High proportion of mixed-species Plasmodium infections in India revealed by PCR diagnostic assay

Authors


Corresponding Author Aparup Das, Evolutionary Genomics and Bioinformatics Laboratory, Division of Genomics and Bioinformatics, National Institute of Malaria Research, Sector 8, Dwarka, New Delhi 110 077, India. Tel.: +91 11 25307 322; Fax: +91 11 25307 377; E-mail: aparup@mrcindia.org

Summary

Accurate diagnosis is the key to effective treatment and control of malaria. We screened 180 microscopically diagnosed Indian malaria-positive blood samples for pure and mixed infections by Plasmodium falciparum and Plasmodium vivax. An unusually high proportion of mixed infections was detected, signifying the sensitivity of PCR assay over traditional microscopic diagnosis.

Abstract

Forte proportion d’infections mixtes d’espèces Plasmodium en Inde révélée par test de diagnostic PCR

Le diagnostic précis est la clé pour un traitement efficace et pour la lutte contre la malaria. Nous avons examiné 180 échantillons de sang provenant de l’Inde, diagnostiqués microscopiquement comme positifs pour la malaria, afin de détecter des infections uniques et mixtes àPlasmodium falciparum et P.vivax. Une proportion exceptionnellement élevée d’infections mixtes a été détectée, ce qui confirme la sensibilité plus élevée de la PCR par rapport au diagnostic microscopique traditionnel.

Abstract

En India, una alta proporción de infecciones mixtas con dos especies de Plasmodium se diagnóstica mediante PCR

Un diagnóstico preciso es la clave para un tratamiento efectivo y el control de la malaria. Se han estudiado 180 muestras de sangre de Indios diagnosticados con malaria mediante microscopía, con el fin de

evaluar si se trataba de infecciones puras o mixtas con Plasmodium falciparum y P. vivax. Se detectó una proporción inusualmente alta de infecciones mixtas, demostrando la mayor sensibilidad de la prueba de PCR sobre el diagnóstico microscópico tradicional.

Introduction

Malaria is an often-deadly infectious disease with global extent. Although significantly fewer malaria cases have been reported in some endemic countries in recent years (WHO 2008), efforts to curb the disease overall have so far been a feeble success. Despite serious efforts to control malaria in India, about one million cases are reported annually (National Vector Borne Disease Control Program http://www.nvbdcp.gov.in). Further, India is one of the very few malaria-endemic countries where the two most prevalent human malaria parasites, Plasmodium falciparum and Plasmodium vivax, occur almost equally (Singh et al. 2009). This enhances the possibility of mixed parasitic infections in a single individual. However, few studies have reported data on the prevalence of mixed-species infections by these two parasites in India (Bharti et al. 2007; Singh et al. 2008).

Malaria diagnosis mainly relies on microscopic examination of finger-prick blood samples of malaria-symptomatic individuals, because microscopy is inexpensive and easy for clinical diagnosis (Chotivanich et al. 2007). But in mixed infections, the tendency of one parasite to dominate the other (Sethabutr et al. 1992) lowers the efficiency of microscopic detection of two species in the same sample. Although, rapid diagnostic tests based on Plasmodium species-specific antibodies are useful and easy tools for field surveys, these tests are unable to differentiate between P. falciparum single infection and co-infections with other Plasmodium species (http://www.malariasite.com/MALARIA/rdts/htm; Mayxay et al. 2004). By contrast, a variety of diagnostic techniques for malaria based on the detection of species-specific DNA sequences by PCR amplification have been developed and are generally considered to be robust and sensitive (Snounou et al. 1993; Hänscheid & Grobusch 2002).

To determine the extent of mixed parasitic infections, as well as the specificity and sensitivity of microscopic assay, we evaluated by PCR assay 180 microscopically diagnosed malaria cases collected randomly from six states covering nearly all malaria-endemic regions of India (Figure 1).

Figure 1.

 Map of India indicating the sample collection sites.

Materials and methods

All 180 blood samples from the states of Asom, Karnataka, Odisha, Goa, Tamil Nadu and Chhattisgarh were examined as Giemsa-stained thick and thin blood smears under light microscope following the standard protocol when collected in the field. The details of the total number of samples from each location are shown in Table 1. Samples were collected from March to May and from September to November (peak transmission seasons for malaria infections in India) in 2007 and 2008 (Table 1). The microscopy-based diagnoses were performed by expert technicians of the National Institute of Malaria Research (NIMR) field units located at Sonapur (Asom), Bengaluru (Karnataka), Rourkela (Odisha), Raipur (Chhattisgarh), Chennai (Tamil Nadu) and Goa (Figure 1). All necessary clearances for subsequent steps were obtained from the ethical committee of the National Institute of Malaria Research, and informed consents received from each patient.

Table 1.   Details of sample collection sites in India with diagnostic results with microscopy and PCR assay
Location (State) of sample collectionPopulation coordinatesTime of collectionDifferent diagnostic methodsTotal samples
MicroscopyPCR assay
Plasmodium falciparumPlasmodium vivaxP. falciparumP. vivaxMixed (Pf + Pv) (percentage)
P. falciparum endemic regions
 Sonapur  (Asom)26°07′0″ N
91°58′60″ E
March–May 200720 57 24431 (40.25%)77
 Raipur  (Chhattisgarh)21°14′ N
81°38′ E
November–December 200717 010 0 7 (41.17%)17
 Rourkela  (Odisha)22°07′ N
84°32′ E
February–March 2008 0 16 1 6 9 (56.25%)16
P. vivax endemic regions
 Panaji  (Goa)15°28′60″ N
73°49′60″ E
September–November 2008 0 16 014 2 (12.5%)16
 Chennai  (Tamil Nadu)13°4′60″ N
80°16′60″ E
September–November 2008 0 10 0 3 7 (70%)10
 Bengaluru  (Karnataka)12°58′60″ N
77°34′60″ E
September–November 2008 0 44 01826 (59.1%)44
Total  37 (20.5%)143 (79.5%)13 (7.22%)85 (47.2%)82 (45.55%)180

After microscopic confirmation, finger-prick blood samples from malaria-positive patients were spotted on Whatman filter paper (5–7 spots per filter paper) and kept at 4 °C until further processing. Genomic DNA was extracted using QIAamp mini DNA kit (Qiagen, Germany) according to standard protocol. The PCR assay used nested PCR (Johnston et al. 2006) as follows: in the first step of nested PCR, a pair of Plasmodium genus-specific primers [rPLU5 (5′CCTGTTGTTGCCTTAAACTTC3′) and rPLU6 (5′TTAAAATTGTTGCAGTTAAAACG3′)] amplified a 1100-bp PCR product from the rRNA small subunit gene (18S rRNA) (Figure 2). In the second step, two primer pairs were used: one specific to P. vivax, [rVIV1 (5′CGCTTCTAGCTTAATCCACATAACTGATAC3′) and rVIV2 (5′ACTTCCAAGCCGAAGCAAAGAAAGTCCTTA3′)], and the other specific to P. falciparum [rFAL1 (5′TTAAACTGGTTTGGGAAAACCAAATATATT3′) and rFAL2 (5′ACACAATGAACTCAATCATGACTACCCGTC3′)]. The P. vivax-specific primers amplify a 120-bp region of Plasmodium 18S rRNA, whereas P. falciparum-specific primers amplify a 205-bp region. Thus, pure P. vivax and P. falciparum samples show a clear band at 120 and 205 bp, respectively, whereas the mixed-infected samples show bands at both positions (Figure 2).

Figure 2.

 Agarose gel electrophoresis of PCR products obtained after 1st and the 2nd step of the nested PCR. Lane M, 100 bp ladder; lanes 1–5, amplified PCR products from parasite genomic DNA obtained after 1st step using Plasmodium genus specific set of primers (1100 bp); lanes 6–11, 30 and 32 P. vivax single infections (120 bp product); lane 12, negative control with human DNA; lane 13, negative control without DNA; lanes 14–16, P. falciparum single infections (205b product) and lanes 17–29, 31 and 33–36 mixed infections of both P. vivax (120 bp) and P. falciparum (205 bp).

To rule out false positives, two negative controls were used, one lacking DNA and the other using the genomic DNA of a healthy individual (BG) from a malaria non-endemic region (Jammu and Kashmir state) of India as a template. Furthermore, to confirm the species-specificity of the two DNA fragments, both fragments from five mixed-infected samples (from Bengaluru) were gel-purified using the Qiagen Gel Extraction Kit (Qiagen, Germany) and sequenced in forward and reverse directions (2X coverage) using species-specific sequencing primers similar to those used for PCR amplification. DNA sequencing was performed using Big Dye Terminator Chemistry in an in-house 96-capillary DNA analyzer (ABI 3730XL). DNA sequences reported in this article can be found in GenBank under accession numbers HM014352 to HM014361. The sequences were aligned with the published reference sequences of both P. falciparum (3D7) and P. vivax (Sal-1), separately using clustalW software (Figures 3a,b), which yielded perfect alignments for both fragments. However, at certain nucleotide positions, e.g. the 49th nucleotide position of the P. falciparum (Figure 3a), the 88th and 110th positions of the P. vivax fragment (Figure 3b), single nucleotide polymorphisms (SNPs) were detected in some of the isolates from Bengaluru. Species specificity of the two sequenced fragments was further confirmed by homology search of the NCBI web repository of sequences (http://www.ncbi.nlm.nih.gov/), using BLASTN with default parameters.

Figure 3.

 The DNA sequence alignment between the reference sequences and five isolates from Bengaluru (a) in P. falciparum (b) in P. vivax. The topmost sequences of both the alignments are reference sequences of the respective species.

Results

Of 180 blood samples tested for mixed infection using PCR, 85 were purely P. vivax (47.2%) and 13 (7.22%) purely P. falciparum infections. By contrast, microscopic diagnoses of these samples before PCR assays were conducted which indicated 143 (79.5%) as purely P. vivax and 37 (20.5%) as purely P. falciparum infections. Interestingly, 82 of 180 samples (45.5%) were found to be mixed P. falciparum/P. vivax infections by PCR assay, while microscopic examinations did not detect a single case of mixed infection. The incidences of mixed infection varied across different localities, but perfectly correlated with reported P. falciparum and the P. vivax endemicity in India (Table 1); overall, the highest proportion of mixed infections (70%) was seen in Chennai samples and the lowest (12.5%) in Goa samples. The Bengaluru samples were found to contain 59.1% mixed infection, closely followed by Rourkela samples (56.25%) from Odisha (Table 1).

Discussion

The high rate of misdiagnosis by microscopy the primary means of malaria diagnosis in India could be because of the presence of indistinguishable initial blood stages of parasite species and/or very low level of parasitemia as a consequence of either competition between the two species at the blood cell level (Mcqueen & McKenzie 2006), cross-species immunity (Maitland et al. 1997) or both. Furthermore, microscopists often are inclined to identify only one species based on pre-conceived notions of the malaria-epidemiological setting of the locality, leaving the rare species undetected.

To test these possibilities, we microscopically re-examined 11 slides corresponding to mixed infections detected by PCR assay and visually detected only two cases of mixed infections. Initial misdiagnosis of these two slides (18.2%) can thus be ascribed to human error; misdiagnosis of the remaining nine slides (81.8%) might be the consequence of either insensitivity of microscopy-based diagnosis, differential pathogenic behaviour of the two parasite species or both. For example, in P. vivax-infected patients, most of the life cycle stages ring, trophozoite, schizont and gametocytes circulate in human peripheral blood, but in P. falciparum infection, only ring-stage and gametocytes are present in peripheral circulation. Immature blood stages (e.g., ring stages) of P. falciparum and P. vivax look very similar under the microscope (http://www.malariasite.com/MALARIA/diagnosisofmalaria.htm), and overlooking one species on this basis is quite possible (Mayxay et al. 2004). Thus, if P. falciparum gametocytes are not numerous in the sample, it increases the chance of diagnosing a mixed infection as a single infection of P. vivax, which may explain the present results.

Our study clearly reveals a very high percentage of mixed malaria infections because of P. falciparum and P. vivax in India, similar to what has been found in many other malaria-endemic regions (e.g. in South America and Africa) (Postigo et al. 1998; Mehlotra et al. 2000). It has been postulated that a higher overall prevalence of P. vivax and P. falciparum is associated with fewer mixed-species infections (McKenzie & Bossert 1997), with either P. falciparum or P. vivax inducing cross-species immunity in the host (Maitland et al. 1997). However, this does not seem to be the case here, as high rates of mixed infection were observed in all six population samples. Further, we partitioned the six sampling sites into two sets (three sites each) on the basis of P. falciparum or P. vivax endemicity as revealed by epidemiological surveillance data (http://www.nvbdcp.gov.in), but could detect little difference between them in the proportion of mixed infections (average of 42.7% in P. falciparum-endemic vs. 50% in P. vivax-endemic areas).

Although the size of samples used in this study is small and uneven among the different localities, it is apparent that location-specific variation in the proportion of mixed infection does exist in India. Interestingly, the lowest (12.5%, Goa) and the highest (70%, Chennai) incidences of mixed infection were detected in P. vivax endemic localities; on the other hand, 40–56% mixed infections were recorded among the three P. falciparum-endemic localities in India. Surprisingly, an average of 59% of pure P. vivax infections was detected in samples collected in P. falciparum-endemic areas, whereas none of the samples from P. vivax-endemic areas were found to contain any pure P. falciparum infections. Although factors responsible for such a high rate of pure P. vivax infection in P. falciparum-endemic areas are unknown, this might be because of the cyclic seasonal changes of infection by different malaria parasites in any particular Indian locality (Bhattacharya et al. 2006).

Further, an impact of eco-geographical variations on the prevalence and transmission of the parasite species (Richie 1988) cannot be ruled out. The pattern of infection (single/mixed) is also determined by the occurrence and the ability of the vector species to be infected by different parasite species simultaneously (Mayxay et al. 2004). In India, nine different Anopheles species designated as primary vectors of malaria, with variable prevalence and transmission ability, are distributed in different geographical regions (Singh et al. 2009).

Accurate identification of the malaria parasite species is important not only for successful treatment, but also to design and develop effective malaria control measures and accurate malaria-epidemiological monitoring. Incorrect malaria diagnosis is a severe public health concern; ineffectively treated P. vivax infections can lead to relapses, and undiagnosed P. falciparum infections can turn severe and fatal. Further, misidentification of malaria parasites could lengthen the parasite clearance time and lead to recrudescence (Mayxay et al. 2004) and drug resistance (de Roode et al. 2004). This is because some antimalarials are less effective in certain Plasmodium species compared to others. For instance, it is suggested that monotherapy with sulphadoxine–pyrimetahmine (SP) should not be used in cases of P. falciparum and P. vivax mixed infections, as the latter species is intrinsically resistant to sulphadoxine and readily develops resistance to pyrimethamine (Tijitra et al. 2002). Moreover, incorrect treatment could also affect the local sensitivity pattern of the parasite species to the drugs (Mayxay et al. 2004). However, the impact of mixed infections on human health is still controversial; some workers in the field believe that there are benefits from mixed infections, whereas others consider them to be detrimental (Mayxay et al. 2004).

Conclusion

Our study affirmed that the PCR technique holds enormous promises in malaria diagnosis and seems to be much more robust and accurate than microscopic diagnosis. Although wide-ranging studies covering all malaria transmission regions with increased sample sizes, partitioned by different transmission seasons in India, would be needed for a more accurate picture of mixed malaria parasitic infection. The alarmingly high proportion of mixed malaria parasitic infections detected in this study should be kept in mind before applying malaria interventions.

Acknowledgements

We thank all Officers-in-Charge (OICs) and expert technicians in the NIMR field stations of Panaji, Guwahati, Bengaluru, Chennai, Rourkela and Raipur for their help in sample collection and microscopic identification, and Indian Council of Medical Research (ICMR), Delhi, India for an intramural funding to AD and a JRFship (NET) to BG. We are grateful to the two anonymous reviewers for critical and constructive comments on an earlier version of the manuscript and to Dr. Jane Carlton and Dr. Steven Sullivan (New York University Langone Medical Center) for suggestions, scientific comments and language modifications of the manuscript.

Ancillary