Date Published: August 12, 2010
Publisher: Hindawi Publishing Corporation
Author(s): Sansanee Noisakran, Nattawat Onlamoon, Pucharee Songprakhon, Hui-Mien Hsiao, Kulkanya Chokephaibulkit, Guey Chuen Perng.
Dengue has been recognized as one of the most important vector-borne emerging infectious diseases globally. Though dengue normally causes a self-limiting infection, some patients may develop a life-threatening illness, dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS). The reason why DHF/DSS occurs in certain individuals is unclear. Studies in the endemic regions suggest that the preexisting antibodies are a risk factor for DHF/DSS. Viremia and thrombocytopenia are the key clinical features of dengue virus infection in patients. The amounts of virus circulating in patients are highly correlated with severe dengue disease, DHF/DSS. Also, the disturbance, mainly a transient depression, of hematological cells is a critical clinical finding in acute dengue patients. However, the cells responsible for the dengue viremia are unresolved in spite of the intensive efforts been made. Dengue virus appears to replicate and proliferate in many adapted cell lines, but these in vitro properties are extremely difficult to be reproduced in primary cells or in vivo. This paper summarizes reports on the permissive cells in vitro and in vivo and suggests a hematological cell lineage for dengue virus infection in vivo, with the hope that a new focus will shed light on further understanding of the complexities of dengue disease.
Dengue is one of the most important mosquito-borne viral diseases affecting humans, with over half of the world’s population living in areas at risk. Originally, dengue virus infections occurred mainly as epidemics in tropical and subtropical countries. But over time, with increasing globalization and the geographic spread of inhabitants of Aedes aegyti and Aedes albopictus mosquitoes, the dominant vectors for dengue virus transmission, dengue virus infection has steadily penetrated every corner of the world [1, 2]. Dengue virus has four serotypes, and each of them can cause a spectrum of diseases ranging from asymptomatic, mild febrile (dengue fever, DF) to a life-threatening illness, dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS). Approximately 50 to 100 million people contract dengue fever annually, and about 200,000 to 500,000 of these are DHF/DSS, which has a mortality rate about 1%–5%, mainly in children under 15 years of age .
Dengue viruses, similar to other flaviviruses, possess a positive single-stranded RNA genome packaged inside a core protein, which is surrounded by an icosahedral scaffold and encapsidated by a lipid envelope. Its 11 kb genome functions similar to mRNA, encoding a polyprotein which upon translation is cleaved into three structural proteins (C, prM/M, and E) and seven nonstructural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) by viral or host proteases. Since dengue viral genome can function as mRNA, if the viral RNA can be delivered into a cell’s cytoplasm through biologically active vesicles, translation and genome synthesis can occur accordingly .
Viremia is a common clinical manifestation in several severe viral infections. However, dengue viremia is unique because in endemic regions, where majority of the population has demonstrable neutralizing antibody to all four dengue serotypes , viremia still occurs in some of these populations upon bitten by mosquitoes carrying infectious dengue virus. The reasons why certain individuals developed clinical illness are not known, although an individual’s genetic background, the interval between reinfection, sequence of infection by specific serotype, and quality of immune responses may partially account for the differences [4, 8]. Since identifying the permissive cell lineage(s) in vivo may uncover the underlying mechanisms leading to DHF/DSS and aid in vaccine and antiviral drug development, the source(s) of circulating virus in acute dengue patients has been the central focus for several decades. In spite of the efforts made to identify these cell(s), the question remains elusive.
In vitro, numerous primary cell lineages and established cell lines have been studied for their relative permissiveness for dengue virus infection, including endothelial and fibroblast cells, myeloid-derived cells, and lymphocytes [9–17]. Although some of the cells defined in vitro could be permissive cells for dengue virus replication in vivo [18–21], the actual phenotypes of these cells have not been delineated or defined in detail. Consequently, conflicting reports abound in the literature.
Dengue disease is introduced to its hosts by the bite of mosquitoes carrying infectious virus. The first obstacle that the mosquito encounters is the physical barrier of the skin, which is composed of several layers of keratinocytes interspersed with a network of capillaries (Figure 1). Keratinocytes are on the outermost epidermal layer of the skin, are endowed with Toll-like-receptors (TLR) , and may be considered a component of the primary innate immune system. Langerhans cells mainly reside in the thin layer of the epidermis, which does not contain capillaries, while dendritic cells are predominantly in the thicker dermis layer, which is filled with capillaries. Although Langerhans cells, in general, have the same phenotype as dendritic cells, and is impossible to distinguish activated Langerhans cells from dendritic cells by morphological appearance, numerous studies indicate that biological activities are discernible between these two cell types [41–44]. Many interesting questions can be asked. How does dengue virus interact with skin cells during mosquito probing prior to penetration? How deep does the mosquito fascicle penetrate into the skin? How does dengue virus behave upon contacting epidermal and dermal innate immune cells after the mosquito fascicle penetrates? And how does dengue virus get deposited and disseminated during the engorgement period while the mosquito imbibes the blood? The answers to these questions can elucidate how the fates of the cells on or in the skin are orchestrated.
Gordon and Lumsden, the authors of a historical in vivo frog’s web paper in 1939, observed that the mosquito’s proboscis is flexible and predominantly imbibes blood directly from the capillary and only occasionally from the pools formed in the tissues by the leakage of blood from the capillary previously lacerated by the mosquito’s proboscis . This study is later confirmed in mice ear and human beings implementing the same experimental designs [46, 47]. The dimensions of an Aedes aegypti fascicle are typically 1.8 mm in length with an internal radius of 10 μm  and typically engorge a blood meal of 4.2 μl in 141s . It is estimated that during imbibing, approximately 50% (~0.9 mm) of the fascicle penetrates into skin , suggesting that the location of blood drawn from is the capillary-rich dermis layer, implicating that pathogens may be directly injected into the blood.
Mosquito probing, penetration, and feeding on the surface of the skin is easily interrupted by the movement of the host. Unsuccessful imbibing may result in a certain amount of virus deposited on the outermost layers of skin, where keratinocytes, Langerhans, and dendritic cells may encounter the virus. The delicate alarm system of the skin can sense the probing of the mosquito and the penetration of the fascicle, potentially initiating a signaling cascade and the activation of defense mechanism. Thus, if these dendritic cells are permissive as others suggested [27–33], we would anticipate quite high incidence of the dengue cases in endemic regions during the rainy season. The critical role of these antigen presenting cells (APCs) is to ingest foreign particles including viruses, process these materials while migrating to the regional lymph nodes. Here, the APCs can present the foreign proteins to other immune cells, such as T cells, to initiate the cascade of the adaptive immune responses, including antibody production. Dendritic cells, therefore, may be more important for the induction of the host’s defense. Importantly, it is of benefit to the host that the virus be engulfed and processed in order to generate an adequate immune response against the invading pathogen and protect the host from further infection. Since such phagocytic cells are the first line of defense in our body, this may perhaps explain why a majority of dengue cases are asymptomatic.
Since dengue viral antigens are detectable in adherent cells obtained from the peripheral blood of dengue patients, monocytes and/or macrophages have been an assumptive target cell for more than three decades. With the high level of interest in the pathogenesis of DHF/DSS, intensive efforts have been made to identify the infected monocytes and/or macrophages in the peripheral blood of infected patients, and some suggestive successes have been documented. However, dengue is a timing disease. Specimens collected from dengue patients are often after the onset of clinical manifestations; therefore, the intervals prior to symptoms developed are different among individuals and are likely at the peak of dengue viremia, and autopsy samplings are always at the convalescent stage or later. Within the context, identifying a cell that is positive for dengue viral antigens in collected specimens requires meticulous investigations and cautious interpretations. Although recently researchers are attempting to address the issue with small animal models, such as the AG129 mice experimentally infected with dengue virus, the major pitfall of this model is that mice have a defective interferon response, which has been shown to play a very critical role in controlling virus replication and proliferation. Consequently, dengue viral RNA or antigens are observed in almost all the cells and organs that have been investigated [18, 21]. Within the same content, this same group investigated the autopsy tissues from patients who died of dengue virus infection. The authors showed that human tissues and the corresponding mice AG129 tissues were positive for dengue virus NS3 antigen, concluding that these cells propagated virus. However, the phenotypic markers of the cells that were positive for dengue viral antigen were not confirmed, and thus a conclusion was drawn based upon the similarities between humans and mice. Also, a new finding suggests that liver sinusoidal CD31+ endothelial cells in AG129 mice are positive for dengue viral antigen and can support the antibody-mediated infection . However, evidence indicates that there are many differences in immunological and antiviral responses between humans and mice [51–53]. Thus, clarifications of the role of monocytes and macrophages in dengue virus infection in vivo are urgently needed. This notion is also applied to the paper published by Jessie et al. , in which the cell phenotype markers in those cells positive staining for either dengue viral antigens or RNA, were not confirmed.
Retrospective literature reviews reveal that in bone marrows aspirated during the recovery stage or immediately after death, phagocytic clasmatocytes contain normoblastic, lymphocytic, granulocytic, erythrocytic, and platelet-like remnants in their cytoplasm [60–62]. Infected leukocytes (or monocytes) are frequently present on the last day, at the end of viremia, or the day after the disappearance of the virus from the plasma , suggesting that leukocytes may play an essential phagocytic role in the clearance of circulating virus. Recently, the phagocytic phenomenon has been confirmed in dengue hemorrhagic nonhuman primate model . Due to difficulties and inconsistencies in identifying the cell lineages responsible for dengue viremia at the acute stage, monocytes and/or macrophages are gradually being assumed as the main cells for dengue virus propagation for the following reasons: (i) like the cells that can propagate the virus, they can adhere to cell culture flasks [63, 65], (ii) they are capable of phagocytosis [23, 66], and (iii) infrequently observed dengue viral antigens in cells with a similar morphology in tissues obtained postmortem [20, 67, 68]. These observations then led to the postulated hypothesis of antibody-dependent enhancement (ADE)  in an attempt to explain the epidemiological observation in which secondary infection with subsequent heterologous dengue serotypes is a risk factor for DHF/DSS . The ADE theory is used to explain the severe dengue virus infection; antibody to the first infection may not be sufficient enough to neutralize a heterologous infection, and this partial cross-reacting antibody (or subneutralizing antibody) may promote Fc-bearing cells such as monocytes and macrophages to opsonize the virus, leading to increased virus production.
The reason why dengue viruses are capable of infecting a wide range of immortalized cell lines, such as myeloid-originated, B, T, fibroblast, and endothelial cells but yet comparatively poor at replicating in primary cells is currently unknown. Perhaps, it is likely that cell factors that are altered in immortalized cell lines contribute to this differential permissiveness. Immortalized cell lines are normally transformed with viruses, such as SV40 or EBV, which encode viral gene products that have an effect on interferon-signaling. Interestingly, among the cell mediator repertoire, interferon expression appears to be a very crucial element limiting the propagation of dengue virus [14, 82, 83]. In addition, defects in interferon signaling pathway has been shown in cancer cells, such as lymphoma and leukemia and established immortalized cell lines [84–88]. This line of evidence may, to some extent, explain why cell lines, such as Vero and K562 cells, which lack a functional interferon system, are highly permissive to dengue virus infection. In addition, activation of interferon-stimulated genes are the constant findings in cells with relatively poor permissive for dengue virus [14, 89, 90] and in specimens obtained in dengue-virus-infected humans and rhesus monkeys [89, 91, 92]. Within the same content, it is interesting to review what has been investigated in paucity of dengue animal models.
Currently, no perfect animal model that recapitulates the cardinal features of human DHF/DSS is available, even though a recent dengue hemorrhagic monkey model appears to be promising for dengue hemorrhagic investigation . Since understanding the mechanisms leading to viremia and disease is necessary for vaccine and antiviral drug development, efforts have been made to search and/or generate a suitable dengue animal model. The readers should refer to recent review articles on the subject in smaller animals [93, 94]. This paper focuses mainly on why dengue viremia is seen in these animal models.
Studies over the years with specimens collected from the peripheral blood of dengue patients reveal that virus can be recovered or detected in a variety of cells. However, a general consensus concerning which cell lineages are involved in dengue viremia has never been conclusive, partly due to the variation of timing in specimen collection. Upon admission to the hospital with clinical symptoms, patients are always several days after the infection and frequently at the peak or downturn in viremia. By that time, a complex network of immune responses initiated and is in the action of viral clearance. Perhaps, this may explain why immune cells are commonly associated with the detection and/or isolation of virus in dengue patients . Thus, the cells that are infected early, before the peak in viremia, and accounting for dengue viremia are still unknown.
A new lineage of cell—MEP or CD41+CD61+ cells, such as megakaryocytes and/or platelets—is suggested for a potential cell accounting for dengue viremia in vivo. The objective of the authors is to draw scientific attention to the highly fragile cell with unusual biological properties in acute dengue virus infection. After all, hemostatic defects in DHF appear to be a major clinical finding. Our aim is to foster more detailed investigations of the MEP or CD41+CD61+ cells in specimens collected from acute dengue patients, which conceivably will not only provide a piece of valuable information of the mechanisms associated with DHF/DSS, but will also pave a new way on the formulation of effective candidate vaccines or antiviral drugs development.