Sterile protection against Plasmodium vivax malaria by repeated blood stage infection in a non-human primate model

The malaria parasite Plasmodium vivax remains a major global public health challenge, causing major morbidity across tropical and subtropical regions. Several candidate vaccines are in preclinical and clinical trials, however no vaccine against P. vivax malaria is approved for use in humans. Here we assessed whether P. vivax strain-transcendent immunity can be achieved by repeated infection in Aotus monkeys. For this purpose, we repeatedly infected six animals with blood stages of the P. vivax Salvador 1 (SAL-1) strain until sterile immune, and then challenged with the AMRU-1 strain. Sterile immunity was achieved in 4/4 Aotus monkeys after two homologous infections with the SAL-1 strain, while partial protection against a heterologous AMRU-1 challenge (i.e., delay to infection and reduction in peak parasitemia compared to control) was achieved in 3/3 monkeys. IgG levels based on P. vivax lysate ELISA and protein microarray increased with repeated infections and correlated with the level of homologous protection. Analysis of parasite transcriptional profiles across inoculation levels provided no evidence of major antigenic switching upon homologous or heterologous challenge. In contrast, we observed significant transcriptional differences in the P. vivax core gene repertoire between SAL-1 and AMRU-1. Together with the strain-specific genetic diversity between SAL-1 and AMRU-1 these data suggest that the partial protection upon heterologous challenge is due to molecular differences between strains (at genome and transcriptome level) rather than immune evasion by antigenic switching. Our study demonstrates that sterile immunity against P. vivax can be achieved by repeated homologous blood stage infection in Aotus monkeys, thus providing a benchmark to test the efficacy of candidate blood stage P. vivax malaria vaccines. Author summary Plasmodium vivax is the most widespread human malaria parasite. Elimination efforts are complicated due to the peculiar biology of P. vivax including dormant liver forms, cryptic reservoirs in bone marrow and spleen and a large asymptomatic infectious reservoir in affected populations. Currently there is no vaccine against malaria caused by P. vivax. Here we induce sterile immunity by repeated P. vivax infection with the SAL-1 strain in non-human primates. In contrast, heterologous challenge with the AMRU-1 strain only provided partial protection. Antibody levels against a crude antigen and a protein microarray correlated with the level of homologous protection. Parasite transcriptional profiles across inoculation levels failed to show major antigenic switching across SAL-1 infections or upon heterologous challenge, instead suggesting other mechanisms of immune evasion. Our study demonstrates that sterile immunity against P. vivax can be achieved by repeated blood stage infection in Aotus monkeys, thus providing a benchmark to test the efficacy of candidate blood stage P. vivax malaria vaccines.

AMRU-1 and SAL-1, respectively, compared to over 1000 in the PvP01 reference strain extensive pir gene switching either in SAL-1 or AMRU-1 parasites. Rather, there appear to 3 2 4 be significant strain-specific differences in both core and pir expression between SAL-1 and expression across SAL-1 challenges or upon AMRU-1 challenge ( Figure 6C).

2 7
Altogether, the transcriptional analysis does not indicate that AMRU-1 parasites by antigenic switching. Thus, the lower protection observed after the heterologous challenge 3 3 0 may be due to major genetic and hence antibody epitope variation between these two 3 3 1 geographically separated strains (50). prominence as an alternative to subunit vaccines (65, 66). One major advantage of 3 4 1 vaccination using whole blood stage parasites is the multiplicity of immunogenic antigens, including those conserved across strains that may be able to induce strain transcendent  To assess whether strain-transcendent immunity can be achieved by repeated blood stage in Aotus monkeys after only two infections. In contrast, Aotus monkeys infected with P.
falciparum needed between three to four (69) and six to seven (38) repeated infections, 3 5 1 respectively, to achieve sterile immunity. This is consistent with previous observations made 3 5 2 during malariotherapy in patients with neurosyphilis demonstrating that immunity to P. falciparum is acquired more slowly than to P. vivax or P. malariae (25). Interestingly,

5 4
Saimiri sciureus boliviensis monkeys immunized with irradiated sporozoites of P. vivax SAL- when challenged with SAL-1 blood stage parasites, suggesting that humoral immunity is a 3 5 8 correlate of protection against repeated blood stage infections.

5 9
Furthermore, our study demonstrates that the sterile immunity achieved after repeated infection with a homologous strain was only partially protective after a heterologous 3 6 1 challenge (i.e., delay to infection and reduction in peak parasitemia compared to control). antigens and antibody breath in sera from these animals increased with each inoculation level 3 7 5 and were statistically significantly different between inoculation levels I and III, when the 3 7 6 animals achieved sterile immunity to a homologous SAL-1 challenge. Among the most 3 7 7 significant asexual blood stages antigens detected by the protein microarray were ETRAMP PVX_099980_s2 located on chromosome 7, the latter, a leading vaccine candidate that has 3 8 0 been identified as a major determinant of strain-specific protective immunity (85).

8 1
In this study, animals with sterile immunity to a P. vivax SAL-1 homologous with essential parasite phenotypes such as red cell invasion, rosetting or cytoadherence.

8 5
Maintaining genetic diversity enables immune evasion, as suggested in recent genomic 3 8 6 studies of P. vivax parasites from distinct geographic origin such as SAL-1 and AMRU-1 (50, 3 8 7 86). Finding conserved and cryptic (not exposed to the immune system) epitopes involved in 3 8 8 essential phenotypes that could be targeted by strain transcending neutralizing antibodies 3 8 9 represents a possible way forward (87). In contrast to P. falciparum that utilizes the variant PfEMP1 antigens to induce cytoadherence and avoid splenic clearance of blood stage were observed across all the different inoculation levels, whether homologous or  challenges may therefore be due to major genetic differences and hence antibody epitope In conclusion, our study demonstrates that sterile immunity against P. vivax can be  Twelve laboratory bred (lab-bred) adult male and female "spleen-intact" Aotus l. study (91). The animals were cared and maintained as described elsewhere (92). Isolates of P. (95) were used. This study can be considered as exploratory (i.e. looking for patterns of 4 3 5 response rather than hypothesis testing (96)), hence the number of subjects used in the only 4 3 6 group studied is typical of such exploratory research with humans (35, 97) and NHP (38).

3 7
Briefly, each frozen stabilate of SAL-1 and AMRU-1 was thawed, washed three times with incomplete RPMI medium, and resuspended in 1 ml of RPMI medium. This suspension RPMI to get a total inoculum of 50,000 parasites/ml. All animals received 1 ml of the gastric intubation to end the experiment. For this study patency was defined as the first of three consecutive positive days after 4 7 3 inoculation. Clearance was defined as the first of three consecutive negative days. Recrudescence was defined as the first of three consecutive positive days after a period of infection. For this study we classified anemia based on the hematocrit % as mild (Hct% = 31-36), baseline or hemoglobin was < 8 gm/dL, platelets were < 50 x 10 3 /µL or the animals remained  antigen-antibody reactivity, the plates were then read immediately at 492 nm in a ELx808

0 2
Plate reader (BioTek ® , Winooski, VT, USA). washing, samples were diluted 1:2, added to the plate and incubated for 2hrs. Next, plates  Protein microarray and hybridization. The construction of the protein microarray was 5 2 0 conducted using methods as described elsewhere (100). Briefly, coding sequences were PCR 5 2 1 amplified from P. vivax SAL-1 genomic DNA and cloned into the PXT7 plasmid using incubated with the protein arrays overnight at 4 0 C, followed by incubation with a goat anti- wavelength of 594nm (101).

3 1
Protein microarray data processing and analysis. Raw median fluorescent intensity was local 5 3 2 background corrected using the normexp function (offset = 50, method = "mle", limma R 5 3 3 package). All data was log transformed (base 2) and normalized as a ratio of the signal for inoculations. Pearson's correlations were performed for available ELISA titers and antigen 5 4 2 breadth. All statistics and plots were done using R unless otherwise specified. and AMRU-1, respectively. For annotation, the assemblies were loaded into Companion (48), annotation can be found at http://cellatlas.mvls.gla.ac.uk/Assemblies/.

7
For the Gephi analysis, we extracted all the genes annotated as pir from the two 5 5 8 Companion runs, merged them with the pir genes of PvP01 and performed an all-against-all global identity cut-off was set to 32% and the Fruchterman Reingold algorithm was run. using OligoRankPick (104)(oligo size=60, %GC=40). The oligos that were overlapping with 5 6 7 core genome oligos from the existing P. vivax microarray (47) were removed (12 for SAL-1 1 probes, amongst which 8 match two SAL-1 genes and 5 ma tch two AMRU-1 genes (Full 5 7 0 list as Table S5).

7 1
RNA preparation and microarray hybridization. Cell pellets from the blood samples collected was extracted and processed to be run in a customized microarray assay detecting both core 5 7 4 and subtelomoeric genes. The previously described microarray hybridization protocol was 5 7 5 used for this study, with several modifications (105). In brief, 100 ng of cDNA was used for reference pool were then hybridized together on customized microarray chip using Microarray analysis. To quantify microarray data signals, intensities were first corrected 5 8 6 using an adaptive background correction using the method "normexp" and offset 50 using the 5 8 7 Limma package in R(106). Next, we performed within-array loess normalization followed by  Plasmodium vivax reference sequence with improved assembly of the subtelomeres reveals specific antibody blocks binding of infected erythrocytes to amelanotic melanoma cells.