The authors have declared that no competing interests exist.
Current address: BioMatrix International, Princeton, Minnesota, United States of America
New regulatory and consumer demands highlight the importance of animal feed as a part of our national food safety system. Porcine epidemic diarrhea virus (PEDV) is the first viral pathogen confirmed to be widely transmissible in animal food. Because the potential for viral contamination in animal food is not well characterized, the objectives of this study were to 1) observe the magnitude of virus contamination in an animal food manufacturing facility, and 2) investigate a proposed method, feed sequencing, to decrease virus decontamination on animal food-contact surfaces. A U.S. virulent PEDV isolate was used to inoculate 50 kg swine feed, which was mixed, conveyed, and discharged into bags using pilot-scale feed manufacturing equipment. Surfaces were swabbed and analyzed for the presence of PEDV RNA by quantitative real-time polymerase chain reaction (qPCR). Environmental swabs indicated complete contamination of animal food-contact surfaces (0/40 vs. 48/48, positive baseline samples/total baseline samples, positive subsequent samples/total subsequent samples, respectively;
Federal regulations recognize animal feed as food and an important part of our national food supply. Recent changes in legislation through the Food Safety Modernization Act, along with evolving consumer demands, are placing greater emphasis on the role of animal food in the farm-to-fork food safety system [
The animal food manufacturing portion of the experiments was conducted at the Kansas State University Cargill Food Safety Research Center (FSRC; Manhattan, KS), a 3-story biosafety level 2 biocontainment laboratory and animal food manufacturing facility containing pilot scale animal food manufacturing equipment. Procedures were approved by the Kansas State University Institutional Biosafety Committee (Approval No. 929.3). All manufacturing procedures were replicated three times. Decontamination occurred before and after each replicate to establish baseline and confirmed negative by the absence of PEDV RNA on animal food-contact and non-food contact surfaces as measured by qPCR as previously described [
The portion of the experiment evaluating infectivity in animals was conducted at Iowa State University. Procedures were approved by the Iowa State University Institutional Animal Care and Use Committee (Approval No. 1-16-8168-S).
Virus isolation, propagation, and titration were performed in Vero cells (ATCC CCL-81) as previously described [
A corn-soybean meal-based diet with a composition typically fed to adult swine was manufactured at the Kansas State University O.H. Kruse Food Technology Innovation Center (Manhattan, KS) (
Ingredient, % | Composition |
---|---|
Corn | 79.30 |
Soybean meal, 46.5% CP | 15.70 |
Choice white grease | 1.00 |
Calcium phosphate (monocalcium) | 1.40 |
Limestone | 1.15 |
Salt | 0.50 |
L-Threonine | 0.03 |
Trace mineral premix |
0.15 |
Sow add pack |
0.50 |
Vitamin premix |
0.25 |
Phytase |
0.02 |
Total | 100.00 |
Formulated analysis |
|
DM | 91.4 |
CP | 17.1 |
Crude fiber | 3.7 |
Ca | 0.78 |
P | 0.52 |
Fat | 3.5 |
aEach kilogram of premix contains 73 g Fe, 73 g Zn, 22 g Mn, 11g Cu, 0.198 mg I, and 0.198 mg Se.
bEach kilogram of premix contains 4,409 IU vitamin E, 44 mg biotin, 992 mg pyridoxine, 331 mg folic acid, 110,229 mg choline, 40 mg chromium, 9,920 mg L-carnitine.
cEach kilogram of premix contains 4,409,171 IU vitamin A, 551,146 IU vitamin D3, 17,637 IU vitamin E, 1,764 mg menadione, 3,300 mg riboflavin, 11,023 mg d-pantothenic acid, 19,841 mg niacin, 15 mg vitamin B12.
dHigh Phos 2700 GT, DSM Nutritional Products, Parsippany, NJ.
eOne sample was analyzed by Ward Laboratories Inc., Kearney, NE.
The previously-prepared 5 kg of inoculum was added to 45 kg of virus-free animal food in a 0.113 m3 electric paddle mixer (H. C. Davis Sons Manufacturing; Model SS-L1; Bonner Springs, KS) to form the positive control, and was mixed and discharged as described above. Four sequenced 50 kg batches (Sequence 1 to 4) of virus-free animal food were mixed and discharged after the positive control without any cleaning or decontamination between batches to mimic commercial animal food production conditions.
Prior to and after each batch of feed being manufactured, environmental surfaces were swabbed using large foam-tipped disposable swabs (World Bio-Products LLC, Woodinville, WA) that were pre-wetted with 2 ml of phosphate buffered saline. To collect samples, a clean pair of disposable gloves was worn, each swab opened aseptically, and rubbed across the desired surface. Swabs were then capped and placed in a cooler with ice until analyzed.
Designated locations were sampled as illustrated in
Designated areas swabbed for PEDV qPCR analysis include high and low foot traffic areas (concrete), drain (concrete), garage door (metal), pellet mill (equipment), table ledge (metal), conveyer (equipment), and food mixer (equipment). Not shown are rubber boot bottoms (rubber).
Eighteen pigs were purchased from a conventional breeding farm and delivered to the Iowa State University Laboratory Animal Resource (LAR) facilities. All pigs were administered an intramuscular dose of ceftiofur (Exede; Zoetis, Florham Park, NJ) per label instructions upon arrival and confirmed negative for PEDV, porcine delta coronavirus (PDCoV), transmissible gastroenteritis virus (TGEV) and porcine rotaviruses (groups A, B, and C) by virus specific qPCR on rectal swabs. In addition, pigs were confirmed PEDV antibody negative by fluorescent foci neutralization serologic analysis performed at South Dakota State University Veterinary Diagnostic Laboratory (SDSU VDL).
A bioassay was conducted 11 months after animal food preparation and sample collection. A total of 6 rooms (3 pigs per room) were assigned to swabbed dust samples collected from the conveyer after production of each animal food treatment (1 negative control room and 5 challenge rooms). Pigs were blocked by weight, then randomly divided into groups of 3 per room. Rooms had independent ventilation systems and solid flooring that was minimally rinsed to reduce PEDV aerosols. Pig were fed liquid milk replacer (Esbilac; PetAg, Hampshire, IL) and commercially pelleted diet (All Natural Starter 2; Heartland Co-op, Alleman, IA). Pigs had ad libitum access to food and water at all times.
After 2 days of acclimation, each pig was administered the dust suspension from swabbed surfaces by orogastric gavage using an 8−gauge French catheter and 60 ml syringe (8 ml/pig), which marked day 0 post inoculation (0 DPI). The 8 ml aliquot combined eight 1-ml dust suspensions sampled from 4 buckets and 4 adjacent belt areas after manufacturing each food treatment from one replicate. Thus, each pig represented 1 of 3 replicates per treatment and each room represented each treatment.
Rectal swabs were analyzed from all pigs on -2, 0, 2, 4, 6, and 7 DPI. Swabs were submerged into 1 ml phosphate buffered saline (PBS, 1 × pH 7.4) immediately after collection and submitted to Iowa State University Veterinary Diagnostic Laboratory (ISU VDL) for PEDV RNA by qPCR. All pigs were euthanized at 7 DPI for necropsy by intravenous overdose of pentobarbital sodium solution as per label instructions (Fatal-Plus; Vortech Pharmaceuticals Ltd, Dearborn, MI). At necropsy, an aliquot of fresh cecal contents was submitted for PEDV qPCR to ISU VDL.
Dust samples from swabs were tested at Kansas State University Molecular Diagnostics Development Laboratory (Manhattan, KS) for PEDV using a PEDV spike (S) gene-based qPCR. Nucleic acids were extracted from a 50 μL sample of supernatant. Automated extraction was carried out on a KingFisher magnetic particle processor (Thermo Scientific, Waltham, MA) using a MagMAX-96 Viral RNA Isolation Kit (Life Technologies, Grand Island, NY). All manufacturer’s instructions were followed, with the exception of a final elution volume of 60 μl. Each 96-well extraction run included an extraction positive control (PEDV stock virus) and an extraction negative control (1x PBS). Four μl of RNA template was used in qPCR setup in a 20 μl reaction using a real time RT-PCR kit (Path-ID Multiplex One-Step RT-PCR Kit; Thermo Scientific, Waltham, MA). Amplification reactions were conducted on a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA). The thermal cycling parameters were: 10 min reverse transcription at 48°C, 10 min of reverse transcriptase inactivation/initial denaturation at 95°C followed by 45 cycles of 10 sec at 95°C and 40 sec at 60°C.
Animal samples and samples for bioassay were tested for PEDV using a previously described PEDV nucleocapsid (N) gene-based qPCR [
Swabs were categorized as animal food-contact and non-animal food-contact surfaces. Within animal food-contact surface, Ct analysis of the metal mixer, plastic conveyer buckets, and rubber conveyer belt were performed using PROC GLIMMIX (SAS Institute, Inc., Cary, NC). Within animal food-contact surface, the statistical model evaluated the effect of treatment (negative, positive, sequence 1, sequence 2, sequence 3 and sequence 4) and surface (metal mixer, plastic conveyer buckets, and rubber conveyer belt) and the associated interaction. Each swab was classified from treatment and surface type. The LSMEANS procedure compared surface type among treatments within animal food-contact surfaces by pairwise comparison. The non-animal food-contact surfaces were reported in the results text using descriptive statistics; non-animal food-contact swabs were organized by surface type (metal garage, metal tabletop, concrete floor, and rubber boot bottoms worn during the experiment) among treatments. Samples considered negative by qPCR were evaluated as a value of 45 in the statistical model. Results were considered significant at
As expected, all animal food-contact negative control swabs were qPCR negative (
Treatment | ||||||
---|---|---|---|---|---|---|
Item | Negative | Positive | After sequence 1 | After sequence 2 | After sequence 3 | After sequence 4 |
Contact Zone, Detectable RNA/Total |
||||||
Animal food-contact | ||||||
Metal mixer |
0/9 | 9/9 | 9/9 | 6/9 | 4/9 | 4/9 |
Plastic conveyor bucket |
0/12 | 12/12 | 12/12 | 12/12 | 12/12 | 12/12 |
Rubber conveyor belt |
0/12 | 12/12 | 12/12 | 12/12 | 12/12 | 12/12 |
Swab, Ct |
||||||
Metal mixer | 45.0 |
29.2 |
33.9 |
38.2 |
40.7 |
40.5 |
Plastic conveyor buckets | 45.0 |
30.8 |
32.1 |
34.2 |
32.8 |
32.1 |
Rubber conveyor belt | 45.0 |
30.8 |
31.5 |
31.5 |
32.2 |
32.1 |
abcdefgSuperscripts within a row that do not share a letter differ
‡Count of swabs with detectible PEDV RNA/number of swabs analyzed.
¶Metal includes one sample each from the mixer paddle, mixer interior lid, and mixer interior bottom.
#Plastic includes one swab each from 4 randomly chosen interior conveyor buckets.
All non-animal food-contact surface baseline swabs were qPCR negative. Non-animal food-contact swabs were analyzed by surface type (metal garage, metal tabletop, concrete floor, and rubber boot bottoms worn during the experiment). Unexpectedly, in 1 of 3 repetitions, 1.7% of non-animal food-contact surface swabs were qPCR positive after the negative treatment was manufactured, although the animal food was qPCR negative. For all repetitions, after the positive treatment and after sequence 1, 89% of non-food-contact surface swabs were qPCR positive. After sequence 2, 94% of non-food-contact surface swabs were qPCR positive. After sequence 3, 89% of non-food-contact surface swabs were positive that again increased to 94% after sequence 4. The percentage of positive swabs from non-animal food-contact metal surfaces (metal garage and tabletop) varied, whereas non-animal food-contact concrete floor and rubber boot bottoms remained the same (67%, 67%, 83%, 67%, 83%; after positive, after sequence 1, after sequence 2, after sequence 3 and after sequence 4, respectively vs. 100% after positive and sequence 1 to 4, respectively). Dust suspensions from animal food-contact surfaces were challenged in pigs and failed to produce infectivity.
The recent enacting of the Food Safety Modernization Act (FSMA) requires animal food manufacturers to identify and control animal food safety hazards because feed is considered animal food and a part of the human food safety system [
For these reasons, an established protocol for monitoring viral transmission is needed to model animal and human food hazards if additional pathogenic viruses are discovered in our food supply. This is the first study of its kind to fully observe environmental contamination of an animal food-manufacturing facility during a proposed control method after manufacturing viral-inoculated swine food. Objectives were met by monitoring the extent of virus contamination in an animal food manufacturing facility and investigating a control method to decrease virus contamination on animal food-contact surfaces.
In general, environmental contamination of a virus in any food manufacturing facility has not been well-documented. In human food, norovirus is a known cause of foodborne illness with contamination presumed at point-of-service [
The results from this study clearly demonstrate the extent of the widespread viral contamination that occurs in an animal food manufacturing facility after production of virus-inoculated animal food. All of the animal food-contact surfaces and most of the non-animal food-contact surfaces were qPCR positive when swabbed after the contaminated animal food was manufactured and remained qPCR positive after multiple batches of animal food were mixed and conveyed. Therefore, it seems that the proposed mitigation technique (feed batch sequencing) did not mitigate environmental PEDV contamination. Additionally, detectible PEDV seemed to persist on some animal food-contact surfaces, such as plastic and rubber conveyors, more than others such as metal. Previous studies have investigated the survivability of virus on inanimate surfaces and determined viral persistence in the environment can be affected by several factors including surface type [
In pet food manufacturing, equipment surfaces are easy-to-clean with non-porous equipment surfaces selected in order to prevent biofilms or the prevalence of
Alternatively, the difference in rate of contamination between the metal mixer or plastic and rubber in the conveyor may be due to equipment design. For example, mixers are typically designed to self-clean with little residual material from one batch to the next compared to conveyors. This is particularly true of bucket elevators, which is the conveyor type used in this experiment. The large rubber belt of a bucket elevator is suspended vertically, and plastic buckets convey feed upward until the feed is flipped from the buckets into a discharge chute. The boot pit, which is the area at the bottom of the bucket elevator, must be large enough for buckets to clear the bottom without coming into contact with the guard or cover. This area typically fills with residual feed and may lead to batch-to-batch cross contamination, which has been demonstrated by carryover of animal drugs [
This research concludes that differences exist in viral contamination rates on different equipment surfaces, which may be due to differences in surface type, equipment design, or other phenomena. Regardless of the source of these differences, animal food manufacturing facilities at risk for PEDV contamination should consider these findings when choosing manufacturing equipment. The results of the current experiment are applicable to other species of animal food and to human food manufacturing facilities because entry of a viral pathogen may cause widespread contamination that is difficult to eliminate. Even with wet chemical cleaning and facility heating, PEDV proved difficult to decontaminate from our facility [
In the current study, environmental surfaces were swabbed for dust after production of PEDV inoculated animal food and animal-food contact surfaces were evaluated for infectivity. A previous proof-of-concept-study demonstrated that animal food dust can be infectious [
Another result from this study is that some non-food contact swabs from a repetition were qPCR positive after the negative animal food was manufactured, although importantly, animal food tested was qPCR negative. We hypothesize this genetic material remained on the boot due to inadequate cleaning after a previous replicate and was tracked then detected on the concrete floor. Due to the chemical cleaning between repetitions, the viral material should not have been infective [
As the current study demonstrates, widespread contamination of PEDV occurred and was detected on most surfaces. Material collected from dust collection systems and sweepings should be collected and disposed instead of added to the product flow as per traditional measures [
In conclusion, this study clearly demonstrates widespread contamination occurred in an animal food manufacturing facility after PEDV swine food production. Furthermore, the proposed mitigation method of feed batch sequencing was not effective to reduce environmental contamination, although the potential impact of PEDV contamination and importance to prevent virus entry in such facilities was better understood. It is concerning once an animal food manufacturing facility is contaminated with PEDV, it appears to harbor PEDV until chemically cleaned. This research indicates animal food manufacturing facilities potentially contaminated with PEDV can be a central point for virus transmission and the quantification for this risk should be assessed. As a result, the practicality of decontamination is a new challenge facing our animal food manufacturing facilities.
The authors would like to thank Iowa State University Veterinary Diagnostic Laboratory for processing animal samples.