Megakaryopoiesis and thrombopoiesis

Megakaryocytes are large haematopoietic cells ranging from 20 to 100 μm. The name ‘mega’ (large) – ‘karyo’ (nucleus) reflects their appearance, with a large and multilobulated nucleus that encompasses most of their granular cytoplasm. The nuclei of mature megakaryocytes are characterized as hyperploid, with an average ploidy of 16 N DNA, but could go up to 128 N (Ref. Reference Noetzli, French and Machlus20). Megakaryocytes are specialized cells that serve as the precursor in platelet biogenesis, where the granulated cytoplasm will pinch off, releasing up to 10⁴ platelets (Ref. Reference Pang, Weiss and Poncz21).

Megakaryopoiesis, or the process of producing mature megakaryocytes, is orchestrated by several factors that play different roles at different stages of megakaryopoiesis. It begins with haematopoietic stem cells (HSCs) differentiating into common myeloid progenitors (CMPs). CMPs will then differentiate into megakaryocyte–erythrocyte progenitors (MEPs). MEPs will continue to differentiate along the megakaryocytic lineage to produce megakaryocyte progenitors (MKPs) or megakaryoblasts and eventually form mature megakaryocytes to produce platelets (Refs Reference Geddis22, Reference Couldwell and Machlus23).

Thrombopoietin (TPO) is a key factor that binds to the c-Mpl receptor to activate several signalling pathways involving Janus kinase, signal transducer and activator of transcription protein 3 and 5 (STAT3 and STAT5), p38 mitogen-activated protein kinase (p38 MAPK), extracellular signal-regulated kinase, phosphoinositide-3 kinase (PI3K) and protein kinase B (AKT) to initiate megakaryopoiesis (Ref. Reference Kaushansky24). Activation of these signalling pathways increases the expression of transcription factors such as GATA1, FOG1, FLI1, MYB and nuclear factor erythroid 2 (NF-E2), which upregulate genes crucial for megakaryocytes such as CD41, CD42b and CD61 (Refs Reference Noetzli, French and Machlus20, Reference Kaushansky24). TPO also increases the number of MKP cells, induces polyploidy and promotes megakaryocyte maturation (Ref. Reference Kaushansky24).

In addition to TPO, various cytokines such as the interleukin family (IL-1b, IL-3 and IL-6) work in synergy with TPO to promote the proliferation of MKP cells (Ref. Reference Noetzli, French and Machlus20). IL-1b has also been shown to upregulate the gene and protein levels of TPO, c-Jun, c-Fos, GATA-1 and NF-E2 in a dose-dependent manner, which likely serves as the basis for thrombocytosis during inflammation (Ref. Reference Chuen25).

Some cytokines and factors influence megakaryopoiesis independent of TPO-mediated signalling. For example, IL-1a induces proplatelet shedding into the bone marrow during megakaryocyte maturation (Ref. Reference Noetzli, French and Machlus20). C-C motif ligand 5 (CCL5) or regulated on activation, normal T cell expressed and secreted chemokine (RANTES) can also increase megakaryocyte ploidy and platelet production through CCR5 signalling (Ref. Reference Machlus26). A study has identified an enzyme, tyrosyl-tRNA synthetase variant (YRSACT), that can directly convert HSCs into MKPs to enhance megakaryopoiesis and thrombopoiesis of human induced-pluripotent cells that lack TPO signalling in vitro and in vivo (Ref. Reference Kanaji27).

Thrombopoiesis is the process of platelet formation and is highly dependent on the development of mature megakaryocytes during megakaryopoiesis (Ref. Reference Severin, Ghevaert and Mazharian28). During this process, cytoskeletal proteins, membrane and granulated cytoplasm of mature megakaryocytes undergo extensive remodelling and form pseudopodial projections called proplatelets (Ref. Reference Geddis22). The proplatelets, characterized by long, thin shafts with swelling at the tip, are then released into the blood vessels (Ref. Reference Kaushansky29). The production of proplatelets requires intracellular factors such as the NF-E2 transcription factor and extracellular factors such as estradiol (Ref. Reference Nagata30) and shear force (Ref. Reference Thon31). TPO, in addition to its central role in megakaryopoiesis, is also a crucial factor in thrombopoiesis (Ref. Reference Kaushansky24). As previously mentioned, TPO-mediated signalling elevates the level of the NF-E2 transcription factor, which is paramount in megakaryocyte maturation (Ref. Reference Noetzli, French and Machlus20). NF-E2 is also paramount in thrombopoiesis, as knocking out this factor resulted in severe thrombocytopenia because of impaired thrombopoiesis but no effects on megakaryopoiesis, as the megakaryocytes in these mice, though morphologically comparable with normal megakaryocytes, were unable to generate proplatelet extensions (Ref. Reference Lecine32). Thrombopoiesis and its critical factors and regulators have been extensively reviewed elsewhere (Refs Reference Noetzli, French and Machlus20, Reference Noh33, Reference Tijssen and Ghevaert34).

Dengue impairs megakaryopoiesis and thrombopoiesis

Several viruses can disrupt megakaryocyte proliferation and maturation. Herpesviruses such as human cytomegalovirus and human herpes simplex virus impair megakaryopoiesis by causing apoptosis (Ref. Reference Raadsen35). Pathogenic Hantaan orthohantavirus has been reported to infect and replicate in mature megakaryocytes by hijacking CD61 surface protein (Ref. Reference Lutteke36).

DENV causes platelet reduction in patients, beginning typically on day 2 before the onset of the critical phase and persisting until days 6–7 (Ref. 1). One mechanism of thrombocytopenia is megakaryocyte infection and death. DENV has been shown to efficiently infect human megakaryocyte cell lines, primary human megakaryocytes or progenitors and megakaryocytes in humanized mice (Refs Reference Vogt37, Reference Noisakran38, Reference Lahon, Arya and Banerjea39).

MEG-01 is one of the most common megakaryoblastic leukaemia cell lines in thrombocyte studies. Lahon et al. reported an effective DENV infection and replication in TPO-treated MEG-01 cells, compromising the PI3K/AKT/mTOR pathway that is essential in megakaryocyte survival and maturation (Ref. Reference Lahon, Arya and Banerjea39). DENV infection also caused significant cell death and repressed the expression of megakaryopoiesis-related transcription factors, namely GATA-1, GATA-2, NF-E2 and mature megakaryocyte marker CD61 (Ref. Reference Lahon, Arya and Banerjea39). Another study verified the susceptibility of MEG-01 cells to DENV infection, but reported a reduced DENV replication in phorbol-12-myristate-13-acetate (PMA) pre-treated MEG-01 cells when compared with control and MEG-01 cells treated with PMA after DENV infection (Ref. Reference Banerjee40). In this study, Banerjee et al. showed that viral RNA copy number was significantly low in PMA-pre-treated cells at 2 dpi compared with control and PMA treatment post-infection samples. The authors stated that these PMA-differentiated MEG-01 cells were refractory against DENV infection/replication (Ref. Reference Banerjee40). However, this statement contradicts several reports showing high DENV infection in mature megakaryocytes (Refs Reference Vogt37, Reference Noisakran38, Reference Attatippaholkun41). A plausible explanation for the lower DENV replication in PMA-pre-treated cells could be because of PMA activity that activates protein kinase C (PKC), which in turn, inhibits DENV NS5 and suppresses DENV replication (Ref. Reference Noppakunmongkolchai42). In contrast, PMA treatment post-DENV infection substantially increased DENV replication in MEG-01 cells (Refs Reference Banerjee40, Reference Kaur43). Similar observations were reported in K562 cells treated with PMA after DENV infection (Ref. Reference Kaur43). PMA treatment post-DENV infection significantly increased DENV replication and infectious progeny production without influencing viral entry. In this study, PMA suppressed cellular reactive oxygen species (ROS) production by upregulating antioxidant NFE2L2 transcription factor to allow virus replication (Ref. Reference Kaur43). It is also possible that the maturation process with highly active transcriptional and translational machineries inadvertently create an intracellular milieu suitable for viral replication and virion production (Ref. Reference Singh44).

In addition to cell lines, primary CD34⁺ MKPs obtained from umbilical cord blood are also prone to DENV infection, leading to disrupted colony formation and elevated apoptosis (Ref. Reference Basu45). Similarly, another study reported that umbilical cord blood cells were highly susceptible to DENV infection, and these CD34⁺ cells serve as the reservoir of viral progenies (Ref. Reference Vats46). Interestingly, these progenies were mostly latent.

In a humanized mouse model, a higher percentage of DENV infection in mature CD41a^+/− + CD42b⁺ megakaryocytes (35%) was observed when compared with immature CD41a⁺ + CD42b⁻ megakaryocytes (1.5%) (Ref. Reference Vogt37). However, the number of mature human megakaryocytes in this model was greatly diminished (Ref. Reference Vogt37). Nevertheless, this finding is consistent with another study where DENV displayed a selective tropism for CD42-expressing MEG-01 cells compared with CD42⁻ or CD41⁺ cells (Ref. Reference Attatippaholkun41). Another study, using patient samples and rhesus monkeys, showed CD61+ cells were susceptible to DENV infection (Ref. Reference Noisakran38).

In addition to impaired megakaryopoiesis, DENV infection can hamper thrombopoiesis by significantly reducing proplatelet formation in PMA-treated MEG-01 cells (Ref. Reference Banerjee40). The exact mechanism of this effect is currently unclear but may involve the NF-E2 transcription factor, a crucial thrombopoiesis factor. DENV infection markedly reduced NF-E2 protein expression in mature MEG-01 cells (Ref. Reference Lahon, Arya and Banerjea39). Another potential mechanism may involve the sphingosine 1-phosphate receptor (S1pr). S1pr1, one of the sphingosine-1-phosphate receptor subtypes, has been studied in dengue infection settings but mostly on its roles in regulating vascular permeability where sphingosine 1-phosphate (S1P) levels were found markedly reduced in acute dengue patients (Ref. Reference Gomes47). S1pr1 was reported to be crucial in platelet production, where knocking out of S1pr1 in mice resulted in severe thrombocytopenia concomitant with significant reductions in proplatelet formation and proplatelet shedding (Ref. Reference Zhang48). Thus, the S1P–S1pr1 axis may also be implicated in dengue-mediated thrombopoiesis impairment.

A study unravelled a novel role of DENV NS3 protease in cleaving GrpE protein homologue 1 (GrpEL1), a cochaperone of mitochondrial heat shock protein 70 (mtHsp70) (Ref. Reference Gandikota49). GrpEL1 and its orthologue, GrpEL2, form a subcomplex that is pivotal in maintaining the stability of nucleotide exchange factor for mtHsp70, thus modulating critical mtHsp70 functions such as importing mitochondrial pre-protein into the matrix (Ref. Reference Srivastava50). Overexpression of NS3 protease reduced GrpEL1 protein level in vitro. More importantly, GrpEL1 protein level was significantly reduced in dengue patient samples, especially in dengue haemorrhagic fever (DHF) and dengue shock syndrome (DSS) patients, which correlated with thrombocytopenia level in these patients (Ref. Reference Gandikota49). Reduced GrpEL1 compromised mitochondria health and functions, leading to thrombocytopenia (Ref. Reference Gandhi51). Another form of DENV protease, NS2B-NS3 protease, which localized to the nucleus, could cleave erythroid differentiation regulatory factor 1 (EDRF1) transcription factor (Ref. Reference Gandhi51). EDRF1 is critical in platelet formation as it regulates the levels of GATA1 and spectrin, a cytoskeletal protein essential for membrane re-organization during platelet biogenesis (Ref. Reference Patel-Hett52). Overexpression of NS2B-NS3 protease significantly diminished EDRF1 and GATA1 protein levels. Similarly, EDRF1 level was also reduced in dengue patients, especially in thrombocytopenic dengue haemorrhage patients and dengue shock syndrome patients (Ref. Reference Gandhi51). Thus, NS3 contributes to thrombocytopenia by cleaving EDRF1 to impair thrombopoiesis, and cleaving GrpEL1 to cause mitochondrial dysfunction and disrupt platelet formation.

In short, DENV can infect megakaryocytes at different stages of cell maturation and impair platelet production, leading to reduced platelet formation.

Platelet activation

The roles of platelets during haemostasis and thrombosis events begin with a series of activation processes that include platelet adhesion, network extension and thrombus formation involving multiple factors, as summarized in Table 1 (Ref. Reference Thomas and Storey53). Briefly, in the event of high shear or vascular injury, nearby platelets will be exposed to von Willebrand factor (vWF) on the vessel wall (Ref. Reference Freedman54). The vWF will serve as a bridge to connect platelet glycoprotein (GP) Ib receptor complex on the platelet's membrane with the collagen layer, allowing platelet GPVI receptors to adhere directly onto the collagen matrix for a strong adhesion (Ref. Reference Estevez and Du55). The vWF also facilitates the activation of circulating platelets by connecting the GP IIb/IIIa receptor on the activated platelets to the GP Ib on the circulating platelets to activate them. Bound fibrinogen will strengthen the crosslink of the GP IIb/IIIa receptor on both activated platelets (Ref. Reference Jennings56). The activated platelets undergo structural changes with the formation of pseudopodia and initiate the release of platelet granules (Ref. Reference Jennings56). The released granules contain chemokines and cytokines to signal other circulating platelets to interact with the adhering platelets and form extensions of activated platelet network (Ref. Reference Herter, Rossaint and Zarbock57).

Table 1. Factors in platelet activation and aggregation

Platelets contain α- and dense granules. α-Granules are the most abundant secretory organelles in platelets, and they contain adhesive proteins, cytokines and chemokines crucial for platelet adhesiveness and thrombus formation (Ref. Reference Heijnen and van der Sluijs65). These chemokines such as RANTES, CXCL1, platelets factor 4 (PF4 or CXCL4) and IL-6 interact with platelet surface receptors and immune cells (Ref. Reference Smith66). RANTES (CCL5) binds to chemokine receptors CCR1, CCR3 and CCR5 on platelets to stimulate other platelets and recruit monocyte to the inflamed endothelium (Ref. Reference Gear and Camerini67). PF4 also facilitates platelet adhesion and mediates leucocyte interaction (Ref. Reference Fox68).

Dense granules, on the other hand, contain small molecules such as adenosine diphosphate (ADP), adenosine triphosphate (ATP), calcium ions serotonin and phosphates (Ref. Reference Heijnen and van der Sluijs65). ADP is crucial in platelet activation through its interaction with two G-protein-coupled receptors (P2Y1 and P2Y12) to initiate platelet aggregation and provide a feedback mechanism to increase the secretion of thromboxane A2 (TxA2) and other agonists (Ref. Reference Offermanns69). Studies showed that the absence of P2Y12 in humans resulted in haemorrhage, and lacking P2Y1 in mice prolonged the bleeding time (Ref. Reference Gachet70). In addition to procoagulant activity, ADP also mediates the release of TxA2 where the interaction of TxA2 with TxA2 receptor (TP) attracts other platelets to bind to the adherent platelets, forming a stable thrombus (Ref. Reference Jennings56). Thrombin plays essential roles in plug formation through protease-activated receptor 1 and 4 (PAR-1 and PAR-4) that are coupled to Gq and G12/G13 proteins, respectively (Ref. Reference Chen71). In addition, hormones such as serotonin activate platelets through 5-hydroxytryptamine 2A (5HT-2A) receptors and promote platelet aggregation (Ref. Reference Marcinkowska72).

Platelet death and clearance

Platelets generally have a life span of 5–10 days before clearance from the body (Ref. Reference Tanaka, Bolliger, Hemmings and Egan73) via several mechanisms such as platelet apoptosis, antibody-mediated phagocytosis in the spleen, removal by the Kupffer cells in the liver via lectin-glycan recognition or massive platelet release because of blood loss (Refs Reference Grozovsky, Hoffmeister and Falet74, Reference Quach, Chen and Li75). Ageing platelets express a higher level of phosphatidylserine (PS), a type of phospholipid that activates apoptosis (Ref. Reference Hou76). Other pro- and anti-apoptotic factors such as Bak, Bax and Bcl-XL are also involved in platelet apoptosis (Ref. Reference Quach, Chen and Li75). Thrombin-activated platelets undergo metabolic exhaustion marked by mitochondrial depolarization, accumulation of ROS and ATP depletion, followed by platelet dysfunctional and disintegration, calpain-activation and eventually platelet fragmentation (death) (Ref. Reference Kim77). Activated and apoptotic platelets are also exposed to clearance by macrophages. Desialylation of platelet via neuraminidase or vWF action induces platelet clearance and acute thrombocytopenia (Ref. Reference Quach, Chen and Li75). Platelets are also cleared from the body through platelet plug clearance (Ref. Reference Ojha16).

Dengue increases platelet activation and apoptosis

Invading pathogens such as bacteria, parasites and viruses activates platelets via pattern recognition receptors such as Toll-like receptors (TLRs) and C-type lectin receptors (CLRs), complement Fc receptors (FcR), major histocompatibility complex (MHC) class I and many more (Refs Reference Attatippaholkun41, Reference Li, Li and Ni78, Reference Fard79). Excessive platelet activation and apoptosis, marked by elevated P-selectin and annexin V level, respectively, was observed in dengue patients during the febrile and defervescence phases, causing thrombocytopenia in these patients when compared with healthy controls, non-dengue febrile patients and non-thrombocytopenic dengue patients (Ref. Reference Hottz64). Similarly, increased platelet activation concurrent with thrombocytopenia was observed in dengue patients during the critical phase (days 4–6) (Ref. Reference Ojha16). Dengue-induced thrombocytopenia was replicated in vivo when mice injected with DENV2 suffered thrombocytopenia concomitant with elevated platelet activation (Ref. Reference Masri63). Several mechanisms for platelet activation and death mediated by the DENV are summarized below.

DENV NS1. Following infection, DENV hijacks the translational machinery to produce new dengue proteins such as dengue non-structural 1 (NS1) protein and new virions. NS1 binding to TLR4 on primary human platelets enhanced platelet activation marked by enhanced P-selectin expression, causing platelet apoptosis (Ref. Reference Chao58) and the released of granule-stored chemokines such as RANTES, macrophage migration inhibitory factor (MIF) and PF4 (Ref. Reference Quirino-Teixeira80). Full-length NS1 and all NS1 domains, especially the wing domain, could activate platelets, but only the full-length NS1 could induce platelet aggregation in the presence of platelet agonists (Ref. Reference Garcia-Larragoiti81). The increased P-selectin expression on NS1-activated platelets promoted platelet adherence to endothelial cells and macrophages, compromising endothelial permeability and increasing platelet phagocytosis, respectively (Ref. Reference Chao58). DENV NS1 also amplified thrombo-inflammatory responses and induced the synthesis of pro-IL-1β but not its secretion (Ref. Reference Quirino-Teixeira80). IL-1β secretion from NS1-activated platelets was mediated by nucleotide-binding domain leucine-rich repeat-containing protein (NLRP3)-caspase-1 inflammasome (Ref. Reference Quirino-Teixeira80). More importantly, the secretion of pro-inflammatory cytokines such as IL-1β and MIF contributed to vascular permeability and tissue injury (Ref. Reference Pan82). DENV NS1 also sensitized the platelets to aggregation where exposure to DENV NS1 alone could not stimulate platelet aggregation, but co-treatment of NS1 with subthreshold levels of platelet agonists markedly increased platelet aggregation which contributed to thrombocytopenia (Refs Reference Chao58, Reference Garcia-Larragoiti81).

DENV envelope protein. Dengue can infect cells via two CLRs, specifically DC-SIGN (dendritic cell-specific intercellular adhesion molecule-3-grabbing nonintegrin) and CLEC-2 (C-type lectin-like-receptor). The N-glycans on the DENV envelope protein can bind to DC-SIGN, causing virus adsorption onto the plasma membrane (Ref. Reference Alen83). The binding of DENV to DC-SIGN was shown to increase platelet activation, cause mitochondria dysfunction and trigger caspase-dependent apoptosis (Ref. Reference Hottz64).

As previously mentioned, co-stimulation of NS1-activated platelet with suboptimal ATP concentration upregulated IL-1β secretion, possibly through NLRP3 inflammasome (Ref. Reference Quirino-Teixeira80). Similarly, DENV envelope domain III (EIII) also activated platelet, and caused thrombocytopenia in vivo primarily through NLRP3-mediated pyroptosis (Ref. Reference Lien84). EIII also impaired clotting time, and interestingly, inhibition of this response pathway via NLRP3 inhibition or EIII blocking could significantly improve the conditions (Ref. Reference Lien84).

Other factors. Proteomics analysis comparing platelets from dengue patients with healthy controls showed 167 differentially abundant proteins, with most of the proteins involved in ‘antigen processing and presentation’ and ‘platelet activation’ processes (Ref. Reference Trugilho85). As expected, P-selectin, PF4 and RANTES were among the highly upregulated factors in dengue (Ref. Reference Trugilho85). Histones H2A and HLA class I were also upregulated in dengue-infected platelets. The binding of circulating histone H2A to the TLR4 receptor on platelets caused platelet activation that could be completely blocked by anti-histone H2A monoclonal antibody (Ref. Reference Trugilho85).

CLEC-2 also plays a pivotal role in platelet activation. Elevated CD62p and CD63 expression after DENV infection was abolished in human platelets treated with CLEC2-mAb and clec2^−/− mouse platelets (Ref. Reference Sung, Huang and Hsieh86). More importantly, the study reported that activated platelets via CLEC2 produced extracellular vesicles that activated neutrophils and macrophages via CLEC5A and TLR2 and contributed to the formation of neutrophil extracellular traps (NETs). NET level in the plasma was reported to inversely correlate with platelet count and positively correlate with P-selectin expression during days 7–13 of disease in acute dengue patients (Ref. Reference Garishah87).

Mast cell-derived serotonin was reported to be elevated during dengue infection in vitro and in vivo and activated the platelets through the 5HT-2 receptor, which was then demonstrated to be phagocytosed (Ref. Reference Masri63). Inhibiting the 5HT-2 receptor reduced dengue-induced thrombocytopenia and platelet activation (Ref. Reference Masri63). However, although the roles of serotonin in platelet activation, aggregation and phagocytosis were sufficiently established in vivo (Ref. Reference Masri63), an earlier metabolomic study showed a significant reduction of serotonin in dengue high fever patients compared with dengue fever patients, especially during the febrile and defervescence phases (Ref. Reference Cui88). The study also proposed that reduced serotonin coupled with elevated interferon gamma may serve as a biomarker of severe dengue. The conflicting reports on serotonin levels post-dengue infection could be because of the different hosts where, as opposed to rodents, mast cells are not the primary source of serotonin in humans (Ref. Reference de Mast and de Groot89). Thus, the role of serotonin in dengue and dengue-mediated thrombocytopenia requires further investigation.

Another important mechanism of platelet activation is through antibody-dependent enhancement or ADE. This mechanism involves the interaction between virus-immunoglobulin (Ig) G complexes and FcR-bearing cells (Ref. Reference Koupenova, Livada and Morrell90). Unlike other factors, ADE is primarily active during secondary DENV infection as the IgG antibody secreted during the first infection binds to the second-generation virus and permits faster and more severe infections via FcR expressed on immune cells and platelets (Ref. Reference Teo91). A previous study showed that immature dengue virion becomes highly infectious in the presence of anti-prM antibodies that facilitate viral entry and enhance the intracellular processing of prM to M protein (Ref. Reference Rodenhuis-Zybert92). Another study demonstrated that DHF/DSS patients showed elevated afucosylated IgGs with strong affinities towards FcγRIIA and FcγRIIIA receptors expressed on the platelets, leading to platelet reduction. The level of afucosylated IgGs correlated with thrombocytopenia severity in these patients and was a significant risk factor for thrombocytopenia (Ref. Reference Wang93). As FcRs are expressed on immune cells, it was postulated that thrombocytopenia observed in this study could be because of enhance platelet activation, platelet sequestration and/or antibody-dependent cell cytotoxicity (Ref. Reference Wang93). Additionally, afucosylated IgGs could also serve as a prognostic factor of dengue disease severity (Ref. Reference Bournazos94).

Dengue increases platelet clearance

The activated and apoptotic platelets are subsequently cleared from the body in several ways. Phagocytosis of apoptotic platelets by macrophages in acute and early convalescence dengue patients was reported to be 2.5–3.5 times higher than the platelets from healthy controls (Ref. Reference Alonzo95). Phagocytosis was substantially reduced when these patient-derived platelets were pre-treated with D89E mutant protein that masked the PS on the apoptotic platelets (Ref. Reference Alonzo95). Phagocytosis of DENV2-activated platelets by primary human monocytes was also reported in vitro (Ref. Reference Ojha16).

The vWF could also mediate platelet clearance. vWF, elevated in thrombocytopenic dengue patients, increased platelet desialylation, thus exposing the platelet to clearance via Ashwell–Morell receptor-mediated pathway (Ref. Reference Riswari60). DENV-activated platelets also showed high levels of bound complement factor C3 and IgG, suggesting elevated platelet lysis, clearance and increased thrombus formation that caused platelet depletion in the plasma (Refs Reference Ojha16, Reference Wang93). Immune-mediated platelet clearance and lysis during dengue infection have been reviewed here (Ref. Reference Hottz96).

The events and processes described in the sections ‘Dengue reduces platelet production’ and ‘Dengue increases platelet activation and clearance’ are not mutually exclusive and may simultaneously occur to exacerbate dengue symptoms. Figure 2 summarizes these events. Since the pathophysiology and progression of dengue fever vary among patients, it is challenging to determine the predominant mechanism for thrombocytopenia in patients. Nonetheless, research in this field has improved disease management. An emerging application of knowledge in this field is immature platelet fraction (IPF), which will be briefly discussed in the next section.

Figure 2. Summary of DENV-mediated thrombocytopenia. DENV causes thrombocytopenia in several ways. (1) DENV impairs megakaryopoiesis. DENV infects and causes apoptosis of megakaryocytes and the progenitor cells. DENV also prevents megakaryocyte maturation. (2) DENV impairs thrombopoiesis by reducing megakaryocytes and interfering with platelet formation. (3) Secondary DENV infection causes elevated platelet activation via IgG–FcR connection in the ADE process. (4) DENV increases platelet activation. (5) DENV causes cytokines, chemokines and other factors to be released from infected platelets and nearby endothelial cells. (6) DENV infects and causes platelet apoptosis. (7) ADE, elevated platelet activation and secretion of various factors cause the platelets to coagulate. (8) Coagulated platelets and apoptotic platelets are cleared from the circulation by phagocytes. Regular arrows (→) in green indicate stimulatory modifications. Blunt-ended arrows (˧) in red indicate inhibitory modifications.