Optimizing the dengue virus infection mouse model: Comparing different backgrounds and infection route for enhanced stability

Abstract

Background

In recent decades, the global incidence of dengue fever has been steadily increasing, with continuous geographical expansion. Researchers have successfully modeled most clinical symptoms of human dengue fever using interferon type I (IFN‐I) or combined IFN‐I/II receptor knockout mice infected with dengue virus (DENV). However, this model requires further optimization to better support related studies.

Methods

This study aimed to establish a stable dengue infection model by evaluating the effects of different genetic backgrounds and injection routes on DENV infection in interferon receptor knockout mice. We first infected various strains of interferon receptor‐deficient mice with DENV and compared their susceptibility based on clinical symptoms, viremia levels, organ indices, histopathological findings, and vascular leakage markers. Subsequently, we selected the most susceptible strain to further investigate the impact of different injection methods on infection outcomes.

Results

We found that BALB/c background mice with type 1 interferon receptor knockout(IFNAR) had the most obvious symptoms. Subsequently, we selected IFNAR−/−BALB/c mice to further explore the effects of different injection methods on dengue virus infection. The results showed that the intraperitoneal injection group had the most severe clinical symptoms, the longest duration of viremia, and the most obvious degree of organ damage.

Conclusion

Through systematic screening and optimization, we established a robust animal model of dengue virus infection via intraperitoneal injection in IFNAR−/− BALB/c mice. This model offers a valuable tool for future dengue research.

Optimized dengue fever animal model: IFNAR^−/− BALB/c mice with intraperitoneal injection. Background screening: Between IFNAR^−/− (BALB/c, C57BL/6) and AG129 mice, IFNAR^−/−BALB/c exhibited the most severe symptoms: (1) highest weight loss and clinical scores; (2) peak viremia (5.44 log10 RNA copies/mL, day 4); (3) significant organ damage (liver/brain/kidney) and vascular leakage. Route optimization: in BALB/c IFNα/β/γR⁻/⁻ mice, intraperitoneal (IP) injection induced: (1) prolonged viremia (day 6); (2) severe multi‐organ pathology (liver necrosis, brain inflammation); (3) neurological signs (limb paralysis). Results: establishes a highly susceptible model for dengue pathogenesis and therapeutic studies.

1. INTRODUCTION

Dengue virus (DENV) is a viral disease endemic in many tropical and subtropical regions of the world and transmitted to humans by Aedes mosquitoes. ¹ According to the World Health Organization (WHO) statistics, since the beginning of 2025, and as of March, over 1.4 million dengue cases and over 400 dengue‐related deaths have been reported from 53 countries/territories in the WHO Region of the Americas (PAHO), South‐East Asia and West Pacific Regions (SEARO and WPRO, respectively), the Eastern Mediterranean WHO Region (EMRO), and Africa (https://worldhealthorg.shinyapps.io/dengue_global/). DENV infection with any of the four dengue serotypes (DENV 1–4) often results in subclinical illness, typically marked by high fever, muscle pain, and skin rash. ² In some cases, as the fever subsides, the condition can progress to severe dengue, which may become life threatening and is characterized by low platelet counts, capillary leakage, bleeding, elevated liver enzymes, and multi‐organ complications. ³ , ⁴

Mouse models, which are small, have a short reproductive cycle, are easy to operate and observe, and have a genome highly similar to humans, are among the key tools for studying dengue fever. ⁵ , ⁶ They are usually used for the study of viral pathogenic mechanisms and the development of antiviral vaccines and drugs. However, immune‐competent mice have natural resistance to certain viruses, making it difficult to establish an effective DENV infection model, which limits the development of subsequent virus research. ⁷ A number of models using immunocompromised mouse strains have been developed. Of these, the AG129 mouse model, which lacks functional receptors for both type I (IFN‐I) and type II (IFN‐II) interferons (IFN), is effective. ⁸ However, as an IFN‐α/β and γ receptor double knockout model, the AG129 mouse model has significant limitations in viral infection research. This model cannot simulate human central nervous system symptoms, such as encephalitis, and its excessive immune deficiency may mask the neurotropism of the virus. ⁹ Moreover, due to severe immune deficiency, the AG129 model has problems such as abnormal T‐cell function and cannot truly reflect the dynamic balance between innate immunity and adaptive immunity, resulting in differences in the viral replication pattern from the clinical reality. In particular, it does not exhibit typical characteristics in the mechanism of vascular endothelial injury. ¹⁰ Later, researchers discovered that type I IFNs have a major impact on DENV replication and began using type I interferon receptor knockout (IFNAR) mice for DENV infection studies. ¹¹ , ¹² These mice not only replicate most dengue‐like symptoms but also retain an intact type II IFN signaling pathway, partially mimicking the human antiviral immune response. ¹³ In addition, IFNAR mice can be produced on various genetic backgrounds, including BALB/c and C57BL/6. ¹⁴

Due to the complexity of the virus and its intricate interactions with the host immune system, ¹⁵ existing mouse models still have certain limitations, and understanding DENV infection and its consequences remains challenging. ¹⁶ Considering different strains of mice have different genetic backgrounds, their biological characteristics and behavioral manifestations are also different, resulting in different susceptibilities to DENV. ¹⁷ In this study, we compared the clinical symptoms, viremia, pathological damage, and vascular leakage of interferon receptor (IFNR) knockout mice infected with dengue fever in IFNAR^−/−BALB/c, IFNAR^−/−C57BL/6, and AG129 mice to screen the optimal background IFNR knockout mice. Based on this, the effects of different injection methods on DENV infection were explored, aiming to develop a mouse model with the best infection effect and provide conditions for building a better dengue mouse model and conducting dengue experiments smoothly.

2. METHODS

2.1. Experimental animals

IFNAR^−/−BALB/c, IFNAR^−/−C57BL/6, and AG129 mice (regardless of sex and aged 4–5 weeks old) were provided by Professor Linzhong Yu of Southern Medical University. They were bred, raised, and managed by the staff of the Animal Experiment Center of Guangzhou University of Chinese Medicine (Laboratory Animal Production License no.: SCXK [Guangdong] 2018‐0034, Use License no.: SYXK [Guangdong] 2018‐0001). The experiments were conducted in a biosafety level 2 (ABSL‐2) laboratory (Yueguang Dongshibei GZL0004).

2.2. Virus and cell

DENV type II (TSV01 strain) was provided by Professor Chen Xulin's team from Jinan University. C6/36 cells were infected with DENV (multiplicity of infection [MOI] = 0.5) for 2 h and cultured for 5–6 days. The cells were refed with fresh medium ~2–3 days after infection. When the cells exhibited obvious lesions, the supernatant was collected and the remaining cells were repeatedly frozen and thawed at −80℃ to release the virus. All the virus liquid was collected and centrifuged at 4000 rpm for 5 min to precipitate cell fragments. It was then filtered with a 0.45‐μm filter membrane, aliquoted, and stored at −80℃.

2.3. Animal models and groups

We first investigated DENV infection in IFN receptor knockout mice from different backgrounds. IFNAR^−/−BALB/c, IFNAR^−/−C57BL/6, and AG129 mice were divided into the normal group and virus group, with four mice in each group. On day 0, 100 μL of DENV (4 × 10⁵ PFU) was intraperitoneally administered, and the same dose of MEM was injected as negative control. The mice were fed specific pathogen free (SPF) chow and sterilized water ad libitum. Another batch of animals were used for modeling as described earlier. On day 7 postinfection, Evans blue was injected via the tail vein to assess vascular leakage.

Next, we further investigated the effects of different injection methods on DENV infection in BALB/c mice with IFNAR. Sixteen IFNAR^−/−BALB/c mice were randomly divided into the normal group, intraperitoneal (IP) injection group, tail vein (IV) injection group, and foot pad injection group, with four mice in each group. The mice in the IP and tail vein groups were injected with 100 μL of DENV solution (4 × 10⁵ PFU), and the mice in the foot pad injection group were injected with 50 μL of DENV solution into the left and right feet. The mice were fed SPF chow and sterilized water ad libitum. All the animal experiments in this study were approved by the Experimental Animal Ethics Committee of Guangzhou University of Chinese Medicine.

2.4. Observation of mouse physical signs

The mice were weighed daily, and their survival or disease status was observed and recorded. Disease scores were based on a scale of 1–5 described by Orozco et al. ¹⁸ , including healthy; mildly lethargic; lethargic, piloerection, hunched back; lethargic, piloerection, hunched back, weak; and moribund. ¹⁹

2.5. RT‐qPCR detection of viremia

On days 2, 4, and 6, 50 μL of blood was collected from the orbital venous plexus of mice in the normal group and the model group and quickly placed in an Eppendorf (EP) tube, and 500 μL of TRIzon reagent was added and shaken. Total RNA was extracted using the Ultrapure RNA Extraction Kit, and then the total RNA was reverse transcribed into complementary DNA using the HiScript II Q RT SuperMix for qPCR Kit. SYBR Green was used as a fluorescent marker, and DENV TSV01‐F/R (CAGATCTCTGATGAATAACCAACG/CATTCCAAGTGAGAATCTCTTTGTCA) was used for reverse transcription real‐time quantitative detection. The reaction volume was 20 μL, and the results were quantified using the standard curve and expressed as log₁₀ copies/mL.

2.6. Organ index test

On day 7 of the experiment, the mice were killed by cervical dislocation. After the abdomen was disinfected, the liver, kidney, brain, spleen, and other organs of the mice were removed. The tissues were gently rinsed with saline to remove blood, and the residual water was absorbed using filter paper. The tissues were weighed and recorded using a precision balance (organ index = organ weight [mg]/body weight [g]).

2.7. Hematoxylin–eosin staining to detect organ damage

The removed organs were fixed with 4% paraformaldehyde for more than 24 h, embedded in paraffin, and sliced. The slices were stained with hematoxylin–eosin (HE), and the histomorphological changes in the organs were observed under an optical microscope to assess the degree of organ damage.

2.8. Evans blue to detect vascular leakage

This experiment was conducted in accordance with the international general operating procedures. ²⁰ , ²¹ , ²² On day 7 of the experiment, 200 μL of Evans blue was injected into the tail vein of mice in each group. The mice were returned to the cages, and the solution was allowed to circulate for 2 h before they were killed by cervical dislocation. The liver, kidney, brain, and other organs were removed and transferred to new EP tubes, and the weights were recorded. Based on the weight of the organs, an appropriate amount of formamide was added (formamide volume [μL] = weight [g] × 3400). All EP tubes were transferred to a biochemical incubator and kept in a constant‐temperature gas bath at 37℃ for 48 h to extract Evans blue. The absorbance at 620 nm was measured using an enzyme reader.

2.9. Statistical analysis

SPSS version 20.0 was used for statistical analysis. The measurement data that met the normal distribution were described by x ± s. If they did not meet the normal distribution, M(P25, P75) was used for description. The two independent measurement data were tested for normality using Shapiro–Wilk test. If they met the normal distribution, the independent sample t‐test was used. If the Levene's variance was significant at >0.05, the significant result assuming equal variance was taken; if the Levene's variance was significant at <0.05, the significant result assuming unequal variance was taken. If the data did not meet the normal distribution test, the Mann–Whitney U nonparametric test was performed. Finally, GraphPad Prism version 8.0 was used for drawing.

3. RESULTS

3.1. DENV exhibited replicating ability in IFNR knockout mice with different backgrounds

Because the genetic background of mice is an important factor influencing DENV infection, we selected AG129, IFNAR^−/−BALB/c, and IFNAR^−/− C57BL/6 mice for investigation. The experimental procedure is shown in Figure 1. IFNAR^−/−BALB/c mice were characterized by rapid weight loss accompanied by disease progression on day 2, including ruffled coat, hunched posture, and reduced activity, but with no obvious signs of neurological disease, such as paresis or paralysis (Figure 2A,B). The IFNAR^−/−C57BL/6 mice exhibited a mild reduction in weight and the disease progressed on day 6 (Figure 2C,D). The weight increase in the normal group and the model group of AG129 mice was almost the same, the difference in changes was not statistically significant, and the disease progressed on day 5 (Figure 2E,F). The results showed that the weight difference in IFNAR^−/−BALB/c mice after DENV infection was the largest, and the clinical symptoms were the most obvious, which was the external manifestation of further replication and infection of DENV.

FIGURE 1.

Schematic diagram of DENV (dengue virus) infection in IFNR (interferon receptor) knockout mice with different backgrounds.

FIGURE 2.

Body weight changes and clinical symptoms of IFNR (interferon receptor) knockout mice with different backgrounds after infection with DENV (dengue virus). (A) Body weight changes in IFNAR^−/−BALB/c mice; (B) clinical scores in IFNAR^−/−BALB/c mice; (C) body weight changes in IFNAR^−/−C57BL/6 mice; (D) clinical scores in IFNAR^−/−C57BL/6 mice; (E) body weight changes in AG129 mice; (F) clinical scores in AG129 mice.

3.2. Viral load in the blood of IFNR knockout mice infected with DENV in different backgrounds

DENV replicates in the blood after infecting mice. By detecting the viral load in the blood, the infection of the virus can be further determined. We detected the viral load on days 2, 4, and 6 and found that on day 2, the viral load in the blood of three background mice was similar. On day 4, the viral load in the blood of IFNAR^−/−C57BL/6 and AG129 mice decreased, whereas that in the blood of IFNAR^−/−BALB/c mice increased significantly. On day 6, the blood viral load of different background mouse model groups showed a decrease, but IFNAR^−/−BALB/c remained at a higher level compared with the other two mice (Figure 3A–C). The peak viral loads of IFNAR^−/−BALB/c, C57BL/6, and AG129 mice were 5.44, 5.01, and 4.83 log₁₀ DENV RNA (copies/mL), respectively. It can be seen that DENV replicates most efficiently and lasts longest in IFNAR^−/−BALB/c mice.

FIGURE 3.

Viral load of IFNR (interferon receptor) knockout mice with different backgrounds. (A) Viral load in the blood of IFNAR^−/−BALB/c mice on days 2, 4, and 6; (B) viral load in the blood of IFNAR^−/−C57BL/6 mice on days 2, 4, and 6; (C) viral load in the blood of AG129 mice on days 2, 4, and 6.

3.3. DENV infection induces organ damage and inflammatory cell infiltration in IFNR knockout mice of different backgrounds

DENV can further replicate in the body and be transported to the whole body through the blood, causing damage to multiple tissues and organs. Next, we further analyzed the effects of DENV infection on organ damage and inflammatory cell infiltration in mice with different backgrounds through organ indexes and histopathology. By testing the liver, kidney, and brain organ indexes of mice with different backgrounds, we found that the organ indexes of the IFNAR^−/−BALB/c mouse model group significantly increased. The kidney and brain indexes of the IFNAR^−/−C57BL/6 mouse model group increased. In the AG129 mouse model group, only the kidney index increased (Figure 4A).

FIGURE 4.

Organ indexes and pathological staining results of IFNR (interferon receptor) knockout mice with different backgrounds infected with DENV (dengue virus). (A) Organ indexes of IFNR knockout mice with different backgrounds after infection with DENV; (B) pathological staining results of liver of mice with different backgrounds; (C) pathological staining results of kidney of mice with different backgrounds; (D) pathological staining results of brain of mice with different backgrounds. The liver index of the IFNAR^−/−BALB/c background mouse model group increased significantly (p < 0.01), and there was no significant difference in the liver index between the IFNAR^−/−C57BL/6 and AG129 mouse model groups. The kidney index of the IFNAR^−/−BALB/c mouse model group increased significantly (p < 0.05), and there was no significant difference in the kidney index between the IFNAR^−/−C57BL/6 and AG129 mouse model groups. The brain index of the IFNAR^−/−BALB/c and C57BL/6 mouse model groups increased significantly (p < 0.05), and there was no significant difference in the brain index between the AG129 mouse model groups. *p < 0.05, **p < 0.01; ns: no statistically significant difference.

We further studied the histopathological changes in the major systemic organs, specifically the liver, kidney, and brain. HE staining is a commonly used staining technique that can dye the cytoplasm red. Therefore, the structure of cells and tissues can be clearly observed, and it is an important technique to clinically determine tissue pathology. We used HE to stain the organ tissues of mice and found that the degree of hepatocyte necrosis was more severe in IFNAR^−/−BALB/c and C57BL/6 background mice, whereas the degree of hepatocyte necrosis in AG129 mice was less severe (Figure 4B). The kidney parenchyma of mice with different backgrounds exhibited different degrees of inflammatory infiltration and necrosis. The renal parenchyma damage was the most severe in mice with IFNAR^−/−BALB/c, whereas the damage was milder in mice with IFNAR^−/−C57BL/6 and AG129 (Figure 4C). Neurons of mice with different backgrounds exhibited varying degrees of inflammatory infiltration and necrosis. IFNAR^−/−BALB/c and C57BL/6 mice exhibited fibrotic changes in brain cells, with pyknosis and dark staining of the nuclei, and the degree of neuronal damage in AG129 mice was the mildest (Figure 4D). These results suggest that IFNAR^−/−BALB/c mice experienced the most severe inflammatory infiltration and organ damage.

3.4. Results of vascular leakage in various organs of IFNR knockout mice with different backgrounds infected with DENV

After the DENV continues to replicate in the organs, it will induce the release of mouse endothelial cells and inflammatory mediators, leading to increased vascular permeability and severe vascular leakage. Subsequently, we used Evans blue staining to detect the vascular leakage level of mice. Under normal physiological conditions, albumin cannot penetrate the vascular endothelium. When vascular permeability increases, the tight‐junction function of endothelial cells is lost, and a large amount of albumin can pass through the blood–brain barrier. The amount of Evans blue that penetrates into the tissue reflects the degree of vascular leakage. ²³ The results showed that after mice were infected with DENV, the brain vascular leakage in mice in each group was mild, and the liver and kidney vascular leakage was severe, and Evans blue staining could be observed with the naked eye. The difference in color of the liver, kidney, and brain of mice in the IFNAR^−/−BALB/c model group was the greatest (Figure 5A). The organs of IFNR knockout mice with different backgrounds exhibited different degrees of vascular leakage (Figure 5B–D). The vascular leakage in the liver and kidney of mice in each group was more obvious than that in the brain. The optical density (OD) values of the liver, kidney, and brain in IFNAR^−/− BALB/c mice after infection with DENV were 0.48 ± 0.08, 0.67 ± 0.13, and 0.12 (0.12, 0.15), respectively, all of which were higher than those in the other two groups (Table 1). There was no statistically significant difference in the level of brain vascular leakage in the AG129 mouse model group. Evans blue exudation in the liver, kidney, and brain of IFNAR^−/−BALB/c mice increased most significantly (p < 0.01). In IFNAR^−/−C57BL/6 mice, only the leakage in blood vessels in the liver was significantly more pronounced (p < 0.01). This suggests that IFNAR^−/−BALB/c mice exhibit the most severe vascular leakage after DENV infection.

FIGURE 5.

Vascular leakage in various organs of mice with different backgrounds. (A) Color comparison of liver, kidney, and brain of mice with different backgrounds after injection of Evans blue; (B) results of vascular leakage in liver, kidney, and brain of IFNAR^−/−BALB/c mice; (C) results of vascular leakage in liver, kidney, and brain of IFNAR^−/−C57BL/6 mice; (D) results of vascular leakage in liver, kidney, and brain of AG129 mice. The level of hepatic vascular leakage in the IFNAR^−/−BALB/c and C57BL/6 mouse model groups was significantly increased (p < 0.01). The level of renal vascular leakage in the IFNAR^−/−C57BL/6 and AG129 mouse model groups was significantly increased (p < 0.05), and the level of renal vascular leakage in the IFNAR^−/−BALB/c mouse model group was significantly increased (p < 0.01). The level of cerebrovascular leakage in IFNAR^−/−C57BL/6 and IFNAR^−/−BALB/c mice was significantly increased (p < 0.05), whereas there was no significant difference in the level of cerebrovascular leakage in AG129 mice. *p < 0.05, **p < 0.01; ns: no statistically significant difference.

TABLE 1.

Results of vascular leakage in the liver, kidney, and brain of mice in each group.

Mouse species	Group	Liver OD value	Kidney OD value	Brain OD value
IFNAR−/− BALB/c	Mock	0.23 ± 0.03	0.36 ± 0.06	0.09 (0.09, 0.10)
	DENV	0.48 ± 0.08**	0.67 ± 0.13**	0.12 (0.12, 0.15)**
IFNAR−/− C57BL/6	Mock	0.28 ± 0.04	0.44 ± 0.04	0.09 ± 0.01
	DENV	0.44 ± 0.05**	0.50 ± 0.03*	0.10 ± 0.01*
AG129	Mock	0.31 ± 0.06	0.29 ± 0.04	0.10 ± 0.01
	DENV	0.43 ± 0.11*	0.47 ± 0.11*	0.10 ± 0.01

Note: (($\overline{\mathsf{x}}$±s)/M(P25, P75), n = 4)

Abbreviations: DENV, dengue virus; IFNAR, type I interferon receptor knockout.

*p < 0.05, **p < 0.01.

3.5. IFNAR ^−/− BALB/c mice were infected with DENV using different injection methods and exhibit clinical signs

Based on the aforementioned studies, we know that the comprehensive dengue‐like symptoms exhibited by IFNAR^−/−BALB/c mice are the most severe. Next, we used IFNAR^−/−BALB/c mice to explore the effects of different injection methods on DENV infection. The model was established as shown in Figure 6. After the virus attack, the weight of mice in each group with different injection methods gradually decreased after the third day, and the weight loss of mice in the IP injection group was the most obvious (Figure 7A). All mouse groups exhibited symptoms such as lethargy, erect hair, and weakness, and the symptoms gradually worsened over time. Some mice in the IP injection group even exhibited pathological conditions such as lower‐limb paralysis and moribundity (Figure 7B). DENV is a neurotropic virus that can infect supporting cells of the central nervous system, causing neurological complications such as encephalopathy and encephalitis, with an incidence of ~0.5%–6.2%, ²⁴ , ²⁵ which is consistent with our observations. Some mice in the IP injection group exhibited lower‐limb paralysis and weakness, which are often associated with central nervous system damage. These findings all emphasize that the clinical symptoms of mice were more severe after IP injection of DENV.

FIGURE 6.

Schematic diagram of DENV (dengue virus) infection in IFNAR^−/−BALB/c mice through different injection methods.

FIGURE 7.

Clinical symptoms of DENV (dengue virus) in IFNAR^−/−BALB/c mice infected with DENV by different injection methods. (A) Body weight changes in IFNAR^−/−BALB/c mice infected with DENV by different injection methods; (B) clinical scores of IFNAR^−/−BALB/c mice infected with DENV by different injection methods.

3.6. IFNAR ^−/− BALB/c mouse viral load in blood after DENV infection via different routes

The replication of DENV in the blood after infection of IFNAR^−/−BALB/c mice was further evaluated by detecting the viral load. By comparing the viral load on days 2, 4, and 6, we found that the viral load of mice in the IP injection group was higher than that of the other two groups. On the sixth day of infection, the viral load of mice in the IV and foot pad injection groups decreased, but the viral load of mice in the IP injection group remained at a higher level (Figure 8A–C). This shows that viremia infection by IP injection lasts longer.

FIGURE 8.

Viral load of DENV (dengue virus) in IFNAR^−/−BALB/c mice infected with DENV by different injection methods. (A) Viral load in the blood of IFNAR^−/−BALB/c mice on day 2 after infection with DENV by different injection methods; (B) viral load in the blood of IFNAR^−/−BALB/c mice on day 4 after infection with DENV by different injection methods; (C) viral load in the blood of IFNAR^−/−BALB/c mice on day 6 after infection with DENV by different injection methods. On day 2 of infection, the viral loads in the IP (intraperitoneal) and foot pad injection groups were higher than those in the IV (tail vein) injection group (p < 0.05). On day 4 of infection, the viral load in the IP injection group remained at a higher level compared with the foot pad and IV injection groups. On 6 of infection, the IP injection group had a significantly higher viral load than the IV and foot pad injection groups (p < 0.05). *p < 0.05, **p < 0.01; ns: no statistically significant difference.

3.7. DENV infection in IFNAR ^−/− BALB/c mice induces tissue with organ damage and inflammatory cell infiltration

Consistent with the earlier experimental results, in this model, the liver and spleen may be the main affected organs. DENV infection of IFNAR^−/−BALB/c mice through different injection methods can cause varying degrees of edema and congestion in the liver, spleen, and brain. We observed that the spleens in the IP and foot pad injection groups were larger than those in the normal group, whereas the spleen size in the IV injection group was no different from that in the normal group. The liver, brain, and kidney indexes of mice in the IP, IV, and foot pad injection groups all increased. However, the liver and brain indexes of mice in the IP injection group increased most significantly (Figure 9A). There was no statistically significant difference in the spleen index between the IV injection groups, which is speculated to be due to the fact that after the tail vein injection of the virus, the virus quickly entered the systemic circulation and did not directly act on the spleen.

FIGURE 9.

Histopathological and organ index results after DENV (dengue virus) infection in IFNAR^−/−BALB/c mice. (A) Changes in liver, spleen, and brain indexes in IFNAR^−/−BALB/c mice infected with DENV by different injection methods; (B) HE (hematoxylin–eosin) staining results of liver of IFNAR^−/−BALB/c mice infected with DENV by different injection methods; (C) HE staining results of spleen of IFNAR^−/−BALB/c mice infected with DENV by different injection methods; (D) HE staining results of brain of IFNAR^−/−BALB/c mice infected with DENV by different injection methods. The liver index increased significantly in the IP (intraperitoneal), IV (tail vein), and foot pad injection groups (p < 0.01). The brain index increased significantly in the foot pad injection group (p < 0.05), and the brain index increased significantly in the IP and IV injection groups (p < 0.01). The spleen index increased significantly in the IP and foot pad injection groups (p < 0.01), and there was no statistically significant difference in the spleen index in the IV injection group. *p < 0.05, **p < 0.01; ns: no statistically significant difference.

Next, we further analyzed the degree of damage to each organ by HE staining. After infection with DENV, a large area of hepatocyte nuclei in the IP injection group was observed to be darker. Even binucleated or multinucleated hepatocytes, accompanied by punctate to focal necrosis, are manifestations of hepatocyte damage. In addition, some hepatocytes condensed through eosinophilic changes, the cell body became smaller, the cytoplasm became dense, the nucleus condensed and disappeared, and eosinophilic bodies were formed, which is a manifestation of cell apoptosis. Focal necrosis with inflammatory infiltration occurred in the IV and foot pad injection groups (Figure 9B). As the largest immune organ in the body, the spleen increases the production and activation of immune cells, including macrophages, B cells, and T cells, when DENV infects the body, leading to splenomegaly. The effects of stimulating the spleen immune response in the IP and foot pad injection groups were significantly better than those in the IV injection group (Figure 9C). Brain neurons in the IP injection group exhibited necrosis with inflammatory infiltration, which may be one of the causes of lower‐limb paralysis in mice. The degree of neuronal damage in the foot pad injection group was milder (Figure 9D). These results indicate that DENV infection intraperitoneally in IFNAR^−/−BALB/c mice had the most significant effect, with the most severe damage and inflammatory infiltration in various organs of the mice.

4. DISCUSSION

4.1. Main research findings and advantages

After DENV infects mice, it usually causes a series of physical changes, reflecting the impact of viral infection on the host immune system. ²⁶ Among them, weight loss in mice may be attributed to factors such as inflammatory response, fever, and decreased appetite. The behavioral changes in mice are manifested as lethargy, reduced exercise, reduced activity, and even pathological states of erect hair and weakness, which may be attributed to the activation of the immune system and systemic inflammatory response (Figures S1 and S2). After DENV infects mice, it replicates in the blood, and the DENV can be transmitted throughout the body with the blood, causing damage to multiple tissues and organs. After the DENV reaches the body's organs through blood circulation, it replicates in large quantities, continues to invade mouse endothelial cells, and induces the release of inflammatory mediators, resulting in increased vascular permeability and severe vascular leakage. ²⁷

Although the widely used AG129 model can support the efficient replication of DENV, it also has limitations such as pathological target deviation, excessive immune deficiency, and lack of neural invasion. Due to the characteristics of the 129/Sv genetic background of AG129 mice, vascular leakage syndrome is mostly limited to the gastrointestinal tract, whereas human dengue patients are more typically characterized by liver damage and plasma leakage into the thoracic and abdominal cavities. ²⁸ The complete lack of type I and type II IFN signals results in the inability of AG129 mice to produce effective neutralizing antibody responses, which severely limits its application in vaccine evaluation. ²⁹ Clinically, about 5% of dengue patients will develop neurological complications such as encephalitis, but the AG129 model rarely reproduces such phenotypes, hindering the simulation of viral central nervous system invasion mechanisms. ³⁰

The IFNAR^−/−BALB/c mouse model in this study showed unique advantages in terms of clinical relevance to organ pathology, controllable degree of immunodeficiency, and dynamic differences in viremia. Compared with the AG129 mouse model, the IFNAR model showed mainly vascular leakage in the liver and kidneys, which is consistent with human dengue liver injury. It was also accompanied by significant splenomegaly and abnormal liver color (Figures S1 and S2), findings that are more consistent with spontaneous spleen rupture and hepatitis injury in human patients. ⁷ More importantly, neurological symptoms such as paralysis of the lower limbs also occurred, reproducing the central nervous system invasion characteristics of DENV in an IFNR knockout model for the first time, filling the gap in the AG129 model in this field. In addition, the IFNAR^−/−BALB/c mice may maintain the ability to recruit some immune cells by retaining the expression of specific chemokines (e.g., CXCL10). This explains why more significant inflammatory infiltration can be observed in this model.

4.2. Limitations and future prospects of the study

Studies have shown that type II IFNs can regulate the polarization of Th1/Th2 cells. ³¹ Although our model retains the type II IFN signal, the absence of type I IFNs may disrupt the balance of cytokines (e.g., IL‐12 and IL‐4), thereby affecting T‐cell differentiation. Although this model enhances the phenotypes of the nervous system and liver, this immune response bias may affect the production of CD8⁺ T cells, thus failing to fully simulate the human antiviral immune response. ³² , ³³ , ³⁴ Additionally, BALB/c mice exhibit a Th2‐biased immune response, characterized by elevated levels of IL‐4 and IgE. ³⁵ This bias may enhance the antibody‐dependent enhancement (ADE) effect of DENV infection, leading to overestimation of the ADE effect and underestimation of the protective immune response. ³⁶

Future research could focus on several directions. First, the expression of specific cytokines (e.g., IL‐12) could be restored through gene editing technology to reshape the Th1/Th2 immune balance, thereby more accurately simulating the human antiviral immune response. Second, for the neuroinvasive phenotype first demonstrated in the model, spatial transcriptomics and immunofluorescence techniques could be combined to analyze the molecular pathways by which DENV breaks through the blood–brain barrier and the response mechanisms of glial cells. Third, combining the immune characteristics data of clinical patients with the results of animal models may help establish a more precise dengue fever immunopathological prediction model, providing a new direction for early intervention in severe cases and potentially further enhancing the translational value of this study.

5. CONCLUSION

Based on the aforementioned results, we successfully established an animal model of IFNAR^−/−BALB/c mice infected with DENV by IP injection, which provides a reliable experimental vector for studying the pathogenic mechanism of DENV and the development of related prevention and control strategies.

AUTHOR CONTRIBUTIONS

Dan Liao: Data curation; methodology; software; validation; writing – original draft. Ming Zhong: Data curation; methodology; software; writing – review and editing. Wenjiang Zheng: Methodology; project administration; resources; supervision. Zhendong Guo: Data curation; investigation; methodology; software. Ye Zhou: Data curation; investigation; methodology. Qiuhong Li: Data curation; methodology; software. Lijuan Qiu: Data curation; methodology. Liangwen Yu: Resources; supervision. Haishan Long: Project administration; supervision. Geng Li: Project administration; resources; supervision; writing – review and editing.

FUNDING INFORMATION

This study was supported by the Project of the Incubation Program for the Science and Technology Development of Chinese Medicine Guangdong Laboratory/ Hengqin Laboratory (HQL2024PZ043), Guangdong Province Natural Science Foundation ‐ Guangzhou‐South China Joint Youth Fund Project (2023A1515110849), Guangdong Province Medical Research Fund Project (B2024112), the Scientific Research Special Project of the Joint Construction Project of High‐level Hospitals between Guangzhou University of Chinese Medicine and the Scientific Research Fund Project (GZYZS2024G09), and Special Project of the Research Platform of Guangdong Provincial Department of Traditional Chinese Medicine (20254040).

CONFLICT OF INTEREST STATEMENT

The authors declare that there are no conflicts of interest.

ETHICS STATEMENT

The animal study was approved by the Animal Ethics Committee of Guangzhou University of Chinese Medicine (protocol code: 20241018005, date of approval: October 18, 2024), conducted in accordance with animal welfare guidelines and the Animal Welfare Law.

Supporting information

Figure S1.

Liao D, Zhong M, Zheng W, et al. Optimizing the dengue virus infection mouse model: Comparing different backgrounds and infection route for enhanced stability. Anim Models Exp Med. 2026;9:103‐114. doi: 10.1002/ame2.70090