Plasmodium falciparum Isolates in India Exhibit a Progressive Increase in Mutations Associated with Sulfadoxine-Pyrimethamine Resistance

ParasitesBlood from patients with fever who were attending malaria clinics in Delhi, Uttar Pradesh (UP), Assam, Goa, and Orissa were screened for malaria parasites by light microscopy after Geimsa staining. About 20 to 50 μl of heparinized blood was collected from those patients who were positive for P. falciparum parasites. Informed consent was obtained from the patients or their guardians, in the case of children, prior to blood collection. The ethical guidelines for blood collection of the Malaria Research Centre were followed.

Mutation-specific PCRParasite DNA was isolated from the clinical samples by a previously described method (10). The P. falciparum 720-bp fragment of the dhfr gene was amplified by using primers AMP-1 (5′-TTT ATA TTT TCT CCT TTT TA-3′) and AMP-2 (5′-CAT TTT ATT ATT CGT TTT CT-3′) (28). The cycling parameters used were as follows: denaturation at 94°C for 30 s, annealing at 45°C for 45 s, and extension at 72°C for 45 s. A total of 45 cycles were carried out. A mutation-specific nested PCR was performed with this primary PCR product to detect the nucleotides at five dhfr codons: codons 16, 51, 59, 108, and 164. Two separate sets of PCRs were carried out for each codon, one for the wild-type allele and one for the mutant allele. In the case of codon 108, three reactions were carried out: one for the wild-type allele and two for mutant alleles. The primers and their sequences were as follows: DIA-3, 5′-GAA TCC TTT CCC AGC-3′; DIA-9, 5′-GAA TCC TTT CCC AGG-3′; DIA-12, 5′-GGA ATG CTT TCC CAG T-3′; SP-1, 5′-ATG ATG GAA CAA GTC TGC GAC-3′; FR-59W, 5′-ATG TTG TAA CTG CAC A-3′; FR-59M, 5′-ATG TTG TAA CTG CAC G-3′; FR-51W, 5′-TTA CCA TGG AAA TGT AA-3′; FR-51M, 5′-TTA CCA TGG AAA TGT AT-3′; SP-2, 5′-ACA TTT TAT TAT TCG TTT TC-3′; DIA-13, 5′-CAA CGG AAC CTC CTA T-3′; DIA-14, 5′-CAA CGG AAC CTC CTA A-3′; DIA-15, 5′-TTT ATG CCA TAT GTG T-3′; DIA-16, 5′-TTA TGC CAT ATG TGC-3′; and SP-3, 5′-TTT AAT TTC CCA AGT AAA AC-3′ (4, 9, 12, 28, 34). Only 15 cycles of the nested PCR were carried out. The cycling parameters for the different sets of primers were as follows. For detection of the nucleotide at codon 16 with primers specific for either the wild type (primers DIA-16 and SP-2) or the mutant (primers DIA-15 and SP-3), denaturation was carried out at 94°C for 30 s, annealing was carried out at 50°C for 30 s, and extension was carried out at 72°C for 45 s. For detection of the nucleotide at codon 51 with primers specific for the wild-type sequence (primers FR-51W and SP-2) or the mutant sequence (primers FR-51M and SP-2), denaturation was carried out at 92°C for 30 s, annealing was carried out at 52°C for 30 s, and extension was carried out at 72°C for 30 s. For detection of the nucleotide at codon 59 with primers specific for the wild-type sequence (primers FR-59W and SP-1) or the mutant sequence (primers FR-59M and SP-1), denaturation was carried out at 92°C, annealing was carried out at 54°C for 30 s, and extension was carried out at 72°C for 30 s. For detection of the nucleotide at codon 108 with primers specific for the wild-type sequence (primers DIA-3 and SP-1) or the mutant sequence (primers DIA-9 and SP-1 for Thr or primers DIA-12 and SP-1 for Asn), denaturation was carried out at 94°C, annealing was carried out at 55°C for 30 s, and extension was carried out at 74°C for 30 s. For detection of the nucleotide at codon 164 with primers specific for the wild-type sequence (primers DIA-13 and SP-1) or the mutant sequence (primers DIA-14 and SP-1), denaturation was carried out at 94°C for 30 s, annealing was carried out at 54°C for 30 s, and extension was carried out at 72°C for 45 s.

For amplification of dhps, a primary PCR was performed to amplify a 1,287-bp fragment of P. falciparum DNA by using primers M3717F (5′-CCA TTC CTC ATG TGT ATA CAA CAC-3′) and 186R (5′-GTT TAA TCA CAT GTT TGC ACT TTC-3′ (38). Forty-five cycles were performed under the following conditions: denaturation at 94°C for 30 s, annealing at 55°C for 45 s, and extension at 72°C for 90 s. A final extension was performed at 72°C for 30 min. Mutation-specific nested PCRs were performed for detection of the nucleotides at codons S436F/A, A437G, K540E, A581G, and A613S/T (38, 39). Nested PCR was performed for 20 cycles under the following conditions: denaturation at 94°C for 30 s, annealing at 50°C for 45 s, and extension at 72°C for 80 s. A final extension at 72°C was carried out for 20 min. The following primer combinations were used: 436S Forward (185F; 5′-TGA TAC CCG AAT ATA AGC ATA ATG-3′) and 436S Reverse (252R; 5′-GGA TTA GGT ATA ACA AAA GGA GCT G-3′), 436F Forward (249F; 5′-GTT ATA GAT ATA GGT GGA GAA TCC AT-3′) and 436F Reverse (218F; 5′-ATA ATA GCT GTA GGA AGC AAT TG-3′). For detection of the 436A mutation, the primer set for wild-type codon 436S was 185F in combination with 252A Reverse (5′-ATT AGG TAT AAC AAA AGG AGC ATA-3′). The remaining primer combinations were 436A Forward (249A; 5′-GTT ATA GAT ATA GGT GGA GGA TCT G-3′) and 436A Reverse (218R), 437A Forward (185F) and 437A Reverse (PWVIR; 5′-TTT GGA TTA GGT ATA ACA AAA GGT G-3′), 437G Forward (PMVIF; 5′-GAT ATA GGT GGA GAA TCC TCT TG-3′) and 437G Reverse (218R), 540K Forward (C070F; 5′-GGA AAT CCA CAT ACA ATG GAA A-3′) and 540K Reverse (218R), 540E Forward (185F) and 540E Reverse (C071R; 5′-CTA GAT TAT CAT AAT TTG TTA GTA C-3′), 581A Forward (216F; 5′-CTA TTT GAT ATT GGA TTA GGA TTT TC-3′), and 581A Reverse (218R), 581G Forward (185F) and 581G Reverse (201R; 5′-AAT AGA TTG ATC ATG TTT CTT CC-3′), 613A Forward (233F; 5′-GGA TAT TCA AGA AAA AGA TTT ATA G-3′) and 613A Reverse (218R), 613S Forward (185F) and 613S Reverse (251R; 5′-TTT GAT CAT TCA TGC AAT GGC T-3′) and 613T Forward (185F) and 613T Reverse (226R; 5′-GAT CAT TCA TGC AAT GGG A-3′). The PCR products were checked on agarose gels.

Statistical analysisThe chi-square test with Yates' correction was applied to determine significant differences between two groups. A P value <0.05 was considered significant.

P. falciparum DHFR and DHPS mutationsA total of 312 P. falciparum-infected blood samples were analyzed for mutations in five codons of the dhfr gene (A16V, N51I, C59R, S108N/T, and I164L) and five codons of the dhps gene (S436F/A, A437G, K540E, A581G, and A613S/T) to assess the level of antifolate drug pressure in India. Among these 312 samples, 285 were PCR positive for dhfr and 234 were PCR positive for dhps. PCR successfully detected both genes in only 207 of 312 samples. The DHFR S108N mutation was most prevalent (89.47%), followed by the C59R mutation (70.17%), in the 285 samples. No isolate had the A16V or the S108T mutation, while the N51I and I164L mutations were found in 18 (6.31%) and 7 (2.45%) isolates, respectively. Thirty (10.52%) isolates had the wild-type sequences at the five DHFR codons. The rate of point mutations was higher in the DHFR sequence (89.47%) than in the DHPS sequence (47.44%). Among the 234 samples PCR positive for dhps, 123 (52.56%) had the wild-type sequences at all five codons. The maximum numbers of mutations were found at codons S436F (24.64%) and A437G (20.94%). Only 17 (7.26%) isolates had the K540E mutation, 10 (4.27%) isolates had the S436A mutation, 7 (2.99%) isolates had the A613T mutation, and 4 (1.7%) isolates had the A581G mutation. No isolate had the A613S mutation.

Seven different alleles in dhfr were observed among the 285 isolates (Fig. 1A). The wild-type DHFR sequence ANCSI was present in only 10.5% of the isolates, whereas the sequence with the double mutation ANRNI was highly prevalent (66%) and the sequence with the triple mutation ANRNL was the least common (1.0%) (mutated amino acids are indicated in boldface). Forty-two (14.7%) isolates had single mutations in the gene for DHFR, while double and triple mutations in the gene for DHFR were seen in 201 (70.5%) and 12 (4.2%) isolates, respectively. The number of different DHPS genotypes was greater than the number of different DHFR genotypes (Fig. 1B). Among the 15 different DHPS genotypes observed, the wild-type sequence (SAKAA) remained prevalent (52.6%) among the 234 isolates. Also, mutants with single DHPS mutations were more common (37.2%) than mutants with double (5.1%) or triple (5.2%) mutations. Among the four types of mutants with single DHPS mutations, the sequences FAKAA (21.8%) and SGKAA (13.3%) were more common than the others. The next most common DHPS sequence included a triple mutation, AGEAA (3.4%). Multiple mutations in the gene for either enzyme were not seen among the Indian isolates.

• Open in new tab
• Download powerpoint

FIG. 1.

Mutations in the loci for DHFR (A) and DHPS (B) of P. falciparum isolates from India. n, number of isolates analyzed. The amino acid sequence is shown at the top of each lane, where mutated amino acids are shown in boldface. The mutated codons are shaded.

The prevalence of the S108N mutation in the DHFR sequence and the S436F and A437G mutations in the DHPS sequence among Indian isolates indicate that these are the key point mutations. Any other mutation should be associated with them (18, 34, 39). However, there were exceptions in the case of DHPS, in which isolates were found to contain independent single K540E and A581G mutations as well as double A581G and A613T mutations (Fig. 1B).

Temporal rise in DHFR and DHPS mutationsThe samples were divided into two groups. Those collected during 1995 and 1996 were categorized into group A, and those collected 5 years later (during 2000 and 2001) were categorized into group B. The rates of mutations in the DHFR enzyme sequence in these two groups are shown in Fig. 2. The inset of Fig. 2 shows a significant increase (P < 0.05) in the number of group B isolates with C59R and S108N mutations. On the other hand, fewer isolates in group B had the N51I mutation (P < 0.05), while no significant difference in the rates of occurrence of the I164L mutation was observed between the two groups. Significantly more isolates in group B had double DHFR mutations, while fewer isolates in this group had the wild-type sequence or single DHFR mutations (Fig. 2). There was an insignificant increase in the number of isolates in group B with triple DHFR mutations. Significantly more (P < 0.05) isolates in group B than in group A had S436F, A437G, and K540E mutations in the DHPS sequence, while there was no significant difference in the rates of the other mutations between the two groups (Fig. 3, inset). Significantly fewer isolates in group B had the wild-type DHPS sequence, whereas more isolates in group B had single DHPS mutations (P < 0.05). Although double and triple DHPS mutations occurred more often in group B, the difference was not statistically significant from the rate of occurrence in group A (Fig. 3).

• Open in new tab
• Download powerpoint

FIG. 2.

Mutation rates in the P. falciparum DHFR enzyme sequence among two groups of isolates (groups A and B). A comparison of the rates of mutation in individual codons between the two groups is shown in the inset. The number of samples in each group is shown in parentheses. The chi-square test with Yates' correction was used to compare values between two groups (*, P < 0.05; **, not significant; ***, not applicable).

• Open in new tab
• Download powerpoint

FIG. 3.

Mutation rates in the P. falciparum DHPS enzyme sequence among two groups of isolates. A comparison of the rates of mutation in individual codons between the two groups is shown in the inset. The asterisks are explained in the legend to Fig. 2.

Among the seven different DHFR sequences shown in Fig. 1A, AICNI and ANCNL were present only in group A and ANRNL was present only in group B, while the other sequences were present in both groups (data not shown). Due to an increased number of DHPS mutations in group B (Fig. 3), the number of different sequences increased from 3 to 15 over the 5 years (data not shown). The DHPS sequence FGKAT was present only in group A, while all other sequences except wild-type sequence SAKAA and the FAKAT sequence, shown in Fig. 1B, were found only in group B. Wild-type sequence SAKAA and the FAKAT sequence with a double mutation were common in both groups.

Regional bias in P. falciparum DHFR and DHPS mutation ratesThe regional distributions of the DHFR and DHPS genotypes are shown in Fig. 4 and Table 1, respectively. Mutants with double DHFR mutation ANRNI were the most predominant in all five states. These mutants were present at the highest proportion in Goa. Isolates from Goa had only three DHFR genotypes, whereas the maximum of seven genotypes was found among the isolates in Assam (Fig. 4). Only isolates from Assam and Orissa were found to contain triple DHFR mutations. The triple DHFR mutation ANRNL was present in isolates from both states, while AIRNI was found only in isolates from Assam. In fact, the AIRNI genotype was the next most common after the ANRNI genotype in Assam. The double DHFR mutation AICNI was present in all states except Orissa, while ANCNL was found only in isolates from Assam and UP. Similar to the different numbers of DHFR genotypes, the minimum number of DHPS genotypes (n = 3) was present in Goa and the maximum number of DHPS genotypes (n = 15) was present in Assam (Table 1). However, the DHPS mutation rate was the lowest among isolates from UP, where the majority of isolates had wild-type sequence SAKAA. Among two of the commonly occurring single DHPS mutations, SGKAA was more common in Goa and Delhi, while FAKAA was more common in Orissa and Assam. None of the isolates from Goa or Delhi sample had the remaining single DHPS mutation SAEAA or SAKGA. The SAKGA genotype was present only in isolates from Assam, while SAEAA was present in isolates from UP, Orissa, and Assam. Surprisingly, none of the isolates from Goa had double or triple DHPS mutations. However, all three different triple DHPS mutations were detected in isolates from Assam, while only one of each of the triple DHPS mutations was detected in isolates from UP and Orissa. In Assam, the triple DHPS mutation (AGEAA) was the next most common mutation after the single DHPS mutation FAKAA.

• Open in new tab
• Download powerpoint

FIG. 4.

Regional distributions of P. falciparum DHFR genotypes among Indian isolates. Mutated amino acids are shown in boldface. n, number of isolates analyzed.

P. falciparum DHFR-DHPS two-locus mutation analysisThe P. falciparum DHFR-DHPS two-locus mutation analysis was carried out with the 207 isolates for which PCR successfully amplified all five codons of each gene. There were a total of 29 different DHFR-DHPS two-locus genotypes among these isolates. Fifteen of these genotypes were not very common, as they were found in only one isolate each (Table 2). The increased rate of mutation of the sequences of both enzymes over a 5-year interval resulted in an increased number of two-locus genotypes in group B. Table 2 shows only 8 different combinations of genotypes in group A but 24 in group B. We also noticed that five of eight combinations of genotypes in group A were absent from group B. On the other hand, the newer combined genotypes emerged in group B during this period (Table 2). There was no correlation between the number of mutations arising in the sequences for DHFR and DHPS (data not shown). It seems that a mutation in the sequence for DHFR occurs first, followed by a mutation in the sequence for DHPS, because group A isolates already had higher rates of mutations in the DHFR sequence than in the DHPS sequence. This is in agreement with information in the literature that DHFR mutations occur first under the influence of SP treatment (20, 40).

Only 8.69% of the 207 isolates had wild-type sequences for both enzymes, while the rest of them (91.31%) had mutations in the sequences of either one or both genes (Table 2). In fact, only 87 (42.03%) isolates harbored mutations in the sequences of both enzymes (types V, VII to XVII, XXI to XXVII, and XXIX in Table 2). The majority of the isolates (58 of 87) harboring mutations in both enzymes had double DHFR mutations plus a single DHPS mutation (types VII to X in Table 2), while 9 isolates had double mutations in each enzyme sequence (types XI to XV in Table 2). Five isolates had a single mutation in each enzyme sequence (type V in Table 2), and six isolates had double DHFR mutations plus triple DHPS mutations (types XVI and XVII in Table 2). Only three isolates had triple mutations in the sequence for each enzyme (types XXVI and XXVII in Table 2). Six isolates had triple mutations in the sequence for DHFR plus single mutations (four isolates; types XXI to XXIII and XXIX in Table 2) or double mutations (two isolates; types XXIV and XXV in Table 2) in the sequence for DHPS.

The regional distributions of the DHFR-DHPS two-locus genotypes are shown in Table 3. The minimum of number of different genotypes was 4 in Goa and the maximum was 24 in Assam, while 9 different genotypes each were detected in Delhi and UP and 11 were detected in Orissa. Assam and UP had the maximum number of isolates of type VI (ANRNI-SAKAA), while Delhi, Orissa, and Goa had the maximum number of isolates of type IV (ANCNI-SAKAA), type VII (ANRNI-FAKAA), and type VIII (ANRNI-SGKAA), respectively. The majority of isolates from these states had the maximum of three total mutations in the sequences for DHFR and DHPS combined; for the isolates from UP, however, the maximum number of mutations was two (Fig. 5, inset). In Assam, however, equal numbers of isolates carried two and three total mutations combined. Isolates with triple mutations in each gene were found only in Assam, with the maximum number of combined mutations being up to six. In fact, 70% (17 of 24) of isolates with more than four mutations in the two enzyme sequences combined were from Assam. When we analyzed the distributions of the combinations of mutations among group A and B isolates, it was observed that the majority of isolates (69%) in group B had a maximum of three or more mutations in the two loci combined, whereas only 4.5% (4 of 89) isolates from group A fell in this category (Fig. 5 and Table 2). Indeed, the majority of isolates in group A had less than three mutations in both the DHFR and the DHPS loci combined. More parasite isolates in group A than group B had the wild-type sequence or a single mutation in either the DHFR or the DHPS locus. The data clearly indicate that there was a shift from lower to higher numbers of two-locus mutations (two to three) over the 5-year period.

• Open in new tab
• Download powerpoint

FIG. 5.

Group and regional (inset) distributions of the total numbers of mutations in the DHFR and DHPS loci. n, number of isolates.