In this analysis, pooled transcriptome data from whole blood cells of malaria patients, which involved MM, NCM and CM were evaluated. Although transcriptome organization is reported to be poorly preserved between brain and blood, there are some brain co-expression modules which demonstrate strong evidence of preservation in blood and could be useful for identifying blood biomarkers for neurological diseases [50]. Therefore, some extrapolations can be made from transcriptome data from blood (which is more readily accessible) to deduce genes and pathways associated with neurological conditions. Previous investigations by Thiam et al. and Boldt et al., which were employed in this analysis had successfully used blood transcriptional analysis to detect changes in gene expression in MM, NCM and CM. In this study, it was discovered from the cluster analysis that the 3 states of P. falciparum infection have some common characteristics between them although certain CM states could be distinctly different. While CM was mainly associated with NCM, there was some clustering between MM and CM. Although, there are standards set to define the different states of falciparum malaria, clinical symptoms of the disease can be protean and non-specific without a clear distinction between MM, NCM and CM. This is manifested clinically as fever, malaise, headache, nausea and vomiting, though MM could rapidly progress into CM if not treated quickly [51, 52]. Furthermore, a study by Taylor et al. showed that the classification of CM in children is sometimes misdiagnosed, where coma has other causes and parasitaemia is incidental [53]. This suggests that the detection of CM-specific characteristics can be challenging, which could result in its clustering with the other states of falciparum malaria.

In addition, genes that are expressed in the 3-P. falciparum-induced conditions were identified. It was discovered that 2876 genes were expressed in common between them and CM surprisingly had the least number of expressed genes. In all, genes associated with host response to parasite invasion, inflammatory response, cellular stress, detoxification of toxic metals and clathrin mediated vesicle assembly were up-regulated. It has been established that entry of malaria parasite into the host elicit immune response, which can lead to oxidative stress [54]. Oxidative stress caused by the release of parasite toxins such as HMZ or inflammatory response to P. falciparum entry have been implicated in cellular stress, especially in severe malaria [5, 55]. Moreover, levels of metals such as copper and zinc are increased in acute malaria, which could be as a result of the host’s inflammatory response against parasite and lysis of erythrocytes [56]. Concurrently, there is trafficking of the virulence protein, P. falciparum erythrocyte membrane protein 1 (PfEMP1) to the surface of infected red blood cells (iRBCs), which involves membranous organelles termed Maurer’s clefts [57]. Clathrin and its adaptor proteins possibly participate in the process hence its upregulation in the 3 disease states [58]. The genes connected to mental retardation and AD were also present in all 3-forms of falciparum malaria. Previous investigations have documented the predisposition of children with CM to develop mental disorders probably as a result of ischaemic and hypoxic neural injury [59, 60]. The elevation of the expression levels of these genes in MM and NCM implies the possible detrimental nature of these P. falciparum infections on the brain as well. Indeed, some studies have documented cognitive deficits in children and adults, who suffered from acute MM and NCM [61–63]. For instance, Hickson et al. detected that children with acute kidney injury in NCM developed neurocognitive impairments in both behavioural regulation and executive function [64]. Additionally, Guha et al. in using an animal model proved that a single incident of MM induced microglial activation, neuroinflammation and behavioural changes accompanied by increase in pro-inflammatory cytokine expression in the brain [65]. Hence, the severity of neurological deficits reported in CM could be attributed to the down-regulation of various genes involved in cell survival, protein synthesis and mitochondrial respiration. Additionally, a study by Bodlt et al. discovered the down-regulation of a number of genes in CM of Gabonese children due to expression of hypoxia-induced genes aryl hydrocarbon receptor (AhRF), GA binding protein transcription factor (GABP) and hypoxia-inducible factor-1 (HIF1) [30]. Brain hypoxia has been documented in CM owing to sequestration of infected erythrocytes, vascular congestion, impaired perfusion and endothelial cell activation [66, 67].

With regards to pathways associated with the different forms of falciparum malaria, KEGG enrichment analysis unveiled the association of MM with the neurodegenerative diseases, PD and PrD. This means MM and some neurodegenerative diseases may share parallel biological processes, which is contradictory to the findings of Thiam et al. [28] and Cabantous et al. [29], where neurodegenerative pathways were found in CM only. Nonetheless, processes involved in cancer, EBV infection and DNA damage repair were the most significant. Some human pathogens, including P. falciparum and EBV, have been implicated as etiologic agents in the development of cancer [68]. The invasion of P. falciparum and subsequent release of HMZ into the host induces inflammation, which facilitates ROS production and cellular stress [69, 70]. ROS production even at low levels can lead to gene and chromosomal mutations via DNA double strand breaks (DSBs), which is one of the most severe forms of DNA damage [71]. DSBs when unrepaired or mis-repaired, can cause stress, cell death, chromosome instability and cancer [72]. EBV infection is also associated with ROS generation [73, 74] and since cancer development can stem from DNA damage, it is not surprising that these pathways were elevated in MM. Again, DNA damage and DNA damage response are hallmarks of neurodegenerative diseases including PD and PrD [75]. The up-regulation of the neurodegenerative pathways in MM suggests a shared cytotoxic mechanism between these diseases and buttresses earlier observation that MM could have implications on brain function.

Similarly, KEGG pathway enrichment assessment was done for NCM and CM. Platinum drug resistance, nucleotide excision repair and glutathione metabolism were associated with NCM. Resistance to anti-malarial drugs is a growing challenge in malarial treatment [76]. The up-regulation of a drug resistance pathway similar to that observed in cancer during NCM implies a parasite facilitated host resistance to anti-malarials. This might be as a result of host cellular modification especially with DNA damage and repair mechanisms, which were also up-regulated in NCM. For instance, multiple efflux and uptake transporters are present in red blood cells and hepatocytes [77, 78]. Increase in and genetically altered drug efflux transporters reduce anti-malarial drug uptake and concentrations in red blood cells [79]. This ensues in minimum effective drug concentration, inhibiting the termination of malaria infection and potentially, the development of resistance. The association of glutathione metabolism in NCM is an indication of glutathione modulation of disease progression particularly with regards to anti-malarial treatment. A study by Zuluaga et al. revealed that amodiaquine treatment failure was connected to erythrocytic glutathione [80]. This is because glutathione competes with amodiaquine for the haem group, which could result in therapeutic failure [80]. Again, the concentration of glutathione was decreased in iRBCs, which is an indication of the antioxidant consumption due to ROS generation [81]. Thus, glutathione though an important antioxidant defense in malaria may likely modulate the host’s response to drug treatment. The pathways that correlated with CM were retrograde endocannabinoid signalling and NAFLD. Previous studies have described the up-regulation of certain genes involved in neurodegenerative diseases including amyloid precursor protein (APP), ubiquilins (UBQLN) and Jun proto-oncogene (JUN) as candidate genes for NAFLD [82–85]. In their study, Karbalaei et al. showed that there were 190 genes expressed in common between NAFLD and AD, and that NAFLD has an undoubted relation to AD [85]. Moreover, the endocannabinoids and their receptors (especially cannabinoid 1, CB1) are known to induce steatosis and lipogenic gene expression resulting in NAFLD [86, 87]. The cannabinoid 2 (CB2) receptor on the other hand regulates neuro-inflammatory responses and affects various macrophage functions, including antigen uptake and presentation and chemokine/cytokine production [88–91]. A study by Alferink et al. reported that mice with a deletion of the CB2-encoding gene (Cnr2⁻/⁻) and immunized with Plasmodium berghei (ANKA strain) erythrocytes showed enhanced survival and diminished BBB disruption in experimental CM [88]. Furthermore, the CB1 receptor antagonist rimonabant, is reported to have neuroprotective properties in some animal models of neurodegenerative disorders [92]. In addition, endocannabinoid molecules were significantly increased in the acute phase of P. falciparum infection in children [93]. Together, these results reveal that genes involved in NAFLD and the endocannabinoid system could serve as potential biomarkers for malaria severity. This could also be an indication that the consequence of CM on the brain may comprise some neurological pathways other than those classically described for neurodegenerative diseases.

Additionally, the GOs up-regulated in the 3 forms of falciparum malaria were assessed and in MM analysis showed biological processes related to biogenesis, host metabolic processes, DNA damage repair activities and anti-inflammation with the up-regulation of IL-10. The imbalance between pro-inflammatory and anti-inflammatory cytokines drives disease severity in falciparum malaria. Boeuf et al. observed a high TNF/IL-10 ratio in NCM patients, whereas MM patients had a relatively balanced IL-10 and TNF levels [94]. Thus, in MM the deleterious effect of P. falciparum invasion seems to be counteracted by an increase in the host’s metabolic and damage control mechanisms against ROS and inflammatory products.

The GOs elevated in NCM revealed destabilization of cellular cytoskeleton, ROS production, DNA damage checkpoint and amyloid-beta clearance. The surge in ROS levels during oxidative stress is implicated in the increase in the activation of MAP3Ks pathways [95]. An aberration from the strict control of MAP3Ks signalling is connected to the development of many neurodegenerative diseases including AD and PD [96]. Activated MAP3K signalling pathways are thought to contribute to neurodegenerative pathogenesis through the phosphorylation of APP and α-synuclein, leading to aggregates formation that trigger neuronal apoptosis [97–98]. Again, an increase in MAPKs signalling induces cytoskeletal abnormalities through unusual phosphorylation and consequent aggregation of cytoskeletal elements [99–101]. For example, the hyper-phosphorylation of the microtubule associated protein, tau results in its aggregation, microtubule destabilization and the formation of neurofibrillary tangles with subsequent degeneration of neurons [102–105]. Also, axonal injury with the accumulation of APP has been reported in post-mortem studies of CM patients, both adults and paediatric cases [106–108]. In a previous study, brain swelling as a consequence of venous congestion by iRBCs was observed in NCM patients on admission [109], pointing to a possible neuronal injury. Accordingly, the amyloid-beta clearance observed in NCM could represent a reversible pathway to neurological damage in NCM. With respect to CM, abnormal signalling, neutrophil degranulation, repression of gene expression and apoptosis were the predominant GO terms. Using an experimental model of CM, a recent study by Kumar et al. reported an increase in D1 and D2 dopaminergic receptor expression, phosphorylated dopamine- and cAMP-regulated phosphoprotein (DARPP), p25, cyclin-dependent kinase 5 (CDK5), Ca²⁺/calmodulin-dependent protein kinase IIα (CaMKIIα) and D1-D2 heteromers [110]. The dysregulation of the dopaminergic receptors led to the impairment and degeneration of medium spiny neurons [110]. Findings from neuroimaging studies have demonstrated that the anatomical and functional grouping of dopaminergic neurons in striatum performs a significant role in the execution of cognition and behavioural consequences in young adults [111, 112]. The cognitive deficit observed in CM survivors therefore could be attributed in part to the dysregulation of the dopaminergic system. Although neutrophils are scarce in the central nervous system under normal physiological conditions, they infiltrate the brain in several pathological conditions, including malaria [113]. The accumulation of these granulocytes lead to the release of neutrophil extracellular traps (NETs) which directly damage the BBB, surrounding neurons and consequently apoptosis [49]. Additionally, Boldt et al. in their findings showed a down-regulation of pre-synaptic synuclein alpha (SNCA), hydroxyacylglutathione hydrolase (HAGH), ankyrin 1 (ANK1), ferro chelatase (FECH), STAT1, homolog A, nucleotide excision repair (RAD23A) and 2,3-bisphosphoglycerate mutase (BPGM) genes in CM patients [30]. Hence, the up-regulation of genes involved in gene repression is an indicator that several genes involved in cellular metabolism and function are down-regulated in CM, which accounts for the severity of the disease as compared to MM.

Most outstanding of the inflammatory processes found associated with the genes up-regulated in CM were neutrophil activation and IL18-signalling. Previous publications have described IL18 as a major player in the pathogenesis of severe malaria through a pathway of elevating IFN-gamma [114]. The activation of this pathway is further supported by the transcription factor IRF1, for which there was a regulatory role in CM via Metascape analysis. The relevance of neutrophils for inflammatory aggravation of malaria has been described by Knacksted et al. [47]. This is mediated by the extrusion of chromatin in the form of NETs upon neutrophil death. In addition, up-regulation of genes linked to herbicide response reiterate the significance of the “plant-like” apicoplast in P. falciparum virulence [115]. The GO term regulation of axon diameter found in the down-regulated genes in CM provides a mechanistic clue of neurological impairments in CM. The involved genes XK and KEL were furthermore reported to induce neuropathological abnormalities, such as giant axons when knocked out in mice [116]. The protein interaction network besides relevant sub-networks of IL-18 signalling and the blood group proteins, membrane transport protein X-linked-Kx (XK) and kell metallo-endopeptidase (KEL) regulating axon diameter showed a large connected area centered by TUBB2A and the metalloproteinases 8 and 9 (MMP8, MMP9). TUBB2A has been reported to be up-regulated in CM by Cabantous et al. [29] as part of the Parkin-Ubiquitin proteasome degradation pathway, which when impaired can be deleterious to dopaminergic neurons [117]. MMP8 and MMP9 which is activated by HMZ have been associated with CM by Polimeni and Prato due to their potential to disrupt the sub-endothelial basement membrane and process and regulate cytokines [16].

A limitation of this study is that, the number of patients enrolled in the investigations were few and hence larger studies are required to determine the pernicious effect of P. falciparum infection on the brain. Furthermore, the study was based only on blood-derived cells and included no brain biopsies, thus implying that some neurological consequences during and after malaria infection are already manifested in the genetic networks in blood-derived cells. However, for the patients included in this study it is not known how many, if any, experienced neurological sequelae subsequent to malaria infection.