Experimental Infection of Domestic Piglets (Sus scrofa) with Rift Valley Fever Virus

ABSTRACT.

Rift Valley fever phlebovirus (RVFV) is a mosquito-transmitted phlebovirus (Family: Phenuiviridae, Order: Bunyavirales) causing severe neonatal mortality and abortion primarily in domestic ruminants. The susceptibility of young domestic swine to RVFV and this species’ role in geographic expansion and establishment of viral endemicity is unclear. Six commercially bred Landrace-cross piglets were inoculated subcutaneously with 10⁵ plaque-forming units of RVFV ZH501 strain and two piglets received a sham inoculum. All animals were monitored for clinical signs, viremia, viral shedding, and antibody response for 14 days. Piglets did not develop evidence of clinical disease, become febrile, or experience decreased weight gain during the study period. A brief lymphopenia followed by progressive lymphocytosis was observed following inoculation in all piglets. Four piglets developed a brief viremia for 2 days post-inoculation and three of these had detectable virus in oronasal secretions three days post-inoculation. Primary inoculated piglets seroconverted and those that developed detectable viremias had the highest titers assessed by serum neutralization (1:64–1:256). Two viremic piglets had a lymphoplasmacytic encephalitis with glial nodules; RVFV was not detected by immunohistochemistry in these sections. While young piglets do not appear to readily develop clinical disease following RVFV infection, results suggest swine could be subclinically infected with RVFV.

INTRODUCTION

Rift Valley fever phlebovirus (RVFV), a mosquito-transmitted phlebovirus (Family: Phenuiviridae, Order: Bunyavirales), is known to infect and cause disease in several species of domestic ruminants including cattle, sheep, and goats.^1–5 Infection in susceptible domestic livestock species results in abortion and acute death in young animals due to liver failure. RVFV has caused large economic losses in endemic areas of eastern and southern Africa, and there has recently been spread of virus and potential vectors into western and northern Africa and the Arabian peninsula with predictable seasonal outbreaks following heavy rainfall.^6–10 The virus is also zoonotic, causing a dengue-like fever and stiff joints in humans with progression to encephalitis, retinitis, and hemorrhagic fever in about 10% of infections.⁷

Young animals are the most susceptible to severe disease; mortality rates in lambs under 2 weeks of age approach 100%.^1,2,4 Abortion rates can also approach 100% and acute death occurs in 20–30% of adult ruminants.^4,7 Human cases are often detected in the farm workers, veterinarians, and abattoir workers who handle tissues from infected animals.¹¹ Recent incursions of this virus onto the Arabian peninsula have raised concerns of a livestock trade ban in Africa.⁶

Infection of domestic swine (Sus scrofa) with RVFV was previously attempted in 1962¹ in three pigs, resulting in no clinical evidence of disease; however, these results are difficult to interpret because critical information, including the age of the pigs used, was not included in the report, and only one strain of RVFV was used. Antibodies against RVFV have been detected in a wild swine species, the desert warthog (Phacochoerus aethiopicus), indicating exposure to RVFV or a closely related virus.^12,13 Further, there are accounts of domestic swine abortions on farms where human cases of RVFV were identified⁷; hence, the true susceptibility of domestic swine, particularly young piglets, remains unknown. Feral swine in the US and Europe, which are genetically similar to domestic swine, are also of concern given their abundance, large geographical distribution, and the high degrees of contact with humans through hunting and livestock through shared grazing areas. Furthermore, a recent in vitro study demonstrated that S. scrofa cell lines are permissive to RVFV infection¹⁴; however, it is uncertain if this translates to in vivo infection. The objective of the current study is to better characterize the susceptibility and pathology of RVFV infection in domestic swine by inoculating young piglets via subcutaneous injection and monitoring for the development of clinical disease, timing and extent of viremia, and viral shedding.

MATERIALS AND METHODS

Animal infection and sample collection.

Eight commercially bred Landrace cross piglets, five females and three males, were acquired at 19 days of age from University of Georgia (UGA) Swine Farm and Research Center. Piglets had been weaned and vaccinated with Circumvent G2 PCV-M, a porcine circovirus vaccine, and then moved to the designated study space in the Animal Health Research Center (AHRC) at UGA in Biosafety Level 3-Ag containment, followed by 1 week of acclimation prior to inoculation. Piglets were housed in groups of 4 in elevated 5′ × 3′ × 6′ pens with free choice feed and water. All animal handling and care was in compliance with approved Institutional Animal Care and Use Committee protocol (A2014 10-015-Y3-A2).

The RVFV strain used for inoculations and all laboratory assays was the ZH501 strain, which was originally isolated from a human during the 1997 outbreak in Egypt.

On post-inoculation day (PID) 0, all piglets were weighed, had blood drawn from the cranial vena cava for serology, hematology, serum biochemistry, coagulation assays, and a rectal temperature reading. Oronasal and rectal swabs were also taken. Piglets were sedated using intramuscular (IM) 2 mg/kg xylazine. Six piglets (three in each pen) were inoculated subcutaneously in the right neck with 10⁵ plaque-forming units (PFU) of RVFV in 1 mL of Minimum Essential Media^™ (MEM; Sigma-Aldrich, Darmstadt, Germany) with added 3% bovine serum albumin, and 5% antibiotics and antimycotics (henceforth referred to as “virus media”). The remaining two piglets (one in each pen) were sham inoculated with 1 mL of virus media in the same location. Piglets were monitored twice daily for clinical signs (lethargy, abdominal pain, inappetence, death), febrile response, standardized body condition scores, respiratory effort, and hydration status. Sham inoculated piglets remained co-housed with inoculated individuals.

On PID 1, 2, 3, 4, 5, 7, and 14, 9 mL of blood was collected from the cranial vena cava of each piglet for virus isolation and titer and serology in an additive-free red top tube. On PID 1, 3, 5, and 7, additional EDTA and citrate tubes were collected for biochemical, coagulation, and hematologic testing. Oronasal and rectal swabs for virus isolation and titer were collected in virus media on PID 1, 3, 5, and 7. All piglets were euthanized and necropsied on PID 14. At necropsy, samples of approximately 0.1 g of liver and spleen were collected fresh in virus media. Liver, spleen, lung, kidney, heart, left and right cervical lymph nodes, and brain were collected in 10% neutral-buffered formalin from each piglet, and one eye was collected in Davidson’s fixative (Poly Scientific R&D Corp, Bay Shore, NY).

Virus isolation.

Blood samples were centrifuged at 1,207 × g for 10 minutes and serum collected on the day of the blood collection. One hundred (100) µL of fresh serum was added to singlicate Vero cell cultures (CRL-1586™, ATCC^®, Manassas, VA) and incubated in 5% CO₂ at 37°C for 5 days. After 5 days, 100 µL of cell suspension were inoculated onto a second Vero culture. Cultures with no cytopathic effect (CPE) after 7 days in the second passage were considered negative. Remaining serum was frozen at −20°C. Tissue samples (∼0.1 g) collected at necropsy were weighed and homogenized in 1 mL virus media (as defined previously) and centrifuged at 1207 × g for 10 minutes. One hundred (100) µL of the supernatant were added to Vero cell cultures and similarly incubated and observed daily for CPE as described previously. Oronasal and rectal swab virus media (100 µL) was similarly cultured on the same day as collection. Cultures showing CPE were tentatively identified as RVFV using the VectorTest RVFV Antigen assay¹⁵ and confirmed using RVFV-specific qRT-PCR.

Viral titers of positive serum samples were determined by plaque assay in Vero cells and expressed as PFU/mL. Cultures were also tested using the VectorTest RVFV Antigen assay and confirmed by qRT-PCR.

RNA extraction and qRT-PCR.

Serum and viral cultures were analyzed by qRT-PCR. Viral RNA was manually extracted using a QIAamp Viral RNA Mini Kit following the manufacturer’s instructions (QIAGEN, Germantown, MD). Quantitative reverse transcription PCR (qRT-PCR) was performed using a commercially available Rift Valley Fever Virus PCR Kit, the target of which is the viral L segment (6.4 kB), according to the manufacturer’s instructions (MyBioSource, San Diego, CA). Reactions were conducted on Applied Biosystems StepOnePlus Real-Time PCR System (Applied Biosystems Inc, Foster City, CA). A standard curve for copy number determination (Ct value) was created using the RVFV Positive Control Template, and a positive and negative (RNase/DNase free water) control were included in each run. The Ct detection threshold of the kit was 0.018797.

Serum neutralization.

Following thawing, serum samples were diluted 1:2 and heat inactivated by incubating at 57°C for 30 minutes, then 2-fold serially diluted and incubated with 100 TCID₅₀ of RVFV in 96-well microtiter plates for 1 hour at 37°C. Duplicate wells were used for each dilution of serum. Wells were then overlaid with 150 µL of Vero cells. Plates were incubated at 37°C and observed for CPE daily for 5 days before neutralization endpoint titers were determined. Titers are expressed as the reciprocal of the highest dilution of serum determined to provide > 50% protection of cells in both wells. Rabbit anti-RVFV serum was used as a positive control (SPB reagent #R146).

Complete blood counts, clotting profiles, and fibrinogen assessment.

Clinical pathology parameters assessed on PID 0, 1, 3, 5, and 7 included leukocytic, lymphocytic, monocytic, and neutrophilic cell counts, hematocrit and platelet counts, activated partial thromboplastin time (aPTT) and prothrombin time (PT) clotting tests, and fibrinogen quantification. For complete blood counts, EDTA blood aliquots were analyzed using the Abaxis VetScan HM5c Hematology Analyzer (Zoetis, Parsippany, NJ). aPTT/PT tests were run on citrated blood using an Abaxis VetScan VSPro (Zoetis, Parsippany, NJ). Citrated blood was then centrifuged at 1,207 × g for 10 minutes and plasma collected for fibrinogen analysis using the Abaxis VetScan VSpro. Results were compared by paired t tests and single factor ANOVA using Excel.

Histopathology and immunohistochemistry.

Tissues collected at necropsy were held in formalin at room temperature for 10 weeks within the AHRC as a biosafety precaution. After confirming that no viable virus remained via virus isolation, formalin-fixed tissues were processed routinely and embedded in paraffin wax. Tissue sections 4-µM thick were then prepared for histopathological examination and stained with hematoxylin and eosin.

Sections of brain, liver, lymph node, and spleen were examined by immunohistochemistry (IHC) for RVFV. Four-µM sections were deparaffinized and rehydrated and antigen retrieval performed in buffer (Antigen Retrieval Citra; BioGenex) at 12°C for 10 minutes followed by protein blocking (Universal block; Biogenex) for 5 minutes and quenching of endogenous peroxidases with 3% H₂O₂ for 20 minutes. Mouse monoclonal anti-RFV antibody (1:1,000, strain ZH-548 M12; Alpha Diagnostics International, San Antonio, TX; Catalog #RVF-NP-12-M) was applied for 1 hour, followed by a biotinylated secondary antibody (Universal Goat link; BioCare), a streptavidin HRP label (BioCare) and DAB (Vector). Sections were counterstained with hematoxylin. This technique was internally validated using liver from naturally aborted sheep and white-tailed deer experimentally infected with RVFV.

RESULTS

Virus isolation, qRT-PCR, and serology.

Virus isolation, qRT-PCR, and serology results are listed in Table 1. Four of the six inoculated piglets experienced a low-titer viremia on PID 1. Three piglets still had a detectable viremia on PID 2. Viral shedding was detected in oronasal swabs on PID 3 from three viremic piglets by both viral cultures and qRT-PCR detection. All inoculated pigs seroconverted by PID 5 and had increased titers on PID 7 and 14. Those pigs with viremia had higher serologic titers. No virus was detected in tissue samples collected at necropsy or in rectal swabs by viral culture.

Table 1.

Comparison of viremia profiles, qRT-PCR Ct values, and serologic response in piglets infected with Rift Valley fever virus (RVFV)

		Day post-inoculation
Piglet ID	Test	0	1	2	3	4	5	6	7	14
Sham-inoculated piglets
507	PT	ND	ND	ND	ND	ND	ND	ND	ND	ND
	PCR	ND	ND	ND	ND	ND	–	–	–	–
	OPT	ND	ND	ND	ND	ND	ND	ND	ND	ND
	OPCR	–	–	–	–	–	–	–	–	–
	SNT	< 4	–	–	–	–	< 4	–	< 4	< 4
505	PT	ND	ND	ND	ND	ND	ND	ND	ND	ND
	PCR	ND	ND	ND	ND	ND	–	–	–	–
	OPT	ND	ND	ND	ND	ND	ND	ND	ND	ND
	OPCR	–	–	–	–	–	–	–	–	–
	SNT	< 4	–	–	–	–	< 4	–	< 4	< 4
RVFV-inoculated piglets
503	PT	ND	3.48	2.3	ND	ND	ND	ND	ND	ND
	PCR	ND	30.69/0.3290	30.00/0.5350	ND	ND
	OPT	ND	ND	ND	POS	ND	ND	ND	ND	ND
	OPCR		–	–	ND	–	–	–	–	–
	SNT	< 4	–	–	–	–	16	–	32	128
504	PT	ND	ND	ND	ND	ND	ND	ND	ND	ND
	PCR	ND	ND	ND	ND	ND
	OPT	ND	ND	ND	ND	ND	ND	ND	ND	ND
	OPCR	–	–	–	–	–	–	–	–	–
	SNT	< 4	–	–	–	–	16		32	32
506	PT	ND	ND	ND	ND	ND	ND	ND	ND	ND
	PCR	ND	ND	ND	ND	ND	–	–	–	–
	OPT	ND	ND	ND	ND	ND	ND	ND	ND	ND
	OPCR	–	–	–	–	–	–	–	–	–
	SNT	< 4	–	–	–	–	8		32	64
508	PT	ND	2.84	ND	ND	ND	ND	ND	ND	ND
	PCR	ND	ND	ND	ND	ND	–	–	–	–
	OPT	ND	ND	ND	POS	ND	ND	ND	ND	ND
	OPCR	–	–	–	31.65/0.1672	–	–	–	–	–
	SNT	< 4	–	–	–	–	16	–	64	128
509	PT	ND	3.95	2.04	ND	ND	ND	ND	ND	ND
	PCR	ND	35.42/0.0117	32.56/0.0884	ND	ND	–	–	–	–
	OPT	ND	ND	ND	POS	ND	POS	ND	ND	ND
	OPCR	–	–	–	30.07/0.5114	–	28.45/1.5924	–	–	–
	SNT	< 4	–	–	–	–	16	–	64	64
510	PT	ND	4.04	2.17	ND	ND	ND	ND	ND	ND
	PCR	ND	ND	33.09/0.0607	ND	ND	–	–	–	–
	OPT	ND	ND	ND	ND	ND	ND	ND	ND	ND
	OPCR	–	–	–	–	–	–	–	–	–
	SNT	< 4	–	–	–	–	16	–	64	256

PT = plaque titration reported as log 10 median PFU per mL serum; PCR = serum culture isolate RVFV qRT-PCR result (Ct/copy number). Ct detection threshold = 0.018797; OPT = oronasal swab plaque titration reported as log 10 median PFU per ml virus media; OPCR = oronasal swab culture isolate RVFV qRT-PCR result (Ct/copy number). Ct detection threshold = 0.018797; SNT = serum neutralization titer to RVFV, reported as reciprocal of the dilution; ND = not detected; POS = virus detected, but titer too low to calculate.

Serum samples from all pigs on PID 0, 1, 2, 3, and 4 and all positive virus cultures were analyzed by qRT-PCR (Table 1). None of the serum samples were positive, even from pigs that had positive viral isolation titers. Viral cultures of serum samples were analyzed by qRT-PCR and two of the four piglets with detectable viremias on PID 1 had detectable RVFV nucleic acid. All of the three piglets with detectable viremias on PID 2 had detectable RVFV nucleic acid. No RVFV nucleic acid was detected by qRT-PCR from the oronasal swabs from the three piglets with evidence of oronasal viral shedding; however, viral nucleic acid was detected in viral cultures of these swab samples.

Clinical disease and pathology.

None of the animals developed clinical disease at any point in the study. Rectal temperatures were consistent throughout the study period (average 102.5, range 101.2–103.1). All piglets gained weight throughout the 14-day study with an average rate of gain of 0.2330 kg/day. There was no significant difference in the average rate of gain between infected and sham-inoculated piglets, or between viremic and nonviremic piglets. At necropsy, one of the inoculated piglets had a slightly larger lymph node on the left side (side inoculated) and more prominent follicular activity histologically. Another inoculated piglet had incidental pleural adhesions on the right cranial lung lobe. Histologically, two of the four viremic piglets had mild lymphoplasmacytic perivascular cuffing and multifocal glial nodules with vacuolation of the neuropil in the brain (Figure 1). No tissues of any piglet had detectable staining with RVFV IHC. Tissues from the two piglets with brain lesions were submitted for porcine circovirus-2 (PCV-2) IHC at the Iowa State University Veterinary Diagnostic Laboratory. No detectable staining for PCV-2 was detected in these sections.

Figure 1.

Histology of the brainstem from viremic piglets (503 and 510), H&E. (A) There was multifocal mild perivascular cuffing of grey matter vessels by lymphocytes and plasma cells (*). (B) There were focal glial nodules (†) with vacuolation of the neuropil and infiltration by lymphocytes and plasma cells. This figure appears in color at www.ajtmh.org.

In inoculated piglets, there was a decrease in average lymphocyte count on PID 1 followed by an increase on PID 3 that remained high until PID 7. The decrease in PID 1 was not significantly different from the count assessed on PID 0 by paired t test, but there was a significant difference between PID 0 and PID 3 (P = 0.0004), and between PID 1 and PID 3 (P = 0.0022). This trend of an initial decrease was seen in the two sham-inoculated animals as well; there was no significant difference between sham-inoculated and inoculated piglets by ANOVA. All piglets, inoculated and sham-inoculated, experienced a decrease in hematocrit between PID 0 and 7 (P < 0.0001). All other parameters assessed had no significant differences between PID 0 and 7 or between piglets.

DISCUSSION

Previous experimental infections and historical data suggested that young piglets would likely not experience severe clinical disease with RVFV infection as is seen in young ruminants.^1,4 This was confirmed in the present trial, in which none of the infected animals developed a febrile response, clinically evident disease, or a decreased rate of weight gain. Inoculated piglets did develop a brief lymphopenia following infection followed by a steady rise in lymphocyte counts. Lymphocytosis in excited, young animals can be considered a normal physiologic response but also could indicate antigenic stimulation following viral infection.¹⁶ This is further supported by all inoculated animals developing antibodies against RVFV following infection. Inoculations in this study were performed without the mosquito vector, thereby excluding potential immunogenic factors associated with feeding that could influence viral transmission, invasion, and host immune response.⁴

Of the six piglets that were inoculated with RVFV, four developed a brief 1–2-day viremia detectable by viral culture of serum samples. RVFV nucleic acid was detected from several of these cultured samples. All positive viral cultures were secondarily tested using the VectorTest RVFV antigen assay.¹⁵ Additionally, RVFV was isolated from the oronasal swabs of three of the four viremic animals, indicating a potential for spread by direct contact or droplets. In this study, the in-contact sham-inoculated piglets did not develop evidence of a contact acquired infection.

Lymphoplasmacytic encephalitis with glial nodules was observed in two viremic animals. While RVFV was not detected in affected tissues by IHC or by qRT-PCR, this inflammatory response may represent a subacute phase of a recent viral infection, potentially at the same time as the observed viremia. Similar histologic findings are found with other neurotropic porcine viruses, including porcine reproductive and respiratory syndrome virus types 1 and 2, porcine circovirus 2, suid alphaherpesvirus 1, teschovirus A, sapelovirus A, and porcine astrovirus.¹⁷ The route of infection may also influence the development of fulminant neurologic disease with RVFV, as has been demonstrated in experimentally infected mouse models that developed fatal encephalitic disease following intranasal inoculation, but subclinical infections with robust immune responses following subcutaneous inoculations.¹⁸

The development of viremia and viral shedding in a subclinical animal is potentially significant from an epidemiologic perspective in that it indicates the possibility for pigs, both domestic and feral, to spread RVFV without detection by existing surveillance systems. Future work to further investigate these possibilities would involve using lower infecting doses and introducing the mosquito vector, which may influence resulting viremic and serologic titers. There is evidence of mosquitoes that feed on infected animals having higher infection rates than mosquitoes that were membrane fed, reinforcing the concept that the insect vector is a crucial component in assessing arboviral infectivity and pathogenicity.¹⁹ The reliability of serologic titer development following exposure in this species would allow for its further assessment as an indication of RVFV exposure. Additionally, given the small number of animals included in this pilot study, further analysis is needed to determine the probability of mosquito infectivity at different viremias,^19,20 as well as the probability of transmission from pig to pig or pig to ruminant via oronasal shedding.

There is discrepancy between the viral culture and qRT-PCR methods used to detect RVFV in this study. Previous studies comparing diagnostics generally recommend that a combination of methods be used.^21,22 Ultimately, ideal testing modalities may vary from species to species, depending on the development of pathologic lesions and the caliber of viral titer expected. Based on the results of this study, we recommend a combination of qRT-PCR, viral isolation, and antigen assays¹⁵ when testing in swine.

In summary, young piglets can develop a brief viremia and seroconversion following inoculation with a high viral dose (10⁵ PFU) of RVFV but do not experience clinical disease as seen in more susceptible ruminant species. Pigs may, however, be a source of virus for mosquitoes or in-contact animals, potentially spreading the virus to more susceptible species populations.