Abstract:
The efficacy of a vaccine is hindered by the antigenic polymorphism elicited by the parasite to
escape the host immune system. Thus, the present study for the first time characterized the
prevailing genetic diversity of two major asexual stage putative vaccine antigens of
Plasmodium vivax, the Apical Membrane Antigen-1 domain II (ama-lDIl) and the Merozoite
Surface Protein-142 (msp-142) in Sri Lanka, where unstable malaria prevails with low
transmission. Parasite isolates (N=217) were collected from patients infected with P. vivax
malaria from two malaria endemic areas, Anuradhapura and Kataragama, and from nonendemic
Colombo. The selected single clone isolates (N=169) identified by a combined
polymerase chain reaction/restriction fragment length polymorphism (PCR/RFLP) of the
Pvmsp-3a locus were amplified at the Pvama-IDII and Pvmsp-142 locus by the standard PCR
technique. Both the local and the worldwide genetic diversity of these two antigens were
analysed.
The ama-lDII gene in Sri Lanka showed meagre meiotic recombination with the enclosure of
single nucleotide polymorphisms (SNPs), where eleven amino acid (a.a) variant positions
defined 21 a.a. haplotypes with 10 unique to Sri Lanka. The HI a.a. haplotype predominant in
Sri Lanka was identical to the reference Salvador I strain, while haplotypes HI4 and HI6
were observed in Sri Lanka, India and Venezuela. Further, 145 globally dispersed isolates
defined 12 a.a. haplotypes (H22-H33), where 7 and 1 haplotypes were unique to India and
Venezuela, respectively. A novel observation where all a.a. sequences locally and globally
were identical at the domain II loop of PvAMA-1 was recorded. The evolutionary relationship
in the phylogenetic tree revealed fewer clustering where most isolates had a very recent
common origin. Evidence of reduced genetic diversity in Venezuela reflected the effects of
recent spread of the parasite to the new world, whereas those from the old world appeared to
reflect a very ancient selectively maintained polymorphism.
At the PvMSP-li9 locus, all a.a. sequences were identical. The analysis of 39 variant a.a.
positions upstream of PvMSP-l^ at the PVMSP-I33 region, documented several novel
observations: i) this region defined 27 a.a. haplotypes with 19 unique to Sri Lanka, ii) 24 of
the 27 PvMSP-142 haplotypes represented 7 basic a.a. sequence types at the hypervariable
region (HVR) while the remaining 3 were generated by interallelic double recombination, iii)
sequences from widely dispersed isolates in the database defined 62 more a.a. haplotypes, 43
of which corresponded to 9 of the 10 HVR types (excluding HVR-T7, unique to Sri Lanka),
iv) two novel HVR types, HVR Til and T12, with a double recombination were derived from
South America and Thailand, respectively, and v) intragenic recombination accounted for a.a.
haplotypes HVR-T3 to T7, and for the generation of H71-H89. T cell epitope polymorphism
arising due to non-synonymous substitutions at PVMSP-I33 may result in differential binding
of the polymorphic peptides to class II MHC alleles, and may evoke different host immune
responses. Hence, die extensive allelic polymorphism evident at Pvmsp-133 was due to
recombination, mutation and positive selection.
viii
This study for the first time reported 33 combined amino acid haplotypes that were generated
by the PvAMA-lDII and the PVMSP-I33 amino acid sequences that demonstrated the
extensive allelic polymorphism present in Sri Lanka.
Subsequently, to evaluate the ensuing strain-specific immunity arising due to extensive
antigenic polymorphism, amino acid sequences of PvAMA-lDII were aligned with the
homologous total (IgM+IgG) antibody responses assayed by an indirect enzyme linked
immunosorbant assay (ELISA) established in-house against seven PvAMA-lDII synthetic
peptides (P01-P07). This analysis was repeated for the a.a. sequences of PVMSP-I42, and
PvMSP-li9 with the previously published antibody responses assayed against PVMSP-I42, and
PvMSP-119 recombinant proteins. A strain-transcending (cross reactive) immune response was
clearly prevalent against peptide P07 of PvAMA-lDII and the PvMSP-142 protein.
Conversely, anti-P04peptide and anti-PvMSP-li9 antibody prevalence precluded strainspecific
immune response against the domain II loop of PvAMA-1 and PvMSP-119,
respectively. More importantly, both these regions clearly elicited a protective antibody
response where an isotype switch was observed from a primary IgM response to a
functionally important cytophilic IgG with increasing exposure to malaria in endemic
residents.
A multi-component vaccine containing a cocktail of parasite antigens would be the final goal
of vaccine development against the asexual stages of Plasmodium. Thus, it is crucial to
identify multiple protective domains or epitopes from different antigens that are protective
against the natural malaria infections. Hence, this study signifies that the highly conserved
domain II loop of PvAMA-1 and the 19 kDa fragment of PvMSP-142 may be suitable as
veritable vaccine components to develop “protective” immunity against P. vivax.