Dengue infection is a mosquito-borne infectious disease, causing a wide spectrum syndrome ranging from benign febrile illness to serious multiple organ failure, also known as severe dengue infection (World Health Organization, 2019). The progression of dengue infection to severe dengue commonly occurs after the febrile phase–between days 3 and 7 from the onset (World Health Organization, 2022). Then, the warning signs that might be associated with the development of severe dengue infection are the parameters of vascular permeability and plasma leakage (Halstead & Cohen, 2015). The impact of adaptive immunity in dengue is well known as a mismatch between the infecting viral serotype and the memory adaptive immunity which exacerbates dengue infection, termed antibody-dependent enhancement (ADE; the interaction between Fc domain of immunoglobulin G (IgG) and Fcγ receptors (FcγRs) on leukocytes (Bournazos et al., 2020) that enhances FcγRs-mediated viral entry mechanism). Accordingly, circulating immune cells and soluble immune components can be detected in peripheral blood and may serve as prognostic markers as a result of the direct consequence of dengue infection.

In contrast, data on the influence of innate immunity in dengue infection are still less. It has recently been suggested that lipopolysaccharide (LPS) (also known as endotoxins) and (1→3)-β-D-glucan (BG) in serum are possibly transferred from the gut (leaky gut syndrome or gut translocation) and may play a major role in the severity of dengue infection through the LPS and BG pro-inflammatory properties, especially through innate immune responses. Serum LPS and BG are proposed as the new prognostic factors at admission to identify dengue patients with poor outcomes (Chancharoenthana et al., 2021). Both LPS and BG are the major cell wall components of Gram-negative bacteria and fungi, respectively, which are the most and second most abundant of organisms in the human gut, respectively (Lagier et al., 2012). In the healthy gut permeability, there is a limited gut translocation of both LPS and BG, partly because the size of these molecules (50-100 kDa) is too large to pass through the normal intestinal tight junction (Benoit et al., 1998; Finkelman, 2020). The injury of the gut tight junction by several causes, including severe dengue infection, induces gut translocation of these molecules allowing the use of these molecules as the indirect biomarkers for gut barrier defect. Moreover, both LPS and BG are foreign microbial molecules for the mammalian host, referred to as “pathogen-associated molecular patterns (PAMPs),” which profoundly activate innate immunity and facilitate adaptive immune responses through several immune cells (Isnard et al., 2021; Tiku & Tan, 2021). Not only LPS and BG, but gut bacterial DNA are also PAMPs with possible gut translocation during leaky gut with a property of innate immunity activation. Although intact bacterial DNAs (genomes) with molecular sizes of 100 to 15,000 kbp (6.5x10⁴–9.8x10⁶ kDa) cannot pass through the gut barrier, bacteria-free DNA is rapidly naturally broken down into smaller sizes of less than 100 bp (lower than 65 kDa) by several processes (depurination or deamination DNA), which can pass across the gut barrier in numerous models. Moreover, exploration of the bacterial genome in the blood (blood bacteriome) are also crudely associated with an alteration in fecal bacteriome (gut dysbiosis) (Kaewduangduen et al., 2022) that might be correlated with the severity of dengue infection as gut dysbiosis between sepsis and non-sepsis are different (Panpetch et al., 2018; Amornphimoltham et al., 2019). Hence, the different patterns of blood bacteriome might be used as another parameter for determining dengue severity.

The early identification of patients with poor prognoses would help physicians provide the proper therapeutic management, and the associated parameters of immune response variability might determine disease severity. Despite the current proposed dengue prognostic parameters, including C-reactive protein polymorphism variants (Mukherjee & Tripathi, 2020), ferritin (Suresh et al., 2020), granulocyte-macrophage colony-stimulating factor (GM-CSF) (Lee et al., 2019), interferon gamma-inducible protein-10 (IP-10) (Chen et al., 2006), and vascular injury biomarkers (Vuong et al., 2021), leaky gut-associated parameters and innate immunity determination are also interesting. Furthermore, very little is known about the physiologic consequences of leaky gut, blood bacteriome, and gut dysbiosis in dengue as host-microbiome during dengue infection which can be beneficial or harmful to the host (González & Elena, 2021). Here, the prospective single-center analysis on patients with dengue infection was conducted to see if the biomarkers of gut translocation, bacteriomes, and innate responses could be used as the prognostic factors in dengue infection. Targeting gut barrier dysfunction and reducing pathogen translocation are emerging strategies for the prevention and treatment of severe dengue infection (Chancharoenthana et al., 2021) and may be useful in other inflammatory diseases.

Here, intestinal permeability impairment (leaky gut) in dengue patients was associated with blood bacteriome alteration and elevated pathogen molecules, lipopolysaccharide (LPS), and (1→3)-β-D-glucan (BG), in serum that activated several immune cells, especially the non-classical monocytes and natural killer (NK) cells.

Although dengue virus (DENV) causes a mild self-limiting illness in the majority of individuals, many cases are annually reported as severe dengue (World Health Organization, 2022) that, at least in some cases, is severe enough to cause leaky gut syndrome with translocation of pathogen molecules from the gut into the blood circulation (gut translocation). Because there might also be some direct impacts of DENV on enterocytes as DENV-V2 could directly infect Caco-2 cells, an intestinal cell line (Kar et al., 2019), the studies on intestinal aspects in patients with dengue are interesting. There was no correlation between dengue severity, as determined by endothelial leakage and bleeding tendency, and serotype of DENV (data not shown). Here, we demonstrated gut barrier defect and gut dysbiosis through several parameters (endotoxemia, serum BG, and LMER) and blood bacteriome, respectively. Gut barrier defect in systemic infections, including viral infection (such as COVID-19) (Sirivongrangson et al., 2020; Saithong et al., 2022) or viral diarrhea (such as Rota viral infection) (Warfield et al., 2006) is previously mentioned. The translocation of pathogen molecules from gut into blood circulation (gut translocation) affects immune responses in several condition (Panpetch et al., 2018; Amornphimoltham et al., 2019). During dengue-induced gut translocation, systemic inflammatory responses against these pathogen molecules (bacterial DNA, LPS, and BG) activates antigen-presenting cells, NK cells, T cells, and B cells (Fasano, 2011), induces profound pro-inflammatory cytokines, and enhances more severe gut translocation, including the viable organisms in some situations with severe gut barrier defect (Elinav et al., 2011; Leelahavanichkul et al., 2016a). Although the high level of LPS in serum induces several acute co-morbidities (mostly the systemic inflammation) through the activation of other immune cells (such as neutrophil extracellular traps) (Saithong et al., 2022) or the presence of other organelles in the blood (such as mitochondrial DNA) (Kaewduangduen et al., 2022), these parameters are not explored in the current study. More studies are needed.

Here, in dengue infection, our previous study has highlighted the importance of enhanced circulating LPS and BG as biomarkers of impaired intestinal permeability in comparison with lactulose-to-mannitol excretion ratio (LMER) (a standard biomarker), which might be associated with the severity of dengue infection (Chancharoenthana et al., 2021). In the present study, we focused on dengue patients with positive LMER test in either acute febrile phase or critical phase regardless of dengue severity. Several inflammatory mediators (Fasano & Shea-Donohue, 2005) along with prominent alteration in circulating bacteriome (Mu et al., 2017) were demonstrated in patients with gut barrier defects. The correlation among LMER, endotoxemia, and serum BG for the determination of gut barrier defects are previously mentioned (Chancharoenthana et al., 2021). Although several characteristics in the critical phase (plasma leakage and bleeding tendency) seem to be more severe than in the febrile phase, systemic inflammation (serum IL-6), the possible major cause of enterocyte injury (Leelahavanichkul et al., 2016b) ( Table 1 ), and gut barrier defect (Amornphimoltham et al., 2019), were similar between these phases of dengue infection. However, the plasma leakage in the critical phase might have more intestinal impacts than the febrile phase that is severe enough to cause intestinal dysbiosis as indicated here by a higher abundance of Bacteroidaceae and Escheriachia in blood bacteriome or the increased lactulose-digestible bacteria by others (Clausen & Mortensen, 1997) (LMER test interference). The direct intestinal immunofluorescent-based histopathology (Boonhai et al., 2021) or other methods of gut barrier evaluation are interesting to use. Nevertheless, an increase in the abundance of these Gram-negative bacteria in the intestine possibly elevates LPS in gut contents and induces a higher level of endotoxemia in severe dengue infection. Perhaps prominent intestinal hypoxia from plasma leakage in patients with the critical phase might be responsible for the selection of pathogens in the critical phase as most pathogens have some virulence factors that enhance their viability in the harsh microenvironments (Walther & Ewald, 2004).

Because both Bacteroidaceae (Martens et al., 2009) and Escheriachia (Lucas et al., 2017) are possible pathogens, the increase of these bacteria in the gut and their DNA in the blood might be a poor prognostic factor of dengue infection. Likewise, a decrease in Bifidobacteriaceae, the beneficial bacteria (Raethong et al., 2021), in dengue (febrile and critical phases) when compared with control also highlighted dengue-induced gut dysbiosis. Perhaps the dysbiosis is started from reduced Bifidobacteriaceae in the febrile phase followed by increased Bacteroidaceae and Proteobacteria in the more severe critical phase which might be another predictor for dengue severity. Although the sequential exploration in the several time-points in the same patients with the disease progression from febrile into critical stages is not performed here, data of blood bacteriome in COVID-19 patients with endotoxemia from gut barrier defects demonstrate the progressive reduction of Firmicutes (the possible beneficial bacteria) DNA in blood at day 3 and day 7 post-COVID-19 compared with the 1st day of the infection (Sirivongrangson et al., 2020). Hence, the progression of gut dysbiosis along with the disease severity is possible and the manipulation of the dysbiosis, using probiotics, might be beneficial. More studies on this topic in patients and mice are needed. Not only bacterial DNA (Kaewduangduen et al., 2022), but LPS and BG (Issara-Amphorn et al., 2018) in serum from dengue-induced leaky gut syndrome also induce pro-inflammatory effect and more profoundly activate monocytes (Wu et al., 2013) and NK cells (Poggi et al., 2019; Chapuy & Sarfati, 2020) compared with patients without gut barrier defect. Indeed, monocytes are important immune cells that recognize pathogen molecules in several inflammatory diseases in either sepsis (Aguilar-Briseño et al., 2020) or non-sepsis (Grip et al., 2007; Mukherjee et al., 2015; Weldon et al., 2015). Although both LPS and BG can stimulate monocytes (Chan et al., 2009; Kew et al., 2017), the simultaneous presence of LPS with BG synergistically activates inflammation (Issara-Amphorn et al., 2018) which might be similar to the condition of dengue-induced leaky gut. On the other hand, the chronic presence of LPS might induce endotoxin tolerance, a reduced LPS responses after the first dose of LPS (Ondee et al., 2019), partly through the interference of cellular energy status (Gillen et al., 2021), which might be associated with ineffective organismal control after viral infection (post-dengue secondary bacterial infection) (Sharma, 2020). Perhaps bacterial DNA, LPS, and BG in dengue patients with leaky gut affects non-classical monocytes and NK cells that alter the pathogenesis of dengue infection ( Figure 4 ). We hypothesize that systemic inflammation (a similar serum IL-6 level in both febrile and critical phases; Table 1 ) (Leelahavanichkul et al., 2016b), together with the direct DENV infection in enterocytes (Kar et al., 2019) induces gut leakage and gut translocation of several pathogen molecules that facilitates systemic inflammation in both stages of the infection. Although serum IL-6 is a biomarker associated with the severity of bacterial sepsis (Calandra et al., 1991), the IL-6 level could not differentiate between dengue in febrile and critical phases. Hence, the systemic inflammation in both febrile and critical phases in patients with gut barrier defects might be more severe than the patients without leaky gut. On the other hand, dengue patients with acute febrile phase demonstrated increased peripheral blood non-classical monocytes (CD14^-CD16⁺) and NK cells (CD56⁺CD16^-) compared with those in the critical phase ( Table 2 ) possibly resulting in better control of the disease severity by innate immunity in the febrile phase over the critical stage. Perhaps the expanded CD14^-CD16⁺ monocyte subsets in the acute febrile phase are responsible for the effective viral control and disease recovery in the later phase (Narasimhan et al., 2019), leading to a lower disease progression rate during the febrile phase at 62% (five from eight cases) compared to 82% (19 from 23 cases) of those with the critical phase. Likewise, prominent NK cells (CD56⁺CD16^-) in the febrile phase compared to the critical phase supported early NK cell responses against DENV (Green et al., 1999; Chau et al., 2008; Keawvichit et al., 2018). A lower rate of disease progression in the febrile phase is possibly due to an impaired NK cell activation during the acute phase of dengue infection (Chau et al., 2008). Although the total abundance of T cells may not represent DENV-specific responsive cells, the increase in the absolute CD4 and CD8 T cells in the febrile phase compared to the critical phase ( Table 2 ) might be a better DENV control of patients in the febrile phase (less severe than critical phase). Taken together, non-classical monocytes, NK cells, and cytotoxic T cells (CD8) may have a key role in the control of the dengue virus in the early stage of patients with severe infection. Dysregulation and aberrant innate immune responses further lead to reduced viral control and severe dengue by inducing vascular leakage and excessive inflammation due to the high levels of inflammatory cytokines (Malavige et al., 2020).

For the clinical translation, the evaluation of gut barrier defects in patients with dengue and the strengthening gut integrity (probiotics, short-chain fatty acid, and/or zonulin alteration) (Seo et al., 2021; Hoilat et al., 2022), LPS absorption (Lipcsey et al., 2020), and BG removal (Ohata et al., 2003) might be the interesting adjuvant therapies that attenuate the disease mortality. More studies are warranted. There were several limitations in our study, including the limited parameters of innate immune analyses (cannot rule out the effect of other cells against dengue). Due to the cross-sectional nature of the analysis with a limited number of patients and blood microbiome evaluation, studies in more time points might be informative. Next, the recruitment of only dengue patients with intestinal permeability that were detected using oral administration-dependent lactulose-to-mannitol excretion ratio (LMER). Then, the very severe dengue infection (shock or hemorrhage) could not be enrolled (unable to do the LMER test). Finally, the study was not directly designed for the prediction of clinical outcomes, although we hypothesized that both innate and adaptive immune parameters could help determine fatal outcomes. Further studies on the topic are needed.

In summary, leaky gut and gut translocation of pathogen molecules might cause sepsis in dengue infection through the activation of non-classical monocytes and NK cells. Blood bacteriome and the manipulations on gut dysbiosis might be useful for biomarker or adjuvant therapy of dengue-induced leaky gut. The increased understanding of immune responses in dengue infection is likely to move toward the rational development and the use of novel approaches for disease modulation in the near future.

The datasets presented in this study can be found in online repositories. The name of the repository and link to the data can be found below: NCBI; PRJNA838435.

The study was approved by the Ethics Committee of the Faculty of Tropical Medicine, Mahidol University (MUTM 2020-071-01). The patients/participants provided their written informed consent to participate in this study.

Conceptualization, WC, SK, WA, and AL; methodology, WC and SK; validation, WC and WA; formal analysis, WC and AL; investigation, WC and AL; resources, WC and WA; data curation, WC and WA; writing—original draft preparation, WC, SK, and AL; writing—review and editing, WC, AL, VT, WP, and PW; visualization, WC, AL, WP, and PW; supervision, AL, VT, WP, and PW; project administration, WC and SK; funding acquisition, WC. All authors have read and agreed to the published version of the manuscript.

This work was funded by Mahidol University–Basic Research Fund: fiscal year 2021 (FRB640032) (Contract No. BRF1-A42/2564) (WC).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.