory power, have been essential in studies of L. monocytogenes sources and contamination routes in food processing plants (4, 56, 66,130). Studies including various molecular subtyping methods have revealed that specific L. monocytogenes subtypes can persist in food processing plants over long periods of time, whereas other subtypes are recovered only sporadically (96, 101, 113, 121). Subtypes that persisted were occasionally isolated from contaminated foods and were sometimes associated with listeriosis outbreaks. Although many field studies have revealed L. monocytogenes persistence in various food-associated environments and laboratory studies have revealed L. monocytogenes characteristics that facilitate persistence, our understanding of L. monocytogenes persistence and its impact on food safety and public health is still fragmented.
PERSISTENCE OF L. MONOCYTOGENES
151
This review includes a discussion of our current knowledge on some key issues regarding L. monocytogenes persistence: (i) persistence in food processing plants and other foodassociated environments (farm, retail, and consumers home environments), (ii) persistence in the general environment (natural and urban environments), (iii) phenotypic and genetic characteristics of persistent strains, (iv) environ mental niches, and (v) public health and economic implications of L. monocytogenes persistence. PERSISTENCE OF L. MONOCYTOGENES The term “ persistence” can acquire a variety of meanings within the context of a bacterial foodborne pathogen. Persistence can describe the long-term survival of a pathogen in a human host (a persistent infection). Although persistent human or animal infections have been described for various pathogens (e.g., persistent infection of the nostrils with Staphylococcus aureus), including foodborne bacterial pathogens (e.g., persistent infection of food handlers with Salmonella) (150, 167), no clear evidence of persistent infection of humans with L. monocytogenes has been published. Persistence also is used to describe long term survival (typically without growth) of a foodborne pathogen in a food or food matrix (often in laboratory incubation experiments), e.g., meat products (57, 75), cheese (94, 115), smoked salmon (60,142), and vegetables (11). Persistence can similarly be used to describe long-term survival of a pathogen in a simple, defined matrix (e.g., soil, water, or stainless steel surfaces) or a complex natural or human-made environment (e.g., a processing plant). In this review, we use the term persistence to describe the presence over time of L. monocytogenes in complex natural or human-made environments. This persistence likely requires both growth and survival in specific compartments of the complex environment. For example, L. monocytogenes could persist in a processing plant in a specific abiotic environment (e.g., a hollow roller in a piece of equipment or an uncleanable part of a slicer) or as a long-term host population inside a plant (e.g., in rodents in a plant lacking appropriate pest control). Differences in criteria used to define persistence complicate determination of persistence in a complex environment (e.g., a food processing plant or a farm). Typically, persistence is defined by repeated isolation on different dates of L. monocytogenes strains that are subsequently identified as identical subtypes (as determined by phenotypic or genotypic methods). However, some subtyping methods have limited discriminatory power, either for L. monocytogenes as a species or for specific subtypes (e.g., serotype 4b strains). Thus, a method with limited discriminatory power may not be useful for determining persistence of the same subtype in multiple sample collections. Some L. monocytogenes subtypes (e.g., specific PFGE types) are relatively common and widely distributed, whereas others appear rare. Fugett et al. (53) used the standard Centers for Disease Control and Prevention (CDC) two-enzyme PFGE protocol for L. monocytogenes and found one PFGE type that had caused
152
FERREIRA ET AL.
outbreaks in Switzerland in 1983 through 1985 and in the United States in 1985; this type also was isolated from multiple farms and urban environments in New York state in 2001 and 2002. Isolation of a very common subtype from multiple sample collections provides weaker evidence of persistence compared with isolation of a rare subtype from multiple sample collections. The relative frequency of a given subtype can be determined with a statistical testing approach developed by Malley et al. (105) that compares the observed frequency of the subtype to the historical frequency of that subtype. Of the five instances in which a subtype was observed more than once in a food processing plant, only three were identified as showing persistence; the other two cases of repeated isolation (two observations of the given subtype) were not significantly different from isolation expected by chance. Appropriate frequency information for subtypes may be very costly to acquire. The relative frequency of particular L. monocytogenes subtypes among human clinical isolates may be available, but the same information may not be available for other source populations (e.g., turkeys and turkey farms, small ruminants, and raw sheep milk cheeses), limiting frequency comparison within understud ied environments. Differentiation between repeated reintroduction of a specific L. monocytogenes subtype into a facility and true persistence in the facility is challenging based on subtyping information alone and almost always also involves infor mation on facility set up and design. To illustrate extremes, isolation of the same subtype over multiple sampling times in a facility that uses only pathogen-free ingredients and is separated from its surrounding environments (e.g., through premises design and constmction and/or implementation of good manufacturing practices) is likely to indicate true persistence. In contrast, isolation over multiple sampling times in a facility that has few measures to prevent introduction of L. monocytogenes from ingredients or surrounding environments (e.g., a farm or a retail facility) may indicate either reintroduction or persistence. In the following examples, criteria used to define persistence differ among studies and often include only isolation of the same subtype in a given facility over at least two or three sampling times. Although these repeat isolations may indicate persistence of the pathogen, we cannot exclude the possibility that in some instances, particularly when subtyping methods with low discriminatory power were used for isolate characterization, repeat isolation of a subtype may not necessarily indicate persistence. We recommend careful evaluation of data on strain persistence and hope to stimulate future efforts to develop and use statistical methods to quantify the likelihood that a set of subtyping data tmly indicates persistence. Persistence of L. monocytogenes in food processing plants. Harvey and Gilmour (61) were among the earliest authors to suggest colonization of food production and processing environments by persistent L. monocytogenes subtypes (Table 1). Multilocus enzyme electrophoresis (MEE) and restriction fragment length polymorphism
J. Food Prot., Vol. 77, No. 1
(RFLP) of L. monocytogenes isolates obtained from raw milk and nondairy food products from various producers and manufacturers revealed the recurrent isolation of L. monocytogenes of the same subtypes. Although lacking environmental samples, the authors suggested that specific subtypes of L. monocytogenes can persist in food processing and farm environments for long periods and recontaminate raw milk. Rprvik et al. (141) first reported persistence of L. monocytogenes subtypes within a fish plant: a salmon smokehouse in Norway (Table 2). MEE of the L. monocy togenes isolates revealed that those belonging to one electrophoretic type (ET-6) were present during the entire 8-month investigation in the environment, in the fish during processing, and in vacuum-packed smoked salmon. In the same year, Lawrence and Gilmour (95) (Table 3) reported characterization of L. monocytogenes isolates from a poultry meat processing plant by random amplification of polymor phic DNA (RAPD). Two of 18 RAPD types found (A and B) persisted throughout a 6- and 5-month period, respec tively, with one type more prevalent in raw poultry products and equipment and the other more prevalent in the cooked poultry processing environment. Samples were contaminat ed by the same RAPD types up to 1 year later. Subsequently, persistence of L. monocytogenes strains for months to several years has been reported in a variety of food processing plants, including those for meat, fish, dairy, and RTE products (Tables 1 through 4). In some studies, environmental sources have been identified as the main source of postprocessing L. monocytogenes contamination of food products within processing plants (124). Persistent L. monocytogenes strains are more often isolated from food processing environments (e.g., drains and equipment), including sites close to food contact surfaces (e.g., dicing machines), rather than from raw materials (99,113,156,179). The recovery of persistent strains from the environment and equipment after cleaning and disinfection emphasizes the risk of growth and establishment of L. monocytogenes, particularly in sites difficult to access, leading to ongoing food product contamination (4, 124, 186). Microbiological testing of the processing environment and equipment is therefore neces sary to detect particular niches of L. monocytogenes and validate the efficiency of sanitation procedures. Persistent strains also can be transferred between facilities through a contaminated environment. For example, a meat dicer harboring L. monocytogenes in an internal niche transferred identical PFGE types to two subsequent processing plants that purchased the contaminated equipment (99). Other authors have emphasized the importance of cross-contamination by raw materials. Berrang et al. (10) reported that three of the four resident L. monocytogenes strains in a chicken processing plant were recovered from the raw product at some point during the 1-year study, suggesting persistent reintroduction of this strain from raw product into the plant environment. Persistence of L. monocytogenes in other foodassociated environments. Although molecular subtyping
J. Food Prot., Vol. 77, No. 1
P E R S I S T E N C E O F L. MONOCYTOGENES
£> c -3 2 8 -a .23 B. 0 S£ £ 8.
°*0
X ON ON
ON ON ON
x 2 T3 C/3 s 1/3 O ’c5—4 /3 ’£ D > o cd T3 s -c c '3 2C /3
•S o 3 VJ 0 3e 2 ^ ■£ 1 ° C ^L 0—t c/3 © O 'c/3 ~ .2 3 c o £ '> 03 CL 3 3d> 3e O 3 , 3 X •S
T3 3cd
T A B L E 1. Selected studies of the persistence ofL. m o n o c y t o g e n e s in dairy processing plants
9O 00
E £cd DC
1>5 .aio ja £ % §
3
T3 O
cd
X ~ ao * “ .o cd X
g> xcd O -n
3
h§C /3 03
O 3 d) 3 t/3 00 ON . « •s .S d > 03 • CL ^ ’ .3
•
Vl ’ jDCL xX c y >,* • X ,cd 4 c— d) ■ £o 3CL
02
M 1"O J
-o 3
3
60 3 '33 03 yo di
) < L > oo ■0-3 dX 3 1 2 3 ■o « o CL 00 Cx0 2 s ,0 -w" 3 C 5. °o •o 1 £ £C/3 dJ s o '3 3 03 d) f.2
X
w
CL
O
E 7
E
GO 3 'a >>
G O
GO
c
a .
2s
vi
X
0 X £ X •&
0
(-H >V
d> 5— < 2
0 C0fl
m
PJ pft .
o Ph CL,
CL
GO 3 *CL •w* O X C4
3 1) _o 3 OGO cd N 'C Id 0
60 3 ‘
5-
>>
2
^ T> y3
CL £
Number of persistent isolates for each subtype is given in parentheses; multiple numbers in parentheses indicate multiple persistent subtypes in a given plant. MEE, multilocus enzyme electrophoresis; PFGE, pulsed-field gel electrophoresis. NA, data not available.
On O
K
ra w onge
> x
^4
15 3
FE R R E IR A E T A L .
J. Food Prot., V o l. 77 , No. 1
O
3
On On ON
C
a
£ .3 Si s X co
o U
Cl- tfc CD o O o •g (D JS o «
1
3
X
P --1 . j-H C/3
Q I CL —
> £ cr *s X 3 o
S.J
I s
1 | Qh ^ x £ d) c -3 0 ^ °
CL O
H
X X O O
£
§
CL o £ H
■g 3 w
E
3 ctf 3 —r r ~ \ T
XL o fc 3 CL £ W c =* & ?' o
.5 5
i
s CJ
oo-
“
3
O
X ^ C O
^
^
8
3 £
5 d> a >. 73
c | 8 o OX) *c
3 X
"
oo X
OX)
U2
cl a) g c p «* Xd> p^ o 5o p .
XI
d> »-
>>
X
D 3 "g .N
OX) L) OX) 3 X c - C - C N OX) • > X > i- c 3 ■ X C X ^ 3 .. u 3 o C CO TO 1/1 TO TO
8„ T3
~ x
CJ. ^
fll* O H
o
CL
5 X CL £
3 7?
o c CL d)
(N O o (N
3 dj CL CL
2 1/3 .52 p
*o _8 ^ 3
CL ~ ^ M
3 > 3 >
o C CL d)
s P CL S O C P OX)
is c
.
OX)
‘cl
o
d) 3
O 3 X O O u.
O
j*
O
co
c H -s
£ 3
CO oo i OX) £ ^ d> — • ^ _c X > OX) d> x 3 .s - i .£ £ « 2 CO Q, oo . 3 “ ^■ 5 E .- P d) 3 X o as 3 o g p d> u cd CL T3 3 c X c OX) P O d> 3 c X o, 3c ’■^ — X ’ n 7j r J 3 « OX) P ..5i 2 o X c .£ •p P o X c3 D , c3 « 3
5’. O — w
3 X
CL g
o o < N
OX)
d> X
c d)
OX) ^
^
§ i 3 X
w> 8
O o o (N
c *c 3 X
p tt> ft d> p
o 3
O O o (N
£
-d P
P
IO 3^
o
NO On
0O
i 0C u Oh
On
DC
£ 2
CO
m
o o
>n
o o
(N
o o
2
,—s o s—^ cni
I
s h* r*>
"g«n
in cn „ ON CN „ in 00 s—■ /
o it (N
eg
X3 S
*3 eg
Cl —
H
P
O cu
O
C eg M £ o fe DT3 & C *3 eg 3
O X OX) eg Ev: VO
r-~ O Z
z
t~-
T3
173 CA 3 CX 73 cd E NW N r_ • a P4 © "O c § E u
P-
*
j5
Eo
E
6 | Z ex
Zlj
&
H
C P TD c ? -o ) T3 W
