Within two decades after its first report in 1954 from Pakistan, Foot and mouth disease has become endemic in the country and poses a serious threat to large as well as small ruminant population. Foot and Mouth Disease (FMD) is prevailing in cattle and buffaloes and is caused by either 0, A, Asia-i serotype of the FMD virus in Pakistan. The present study was undertaken to study the mutation rate of FMD virus and also molecular typing of the strains prevalent in Pakistan was done.

A total of 60 samples from buffalo and cattle were collected from five districts of Punjab including Lahore, Faisalabad, Sialkot, Okara and Sheikhupura. Soon after extraction of their RNA, all of them were reverse transcribed and then subjected to amplification by using different sets of the primers including universal as well as serotype specific primers. Then their VPI portions were amplified by using VP1 specific primers. Among 60 samples, 48 were positive with universal primers. Other 12 samples were not amplified with these primers hence not processed.

Among 48 FMD positive samples, 24 were positive with serotype 0 specific primers, 16 with serotype Asia-i and remaining 8 were positive with serotype A specific primers. After their amplification, the amplicons were run on the gel. These amplicons were extracted by using DNA extraction kit. After their purification, they were sent to Macrogen® (Seopl, Korea) and Centre of Excellence for Molecplar Biology, Pakistan (CEMB) for sequencing. Each amplicon was sequenced thrice and the consensus sequence was established eliminating sequencing errors.

Sequence identity and multiple sequence alignment of molecular sequences (nucleotide and amino acids) were performed with Clustal W algorithm (Thompson et al., 1994). Neighbour joining trees were constructed by using MEGA version 4.0 (Kumar et al., 2004). Nucleotide distance matrices were computed by Kimura two parameter algorithm based on the total nucleotide substitutions and evolutionary trees for VP1 genes were constructed.

For FMDV serotype '0' phylogenetic analysis, 14 VPI sequences from various field isolates were compared with some previously published Pakistani FMD 0 type VP1 specific sequences available with GeneBank and some recently published VP1 sequences reported by countries bordering with Pakistan including India, Iran and Afghanistan Similarly, 12 VP 1 sequences of FMDV serotype Asia-I isolates of this study were compared with previously published sequences and their phylogenetic relationship was established. However, the sequencing results of serotype A were inconclusive and were not included for phylogenetic analysis. Three sequences of three locally available FMD vaccines were also studied and compared with the outbreak strains.

Polymerase chain reaction was optimized with respect to MgCI2, buffer pH, annealing temperature, primer concentration, template concentration, and Taq polymerase. A concentration of 2.5 mM of MgCl2 resulted in the best amplification of the target sequences (Figure 1). The buffer with pH 8.8 yielded the best results (Figure 2) Although, the suggested annealing temperatures for various primers (of various serotypes) ranged from 48 °C to 63 °C, however, a temperature of 56 °C was found to be the best with all sets of primers (Figure 3). The best intensity DNA bands were observed with 0.3 pM concentration of the primers (Figure 4). Moreover, the best cDNA template concentration giving optimum amplification was found to be 3.0 p1 per reaction (Figure 5). Lastly, a concentration of 0.5 U of Taq polymerase was not sufficient for amplification of cDNAs, however, 1.0 U of enzyme was found to yield better amplification (Figure 6).

VP 1 DNA sequences of six previously published Pakistani FMD serotype 0 strains were analyzed phylogenetically with VP 1 DNA sequences of 14 isolates of the study. Serotype 0 isolates of this study distributed themselves into two distinct clusters (Figure 19). First cluster comprised of Sheikhupura 1 and 2, Muridkey 1, Raiwind 1, Nankana 1, Gujranwala 1 and Gujrat I isolates (Figures 19 and 20), whereas the second cluster included Depalpur 1, Sahiwal 1, Okara I, Multan 1, Toba 1, Faisalabad I and Pattoki 1 isolates (Figures 19 and 21). The first cluster was found to be associated with previously published Pakistani isolates of 2006 mostly. However, it also showed association with Afghanistan's isolates of 2004 (Figure 20). The second cluster seemed to be mostly related to previously published Pakistani isolates of 2003 (Figure 21). The overall grouping of the 14 sequences, when compared with each other, depicted a three clustered phylogram (Figure 22). Serotype 0 isolates from Depalpur, Sahiwal, Okara, Multan, Pattoki, Toba Tek Singh and Faisalabad grouped together into a clan and had more than 85% sequence similarity with each other. The second cluster consisted of isolates of Sheikhupura, Nankana, Raiwind and Muridkey. These sequences had more than 86% similarity with each other. The third cluster consisted of only two isolates which were 100 % similar to each other. However the third cluster had only 74 % sequence similarity to cluster I and 73 % sequence similarity when compared with cluster 2.

When the phylogenetic relationships with previously reported isolates of Asia 1 was evaluated, FMD Asia I isolates of this study were found to be scattered into two distinct groups (Figure 16). Group one consisted of isolates of Lodhran, Toba and Hafizabad that were more closely related to Indian isolates sharing more than 98% identity with each other and more than 94 % sequence identity with isolates of Indian 2001 to 2004 (Table 5 and Figures 16 and 17). However, they shared more than 86% sequence similarity with Pakistani isolates of 2002-2005 (Table 5). Group two comprised of isolates of kasur, Lahore, Pakpattan, Okara, Faisalabad, Jhang, Rahim Yar Khan, Bahawalpur and multan alongwith vaccine A and B (Figure 16). The isolates of group 2 were found to be closely associated with previously published isolates of Pakistani and Afghani origin of year 2003 and 2004 (Figures 16 and 18). Collectively, they shared an overall 70% sequence identity with each other. However, isolates of Bahawalpur, Rahim Yar Khan and Multan shared more than 98% similarity with each other, a measurement of close relationship denoting a likely common origin as one clan or dade. Similarly, isolates of Pakpatan, Faisalabad, Okara, Kasur, and Lahore shared 88% sequence identity with each other and qualified as one clade.

Although, overall amino acid sequence similarity of our isolates was not strikingly different from that of the published isolates, however, amino acid substitutions with dissimilar properties were found with a scattered pattern of distribution. For example, 15th amino acid residue which is hydrophilic in the previously published isolates had a substitution with a hydrophobic amino acid residue in our three isolates namely Sheikhupura 2, Muridkey I and Raiwind I (Figure 25). Similarly, 14th amino acid residue which is hydrophobic in nature was found to be replaced with a hydrophilic one in our last five isolates. Amino acid residue number 13 (Figure 25) had a substitution with a hydrophobic residue in some of our isolates etc. etc. It is interesting to note that such substitutions with amino acids having dissimilar properties have also been found, albeit at lower rate, in previously published sequences by many researchers (Figure 25).

A comparison of the deduced amino acid sequences in the critical VP I region of FMD serotype Asia I revealed that most of this study isolates shared very high homology with sequences of Vaccine A. However, the sequences of isolates of Lodhran, Hafizabad and Toba did not match much with that of either vaccines, A or B (Figure 23). Sequences of Vaccine A had a "K" which seemed to be replaced by a "T" in the sequences of most of the isolates. Considering the properties of various amino acids, this change does not signify a major shift in the three dimensional picture of the protein as K is a lysine, a positively charged amino acid, whereas a T is threonine, a hydrophilic amino acid in nature. Next substitution in most of the isolates is a "P" for "A" in comparison to the vaccines. Again, it is not a significant change as both P and A share the same property, hydorphobicity. Similarly a K with an R can be substituted without much change in the overall shape of the protein molecule. Next amino acid substitution is a leucine instead of methionine. Again both are hydrophobic in nature; hence their impact on the overall picture is minute, if at all. However, glycine and arginine are two very different amino acids; the former is a hydrophobic amino acid whereas the latter is positively charged one. Such amino acid substitutions may have the potential to make a major impact in terms of the epitopic differences in the capsids of vaccinal and field viruses. A comparison of the deduced amino acids of FMD serotype 0 isolates also exhibited such changes with the vaccinal virus (Figure 24).

Of the three hyper immune sera raised against three different vaccines in rabbits, only one vaccine induced a measureable immune response yielding good precipitation line against various FMD virus antigens.
In summary, RT-PCR for diagnosis of serotypes A, 0 and Asia 1 of FMDV was optimized and could be used for prompt and precise diagnosis of FMD in the country. Although, RT-PCR data pertains to bovines in the current project, but PCR optimization parameters are equally applicable to FMDV infections in other FMD susceptible animal species such as sheep and goat. The combination of PCR and sequencing of the VP1 gene to detect and analyze FMDV in disease outbreaks is fast (less than 6 hours for PCR and about 24 hours for sequencing), and it can give an accurate immunologic characterization of the virus, thus providing a rational basis for choice of vaccine. In fact, the molecular epidemiology of field isolates is a powerful tool to monitor the circulation of viruses (Saiz et al., 1993).
Secondly, various isolates of serotypes 0 and Asia 1 were sequenced along with some vaccinal strains. Sequence similarity tree analysis indicated that most of our isolates were closely related to previously reported Pakistani isolates and to those of neighboring countries such as India, Afghanistan and Iran. Additionally, amino acid sequence similarity data of major immunogenic site that forms 13G-13H loop in FMDV serotypes revealed that serotype Asia 1 vaccinal strain and Asia 1 isolates of this study possessed high degree of similarity suggesting a likely host immune response against the vaccine that may afford some protection against most field isolates of serotype Asia 1 type. Lastly, of three vaccines tested, only one was found to afford protection against field isolates of FMDV suggesting more work on vaccine issue in the country.