Blood samples (5 ml) were obtained from patients infected with P. vivax who were admitted to S.P. Medical College in Bikaner, India. The collection of patient samples was conducted following informed consent in accordance with hospital guidelines. Approval for sample collection was granted by the hospital ethics committee (No.F.(Acad) SPMC/2003/2395, No.F29(Acad)SPMC/2020/3151). Infection with P. vivax was confirmed by microscopy and rapid diagnostic tests (OptiMal test; Diamed AG, Cressier sur Morat, Switzerland; Falcivax test; Zephyr Biomedical SystemGoa, India) in the hospital. Blood was immediately (within 15 min) subjected to density gradient-based separation (Histopaque 1077, Sigma Aldrich, USA) to isolate peripheral blood mononuclear cells (PBMCs) from the infected blood samples as per the manufacturer’s instructions. Both fractions were then washed twice with phosphate-buffered saline (PBS), lysed with 4 volumes of TRI-reagent, and stored at -80 °C. These samples were maintained within the cold chain during transportation to BITS, where they were processed and evaluated using 18S rRNA gene-based multiplex PCR to ensure the absence of P. falciparum co-infection (Pakalapati et al., 2013). The PCR conditions used for the amplification of the 18S rRNA gene involved initial denaturation at 94 °C for 2 min, followed by denaturation at 94 °C for 1 min 30 s, annealing at 52 °C for 2 min, extension at 72 °C for 3 minutes, after initial denaturation all the reactions were run for 30 cycles. The TRI protocol was used to investigate both RNA and DNA from previous clinical samples (Chomczynski, 1993). The criteria for determining complicated cases were based on the World Health Organization guidelines (World Health Organization, 2010). For CGH array hybridization, isolates from six patients were considered, of which three showed cerebral malarial complications, while the other three were uncomplicated malaria cases (Supplementary Table 1). Cerebral malaria (CM) with a Glasgow coma scale score lesser than 11 was considered a complicated case.

A custom P. vivax 2 × 400 K CGH microarray (AMADID: 25335) was designed on an Agilent platform by Genotypic Technology Private Limited, Bangalore, India, using the genomic location and direction of transcriptional regulation data retrieved from NCBI. The 60mer probes were designed based on the optimal GC% and melting temperature using Genotypic_Probe_Parser.pl (Perl program developed by Genotypic Technology). All oligonucleotides were designed and synthesized in situ according to standard algorithms and methodologies used by Agilent Technologies. The array design comprised 418577 probes tilled across the whole genome, with an average probe spacing of 56 bp. This covers the entire genome, irrespective of intergenic or intronic sequences.

Genomic DNA preparation, labeling, and hybridization of the CGH array were performed at the Agilent certified microarray facility at Genotypic Technology, Bengaluru, India. Total DNA was extracted from the samples following the manufacturer’s protocol (TRI Reagent, Sigma Aldrich, USA). Subsequently, both DNA samples (PVC and PVU) were run on an agarose gel and confirmed to be intact. The purity and concentration of total DNA were determined using a Nanodrop ND-1000 UV-visible spectrophotometer (Nanodrop Technologies, Rockland, USA). The A260/A280 ratio indicated the absence of proteins and RNA. DNA from the three patient samples showing CM manifestations were pooled (PVC) and used as test, while the same was performed for the uncomplicated samples (PVU) as control. This approach was designed to identify CNVs consistently enriched in cerebral malaria compared to uncomplicated malaria isolates, with pooling providing a representative, group-level overview of robust variations associated with severe disease. The samples for Comparative Genomic DNA Hybridization were labeled using an Agilent Genomic DNA Labeling Kit (Part Number: 5190-0453). Two micrograms of each sample were digested using Alu1 and Rsa1. The restricted DNA was then labeled with Cy3 (PVU) and Cy5 (PVC) dUTP using a random primer labeling method. The labeled DNA was then concentrated, and the quality was assessed for yield and specific activity. The labeled DNA samples were hybridized on a collaboratively designed P. vivax 2x400K CGH microarray (AMADID: 25335). 5 micrograms of Cy3 and Cy5 labeled samples were hybridized. Hybridizations were done using the CGH Hybridization kit of Agilent (Part Number: 5190-0404). Hybridization was performed in Agilent’s Surehyb Chambers at 65° C for 40 h. The hybridized slides were washed using Agilent aCGH wash buffers (Part No: 5188-5221/22) and scanned using an Agilent Microarray Scanner G2505C at 3-micron resolution. The microarray data discussed in this manuscript has been deposited in the NCBI Gene Expression Omnibus (GEO) under the GEO series accession number GSE288219.

Image analysis and data normalization were performed by applying the linear dye normalization method using Agilent Feature Extraction Software. Data were further normalized via the centralization algorithm to ensure that log-ratio distributions were centered and comparable across probes. CNVs were then identified using the widely validated Aberration Detection Method II (ADM-2), which integrates probe error information in addition to log-ratio values, providing robust detection even in the presence of noisy probes. ADM-2 is the standard algorithm for Agilent array-CGH analysis and is widely used for high-confidence CNV calling (Delahaye et al., 2012; Chehbani et al., 2022; Bui et al., 2024; Mademont-Soler et al., 2024). To ensure stringent and reliable CNV identification, we applied a minimum requirement of three consecutive probes with an average absolute log₂ ratio ≥0.3 (default threshold 6.0, minProbes = 3, minAvgAbsLogRatio = 0.25), effectively removing single-probe artifacts arising from mismatches between probe and target sequences. Additionally, the Fuzzy Zero correction was enabled to reduce spurious long, low-level aberrations caused by correlated probe errors, such as those resulting from sequence divergence. Only genomic regions meeting all these criteria were considered significant. These thresholds provide both statistical support and effect-size filtering, ensuring high specificity in the detection of true copy number changes.

Amplification/deletion intervals varied from approximately 100bp to 1429 kb. Amplification and deletion were present in all the 14 chromosomes (Supplementary Table 2). Since maintaining a continuous P. vivax culture has been a challenge, therefore many proteins remain uncharacterized. Many of the regions with CNVs have genes encoding these hypothetical proteins.

The longest region showing CNV of approximately 1450 kb is on chromosome 12 (1575586-3004875) covering 316 genes that shows continuous amplification (Figure 2), followed by chromosome 14 (1221441–1677668) with 100 genes and chromosome 10 (496-429938) with 96 genes, both having amplifications.

Regions having genes encoding ABC transporters from different families showed CNVs (Supplementary Table 3). Orthologs of Plasmodium ABC transporters, known to drive drug resistance in microorganisms and human tumor cells, may similarly contribute to drug resistance in Plasmodium (Koenderink et al., 2010). There are three genes encoding ABC transporters (putative) that are in regions showing amplification while two genes were in regions showing deletion. Three of the genes (PVX_097025, PVX_080100, PVX_118100) encoding multidrug resistance proteins that belong to ABC transporter families are within regions of amplifications. Figure 3 shows the continuous amplification in the region containing PVX_097025, highlighting the CNV at this locus. Amplification in an ortholog of the gene PVX_080100 in P. falciparum (PF3D7_0523000) has been reported to be associate with mefloquine resistance (Cheeseman et al., 2009; Menard and Dondorp, 2017) (Rovira-Vallbona et al., 2023). Increased copy number of multidrug resistance gene-1 (mdr1) has been reported to be associated with increased drug resistance in P. falciparum isolates as well as in P. vivax (Imwong et al., 2008; Suwanarusk et al., 2008; Anderson et al., 2009; Rovira-Vallbona et al., 2023). Suwanarusk et al. (2008) showed that pvmdr1 amplification in Thai P. vivax isolates was linked to higher IC₅₀ values for mefloquine and artesunate, while isolates lacking amplification were more susceptible, supporting a functional role of CNVs in drug resistance.

In our data, we have also seen CNVs in regions having chloroquine resistance genes, a putative chloroquine resistance marker protein (PVX_118062 in chr 12) having amplification and putative chloroquine resistance transporters (PVX_087980 in chr 1) having deletion.

There is amplification observed in the gene responsible for producing the enzyme hydroxymethylpterin pyrophosphokinase-dihydropteroate synthetase (PPPK-DHPS). PPPK-DHPS plays a crucial role in the folate pathway of P. falciparum and is a specific target of the drug sulfadoxine (Kasekarn et al., 2004). Sulfadoxine resistance has been reported in both P. falciparum and P. vivax (Marfurt et al., 2008). Amplification in this gene might be playing a role in drug resistance against sulfadoxine.

Regions with genes for surface antigens such as 6-cysteine proteins, tryptophan-rich antigens (TRAgs), serine-repeat antigen (SERA), apical membrane antigen (AMA), have shown CNVs. There are 24 TRAgs showing CNVs and all of them have amplifications. In P. vivax species, TRAgs often occur in clusters along chromosomes, suggesting their proliferation via gene duplication and diversification. Certain members of the P. vivax TRAg family have demonstrated binding properties to red blood cells, and in certain instances, interactions with partner molecules have been documented (Kundu et al., 2023). One of the TRAgs encoded by PVX_088850 which lies within region of amplification in our data, has been reported to be interacting with PvMSP7, suggesting possible establishment or stabilization of protein complex at P.vivax merozoite surface (Tyagi et al., 2016).

Amplification is present in regions with cytoadherence-linked asexual protein (CLAG) genes located on chromosomes 7 and 14. In P. falciparum, CLAG proteins has been shown to be involved in the binding of infected erythrocytes to host endothelial cells, a process termed cytoadherence (Holt et al., 1999; Newbold et al., 1999; Gupta et al., 2015) and involved in nutrient uptake and solute transport (Nguitragool et al., 2011; Gupta et al., 2018a). CNVs are found in regions containing genes associated with merozoite surface proteins (MSP-1, MSP-3, and MSP-7), with eight occurring in amplified regions and six in deleted regions. Using the UCSC Genome Browser, we observed that region having several MSP paralogs lying in close proximity on the chromosome was amplified (Figure 4). Additionally, CNVs are present in 23 genes encoding SERA proteins, all located on chromosome 4 of the P. vivax genome, including four in amplified regions and 19 in deleted regions. They form a multigene family and are conserved among Plasmodium species. Recent studies implicate this gene family in a number of aspects in parasite biology and induction of protective immune response (Arisue et al., 2020).

Amplifications in regions with genes for reticulocyte-binding proteins or PvRBP (PVX_090325 and PVX_090330 in chr 5, PVX_098582 and PVX_098585 in chr 7, PVX_094255 in chr 8, and PVX_121920 in chr 14) are present. A study by Kosaisavee et al. also reported variations in the copy number in the reticulocyte-binding proteins Pvrbp2a and Pvrbp2b by a quantitative real-time SYBR Green PCR assay (Kosaisavee et al., 2012). Additionally, a recent study suggests that copy number variations in PvRBP2 of Plasmodium vivax may contribute to its ability to infect Duffy-negative individuals (Pestana et al., 2024).

Rhoptry proteins, such as rhoptry-associated protein and rhoptry neck protein, have amplifications in their genes, specifically PVX_097590 in chr 10, PVX_117880 in chr 12, and PVX_101485 in chr 14.

Some of the regions with genes for ETRAMPS or early transcribed membrane proteins (PVX_096070 in chr 3, PVX_003565 in chr 4, PVX_088870 and PVX_090230 in chr 5, PVX_086915 in chr 7, PVX_118680 and PVX_121950 in chr 12) were shown to have amplifications. ETRAMPs are expressed during Plasmodium intracellular phases and inserted at the parasite parasitophorous vacuole membrane (PVM) (Brandsma et al., 2022).

There are gains in regions having two genes encoding Pvstp1, which is phylogenetically closely related to P. falciparum SURFIN4.2’, a protein exposed on the parasite-infected erythrocyte (iE) surface, and is thus considered to be exposed on P. vivax-iE (Winter et al., 2005).

Our data showed CNVs in 117 unique vir genes present in the sub-telomeric regions of chromosomes. There are 84 genes that are within amplified regions and 10 in regions with deletions. There are 23 vir genes that have both deletion and amplifications (Supplementary Table 3). The vir (variant interspersed repeat) genes in P. vivax have been proposed to play a crucial role in cytoadherence, facilitating the adherence of infected red blood cells to host endothelial cells. This phenomenon might be contributing to the pathogenesis and severity of malaria, as cytoadherence can lead to microvascular obstruction and tissue damage, exacerbating the clinical manifestations of P. vivax infection (Carvalho et al., 2010; Chotivanich et al., 2012; De las Salas et al., 2013; Marín-Menéndez et al., 2013; Totino and Lopes, 2017). It has also been hypothesized that vir genes play a role in malaria pathogenesis and that P. vivax may exploit the extensive sequence diversity within these genes to enhance virulence (Merino et al., 2006; Gupta et al., 2014; Na et al., 2021).

In our analysis, we have also seen CNVs in genes encoding Plasmodium exported proteins of unknown functions. There are 73 gains and 13 losses in the regions harboring these genes. These proteins are secreted by the parasite into the host erythrocyte, where they modulate various cellular processes. In P. falciparum, some exported proteins have been reported to be involved in cytoadherence, allowing infected red blood cells to sequester in different tissues and evade clearance by the immune system while others are implicated in modifying the host cell membrane, leading to the formation of structures such as Maurer’s clefts, which serve as trafficking hubs for exported proteins. Some exported proteins help permeabilize the RBC to allow nutrients and wastes to be exchanged with the blood plasma to facilitate rapid growth and parasite proliferation (Elsworth et al., 2014).

Some genes encoding for proteins belonging to the ubiquitin system machinery and proteasomal degradation pathway were seen to be in the amplified or deleted regions. 20 genes encoding for proteins of the ubiquitin system machinery and 9 genes encoding for proteasomal degradation pathway have shown CNVs. When the parasite is exposed to host molecule there is an increase in transcription levels of genes encoding for proteins related to the Ubiquitin Proteasome (UPS) System (Pereira et al., 2018). The proteasome is the main engine of Plasmodium protein degradation. Protein ubiquitylation plays key roles in cell biology, for example in tagging proteins for proteasomal degradation, or targeting proteins to distinct subcellular locations. The transition from intracellular schizont to extracellular merozoite stages in the asexual blood stage cycle is associated with a general increase in ubiquitylation levels (Green et al., 2020). Ubiquitin modification is associated with altered parasite susceptibility to multiple antimalarials (Ng et al., 2017).

Amplification in 12 genes encoding Phist proteins are present. Studies in P. falciparum have shown that these proteins are involved in different processes including surface display of PfEMP1, gametocytogenesis, changes in cell rigidity. Some phist genes have been shown to be differentially expressed in cerebral malaria and pregnancy-associated malaria, indicating a significant role of Phist in these complications (Warncke et al., 2016).

Approximately 80 genes responsible for encoding proteins forming ribosomal complexes have exhibited copy number variations. This observation aligns with expectations, given the parasite’s multifaceted life cycle necessitating a robust protein synthesis machinery.

Another important amplification is seen in the gene encoding for putative plasmepsin IV (PVX_086040). Plasmepsin IV (Plm IV), an aspartic protease within the food vacuole of the malaria parasite Plasmodium falciparum, participates in the degradation of host hemoglobin by the parasite (Gutiérrez-de-Terán et al., 2006).

CNVs have been seen in the intergenic region as well (Supplementary Table 3). CNVs in intergenic regions can also affect gene expression, although indirectly. While intergenic regions do not contain genes themselves, they often harbor important regions such as promoter and other regulatory elements such as enhancers and insulators, which play crucial roles in regulating the expression of nearby genes (Nelson et al., 2004). When CNVs occur in intergenic regions, they can disrupt the regulatory landscape by altering the number or arrangement of regulatory elements. This disruption can affect the expression of neighboring genes by modifying the accessibility of their regulatory elements to transcription factors and other regulatory proteins.

Additionally, CNVs in intergenic regions can sometimes influence gene expression through long-range interactions between regulatory elements and target genes. These interactions can occur through chromatin looping and other mechanisms, allowing distal regulatory elements to exert control over gene expression. Disruption of these long-range interactions by CNVs can consequently impact gene expression levels. In our study, several CNVs were located within intergenic regions upstream of protein-coding genes. While the precise functional impact of these CNVs remains to be determined, their proximity to genes encoding surface antigens, drug resistance proteins, and stress-related proteins suggests that they could potentially influence gene regulation. However, experimental studies are required to clarify whether these intergenic CNVs play a regulatory role in P. vivax.

Although direct functional validation of intergenic CNVs in Plasmodium vivax has not yet been performed, evidence from malaria research and studies in other human diseases suggests that intergenic regions can have important regulatory roles. In P. falciparum, intergenic regions can give rise to the transcription of regulatory lncRNAs, which may influence gene expression, sexual development, and parasite transmission (Batugedara et al., 2023). Similarly, in humans, CNVs in non-coding regions have been shown to alter gene regulation and impact phenotypic outcomes (Zhang and Lupski, 2015). By analogy, intergenic CNVs in P. vivax may also influence nearby gene expression and contribute to parasite adaptability, highlighting a potential mechanism by which non-coding genomic variation could affect parasite biology. Future studies are required to experimentally validate these effects.