Malaria Journal BioMed Central Methodology

Background Malaria is a very important cause of anaemia in tropical countries. Anaemia is assessed either by measurement of the haematocrit or the haemoglobin concentration. For comparisons across studies, it is often necessary to derive one measure from the other. Methods Data on patients with slide-confirmed uncomplicated falciparum malaria were pooled from 85 antimalarial drug trials conducted in 25 different countries, to assess the haemoglobin/haematocrit relationship at different time points in malaria. Using a linear random effects model, a conversion equation for haematocrit was derived based on 3,254 measurements from various time points (ranging from day 0 to day 63) from 1,810 patients with simultaneous measurements of both parameters. Haemoglobin was also estimated from haematocrit with the commonly used threefold conversion. Results A good fit was obtained using Haematocrit = 5.62 + 2.60 * Haemoglobin. On average, haematocrit/3 levels were slightly higher than haemoglobin measurements with a mean difference (± SD) of -0.69 (± 1.3) for children under the age of 5 (n = 1,440 measurements from 449 patients). Conclusion Based on this large data set, an accurate and robust conversion factor both in acute malaria and in convalescence was obtained. The commonly used threefold conversion is also valid.


Background
Plasmodium vivax accounts worldwide for an estimated 70-80 million cases of malaria per year [1]. In some countries such as Brazil, P. vivax was responsible for approximately 79% of the 389,736 cases of the disease reported in 2001 [2].
Laboratory methods are important tools for the control of the disease progress. They can be useful for the individual diagnosis or for the patients' follow-up after specific antimalaria treatment. Conventional light microscopy has been widely used for malaria diagnosis. However, outside areas where malaria is endemic, it is rarely performed. Also, this method is time consuming and requires trained personnel.
In attempts to overcome these problems several rapid diagnostic tests have been developed recently. These tests detect specific proteins such as HRP2, a histidine rich protein 2, or pLDH, lactate dehydrogenase, in unfractionated blood of patients with malaria. However, these assay do not differentiate P. vivax from Plasmodium malariae or Plasmodium ovale infections, and vary in their sensitivity and specificity (reviewed in reference [3]).
Molecular diagnosis by PCR is described as the most sensitive and specific method for Plasmodium detection. Genus-and species-specific primers have been used to amplify Plasmodium ssrRNA genes of the four human malaria parasites and to detect mixed infections [4][5][6]. However, this methodology is costly and requires trained personnel for its implementation.
Detection of antibodies by immunofluorescence or ELISA has been used for seroepidemiology of malaria. However, in the case of P. vivax, the difficulty of blood stage cultivation has been hindering the use of these methodologies. Production of recombinant proteins through the techniques of genetic engineering may provide sufficient P. vivax blood stage antigen(s) for the establishment of specific serological assays.
In the course of immuno-epidemiological studies using recombinant proteins based on the sequence of the Merozoite Surface protein-1 (MSP1) of P. vivax, we found that a recombinant protein representing the 19 kDa region of MSP1 (MSP1 19 ) was highly immunogenic during natural infection in humans [7,8]. This recombinant protein was recognized by antibodies of a large fraction of Brazilian individuals recently exposed to P. vivax [7]. This observation was subsequently confirmed in studies performed in Korea, where more than 90% of P. vivax-infected individuals displayed specific antibodies to P. vivax MSP1 19 [9]. Also important was the observation that the P. vivax MSP1 19 gene displays very limited allele polymorphism in different regions of the world that does not restrict recognition by human antibodies [10,11]. Together, these results suggested that it could be possible to develop an ELISA using a single recombinant protein based on the P. vivax MSP1 19 . This relatively simple and inexpensive technique could be of great applicability for epidemiological studies, the screening of blood donors and the serological diagnosis of malaria caused by P. vivax.
In the present study three purified recombinant proteins produced in E. coli (GST-MSP1 19 , His 6 -MSP1 19 and His 6 -MSP1 19 -PADRE) and one in Pichia pastoris (yMSP1 19 -PADRE) were compared for their ability to be recognized by IgG antibodies of individuals with patent P. vivax infection. Serological evaluation was performed with serum samples collected from individuals living in areas endemic for malaria in the north of Brazil and compared to serum samples from individuals never exposed to P. vivax malaria.

Study population
Sera from 430 individuals were used in this study, of which 200 were from patients with patent P. vivax malaria and 230 from persons not exposed to P. vivax malaria (negative controls). The blood samples from P. vivax patients were collected in the state of Pará, in the north of Brazil, in areas endemic for malaria: 103 from Belém, 21 from Marabá, 20 from Itaituba, 20 from Tailândia, 36 from Igarapé-Açu. Patent infection was documented by microscopic analysis of Giemsa-stained blood drops. These samples were obtained between January 1996 and July 1999 with informed consent of all individuals and kept at -20°C.
The 230 individuals non-exposed to P. vivax malaria included: (i) 49 donors to blood banks in the city of São Paulo, an area where malaria is not endemic (negative controls); (ii) 108 blood bank donors with unrelated infectious diseases, detected serologically; among them 21 with Chagas Disease, 21 with syphilis, 19 with HBV, 21 with HCV, 14 with HTLV and 12 with HIV; (iii) 10 individuals positive for antinuclear antibodies (ANA) and 10 for rheumatoid factors (RF); (iv) 53 individuals living in distinct areas in West Africa where malaria caused by Plasmodium falciparum is endemic: a) 26 adults without clear symptoms of malaria or other infectious disease from Senegal (n = 19) and Gambia (n = 7), b) seven children from Gambia during patent infection by P. falciparum, c) 20 malaria immune adults from Ghana. Aliquots of these samples were kindly supplied by Dr. Marcelo Urbano Ferreira (Universidade de São Paulo, São Paulo) and Dr. Silvia Di Santi (Superitendência de Controle de Endemias, SUCEN, São Paulo). Clinical and laboratory data were reported elsewhere [12][13][14].

Recombinant proteins
The recombinant proteins used in the present study represent amino acids 1616-1704 of the MSP-1 (Belem strain) of P. vivax. These proteins were expressed in E. coli (GST-MSP1 19 , His 6 -MSP1 19 and His 6 -MSP1 19 -PADRE, ref. 15) or Pichia pastoris (yMSP1 19 -PADRE, ref. 16). The generation of the plasmids and expression/purification of the recombinant proteins produced in E. coli or Pichia pastoris were performed as described in references 15 and 16, respectively.

ELISA for detection of human antibodies
Human IgG antibodies against MSP1 19 were detected by ELISA as described [7]. ELISA plates were coated with 200 ng/well of each recombinant protein. The amount of recombinant protein gave the same OD 492 when we used an anti-MSP1 19 MAb [17]. Fifty µl of each solution were added to each of 96 well plates (High binding, Costar). After overnight incubation at room temperature (rt), the plates were washed with PBS-Tween (0.05%, v/v) and blocked with PBS-milk (5% w/v) for two hours at 37°C. Serum samples were diluted 1:100 in this same solution and 50 µl of each sample was added to each well in duplicate. After incubation for two hours at rt and washes with PBS-Tween, 50 µl a solution containing peroxidase-conjugated goat anti-human IgG (Fc specific) diluted 1:10,000 (Sigma) were added to each well. The enzymatic reaction was developed by the addition of 1 mg/ml of o-phenylenediamine (Sigma) diluted in phosphate-citrate buffer, pH 5.0, containing 0.03% (v/v) hydrogen peroxide, and was stopped by the addition of 50 µl of 4 N H 2 SO 4 . Plates were read at 492 nm (OD 492 ) with an ELISA reader (SLT SPEC-TRA, SLT Labinstruments, Austria). The individual values of OD 492 obtained for the recombinant proteins of MSP1 19 were corrected by subtraction of the individual values of OD 492 obtained against glutathione S-transferase (GST).

Statistical analysis
OD 492 from different samples were plotted using computer graphics software (GraphPad Prism, Version 3.0, San Diego, California). Cutoff values of each recombinant protein in ELISA were calculated as the mean OD 492 of sera from 49 blood donors plus five standard deviations (SD). The values for sensitivity and specificity were estimated as described [18] with microscopy used as the gold standard. The Kruskal-Wallis test was used to test the significance of differences between the group values. Differences between proportions were analyzed by a Chi-square test.

Reactivity of recombinant proteins expressed in distinct vectors with serum samples from individuals with patent P. vivax
Four recombinant proteins obtained as earlier described in references 15 and 16 (Table 1)  When we compared the OD 492 values from serum samples of P. vivax infected individuals, we observed that values obtained for the recombinant proteins His 6 -MSP1 19 and His 6 -MSP1 19 -PADRE were significantly higher than those obtained for GST-MSP1 19 (Fig. 1, P < 0 Sera from the few individuals who tested negative towards all four recombinant proteins were also tested for the recognition of recombinant proteins representing the N-terminal region of P. vivax MSP1, MSP3α (C-terminal) or MSP3β(N and C-terminal and entire protein, references 19 and 20). All of them failed to recognize any of the recombinant protein tested (data not shown). However, we detected antibodies IgM anti-MSP1 19 in 90% (9/10) of these IgG negative individuals. If we consider the results of detection of antibodies IgG and IgM for MSP1 19 , the sensitivity of the assay was of 99.5%. These 10 individuals were primo-infected and therefore the presence of IgM but not IgG may reflect a delay in the immunoglobulin class switch.
From the 200 individuals that we evaluated, 111 were primo-infected. One hundred and one of them had specific IgG for the recombinant proteins that we tested. Only 10 were negative for IgG and, as mentioned above, 9 of them had specific IgM.

Evaluation of the specificity of the recombinant proteins tested with serum samples from individuals exposed to P. falciparum, from unrelated diseases, or healthy donors
The specificity of the assay using recombinant P. vivax antigens was examined with 230 serum samples from individuals without previous history of P. vivax malaria, including individuals exposed to P. falciparum malaria, individuals with unrelated diseases or healthy individuals. Due to limitations on the volume of each sample available, sera from African individuals exposed to P. falciparum were tested only against recombinant protein His 6 -MSP1 19. For the calculation of the specificity of the assay using the recombinant protein His 6 -MSP1 19 the results that was obtained did not include sera from African individuals. The specificity values determined with sera from healthy individuals and sera from individuals with other infectious diseases, were 98.3% (GST-MSP1 19 ), 97.7% (His 6 -MSP1 19 and His 6 -MSP1 19 -PADRE) and 100% (yMSP1 19 -PADRE). In figure 2, we compared the OD 492 values from serum samples of P. vivax infected individuals, individuals exposed to P. falciparum and individuals with unrelated diseases.

Discussion
Due to the difficulties in cultivating blood stages of P. vivax, serological diagnosis of patent P. vivax malaria can best be accomplished with the use of recombinant proteins. In the present study, we compared for purified  On the other hand, the persistence of the IgG antibodies specific for MSP1 19 is still a matter that has to be further evaluated. In previous studies we determined that the antibody titers decreased relatively rapidly after treatment [8]. However, studies are underway using these recently generated recombinant proteins.
Distribution of the OD 492 data for sera from individuals with patent P. vivax malaria, from individuals exposed to P. falciparum, from individuals with unrelated diseases or healthy controls  An important observation from our study was the fact that sera from African individuals naturally exposed to P. falciparum failed to cross-react with the recombinant protein His 6 -MSP1 19 ( Figure 2). As a control, 39 of these sera were also tested against a recombinant protein derived of the MSP-2 of P. falciparum. The percentage of responders was 69.2% (data not shown). The lack of cross-reactivity may have several implications and should be further evaluated in the light of the current knowledge that both MSP1 19 share similarities in their predicted tertiary structures [21]. For epidemiological studies, for the screening of blood donors and the serological diagnosis of P. vivax malaria, the lack of cross reactivity can be a major advantage. Nevertheless, the absence of cross-recognition of P. vivax MSP1 19 by antibodies from P. falciparum-exposed individuals has may also have immunological consequences at the level of acquired immunity and vaccine development in areas where both malarias are prevalent. Detailed studies will be required to determine whether sera from P. vivax infected individuals also fail to recognize recombinant proteins representing the P. falciparum MSP1 19 .
Recently, a direct sandwich ELISA to detect antibodies against the C-terminal region of MSP-1 was proposed as a potential diagnostic method for Plasmodium vivax exposed individuals from Korea [22]. This assay showed a high sensitivity (99.5%) indicating that recombinant proteins containing the C-terminal region of P. vivax MSP-1 may be used in individuals from different parts of the world.
It is also important to mention that it is very likely that serological detection of P. vivax malaria can be further improved by several distinct strategies. We are currently trying to improve the detection level by the combined use of other recombinant proteins based on the sequence of other blood stage antigens of P. vivax such as the Plasmodium vivax Merozoite Surface Proteins-3α and β, Apical Membrane Antigen-1 and the Duffy Binding Protein [19,20,23,24]. Also, we are developing a chemiluminescent enzyme-linked immunosorbent assay that may greatly improve the sensitivity, eliminating the few false negative and false positive samples we had [25,26].
Finally, our results support the notion that recombinant proteins based on the P. vivax MSP1 19 kDa can be useful for the development of a rapid immunochromatographic assay for field studies, small laboratories and blood banks (reviewed in references 3 and 27).

Conclusions
Our study demonstrated that for the Brazilian population, an ELISA using a single recombinant protein based on the P. vivax MSP1 19 can serve as the basis for the development of a valuable serological assay for the detection of P. vivax malaria.