Emergence of African Swine Fever in Sri Lanka, 2024

TL;DR

African swine fever was first detected in Sri Lanka in 2024, with the virus showing genetic similarity to strains in China and Hong Kong.

Contribution

This study reports the first confirmed case of African swine fever in Sri Lanka and identifies the genotype and geographic origin of the virus.

Findings

ASF was detected in Sri Lanka for the first time in 2024.

The virus was found to be closely related to strains circulating in China and Hong Kong.

Histopathology confirmed lymphocyte loss and ASFV antigen staining in infected pigs.

Abstract

African swine fever (ASF) continues to spread, threatening the global swine industry and endangered swine species. Sri Lanka is a tropical island situated south of India in the Indian Ocean. Here, we report the first detection of ASF in Sri Lanka. In September 2024, increased pig mortality was reported across the country, with initial confirmation of porcine reproductive and respiratory syndrome (PRRS). Despite vaccination for PRRS, the mortalities continued to increase and therefore, tissue samples collected from dead pigs were subjected to ASF real-time PCR. ASFV genomic material was detected in most of the samples. The real-time PCR-positive samples were then subjected to genotyping by partial genome sequencing. All p72 and p54 sequences were found to be aligned with ASFV genotype II viruses, and CD2v sequences were found to be aligned with ASFV serogroup 8 viruses. The real-time PCR-positive samples were inoculated onto primary porcine leukocytes for virus isolation, and a selected number of tissues collected from dead pigs were subjected to histopathology. Histopathological studies revealed widespread loss of lymphocytes together with inflammation and extensive staining of ASFV antigens in tissue samples. Hemadsorption (HAD)-positive isolates were obtained from seven clinical samples, and three of them were subjected to whole-genome sequencing. Phylogeographic analysis of the whole-genome sequences showed that the virus is closely related to ASFV strains circulating in China and Hong Kong.

Figures

Full text

1. Introduction

African swine fever (ASF) is a contagious fatal hemorrhagic fever of European domestic and Eurasian wild pigs [1]. It is caused by the ASF virus (ASFV), a large, enveloped, double-stranded DNA virus belonging to the family Asfarviridae [2]. Based on a 478 bp fragment corresponding to the C-terminal end of the p72 gene, all known ASFV strains have been classified into 24 genotypes [3]. All 24 ASFV genotypes have been reported in Sub-Saharan Africa, where ASFV is maintained by a sylvatic cycle of infection between warthogs and Ornithodoros soft ticks [4]. However, it is noteworthy that genotype XVIII has been retired as an archived virus [5].

In June 2007, a highly virulent ASFV genotype II was reported in domestic pigs near the Black Sea port of Poti in the Republic of Georgia [6]. The virus quickly spread across Georgia and into Armenia. By the end of 2007, wild boars in Chechnya and the Russian Federation became infected with the virus and slowly but steadily, spread to the neighboring countries in Eastern Europe [7]. In August 2018, ASFV was detected in domestic pigs in China [8]. The virus rapidly spread across China and neighboring countries, including Vietnam, Cambodia, Laos, Philippines, Myanmar, Timor-Leste, Indonesia, Malaysia, and Thailand, causing severe economic losses to their pig industries [9]. In May 2020, India reported its first ASF outbreaks in the Assam and Arunachal Pradesh states [10]. Subsequently, the virus spread to several Indian states including Southern states such as Kerala, Karnataka, and Tamil Nadu [9,11]. ASF has already spread to free-ranging wild boars, threatening native and endangered pig species in India [11,12,13].

Sri Lanka is a tropical island situated in the Indian Ocean, 54 km south of India. It is 445 km from North to South and 225 km from East to West. In Sri Lanka, swine farms are concentrated in Western Province (Colombo, Gampaha, and Kalutara districts) and the adjacent Puttalam district, together traditionally called the “pig belt.” Swine farming provides the primary source of income for approximately 7000 farmers, and it is considered a highly profitable livestock sector gaining increasing popularity in Sri Lanka. In 2023, the total pig population in Sri Lanka was 170,409 [14]. Most of the swine farms in Sri Lanka are small (<50 pigs) to medium in scale (51–100 pigs) and practice swill feeding as commercial animal feed is costly [15]. Swill is collected from many sources, including hotels, restaurants, and institutional canteens. Feeding uncooked swill has been blamed for the past porcine reproductive and respiratory syndrome (PRRS) outbreaks in Sri Lanka [15] and is considered a risk factor and source for the introduction of multiple swine pathogens, including ASFV. In September 2024, an unusually high number of pig mortalities was observed in the country, which was initially confirmed as caused by the PRRS virus (PRRSV). Despite vaccination efforts for PRRS, the pig mortalities increased drastically, causing the local authorities to suspect an outbreak of swine hemorrhagic fevers.

Here, we report the clinical observations, detection, confirmation, pathological lesions, and genetic characterization of the virus responsible for the first outbreak of ASF in Sri Lanka.

2. Materials and Methods

2.1. Samples

In September 2024, an increased number of deaths in pig farms across the country was reported. PRRS was confirmed, and the pigs in the affected areas were vaccinated against PRRS virus (PRRSV). However, it was observed that despite PRRS vaccination, more pig farms were affected. Furthermore, deaths of wild pigs were also reported in Western Province. The gross lesions (splenomegaly and hemorrhages in internal organs) observed in dead pigs were suggestive of ASF (Figure 1); therefore, tissue and blood samples collected from dead/dying domestic and wild pigs in October with archived samples collected in September were sent to the Animal Virus Laboratory (AVL), Veterinary Research Institute, Polgolla, Sri Lanka.

2.2. Real-Time PCR

At the AVL, Polgolla, Sri Lanka, nucleic acid was extracted from whole-blood and tissue samples using the PowerPrep^TM^ Viral DNA/RNA extraction Kit Ver 1.0 (Kogenbiotech, Seoul, Reoublic of Korea, cat no: E0016) and subjected to a commercially available ID Gene™ African Swine Fever Duplex PCR assay (Innovative Diagnostics, Grabels, France) on a qTower3G real-time PCR instrument (Analyticjena, Jena, Germany). At the NCFAD, nucleic acid was extracted from whole-blood and tissue samples using 5X MagMax^TM^ Viral/Pathogen Nucleic

Acid Isolation kit utilising the KingFisher^TM^ Flex Purification system (Thermo Fisher Scientific, Waltham, MA, USA). The extracted nucleic acid was then subjected to a Tignon quantitative real-time PCR assay, which targets a highly conserved region of the ASFV p72 open reading frame [16]. A real-time PCR assay specific to β-actin (Moniwa assay) was performed to verify efficient nucleic acid extraction and amplification [17]. TaqMan™ Fast Virus 1-Step Master Mix (Thermo Fisher Scientific) was used to prepare reactions, and amplification was conducted using the Bio-Rad CFX96 instrument (Bio-Rad, Mississauga, ON, Canada), using recommended thermal cycling conditions for the TaqMan™ Fast Virus 1-Step Master Mix (50 °C for 5 min, 95 °C for 20 s, followed by 40 cycles of denaturation at 95 °C for 3 s and annealing/extension at 60 °C for 30 s). A positive control well was included, a regression analysis of the positive control was performed to calculate a threshold cutoff for all experimental wells, and cycle threshold (Ct) values were determined for each sample.

2.3. Conventional PCR

Conventional PCR was used to confirm the positive real-time PCR results, according to the NCFAD protocol. Briefly, the full-length p72, p54, and CD2v genes and the variable CVR and IGR regions were amplified from real-time PCR-positive samples using LongAmp^®^ Taq 2X Master Mix (New England Biolabs, Ipswich, MA, USA) before Nanopore sequencing (Table 1). The primers ASF_1500_FOR and CVR2 were used to amplify the full-length p72 gene and the adjacent CVR region in the B602L gene. The cycling conditions were 94 °C for 30 s initial denaturation, 40 cycles of 94 °C for 30 s, 55 °C for 30 s, and 65 °C for 4 min, followed by a final extension at 65 °C for 10 min. For amplifying the full-length p54 gene, the primers ASF_GENO_PPA89 CISA and ASF_GENOp_PPA722 CISA were used with the following conditions: 94 °C for 30 s, 40 cycles of 94 °C for 30 s, 54 °C for 30 s, and 65 °C for 1 min, followed by a final extension at 65 °C for 10 min. To amplify the IGR, primers ECO1A and ECO1B were used with the following cyclic conditions: 94 °C for 30 s initial denaturation, 40× 94 °C for 30 s, 59 °C for 30 s, and 65 °C for 30 s, followed by a final extension at 65 °C for 10 min. Amplicons were run on agarose gels and purified using SPRI select beads (Beckman Coulter, Brea, CA, USA).

2.4. Amplicon Sequencing and Analysis

For amplicon sequencing, the rapid barcoding kit SQK-RBK114.96 (Oxford Nanopore Technologies, Oxford, UK) was used and the library preparation was done following manufacturer’s instructions. Sequencing was done on a minION Mk1B using minION flow cells (Both from Oxford Nanopore Technologies) with barcode trimming and high-accuracy base calling. Analysis was done on Geneious Prime 2024.0 (https://www.geneious.com/ accessed on 23 December 2025), in which the reads were first trimmed using BBDuk, a Geneious Prime plugin for trimming reads (https://www.geneious.com/plugins/bbduk accessed on 23 December 2025), then mapped to an in-house database that contains sequences from the NCBI of each of the genomic regions of interest.

2.5. Virus Isolation

Attempts to isolate live virus were made on samples with Ct values of less than 30. NCFAD’s standard(African swine fever: Virus Isolation by Inoculation of Primary Pig Leucocyte Cultures, Version 4.0, October 4, 2024) operating protocol for ASFV isolation was followed. Briefly, whole-blood sample was collected from a healthy pig via heparin-coated blood collection tubes. Porcine primary leukocytes (PPL) were separated from the blood using 6% w/v dextran (Sigma-Aldrich Canada Co., Oakville, ON, USA) solution. The PPL were re-suspended at 10^6^ cells/mL in RPMI (Thermo-Fisher Scientific) supplemented with 5% fetal bovine serum (Thermo-Fisher Scientific), 1× Glutamax (Thermo-Fisher Scientific), and 5 mg/mL gentamicin (Gibco, Grand Island, NY, USA). Washed red blood cells (RBC) from the same animal were added to the leukocyte cultures at a final RBC concentration of 0.4% v/v. The leukocyte + RBC suspension was then plated in 24-well plates at 1 mL per well. Following 48 h at 37 °C in a 5% CO_2_ incubator, 10-fold dilutions of 10% tissue suspensions were inoculated into duplicate PPL wells (200 uL/well). The plates were incubated for 5–7 days at 37 °C in a 5% CO_2_ incubator and examined daily for the presence of hemadsorption (HAD) and cytopathic effects.

2.6. Whole-Genome Sequencing

Nucleic acid was extracted using the 5x MagMaxTM Viral/Pathogen Nucleic Acid Isolation kit and the KingFisher^TM^ Flex Purification system (Thermo Fisher Scientific). Nucleic acid was then used as input for Rapid PCR barcoding kit SQK-RPB114.24 (Oxford Nanopore Technologies) and sequenced on a minION R10.4.1 flowcell (Oxford Nanopore Technologies). For Illumina sequencing, a custom myBaits^®^ ASF target capture kit (Daicel Arbor Biosciences, Ann Arbor, MI, USA) was used to enrich the ASFV genome in the sample. The custom myBaits^®^ ASF target capture kit was designed based on 1614 ASFV sequences (19 sequences ranged from 170,101 bp to 193,886 bp, and the remaining 1595 sequences ranged from 84 bp to 55,098 bp). The libraries were prepared using Nextera XT DNA Library Preparation Kit (Illumina, San Diego, CA, USA). An Illumina MiSeq instrument was used for sequencing with a V2 flow cell with a 500-cycle reagent cartridge (Illumina). For assembly, Chopper was used to trim the nanopore reads with a quality cut off of 10 and a read length cut off of 200 [22]. Next, HoCoRt was used to specifically remove host DNA reads [23]—in this case, swine. Lastly, Flye was employed for de novo assembly [24]. The Illumina and nanopore reads were then mapped to the Flye-assembled genome to construct the final genome. Based on the mapped reads, ambiguous nucleotides were manually corrected.

2.7. Phylogenetic Analysis

Consensus sequences for each genomic region—p72, p54, and CD2v—were aligned separately using MAFFT v7.471 [25], with available partial sequences from NCBI. The alignments were then used to create the maximum-likelihood (ML) trees using IQ-Tree v2.3.4 [26], as well as FigTree v1.4.4 [27] to produce the final midpoint rooted trees.

A total of 95 ASFV genotype II genome sequences were retrieved from the NCBI’s GenBank. Laboratory-derived strains and those with incomplete sequence, location, or date metadata were excluded from the dataset, leaving a subset of 59 genomes. These data were combined with the whole-genome sequences generated in this study and aligned with MAFFT v7.471 using automatic algorithm detection. Whole-genome alignments were trimmed of the regions flanking the first and last open reading frames and used to infer a maximum-likelihood phylogenetic tree with IQ-TREE (v2.3.4). Phylogenetic trees were inferred using the best fitting model of nucleotide substitution, as determined by ModelFinder [28], and node support was generated with 5000 ultrafast bootstrap replicates. The bootstrap consensus tree was re-rooted on the best fitting root using augur-refine from the Augur bioinformatics toolkit [29]. Ancestral location estimates were reconstructed as discrete characters across all internal tree nodes using augur-traits from the Augur bioinformatics toolkit. Viral transmissions between discrete locations were inferred with StrainHub (https://strainhub.io accessed on 23 December 2025) and visualized with Spread.gl [30].

2.8. Histopathology and Immunohistochemistry

For histopathologic examination, tissues were fixed in 10% neutral phosphate buffered formalin, sectioned at 5 µm, and stained with hematoxylin and eosin (H&E). For IHC, paraffin tissue sections were quenched for 10 min in aqueous 3% hydrogen peroxide, and the epitopes were retrieved using Glyca Target Retrieval solution (made in-house) in a Decloaking Chamber™ (Biocare Medical, Pacheco, CA, USA). The monoclonal anti-ASFV antibody F88ASF-55 [31] was applied to the sections at a 1:500 dilution for 30 min. The sections were then visualized using EnVision+ Single Reagent HRP Mouse kit (Agilent Technologies Canada Inc., Mississauga, ON, Canada) and reacted with the chromogen diaminobenzidine + (DAB+) substrate (Agilent Technologies Canada Inc.). Finally, the sections were counter-stained with Gill’s hematoxylin (made in-house).

3. Results and Discussion

The samples collected in October with the signs suggestive of ASF were positive for ASFV genomic material according to real-time PCR at the AVL, Polgolla, Sri Lanka. The lung sample collected on 3 September 2024 from a dead pig in a medium-size family farm in Western Province was also tested positive for ASFV genomic material via real-time PCR. The farm was located close to many tourist hotels and restaurants in Uswetakeiyawa, a village famous for its beaches north of Colombo, the capital of Sri Lanka. Additional samples submitted from small and medium pig farms experiencing fever, anorexia, bleeding, respiratory distress, and increased mortality in Western and Uva Provinces also tested positive for ASFV genomic material.

These samples, together with additional tissue samples collected in late October and early November 2024 from seven different provinces in Sri Lanka (Figure 2), were sent to the World Organization for Animal Health (WOAH) Reference Laboratory for ASF and classical swine fever (CSF) at the National Centre for Foreign Animal Disease (NCFAD), Winnipeg, Canada, for confirmation. In addition to the fresh tissues, paraffin blocks of formalin-fixed tissues collected from a selected number of animals at the Faculty of Veterinary Medicine & Animal Science (FVMAS) were also sent to the NCFAD for histopathological examination.

At the NCFAD, all the samples were subjected to the Tignon real-time PCR, and all samples except two tested positive for ASFV genomic material (Table 2). Many of the farms affected were small to medium in scale in Western Province, and the disease appeared to quickly spread to domestic pigs in nearby provinces and to wild boars. The two negative samples came from a small pig farm in Uva Province and one of the wild boar bone marrow samples from Western Province. Three out of four wild boar samples tested positive for ASFV genomic material. The positive samples included a bone marrow and a lung sample collected from wild boars in Western and Sabaragamuwa Provinces and a bone marrow sample collected from a wild boar in Yala National Park. Yala National Park is the most visited and second largest national park, located in the southeastern region of Sri Lanka, spanning both the Southern and Uva Provinces.

From the real-time PCR-positive samples, full-length p72 gene (B646L), p54, CD2v, CVR, and IGR regions were amplified using conventional PCR. Conventional PCR for full-length B646L and CVR was successful in 20 samples, p54 in 19 samples, CD2v in 15 samples, and IGR in 20 samples. The amplicons were subjected to Nanopore sequencing. All p72 and p54 sequences from Sri Lanka aligned with genotype II viruses, and CD2v sequences aligned with serogroup 8 viruses (Figure 3). When B646L sequences were subjected to BLAST (Version 2.16.0) search, they closely matched with ASFV genotype II isolates from Russia (GenBank #KP843857) and China (GenBank #MW521382). All CVR sequences were BNDBNDBNA, and IGR sequences belonged to variant 2.

The samples with Ct values below 30 were inoculated onto PPL, and hemadsorption (HAD)-positive isolates were obtained from seven clinical samples. No virus could be isolated from wild pigs, most likely due to poor sample quality.

The formalin-fixed tissues were processed and subjected to hematoxylin and eosin (H&E) staining and immunohistochemistry (IHC) using the monoclonal antibody F88ASF-55. Widespread apoptosis of uninfected lymphocytes and inflammation were observed with H&E. ASFV structural protein pA137R was extensively stained and visible in all the tissues observed by IHC (Figure 4).

Two of the seven isolates, ASFV/SL/22/2024 (isolated from a whole-blood sample collected from a sick pig on 29 October 2024 from Udaperadeniya, Central province) and ASFV/SL/4/2024 (isolated from a lung sample collected from a dead pig on 26 September 2024 from Padukka, Western Province), were subjected to whole-genome sequencing and deposited in GenBank (accession numbers PV423386 and PV423385, respectively). For ASFV/SL/22/2024, the Illumina MiSeq generated 5,582,718 reads, and Nanopore GridION generated 877,200 reads. For ASFV/SL/4/2024, Illumina produced 2,133,768 reads and Nanopore generated 567,200 reads. Nanopore reads were trimmed, and low-quality reads were removed via Chopper. HoCoRT was used to remove swine DNA reads. The resulting reads were used for de novo assembly using Flye. Illumina and Nanopore reads were then mapped to the de novo consensus sequence to generate the final genome. ASFV/SL/22/2024 generated a 191,971 bp genome with a GC content of 38.3%. ASFV/SL/4/2024 generated a 191,972 bp genome with a GC content of 38.3%. These sequences were put into BLAST, and the top hits were ASFV/Timor-Leste/2019/1 (MW396979), with a percentage identity of 99.97% (100% coverage), and ASFV JS (OR180113), with a percentage identity of 99.96% (100% coverage).

A whole-genome phylogenetic tree revealed that the Sri Lankan-derived isolates are nested between those collected in China between 2018 and 2021 (Figure 5A). Chinese- and Sri Lankan-derived isolates cluster together, partly due to a shared deletion in a conserved genomic region (Figure 5C). Ancestral reconstruction of isolate locations supports a possible transmission of the virus from China to Sri Lanka (Figure 5B).

4. Conclusions

In this study, we report the detection, confirmation, and molecular characterization of ASFV genotype II in both domestic and wild pigs in Sri Lanka for the first time. The incursion of ASFV likely occurred through contaminated swill collected from the air/seaports or swill contaminated with illegally imported contaminated pork and/or pork products. Importation of pork and pork products from ASF-reported countries is banned in Sri Lanka. The disease has rapidly spread across the island, most likely due to poor biosecurity, the feeding of uncooked swill, and/or the movement of pigs and fomites between farms. Disposal of carcasses into water bodies by farmers and torrential rains and flooding may also have contributed to the aggressive spread of the virus via contaminated water. In Western Province, backyard farms are highly congested and located close to common water bodies and wildlife.

Sri Lankan wild pig, Sus scrofa cristatus, is widely spread from the mountains of the upcountry to the coastal areas of the island. Poor biosecurity in small- to medium-scale pig farms and the presence of ASFV in wild boars will make the control and eradication of ASF from Sri Lanka extremely challenging. Sri Lanka is home to the soft tick sand tampan (Ornithodoros savignyi), which is abundant in the country and feeds on various mammals, including pigs. This tick species is a proven vector for ASF [32]; therefore, the possible role of the sand tampan as a vector for ASFV in Sri Lanka, especially in the national parks, warrants further investigation.