REPORT

Dengue is the most significant arthropod-borne viral disease in the world. It is estimated, fundamentally existing in tropical and subtropical areas of the world, 100 million infections and more than 20 000 deaths are caused by dengue every year. This disease is caused by any of the four dengue virus (DENV) serotypes, an RNA virus of the family Flaviviridae. These viruses can be produced from a benign infection, named dengue fever (DF) to the most severe forms, dengue hemorrhagic fever and dengue shock syndrome (DHF/DSS). The most important pathophysiological event in DHF/DSS is plasma leakage that causes pleural effusion, ascites and occasionally death (1, 2).

DHF/DSS pathophysiology is not entirely understood. However, a number of studies associate severe dengue with DENV sequential infections by a phenomenon named antibody-dependent enhancement, high viral loads, complement activation, and cytokines storm (1-3). Although apoptosis (a form of programmed cell death) has been involved in dengue pathogenesis (1, 4), the mechanisms in humans have not been extensively explored (5, 6).

In the present study, different formalin fixed and paraffin embedded tissues from fatal DENV cases from two Cuban dengue epidemics were evaluated. Considering the growing evidence of apoptosis association with dengue pathogenesis (1, 4-6), we considered that detecting apoptotic cell death in these tissues with the goal of exploring the possible contribution of apoptosis to DHF/DSS pathophysiology is important.

RESULTS AND DISCUSSION

Different formalin fixed and paraffin embedded tissues from 8 autopsies of adult subjects who had died of DHF/DSS were examined. The cases were classified as grade IV DHF/DSS according to WHO guidelines (2). These individuals were previously diagnosed with a secondary dengue infection by serology, viral isolation in mosquito’s cell culture and/or molecular detection of DENV genome. The fatal cases were from DENV-2 Santiago de Cuba epidemic in 1997, and DENV-3 epidemic in 2001 from Havana city (2). Six autopsies were evaluated from DENV-2 epidemic and two autopsies from DENV-3 epidemic. Our study included liver, brain, intestine, lung and heart tissues. Medical records were reviewed, and microscopic/macroscopic pathological findings were collected.

Immunohistochemical assay using an indirect immunoperoxidase method with mouse monoclonal and polyclonal anti-DENV antibodies produced at our institution detected viral antigens in tissues sections (7). The detection of apoptotic cells was carried out through the observation of characteristic morphological changes of apoptosis in cells and by the TUNEL reaction assay (in situ Cell Death Detection Kit, Roche Applied Science) according to the manufacturer’s instructions. Positive and negative controls of both techniques were performed as it had previously been described (Figure 1A) (5, 6).

In this study, the presence of DENV antigens was detected in several hepatic midzonal necrotic areas, including some Kupffer cells, as had been documented earlier (5, 6). Six out of the 8 studied cases had apoptotic cells, 5 subjects from DENV-2 epidemic and one case from DENV-3 epidemic. Cells with a positive TUNEL reaction and morphological changes of apoptosis were observed in sections from the liver, brain, intestine and lung (Figure 1 B, C). These results significantly increased the number of human tissues that had been involved with apoptosis in den- gue disease and highlighted that our work is the largest apoptosis study carried out so far in tissues from dengue cases. In this respect, it is very relevant to mention that apoptosis in tissues from DENV cases had only been previously demonstrated in liver cells from an adult traveler in France that died of DENV-2 disease (5) and in five Vietnamese children with fatal DENV-3 infection (6).

The increased risk of severe dengue in individuals suffering from sickle cell anemia has previously been reported in Cuban dengue epidemics (9). Case 1 had sickle cell anemia, and apoptosis in cerebral tissues was demonstrated. We cannot completely exclude that abnormal characteristic S hemoglobin of sickle cell anemia has played a role. This postulate takes into account the possible induction of neuronal apoptosis by a neurotoxic mechanism of S hemoglobin that has been described previously in an in vitro study (10). Interestingly, our study demonstrated apoptotic endothelial cells from the intestinal microvasculature (case 3) and pulmonary alveoli (case 7). It has been reported that DENV infection of human cell lines of endothelial cells induce apoptotic cell death (4). However, it has also been postulated that immune mechanisms through proinflamatory mediators in serum as TNF-α and anti-NS1 (non structural protein 1 of DENV) antibodies that cross-react with surface proteins on endothelial cells may induce the apoptosis of these cells during DHF/DSS. These in vitro studies suggest that apoptotic cell death of endothelial cells might be associated with plasma leakage (1).

Taking into consideration our results we cannot rule out completely that DENV replication triggers apoptosis in endothelial cells, however viral antigens were not detected in lung and intestinal tissues by immunohistochemistry in our study (Table). In a previous work, the detection of DENV antigens was documented in the lung microvascular endothelium; yet, the in situ hybridization did not demonstrate the presence of viral RNA (11).

A recent study has associated severe dengue through plasma leakage with high levels of C5b-9 complex causing complement activation by DENV NS1 (3). Therefore, if we consider that all the studied cases had DHF/DSS, the C5b-9 complex could be involved in apoptosis induction of endothelial cells as has been demonstrated in a previous experimental study (12).

In this work, apoptosis was demonstrated in microvascular endothelial cells of lung and intestine. This is the first confirmation of apoptosis of endothelial cells in tissues from dengue cases, a prior event only described in in vitro studies and laboratory animals (4). Although this finding was not shown in all the examined fatal cases with plasma leakage, this phenomenon may be associated to pulmonary edema described in case 7, and intestinal edema and ascites found in case 3 (Figure 2). Additionally, we cannot exclude that apoptotic endothelial cells of intestinal microvasculature of case 3 (Figure 1 B) are related to gastrointestinal bleeding found in this individual (Figure 2), if we take into consideration a very recent mouse model of hemorrhage in dengue infection that demonstrated the apoptotic endothelial cells of subcutaneous tissues were associated with local tissue hemorrhage (13).

It is significant that case 3 suffered from type 2 diabetes mellitus. Remarkably, precedent reports in type 2 diabetes mellitus (14) have demonstrated similar dysfunctions of endothelial cells and increased production of the same citokines showed in patients with dengue disease. These earlier diabetes mellitus studies (14) suggest that the increased risk of severe dengue in patients with type 2 diabetes mellitus may be related with abnormalities of microvascular endothelium and citokines hiperproduction as probably happened in studied case 3. In this regard, the likely increased risk of severe dengue in diabetics was reported for the first time during Cuban dengue epidemics (2). However, it is interesting that an increment of dengue severity has also been documented recently in diabetics from Southeast Asia (15).

In this study, the presence of apoptotic white blood cells was demonstrated within the pulmonary microvasculature of case 7. This finding supports previous reports documenting a higher number of apoptotic peripheral blood mononuclear cells in DHF than in DF (Figure 2). Additionally, very recent works of transcriptomes of peripheral blood mononuclear cells obtained during acute DENV infection indicate an activation of genes with pro-apoptotic effect (1).

RELEVANCE OF THE STUDY

So far, the present work is the largest apoptosis research performed in human tissues of DENV infected individuals. To our knowledge, this work shows for the first time apoptosis in white blood cells, brain cells, intestine and pulmonary endothelial cells from microvasculature in DENV cases. The apoptosis of microvascular endothelial cells may be associated with plasma leakage and hemorrhage during DHF/DSS. This study suggests a risk of severe dengue likely increased in association with apoptosis in patients that suffer from type 2 diabetes mellitus and sickle cell anemia. In addition, apoptotic cerebral cells might be related to hypoxic and ischemic mechanisms during DHF/ DSS. At present, a specific anti-viral therapy to dengue is not available; however, our investigation contributes to clarify the possible implications of anti-apoptotic treatment in severe dengue.

Summarizing, this is the largest apoptosis study in tissues of DENV infected subjects and the first work ever demonstrating apoptotic cell death in white blood cells, brain cells, intestinal and pulmonary microvascular endothelial cells from DENV cases. These results suggest that apoptosis may contribute to DHF/ DSS pathophysiology.

ACKNOWLEDGMENTS

The novel findings of this work obtained the Cuban National Academy award in 2009 and a number of colleagues were involved. We are very grateful for the assistance of researchers from the Virology Department at the Pedro Kourí Tropical Medicine Institute. They are Gissel García, Beatriz Sierra, Susana Vázquez, Maritza Pupo-Antúnez, Mayra Muné, Lídice Bernardo, and Luís Sarmiento. We also would like to thank the anatomic pathology support of Dinorah Betancourt and María C López.