Elevated Soluble HLA-G Levels Associate with Dengue Severity in Vietnamese Patients

Statements of the problem : Dengue pathogenesis is complex and not fully understood. The virus may evade the immune system through mechanisms like the increased expression and/or secretion of immune inhibitory molecules. This study examines the association of the soluble human leukocyte antigen G (sHLA-G) with dengue in hospitalized patients. Method of study: A total of 238 dengue patients and 118 healthy controls were recruited. Dengue-infected patients were confirmed by real-time RT-PCR and clinically categorized into different severity groups. Laboratory parameters were assessed on admission. Plasma sHLA-G levels were measured by a commercial ELISA. Results: sHLA-G levels were significantly higher in dengue patients (median [range]: 42.7 [7.10 - 1300] U/mL) compared to healthy controls (median [range]: 11.1 [4.7 - 620] U/mL) (p<0.001). After adjusting for age, sex and disease severity, a significant association between sHLA-G plasma levels (log-transformed) and the days of illness was observed (β=0.1, p=0.033). Cases requiring strict medical monitoring presented significantly higher sHLA-G level (median [range]: 51.0 [7.17 - 525] U/mL), compared to cases without warning signs (median [range]: 38.0 [7.10 - 1300] U/mL) (p=0.011). While liver enzymes positively correlated with sHLA-G levels in all patients, total lymphocyte counts inversely correlated with sHLA-G levels in severe cases (r=-0.78, p-value=0.002). Conclusions: sHLA-G levels are associated with dengue warning signs and severe cases, suggesting a role in disease pathogenesis. Thus, this soluble protein might be a valuable marker to improve the accuracy of identifying severe cases and support clinical monitoring.

The dengue virus (DENV) is responsible for massive outbreaks of viral haemorrhagic fever worldwide, particularly in tropical and subtropical regions (1). Vietnam bears an annually high incidence of dengue with a significant impact on the health and economic system (2). Clinically, the disease can range from non-specific acute fever to haemorrhagic manifestations with altered haemostasis, possibly leading to shock or multi-organ failure (3). The pathogenesis of dengue is complex and has not yet been elucidated in detail, as it depends on both viral and host factors, with the host immune system playing a decisive role. After infection, dendritic cells (DC) presenting the DENV antigen recruit monocytes and macrophages to the site of infection. Instead of effectively eliminating the virus, these recruited immune cells - important components of the innate immune response - become primary targets for DENV (4). This step facilitates viral replication and the spread of the virus to other host immune cells in the lymph nodes, bone marrow, spleen and liver. The molecular and immunological mechanisms, by which DENV subverts the host’s immune response to viral antigens, are not fully understood. One possible explanation is that DENV preferentially infects DC, monocytes and macrophages thereby impairing the development of an efficient early antiviral response. During viral replication, DENV secretes non-structural proteins (NS), specifically NS1, NS2B and NS4, which disrupt the signalling pathway of type I interferon (IFN), a critical component of the innate immune defence (5). This disruption delays the host’s antiviral response and allows the virus to persist and propagate. In addition, secondary DENV infections can lead to antibody-dependent enhancement, in which non-neutralising antibodies facilitate viral replication and thus increase the disease burden for the host (5). Furthermore, the production of immune checkpoints (ICP) during DENV infection can impair the host’s immune response (6, 7). These molecules, known for their role in triggering T cell dysfunction and inhibiting other immune cells, are associated with impaired immune defence in various diseases, including cancer and infectious diseases (7, 8). One of the potential immune inhibitors is the HLA-G ligand (Human Leukocyte Antigens G), which is a ligand for receptors in B cells, T cells, monocytes, DC and subsets of natural killer cells (NK) (9). HLA-G molecules are involved in the inhibition of NK cell activity, the maturation of CD4+ T lymphocytes and DC, the apoptosis of CD8+ cytotoxic T cells and the development of regulatory T cells (Tregs) (9). The expression and regulation of HLA-G is highly dynamic: four membrane-bound forms (HLA-G1 to G4) and three soluble, secreted forms (sHLA-G5 to G7) generated by alternative splicing of the primary transcript (10). The HLA-G1 transmembrane isoform can produce a soluble form (sHLA-G), by proteolytic shedding, which retains all the functions of the membrane counterpart, potentially expanding immunoregulatory activities on a systematic scale. Studies have reported an upregulation of HLA-G antigens upon DENV infections suggesting a role of the ICPs in the pathogenesis of dengue (11). Furthermore, it has also been observed that the level of sHLA-G is modulated at different stages in arboviral infections (6). While the role of HLA-G in organ transplantation, pregnancy and cancer is well documented, its involvement in viral infections has not yet been sufficiently studied. This study examines sHLA-G levels in dengue patients to assess their association with disease severity. Ethical approval statement All study participants provided signed informed consent before enrolment. The Institutional Review Board of the 108 Military Hospital and the University of Tübingen approved the study, titled “Host and Viral Factors Influencing Dengue Severity and Susceptibility” (Ethics Approval No. 274/2022B02). The study complies with the Nagoya Protocol and authorization for the use of genetic resources in Germany was obtained from the Vietnamese Ministry of Natural Resources and Environment (Reference No. 2995/QĐ-BTNMT). All procedures followed GCP/GCLP guidelines. Study population Samples were collected during two consecutive seasonal outbreaks in northern Vietnam, spanning September to November in 2021 and 2022. The study population consisted of 238 civilian patients suspected of having dengue who agreed to be enrolled in the study and were admitted to the 108 Military Central Hospital in Hanoi. The dengue diagnoses followed the World Health Organisation diagnostic’s criteria (3) (https://apps.who.int/iris/handle/10665/44188), as adopted by the Vietnamese Ministry of Health. The inclusion criterion are patients presenting fever with at least two of the clinical sign/symptoms suggesting dengue (nausea/vomiting, rash, body aches and pains, tourniquet test positive) and/or positive for at least one of the indirect diagnostic methods (serological rapid test), as recommended and detailed in the WHO guideline 2009 (3) . Patients with bacterial or other viral infections, chronic diseases, or haematological disorders were excluded. A total of 118 Vietnamese healthy blood donors were recruited and considered as healthy controls in the study. Blood samples were collected upon admission, and plasma was separated and stored at -70°C until further use. PCR confirmation of dengue Total viral RNA was isolated from 140µL of patient plasma utilizing the QIAmp Viral RNA Mini Kit (Qiagen GmbH, Hilden, Germany) following the manufacturer’s instructions. All samples ( n =238) underwent multiplex real-time PCR analysis for dengue, Zika, and chikungunya viral RNA using the Fast-Track Diagnostics Kit (Siemens Healthcare GmbH, Erlangen, Germany) on a LightCycler480-II (Roche, Mannheim, Germany), following the manufacturer’s guidelines. The testing is performed using internal controls and standards provided. Confirmed dengue cases were identified by the presence of dengue viral RNA through real-time RT-PCR and absence of Zika and chikungunya viral RNA. Dengue severity classification and laboratory assessment In Vietnam, admitted patients were clinically classified into three severity levels according to WHO guidelines (3): dengue without warning signs (DF, n =103), dengue with warning signs (DWS, n =122) and severe dengue (SD, n =13). Clinical presentations were documented upon admission. Laboratory parameters assessed during admission include a complete hemogram: Erythrocyte (RBC), Haemoglobin (Hb), Haematocrit (HCT), Platelet (PLT), Leucocyte (WBC), Neutrophile (NEU), Lymphocyte (LYM), Monocyte (MONO), Eosinophile (EOS), Basophile (BASO), Large Unstained Cells (LUC), and liver enzymes: Aspartate Aminotransferase (AST) and Alanine Aminotransferase (ALT) levels. Measurement of sHLA-G plasma levels Plasma levels of sHLA-G (sHLA-G1 and sHLA-G5) were quantified in 238 patients and 118 healthy controls using the sHLA-G enzyme-linked immunosorbent assay (ELISA) kit (BioVendor–Laboratorní medicína a.s., Brno, Czech Republic), a sandwich enzyme immunoassay for the quantitative measurement of sHLA-G, according to the manufacturer’s instructions, whose detection limit is 0.38 ng/ml. The absorbance was measured by an ELISA reader (Infinite 200 PRO-TECAN, Maennedorf, Switzerland) at 450 nm with the reference wavelength set to 630 nm. The final sHLA-G concentrations were determined from a five-point standard curve using dilutions of calibrator (7.81, 15.63, 31.25, 62.5, and 125 U/mL) purchased by the kit as standard reagent. Results were expressed as Units/millilitre (U/mL). Statistical analysis Data were analysed and visualized using the R software version 4.3.2 (http://www.r-project.org). Clinical and demographic data were presented as median values (with range) for quantitative variables and absolute numbers with percent for categorical variables. The normality of distribution in the quantitative variables was tested using the Shapiro–Wilk test. Categorical data were compared using Chi-square test, while continuous variables were compared using Kruskal-Wallis or Wilcoxon test as appropriate. Dunn test was applied as post-hoc test. Spearman correlation with Bonferroni adjustment was applied to assess the correlation between levels of sHLA-G and other blood parameters. A p-value < 0.05 was considered statistically significant for all statistical comparisons. Multiple linear regression was performed to estimate the relationship between sHLA-G levels and days of illness, adjusted for age, sex and clinical severity. In our study, days of illness are defined as the number of days from the onset of fever until admission. The sHLA-G values were log-transformed prior to inclusion in the regression analysis to reduce skewness and heteroscedasticity.

Demographic and clinical characteristics of dengue patients All patients and healthy controls belonged to the Kinh ethnic group and were residents of Hanoi metropolitan area living in various communes. The patient group comprised 127 male and 111 female, with the median age of 47 years (range: [14-87] years). The healthy control group consisted of 66 male and 52 female with the median age of 45 years (range: [25-56] years). A detailed demographic and clinical data of the recruited patients are presented in Table 1. There are no significant differences in age and sex between different group of patients, or between patients and controls (Table 1). All of the patients were confirmed to dengue by RT-PCR. Patients with DWS and SD admitted to the hospital at the later phase of the disease (median of 5 days), compared to patients who had DF (median of 3 days). Rash and bleeding manifestations were more observed in DWS and SD patients, compared to DF (Table 1). Bleeding manifestations were observed the most in patients with warning signs, predominantly being subcutaneous bleeding. Laboratory parameters of dengue patients Patients’ laboratory parameters are summarized in Table 1. There was no difference in the total leukocyte counts between the three severity grades. DF patients presented lower lymphocyte counts (median [range]: 0.69 [0.17 - 3.56] ×10 6 /μL ) compared to DWS (median [range]: 0.99 [0.22 - 4.09] ×10 6 /μL ) (p=0.001) (Table 1 and Supplementary Table 1). In contrast, neutrophil and platelet counts were significantly higher in DF (NEU median [range]: 2.56 [0.47 - 13.1] ×10 6 /μL ; PLT median [range]: 138 [14.0 - 384] ×10 3 /μL) compared to DWS (NEU median [range]: 1.77 [0.45 - 6.21] ×10 6 /μL ; PLT median [range]: 21.0 [4.00 - 228] ×10 3 /μL) ( Table 1 and Supplementary Table 1) . Platelet counts were also higher in SD patients ( median [range]: 21.0 [4.00 - 228] ×10 3 /μL) compared to DWS ( p<0.001 ). In addition, liver enzymes including AST and ALT were higher in DWS (AST median [range]: 111 [16.0 - 1040] U/L; ALT median [range]: 66.2 [8.20 - 636] U/L) compared to DF (AST median [range]: 46.3 [17.3 - 350] U/L; ALT median [range]: 35.3 [8.00 - 503] U/L) (Table 1 and Supplementary Table 1). sHLA-G plasma levels in healthy controls and in dengue patients The levels of sHLA-G were significantly higher in dengue patients (median [range]: 42.7 [7.10 - 1300] U/ml) compared to healthy controls (median [range]: 11.1 [4.7 - 620] U/ml) (p < 0.001) (Figure 1a). In addition, sHLA-G significantly differs between varying severity: between DWS (median [range]: 45.6 [7.17, 525] U/ml) and DF (median [range]: 38.0 [7.10, 1300] U/ml) (p=0.029); and between SD (median [range]: 63.9 [7.73, 172] U/ml) and DF (p=0.016) (Figure 1b and Table 2). In particular, an association of high sHLA-G levels with cases requiring close medical surveillance (DWS and SD combined as recommended by the WHO (3)) was observed with DWS/SD (median [range]: 51.0 [7.17 – 525] U/ml) compared to DF (median [range]: 38.0 [7.10 – 1300] U/ml) (p=0.011). Multiple linear regression adjusting for age, sex, severity was performed to estimate the association between sHLA-G levels (log-transformed) and days of illness. The results showed a positive association between two variables (β=0.1, p=0.033) (Supplementary Table 2 and Figure 2). Plasma levels of sHLA-G correlated with laboratory parameters in dengue Thirteen laboratory parameters, including WBC, NEU, LYM, MONO, EOS, BASO, LUC, RBC, Hb, HCT, PLT, AST and ALT were included in the analysis (Table 3). While LUC and liver enzymes positively correlated with sHLA-G levels in all dengue patients (LUC: r=0.21, p-value=0.004; AST: r=0.21, p-value=0.001; ALT: r=0.14, p-value=0.038), PLT and NEU demonstrated inverse correlations (PLT: r=-0.21, p-value=0.001; NEU: r=-0.16, p-value=0.013) (Table 3) . In addition, LYM and LUC inversely correlated with sHLA-G levels in severe cases (LYM: r=-0.78, p-value=0.002; LUC: r=-0.81, p-value=0.005) (Table 3) .

The severity of dengue primarily depends on immunopathogenic mechanisms, where an impaired immune response not only contributes to the progression of the disease, but also hinders the elimination of DENV and disease complications (5). HLA-G is known for its tolerogenic role suppressing the activity of effector cells, in particular NK cells and CD8+ cytotoxic T lymphocytes (12) and thus might be potentially involved in the pathogenesis of dengue. Based on this assumption, our study aimed to investigate the association between sHLA-G and dengue in Vietnamese patients. The results suggested that sHLA-G levels were associated with dengue and might serve as a marker for disease severity. In our study, plasma levels of sHLA-G were significantly higher in dengue-confirmed patients, compared to healthy controls. The expression of HLA-G is restricted to certain tissues under physiological conditions, but is upregulated during pregnancy and in pathological conditions such as tumours, viral infections and inflammatory diseases (13). Soluble HLA-G proteins are formed by two mechanisms: alternative splicing and proteolytic release mediated by metalloproteinases. Among two mechanisms, it is likely that proteolytic release is a backup to alternative splicing to control specific immunomodulatory functions (14). The production of sHLA-G is regulated by various factors, including genetic, hormonal, pathological and immunological signals. For example, Park and co-authors reported that proteolytic release is particularly pronounced in pathological conditions, in which mutations occur at the splice sites of the HLA-G gene (14). Furthermore, interleukin-10 (15) and interferon-gamma increase the production of sHLA-G (16) and these two cytokines are known to be associated with dengue and upregulated in severe cases (17). At the cellular level, matrix metalloproteinases (MMPs) have been shown to enhance the release of sHLA-G1 in vitro, thus potentially increase the overall sHLA-G level (18). While down-regulating effects of MMPs on immune responses were reported under pathological conditions (19), excessive MMP activity following infection can lead to immunopathology and promotes the spread or persistence of the pathogen (20). Particularly, MMP-2 and MMP-9 levels were markedly elevated in severe dengue (21) leading to impaired immune responses and increased disease complications in this group of patients. These reports emphasise the potential role of sHLA-G as a mediator in the suppression of immune responses in dengue via different mechanism, which ultimately results in reduced activity of NK cells, activated CD8+ T lymphocytes and regulatory T cells. In contrast, Renata and co-authors observed decreased sHLA-G levels during the acute phase arbovirus infection in Brazilian patients compared to the recovery phase that reflects the stable state of the patient’s immune system (6). However, the majority of patients in this study were confirmed to have Zika and chikungunya infection and additionally presented neurological complications. This could be a possible explanation for the controversial findings in the sHLA-G profile compared to our dengue-infected cohort. Therefore, different viruses may trigger and regulate the sHLA-G production differently. In addition, sHLA-G levels are genetically regulated and different polymorphisms in the HLA-G gene in different populations can influence the HLA-G plasma levels (22, 23). This factor should be taken into account when comparing the results of different studies. However, our results are consistent with other studies demonstrating higher sHLA-G levels upon human cytomegalovirus, herpesvirus, hepatitis virus and SARS-CoV-2 infections compared to healthy controls (24). As previous studies reported, the time is a key factor in determining the clinical severity of dengue (25). Using multiple linear regression, our analysis revealed that days of illness positively associated with sHLA-G plasma levels. These data suggest that sHLA-G plasma levels increase over time during dengue infection in all patients, even after adjusting for disease severity and other factors, such as age and sex. The result could contribute to the understanding of the dynamics of sHLA-G level in dengue and similarly underlines the importance of time since symptom onset as a determining factor for disease severity. However, further study is required to elucidate the mechanism, by which DENV triggers the increased sHLA-G production during the disease course. In addition, our study revealed correlations between sHLA-G levels and important markers of dengue such as NEU, PLT, AST and ALT. These markers are crucial for assessing disease burden and monitoring dengue progression (3, 25). Although the observed correlations were relatively modest, this observation provides haematological and biochemical evidence to support the hypothesis that elevated sHLA-G levels contribute to increased dengue severity. Notably, a strong inverse correlation between sHLA-G levels and total lymphocyte counts were observed in severe cases suggesting a possible role of sHLA-G in regulating lymphocyte numbers. Lymphocytes are critical for antiviral immunity and their depletion, a hallmark of severe dengue and other viral infections, is associated with impaired immune response and increased infection consequences (26). Elevated sHLA-G levels may contribute to lymphocyte depletion by exerting immunosuppressive effects, such as inducing apoptosis or inhibiting lymphocyte proliferation, thereby exacerbating immune dysfunction and influencing disease severity (27). In addition, large unstained cells (LUC) count was inversely correlated with sHLA-G level in severe patients in our study. LUC includes large, activated lymphocytes and other atypical cells such as virocytes, blasts cells, and hematopoietic stem cells. This parameter was proposed as a potential indicator reflecting the immune response status to viral infection (28, 29). A reduction in LUC numbers may indicate impaired immune activation or cytotoxic responses, contributing to infection severity. Correlation of LUC and CD8+ and CD4+ T cells was also noted in HIV patients (30) and LUC was suggested to be a predictive biomarker for hematologic toxicities in cancer and hematologic malignancies (31). The negative correlation between LUCs and sHLA-G observed in severe patients may indicate a reduction in the number of activated lymphocytes in this group upon high sHLA-G levels suggesting a possible role of sHLA-G in attenuating immune defence in these cases. Nevertheless, a significant positive correlation between LUC count and sHLA-G levels was found in the analysis that combined all patients with varying degrees of severity suggesting that LUC counts may respond differently to changes in sHLA-G levels at different stages and different severities of dengue. Notably, the strong and significant correlations between sHLA-G and laboratory parameters, including LYM and LUC, are only observer in severe dengue. Our small number of severe cases may limit the generalizability of the findings but suggests that sHLA-G potentially play an important role in severe dengue. Furthermore, the possibility to assess the level of sHLA-G only at one time-point in each patients limited the understanding of sHLA-G dynamics. A larger cohort may be beneficial to improve the limitation of our study.

Immune inhibitors potentially have an important role in the pathogenesis of dengue. In conclusion, our findings suggest that sHLA-G levels are associated with dengue severity, highlighting its potential as a marker for clinical monitoring and early identification of severe cases.

1. Yang X, Quam MBM, Zhang T, Sang S. Global burden for dengue and the evolving pattern in the past 30 years. J Travel Med. 2021;28(8).2. Hung TM, Clapham HE, Bettis AA, Cuong HQ, Thwaites GE, Wills BA, et al. The Estimates of the Health and Economic Burden of Dengue in Vietnam. Trends Parasitol. 2018;34(10):904-18.3. World Health O. Dengue guidelines for diagnosis, treatment, prevention and control : new edition. Geneva: World Health Organization; 2009.4. Begum F, Das S, Mukherjee D, Ray U. Hijacking the Host Immune Cells by Dengue Virus: Molecular Interplay of Receptors and Dengue Virus Envelope. Microorganisms. 2019;7(9).5. Khanam A, Gutiérrez-Barbosa H, Lyke KE, Chua JV. Immune-Mediated Pathogenesis in Dengue Virus Infection. Viruses [Internet]. 2022; 14(11).6. Almeida RS, Ferreira MLB, Sonon P, Cordeiro MT, Sadissou I, Diniz GTN, et al. Cytokines and Soluble HLA-G Levels in the Acute and Recovery Phases of Arbovirus-Infected Brazilian Patients Exhibiting Neurological Complications. Front Immunol. 2021;12:582935.7. Loacker L, Egger A, Fux V, Bellmann-Weiler R, Weiss G, Griesmacher A, et al. Serum sPD-L1 levels are elevated in patients with viral diseases, bacterial sepsis or in patients with impaired renal function compared to healthy blood donors. Clin Chem Lab Med. 2023;61(12):2248-55.8. He X, Xu C. Immune checkpoint signaling and cancer immunotherapy. Cell Research. 2020;30(8):660-9.9. Carosella ED, Rouas-Freiss N, Tronik-Le Roux D, Moreau P, LeMaoult J. HLA-G: An Immune Checkpoint Molecule. Adv Immunol. 2015;127:33-144.10. Lin A, Zhang X, Zhang RL, Zhang JG, Zhou WJ, Yan WH. Clinical Significance of Potential Unidentified HLA-G Isoforms Without α1 Domain but Containing Intron 4 in Colorectal Cancer Patients. Front Oncol. 2018;8:361.11. McKechnie JL, Beltrán D, Pitti A, Saenz L, Araúz AB, Vergara R, et al. HLA Upregulation During Dengue Virus Infection Suppresses the Natural Killer Cell Response. Front Cell Infect Microbiol. 2019;9:268.12. Contini P, Murdaca G, Puppo F, Negrini S. HLA-G Expressing Immune Cells in Immune Mediated Diseases. Front Immunol. 2020;11:1613.13. Morandi F, Pistoia V. Interactions between HLA-G and HLA-E in Physiological and Pathological Conditions. Front Immunol. 2014;5:394.14. Park GM, Lee S, Park B, Kim E, Shin J, Cho K, et al. Soluble HLA-G generated by proteolytic shedding inhibits NK-mediated cell lysis. Biochemical and Biophysical Research Communications. 2004;313(3):606-11.15. Moreau P, Adrian-Cabestre F, Menier C, Guiard V, Gourand L, Dausset J, et al. IL-10 selectively induces HLA-G expression in human trophoblasts and monocytes. Int Immunol. 1999;11(5):803-11.16. Morandi F, Levreri I, Bocca P, Galleni B, Raffaghello L, Ferrone S, et al. Human Neuroblastoma Cells Trigger an Immunosuppressive Program in Monocytes by Stimulating Soluble HLA-G Release. Cancer Research. 2007;67(13):6433-41.17. Malavige GN, Gomes L, Alles L, Chang T, Salimi M, Fernando S, et al. Serum IL-10 as a marker of severe dengue infection. BMC Infectious Diseases. 2013;13(1):341.18. Rizzo R, Trentini A, Bortolotti D, Manfrinato MC, Rotola A, Castellazzi M, et al. Matrix metalloproteinase-2 (MMP-2) generates soluble HLA-G1 by cell surface proteolytic shedding. Molecular and Cellular Biochemistry. 2013;381(1):243-55.19. Manicone AM, McGuire JK. Matrix metalloproteinases as modulators of inflammation. Seminars in Cell & Developmental Biology. 2008;19(1):34-41.20. Elkington PT, O’Kane CM, Friedland JS. The paradox of matrix metalloproteinases in infectious disease. Clin Exp Immunol. 2005;142(1):12-20.21. Sivasubramanian S, Mohandas S, Gopalan V, Govindan K, Varadarajan P, Kaveri K, et al. Serum levels of matrix metalloproteinases as prognostic markers for severe dengue with plasma leakage. Experimental and Molecular Pathology. 2022;128:104821.22. Alyami A, AlJurayyan A, Alosaimi B, Alkadi H, Alkhulaifi F, Al-Jurayb H, et al. The correlation between soluble human leukocyte antigen (sHLA-G) levels and +3010 polymorphism. Int J Immunogenet. 2024;51(1):39-46.23. Adolf IC, Almars A, Dharsee N, Mselle T, Akan G, Nguma IJ, et al. HLA-G and single nucleotide polymorphism (SNP) associations with cancer in African populations: Implications in personal medicine. Genes & Diseases. 2022;9(5):1220-33.24. Jasinski-Bergner S, Schmiedel D, Mandelboim O, Seliger B. Role of HLA-G in Viral Infections. Front Immunol. 2022;13:826074.25. Recker M, Fleischmann WA, Nghia TH, Truong NV, Nam LV, Duc Anh D, et al. Markers of prolonged hospitalisation in severe dengue. PLoS Negl Trop Dis. 2024;18(1):e0011922.26. Fahrudi Setiawan A, Yueniwati Prabowowati Wajib Y, Handono K, Sakti SP. Role of Lymphocytes and Atyphical Lymphocytes in Dengue Hemorrhagic Fever: A Literature Review. Journal of Medicinal and Chemical Sciences. 2023;7(1):53-8.27. Jasinski-Bergner S, Eckstein M, Taubert H, Wach S, Fiebig C, Strick R, et al. The Human Leukocyte Antigen G as an Immune Escape Mechanism and Novel Therapeutic Target in Urological Tumors. Front Immunol. 2022;13:811200.28. Merter M, Sahin U, Uysal S, Dalva K, Yuksel MK. Role of large unstained cells in predicting successful stem cell collection in autologous stem cell transplantation. Transfusion and Apheresis Science. 2023;62(1):103517.29. Shin D, Lee MS, Kim DY, Lee MG, Kim DS. Increased large unstained cells value in varicella patients: A valuable parameter to aid rapid diagnosis of varicella infection. J Dermatol. 2015;42(8):795-9.30. Vanker N, Ipp H. Large Unstained Cells: A Potentially Valuable Parameter in the Assessment of Immune Activation Levels in HIV Infection. Acta Haematologica. 2013;131(4):208-12.31. Ceylan F, Mehdiyev M, Dede DŞ, Efil SC, Tenekeci AK, Bilgin B, et al. Large Unstained Cells (LUC): A Novel Predictor of CDK4/6 Inhibitor Outcomes in HR+ HER2-Negative Metastatic Breast Cancer. Journal of Clinical Medicine. 2025;14(1):173. Figure 1. sHLA-G plasma levels in dengue patients and its clinical relevance Figure 1a. sHLA-G plasma levels in dengue patients and healthy controls sHLA-G plasma levels were determined using a commercially available ELISA as described in Methods. The data are provided as box plots as log-transformed sHLA-G (U/mL). Figure 1b. sHLA-G plasma levels in dengue patients of varying severity Figure 1c. sHLA-G plasma levels in dengue patients by medical monitoring requirements Figure 2. sHLA-G levels and day of illness Multiple linear regression was performed adjusting disease severity to estimate the association between log-transformed sHLA-G levels and days of illness. Table 1. Characteristics of dengue patients | Demographic data | |||| | Age (years) | 46 [14 - 87] | 49 [17 - 83] | 49 [21 - 82] | 0.627 | | Gender (male/female) | 56/47 | 65/57 | 6/7 | 0.855 | | Clinical manifestation | |||| | Day of disease (days) | 3 [1 - 5] | 5 [1 - 8] | 5 [4 - 7] | < 0.001 | | Headache | 97 (94.2%) | 106 (86.9%) | 12 (92.3%) | 0.192 | | Retro-ocular pain | 54 (52.4%) | 84 (68.9%) | 12 (92.3%) | 0.003 | | Myalgia | 75 (72.8%) | 92 (75.4%) | 10 (76.9%) | 0.637 | | Arthralgia | 62 (60.2%) | 85 (69.7%) | 11 (84.6%) | 0.116 | | Rash | 13 (12.6%) | 70 (57.4%) | 6 (46.2%) | < 0.001 | | Abdominal pain | 0 (0%) | 20 (16.4%) | 5 (38.5%) | < 0.001 | | Vomit | 15 (14.6%) | 31 (25.4%) | 6 (46.2%) | 0.008 | | Lethargy | 0 (0%) | 1 (0.8%) | 4 (30.8%) | < 0.001 | | Edema | 0 (0%) | 26 (21.3%) | 6 (46.2%) | < 0.001 | | Hepatomegaly | 0 (0%) | 3 (2.5%) | 0 (0%) | 0.248 | | Shock | 0 (0%) | 0 (0%) | 5 (38.5%) | < 0.001 | | Respiratory distress | 0 (0%) | 0 (0%) | 6 (46.2%) | < 0.001 | | Bleeding manifestations | 15 (14.9%) | 102 (83.6%) | 10 (76.9%) | < 0.001 | | Subcutaneous | 15 (14.9%) | 82 (67.2%) | 8 (61.5%) | < 0.001 | | Mucosal | 0 (0%) | 64 (52.5%) | 6 (46.2%) | < 0.001 | | Severe | 0 (0%) | 0 (0%) | 4 (30.8%) | < 0.001 | | Laboratory tests | |||| | Erythrocyte ×10 6 /μL | 4.83 [3.66 - 6.26] | 5.14 [3.97 - 7.62] | 5.08 [2.76 - 5.89] | < 0.001 | | Haemoglobin g/L | 146 [110 - 187] | 153 [113 - 190] | 150 [70.0 - 172] | 0.003 | | Haematocrit | 43 [31.8 - 53.4] | 44.7 [34.8 - 60.5] | 43.7 [21.4 - 51.6] | 0.001 | | Platelet ×10 3 /μL | 138 [14.0 - 384] | 21.0 [4.00 - 228] | 29.0 [4.00 - 125] | < 0.001 | | Leucocyte ×10 6 /μL | 4.08 [1.27 - 16.9] | 3.71 [0.96 - 11.6] | 4.70 [1.45 - 10.5] | 0.762 | | Neutrophile ×10 6 /μL | 2.56 [0.47 - 13.1] | 1.77 [0.45 - 6.21] | 1.97 [0.81 - 7.70] | < 0.001 | | Lymphocyte ×10 6 /μL | 0.69 [0.17 - 3.56] | 0.99 [0.22 - 4.09] | 1.03 [0.31 - 2.81] | 0.003 | | Monocyte ×10 6 /μL | 0.39 [0.06 - 1.35] | 0.31 [0.05 - 2.53] | 0.39 [0.10 - 0.74] | 0.189 | | Eosinophile ×10 6 /μL | 0.01 [0 - 0.21] | 0.02 [0 - 0.45] | 0.01 [0 - 0.07] | 0.036 | | Basophile ×10 6 /μL | 0.03 [0 - 1.18] | 0.07 [0 - 1.76] | 0.09 [0.01 - 0.26] | < 0.001 | | LUC ×10 6 /μL | 0.10 [0.02 - 3.49] | 0.36 [0.02 - 5.82] | 0.40 [0.05 - 2.45] | < 0.001 | | Neutrophile % | 65.8 [18.7 - 93.4] | 49.7 [17.3 - 82.3] | 56.5 [26.1 - 76.1] | < 0.001 | | Lymphocyte % | 19.2 [2.5 - 72.8] | 27.9 [6.1 - 56.3] | 21.1 [6.8 - 53.0] | < 0.001 | | Monocyte % | 9.10 [1.6 - 20.9] | 7.80 [2.9 - 36.6] | 8.50 [3.6 - 11.8] | 0.098 | | Eosinophile % | 0.40 [0 - 4.00] | 0.65 [0 - 11.6] | 0.20 [0.10 - 2.30] | 0.002 | | Basophile % | 0.60 [0 - 11.3] | 2.20 [0 - 15.1] | 1.40 [0.60 - 4.70] | < 0.001 | | LUC % | 2.70 [0.4 - 35.0] | 10.0 [1.30 - 55.6] | 6.55 [3.30 - 26.1] | < 0.001 | | AST U/L | 46.3 [17.3 - 350] | 111 [16.0 - 1040] | 186 [31 -11100] | < 0.001 | | ALT U/L | 35.3 [8.00 - 503] | 66.2 [8.20 - 636] | 90.0 [25.6 -2190] | < 0.001 | LUC: Large unstained cells; AST: Aspartate Aminotransferase; ALT: Alanine Aminotransferase. Variables were summarized in Percentage (%) or Median with [range]. P-values were calculated by Chi-square and Kruskal-Wallis test. Table 2. Levels of sHLA-G in dengue patients of varying clinical severities | Dengue without warning signs (n=103) | Dengue with warning signs (n=122) | Severe dengue (n=13) | p value | | | Mean (SD) | 77.8 (174) | 68.1 (68.2) | 76.2 (42.4) | 0.013 | | Median [Min, Max] | 38.0 [7.10, 1300] | 45.6 [7.17, 525] | 63.9 [7.73, 172] | | | Dengue without warning signs (n=103) | Dengue with warning signs and severe dengue (n=135) | Healthy controls (n=118) | || | Mean (SD) | 77.8 (174) | 68.9 (66.1) | 29.1 (67.1) | < 0.001 | | Median [Min, Max] | 38.0 [7.10, 1300] | 51.0 [7.17, 525] | 11.1 [4.70, 620] | P-values were calculated by Kruskal-Wallis test. Table 3. Correlation of sHLA-G levels with laboratory parameters | All Patients | ||||||||||||| | R | -0.04 | -0.16 | 0.11 | -0.05 | 0.03 | 0.13 | 0.21 | 0 | 0.04 | 0.09 | -0.21 | 0.21 | 0.14 | | P-value | 0.557 | 0.013 | 0.097 | 0.488 | 0.7 | 0.051 | 0.004 | 0.975 | 0.518 | 0.184 | 0.001 | 0.001 | 0.038 | | DF | ||||||||||||| | R | -0.2 | -0.21 | 0.03 | -0.13 | -0.23 | -0.03 | 0.09 | -0.08 | -0.09 | -0.01 | -0.1 | 0.06 | 0.01 | | P-value | 0.041 | 0.032 | 0.731 | 0.199 | 0.019 | 0.763 | 0.444 | 0.418 | 0.341 | 0.917 | 0.301 | 0.547 | 0.883 | | DWS | ||||||||||||| | R | 0.14 | -0.04 | 0.18 | 0.11 | 0.26 | 0.2 | 0.2 | 0 | 0.12 | 0.12 | -0.14 | 0.13 | 0.05 | | P-value | 0.132 | 0.645 | 0.06 | 0.259 | 0.006 | 0.032 | 0.049 | 0.992 | 0.225 | 0.195 | 0.112 | 0.166 | 0.575 | | SD | ||||||||||||| | R | -0.44 | -0.23 | -0.78 | -0.28 | -0.01 | -0.39 | -0.81 | 0 | 0.03 | -0.02 | -0.2 | 0.07 | 0.19 | | P-value | 0.135 | 0.448 | 0.002 | 0.352 | 0.976 | 0.188 | 0.005 | 1 | 0.922 | 0.943 | 0.502 | 0.817 | 0.529 | WBC: Leucocyte; NEU: Neutrophile; LYM: Lymphocyte; MONO: Monocyte; EOS: Eosinophile; BASO: Basophile; LUC: Large Unstained Cells; RBC: Erythrocyte; Hb: Haemoglobin; HCT: Haematocrit; PLT: Platelet; AST: Aspartate Aminotransferase; ALT: Alanine Aminotransferase. R: Spearman correlation coefficient. DF: dengue without warning signs; DWS: dengue with warning signs; SD: Severe dengue. Significant p value is highlighted in bold. Supplementary 1. p value from post hoc Dunn test | DF vs DWS | NA | <0.001 | 0.001 | NA | 0.02 | <0.001 | <0.001 | <0.001 | 0.001 | <0.001 | <0.001 | <0.001 | <0.001 | | DF vs SD | NA | 0.727 | 0.258 | NA | 1 | 0.028 | 0.023 | 0.776 | 0.982 | 1 | <0.001 | <0.001 | <0.001 | | DWS vs SD | NA | 0.397 | 1 | NA | 0.322 | 1 | 1 | 0.354 | 0.393 | 0.234 | 1 | 0.497 | 0.373 | WBC: Leucocyte; NEU: Neutrophile; LYM: Lymphocyte; MONO: Monocyte; EOS: Eosinophile; BASO: Basophile; LUC: Large Unstained Cells; RBC: Erythrocyte; Hb: Haemoglobin; HCT: Haematocrit; PLT: Platelet; AST: Aspartate Aminotransferase; ALT: Alanine Aminotransferase. DF: dengue without warning signs; DWS: dengue with warning signs; SD: Severe dengue. NA: not applicable. Supplementary 2. Multiple linear regression for sHLA-G level and day of illness | Intercept | 3.392 | 0.247 | 13.749 | <0.001 | | Days of illness | 0.104 | 0.049 | 2.137 | 0.033 | | Age | -0.001 | 0.004 | -0.186 | 0.853 | | Sex (Male) | -0.132 | 0.122 | -1.086 | 0.279 | | Dengue with warning signs | 0.042 | 0.157 | 0.266 | 0.790 | | Severe dengue | 0.299 | 0.292 | 1.026 | 0.306 | Estimate, Standard Error, t-value and p values were calculated by multiple linear regression. Information & Authors Information Version history Peer review timeline Published Journal of Medical Virology Version of Record5 Sep 2025Published Copyright This work is licensed under a Non Exclusive No Reuse License. Collection

Authors Metrics & Citations Metrics Article Usage 409views 211downloads Citations Download citation Do Duc Anh, Nguyen Trong, Le Huu Song, et al. Elevated Soluble HLA-G Levels Associate with Dengue Severity in Vietnamese Patients. Authorea. 26 January 2025. DOI: https://doi.org/10.22541/au.173788979.94292576/v1 DOI: https://doi.org/10.22541/au.173788979.94292576/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu.