Archives of Virology

, Volume 155, Issue 10, pp 1625–1630

Co-circulation of two extremely divergent serotype SAT 2 lineages in Kenya highlights challenges to foot-and-mouth disease control

Authors

    • Molecular Biology Laboratory, Institute of Environment and Natural ResourcesMakerere University
    • Foot-and-Mouth Disease Laboratory
  • G. J. Belsham
    • National Veterinary InstituteTechnical University of Denmark
  • V. B. Muwanika
    • Molecular Biology Laboratory, Institute of Environment and Natural ResourcesMakerere University
  • R. Heller
    • Department of BiologyUniversity of Copenhagen
  • S. N. Balinda
    • Molecular Biology Laboratory, Institute of Environment and Natural ResourcesMakerere University
  • H. R. Siegismund
    • Department of BiologyUniversity of Copenhagen
Original Article

DOI: 10.1007/s00705-010-0742-9

Cite this article as:
Sangula, A.K., Belsham, G.J., Muwanika, V.B. et al. Arch Virol (2010) 155: 1625. doi:10.1007/s00705-010-0742-9

Abstract

Amongst the SAT serotypes of foot-and-mouth disease virus (FMDV), the SAT 2 serotype is the most widely distributed throughout sub-Saharan Africa. Kenyan serotype SAT 2 viruses have been reported to display the highest genetic diversity for the serotype globally. This complicates diagnosis and control, and it is essential that patterns of virus circulation are known in order to overcome these difficulties. This study was undertaken to establish patterns of evolution of FMDV serotype SAT 2 in Kenya using complete VP1 coding sequences in a dataset of 65 sequences from Africa, collected over a period of 50 years. Two highly divergent lineages were observed to co-circulate, and occasional trans-boundary spread was inferred, emphasizing the value of constant monitoring and characterization of field strains for improved diagnosis and appropriate vaccine application as well as the need for regional approaches to control.

Introduction

Foot-and-mouth disease (FMD) is one of the most economically important infectious diseases of livestock. It is a vesicular disease of domesticated and wild cloven-hoofed animals. The agent of the disease, FMD virus (FMDV), is a single-stranded, positive-sense RNA virus in the genus Aphthovirus, family Picornaviridae. The genome of approximately 8,400 nt encodes a single polyprotein, which is cleaved into four structural (VP1, VP2, VP3, VP4) and several non-structural proteins [10]. FMDV exists as seven immunologically distinct serotypes known as O, A, C, Asia 1 and the Southern African Territories (SAT) 1, SAT 2 and SAT 3. The SAT serotypes are endemic to sub-Saharan Africa, with the African buffalo (Syncerus caffer) reported to play an important role in their epidemiology in Southern Africa [7, 26]. SAT 2 is the most widely distributed of these serotypes throughout sub-Saharan Africa, with viruses in Kenya reported to be the most heterogeneous and having the highest genetic diversity [5].

The history of confirmed SAT2 FMDV in Kenya dates back to 1956, with outbreaks due to this serotype becoming prominent from the late 1960s. It is currently, together with serotype O, the most prevalent serotype in Kenya [14, 29] (records of the Embakasi FMD laboratory). Like type O, SAT 2 outbreaks have been distributed throughout Kenya, with many antigenic variants observed, which has been reflected in the greatest change of vaccine strains used to control FMD outbreaks since vaccination was introduced in the 1950s [14] (records of the Embakasi FMD laboratory). The current FMDV SAT 2 vaccine strain is K52/84, and previous vaccine strains have included Ken 3/57, Tan 5/68, K183/74, R1215 [2] and K65/82. The introduction of new vaccine strains has been undertaken whenever significant antigenic variants were observed among field isolates.

Serotype SAT 2 outbreaks in Kenya have mostly been reported in cattle,and little is known about its circulation patterns in the country, although African buffaloes have been reported to be carriers [1], and serological evidence indicates high prevalence rates among buffalo populations [6]. The high levels of genetic sequence diversity reported for SAT 2 in Kenya complicates diagnosis and vaccination, since the development of suitable serotype-specific primers to identify all field isolates and the selection of appropriate vaccine strains may be quite difficult [5].

The VP1 coding region of FMDV has been extensively analysed to yield valuable molecular epidemiological information [22]. This study was undertaken to establish patterns of evolution of serotype SAT 2 in Kenya using complete VP1 coding sequences from 36 Kenyan isolates in a dataset of 65 sequences from Africa during the period 1948–2007. We specifically infer divergence times and evolutionary rates using genealogy-based coalescent methods.

Materials and methods

Virus isolates

Thirty-two (31 Kenyan and 1 Ugandan) SAT 2 virus isolates for this study, collected between 1981 and 2007 from cattle, were obtained from the Embakasi FMD Laboratory, Nairobi, Kenya. Virus was isolated from clinical material according to standard procedures using a single passage in baby hamster kidney (BHK) cells. The details of the isolates are shown in Table 1. Furthermore, to be able to put the phylogeny into a continental SAT 2 context, 33 other complete VP1 coding sequences representing topotypes defined in other studies [5, 12, 19] and available in GenBank, covering a sampling period from 1948 to 2007, were included in the study.
Table 1

List of the SAT 2 virus isolates included in this study

Isolate code

Date of isolation

District/country

Accession no.

RHO/1/48

1948

Rhodesia

AY593847

KEN/3/57a

1957

Samburu, Kenya

AJ251473

SA/106/59

1959

South Africa

AY593848

KEN/11/60

1960

Kenya

AY593849

MAL/3/75

1975

Malawi

AF367099

ANG/4/74

1974

Angola

AF479417

SEN/5/75

1975

Senegal

AF367140

UGA/51/75

1975

Uganda

AY343963

SUD/6/77

1977

Sudan

AY343939

MOZ/1/79

1979

Mozambique

AF367137

K81/81

1981

Laikipia, Kenya

This study

K46/82

1982

Trans Mara, Kenya

This study

K65/82a

1982

Kiambu, Kenya

This study

ZAI/1/82

1982

Zaire

AF367100

K70/83

1983

Kajiado, Kenya

This study

U267/83

1983

Uganda

This study

K151/83

1983

Nakuru, Kenya

This study

PAL/5/83

1983

South Africa

AF367102

ZIM/7/83

1983

Zimbabwe

AF136607

K34/84

1984

Nakuru, Kenya

This study

K52/84b

1984

Nakuru, Kenya

This study

KEN/1/84

1984

Uasin Gishu, Kenya

AY344505

KEN/2/84

1984

Nakuru, Kenya

AY343941

K37/86

1986

Kiambu, Kenya

This study

K13/87

1987

Kajiado, Kenya

This study

KNP/19/89

1989

South Africa

AF367110

ETH/1/90

1990

Ethiopia

AY343935

GHA/2/90

1990

Ghana

AF479415

K40/90

1990

Kiambu, Kenya

This study

K14/91

1991

Kirinyaga, Kenya

This study

BUN/1/91

1991

Burundi

AF367111

K32/92

1992

Kajiado, Kenya

This study

KNP/1/92

1992

South Africa

AF367114

KNP/132/92

1992

South Africa

AF367115

K3/93

1993

Kajiado, Kenya

This study

ZAM/10/93

1993

Zambia

AF367117

K5/94

1994

Nakuru, Kenya

This study

K25/94

1994

Kajiado, Kenya

This study

K37/94

1994

Nakuru, Kenya

This study

K37/95

1995

Nairobi, Kenya

This study

K39/95

1995

Kajiado, Kenya

This study

K77/96

1996

Nakuru, Kenya

This study

ZAM/10/96

1996

Zambia

AF367121

BOT/18/98

1998

Botswana

AF367123

ERI/12/98

1998

Eritrea

AF367126

NAM/286/98

1998

Namibia

AF367127

NAM/304/98

1998

Namibia

AF367129

UGA/19/98

1998

Uganda

AY343969

ZIM/267/98

1998

Zimbabwe

AF367130

K49/99

1999

Trans Mara, Kenya

This study

KEN/7/99

1999

Kiambu, Kenya

AF367132

RWA/1/00

2000

Rwanda

AF367134

ZIM/1/00

2000

Zimbabwe

AF367136

K13/02

2002

Kericho, Kenya

This study

K120/04

2004

Kiambu, Kenya

This study

K70/05

2005

Nairobi, Kenya

This study

K6/06

2006

Laikipia, Kenya

This study

K12/07

2007

Nyandarua, Kenya

This study

K15/07

2007

Narok, Kenya

This study

K17/07

2007

Kiambu, Kenya

This study

K20/07

2007

Kajiado, Kenya

This study

K42/07

2007

Kericho, Kenya

This study

K59/07

2007

Trans Mara, Kenya

This study

K67/07

2007

Nakuru, Kenya

This study

ETH/2/07

2007

Ethiopia

FJ798161

The GenBank accession numbers for the 32 isolates generated in this study are; HM623678 to HM623709

aPrevious vaccine strain

bCurrent vaccine strain

Viral RNA extraction, cDNA synthesis and amplification

Total RNA was extracted and cDNA synthesized as described previously [3]. The complete VP1 coding region was amplified using a forward primer from the VP3 coding sequence (5′-TTAACTACCACTTCATGTACAC(C/G)G-3′) and the reverse primer FMD-2A34 [11] using PCR reagent volumes and conditions described previously [3], yielding a product of ~1,050 bp. PCR products were visualized, purified and cycle-sequenced employing the same primers as for the PCR amplification, and the other procedures were performed as described previously [3].

Sequence analysis

The sequences that were generated were initially assembled using the Sequencher software 4.8 (Gene Code Corporation, USA). Multiple alignments of the whole dataset were carried out using MUSCLE by log-expectation comparison incorporated within the software program Geneious version 4.6 [9]. Models of evolution for phylogenetic analysis were determined by the hierarchical likelihood ratio tests (LRT) using PAUP* version 4b10 [24] and MrModeltest version 2.2 software [16]. The GTR [18] model with gamma-distributed rates among sites and a proportion of invariable sites was used to co-estimate the phylogenetic relationships, evolutionary rates and demographic histories by applying a Bayesian Markov Chain Monte Carlo (MCMC) method implemented in the BEAST software version 1.4.8 package [8] (http://beast.bio.ed.ac.uk). Preliminary analysis involved testing different demographic models/coalescent priors and the appropriateness of a strict clock versus various versions of relaxed clocks available in BEAST [23]. The constant population size model with an uncorrelated exponential clock was used, and the MCMC chain lengths were established as sufficient by viewing the runs in Tracer software version 1.4 (http://tree.bio.ed.ac.uk/software/tracer/). Uncertainty in the data was reflected in the 95% highest posterior density (HPD) intervals. A maximum clade credibility tree was obtained using Tree Annotator program in BEAST and visualized with FigTree version 1.1.2 software (http://tree.bio.ed.ac.uk/software/figtree/).

To infer amino acid variation in the VP1 coding region in the dataset, the amino acid sequence alignment was obtained using MEGA version 4 software [25]. Evolutionary divergence and mean diversity were also estimated using the Maximum Composite Likelihood method in MEGA4.

Results

Phylogenetic relationships

Complete VP1 coding sequences of 32 (31 Kenyan and one Ugandan) FMDV serotype SAT 2 isolates were generated in this study. In total, 65 (36 Kenyan and 29 from the rest of Africa) VP1 sequences including those from published reports were used to infer phylogenetic relationships, divergence times and evolutionary rates. The maximum clade credibility tree is shown in Fig. 1, with the posterior probabilities for significant branches indicated. Two main clades of Kenyan serotype SAT 2 viruses with more than 20% sequence divergence were identified, with one clade (lineage V, [5]) consisting of viruses from 1956 up to the 1990s and the other clade (lineage I, [5]) consisting of viruses from the 1980s and onwards. The Kenyan viruses of the two clades both have a countrywide distribution. The Kenyan lineage V viruses were related to viruses circulating in the horn of Africa, the Sudan, East Africa (Uganda) and central Africa [5, 19]. The viruses in lineage I (Fig. 1) grouped into two distinct sub-clades, labeled here as IA and IB. Sub-clade IA had exclusively Kenyan viruses from the mid-1980s to the present day, while sub-clade IB had Kenyan viruses from the early 1980s to 2005 together with viruses circulating in the Horn of Africa and East and Central Africa [5].
https://static-content.springer.com/image/art%3A10.1007%2Fs00705-010-0742-9/MediaObjects/705_2010_742_Fig1_HTML.gif
Fig. 1

Maximum clade credibility tree of SAT 2 viruses based on complete VP1 coding sequences inferred using BEAST, assuming a constant size coalescent prior showing lineage divergence since the most recent common ancestor (MRCA). The time axis is shown in years and ranges from the MRCA to the present year. Branches with posterior probabilities ≥50% are labeled

The estimated divergence time from the MRCA was ~350 years before present (ybp) (95% HPD: 170–615). The overall mean substitution rate was 2.42 × 10−3 substitutions/site/year (95% HPD: 1.75 × 10−3–3.12 × 10−3). Branch rate heterogeneity was high, as shown by the coefficient of variation >0.9 obtained using the relaxed molecular clock. Since it has been suggested that recombination may be responsible for non-molecular-clock-like evolution in datasets [20, 21], we tested for evidence of recombination within the VP1 coding region and found none, as determined by the GARD [17] method on the Datamonkey server as well the exploratory methods implemented in RDP 2 beta 0.8 software [13].

Distribution of mutations

Of the 648 nucleotide sites characterized for the whole dataset, a total of 239 (37%) were invariant. At the amino acid level, 92 amino acids (43%) were completely conserved (Fig. 2). Variation was observed in the known hypervariable regions (GH loop and the C-terminus) of VP1, although the RGD motif at amino acid residues 144–146 as well as the cysteine residue at the base of the GH loop (position 134) were completely conserved, in agreement with previous reports [5, 19]. At the VP1/2A cleavage site, the amino acid sequence KQ/LC predominated, although other sequences, including KQ/LL, were observed among the Kenyan isolates.
https://static-content.springer.com/image/art%3A10.1007%2Fs00705-010-0742-9/MediaObjects/705_2010_742_Fig2_HTML.gif
Fig. 2

Sequence alignment of 214 amino acids of the VP1 coding region (plus the first two amino acids of 2A) of 23 SAT 2 FMD viruses. A ‘.’ indicates an amino acid residue identical to that of the sequence K139/81

The net evolutionary divergence among the Kenyan isolates was 23%.

Discussion

The phylogenetic analysis of serotype SAT 2 in this study identified two very different clades of viruses circulating within Kenya that appear to have diverged about 350 years ago. Lineage V is comprised of the earliest reported SAT 2 viruses e.g. KEN/3/57 and KEN/11/60 and seems to have become extinct in the mid-1990s with the last representative isolate being K77/96. This lineage had a wide distribution, as shown by representative isolates from districts in different provinces of the country, e.g. KEN/3/57, K32/92 and K5/94 (Rift Valley province); K65/82 and K14/91 (Central province) and K37/95 (Nairobi province). This countrywide distribution is perhaps a reflection of the extensive livestock movements across the country for trade and grazing [15]. These viruses were related to viruses circulating in the horn of Africa, e.g. ETH/2/07, the Sudan e.g. SUD/6/77, East Africa (Uganda e.g. UGA/19/98) and central Africa e.g. ZAI/1/82, indicative of cross-border introductions and spread. Two previous vaccine strains, KEN/3/57 and K65/82, belonged to this clade, and the apparent disappearance of this lineage may perhaps be a reflection of the success of vaccination in controlling these outbreaks. The inclusion of additional Kenyan sequences covering a wider temporal space has helped to better resolve the genetic relationships of viruses in this grouping identified previously as lineage V: topotypes H, I and J, which were only represented in earlier studies by a limited number of isolates [5].

Lineage I consists of two sub-clades i.e. IA and IB (Fig. 1). While sub-clade IA contains exclusively Kenyan viruses from the mid-1980s to the present day, IB comprises Kenyan viruses from the early 1980s to 2005 together with viruses circulating in the Horn of Africa (ETH/1/90), East Africa (U267/83) and southern Africa (MAL/3/75). Lineage I appears to be related to viruses in southern Africa, in agreement with previous findings suggesting introductions northwards into eastern Africa.

The widespread distribution of very divergent virus strains during the same period was demonstrated by the co-existence of the two lineages in one district, e.g. K5/94 (lineage V) and K37/94 (lineage I), found in Nakuru district in 1994. The identification of two very divergent clades circulating sympatrically for at least dozens of years (and possibly hundreds) without going extinct or outcompeting each other is interesting and highlights the extreme genetic diversity of FMD virus lineages persisting today [5]. This diversity may be a reflection of the complex epidemiological history of FMD and its persistence across the livestock and wildlife interface. Serotype SAT 2 of FMDV has been reported to be prevalent among the African buffalo populations in the Kenyan wildlife sanctuaries where interaction with livestock is frequent [1, 6]. A recent study in Uganda found serological evidence for a high prevalence of SAT 2 FMDV infection amongst buffalo in Ugandan National Parks (Ayebazibwe C and others, unpublished results). However, little is known so far about the role of this interaction in the pattern of viruses circulating in Kenya. Different rates of nucleotide substitution have been observed in SAT 2 virus groups recovered from cattle and wildlife, which perhaps is reflected in the observed different epidemiological roles of these hosts [4, 28]. The nucleotide substitution rates within the SAT2 VP1 coding sequences observed in this study were similar to those reported elsewhere [27] and are higher than those observed for the SAT 1 serotype in eastern Africa (unpublished results by the authors).

The observations in this study emphasize the need for constant characterization of field strains, which is essential for efficient diagnosis and the selection of appropriate vaccine strains. It is apparent from this study that the inclusion of additional genetically related virus isolates from within a country permits a higher level of resolution of the relationships and circulation patterns. Hence, because of the small number of isolates from the other East Africa countries used in this study, more comprehensive studies of trans-boundary spread may require collection of additional samples. Nevertheless, the complex circulation patterns and the trans-boundary spread across the region observed in this study reaffirm the need for regional efforts towards FMD control.

Acknowledgments

We sincerely thank the Director of Veterinary Services, Kenya, for providing the virus isolates used in the study. Dr Sabenzia Wekesa is thanked for the information on the isolates. Teresa Kenduiywo, William Birgen and Eugene Arinaitwe are appreciated for technical assistance. Part of this work was carried out by using the resources of the Computational Biology Service Unit from Cornell University, which is partially funded by Microsoft Corporation. This work was supported by the Danish International Development Agency (DANIDA) under the Livestock-Wildlife Diseases in East Africa Project.

Copyright information

© Springer-Verlag 2010