This study was carried out in accordance with the recommendations of ethical guidelines and regulations for the use of laboratory animals and cells issued by the Institute of Chinese Materia Medica, China Academy of Chinese Medical Sciences, Beijing, China. All animal-related procedures adhered to the protocol approved by the Institutional Animal Care and Use Committee of the Institute of Chinese Materia Medica, China Academy of Chinese Medical Sciences.

Shuanghuanglian injection was obtained from a hospital. Aluminum hydroxide gel (AHG) was purchased from Thermo Fisher Scientific. Ovalbumin (OVA), FITC-dextran (MW 4 × 10⁴), and rhodamine-phalloidin were obtained from Sigma–Aldrich. Fasudil hydrochloride was acquired from Chase Sun Pharmaceutical. Rabbit anti-p-MLC2 (Thr18/Ser19) (3674), rabbit anti-MLC2 (3672), rabbit anti-p-MYPT1 (Thr 696) (5163), rabbit anti-MYPT1 (2634), and rabbit anti-RhoA (67B9) (2117) were purchased from Cell Signaling Technology. Activated RhoA pull down assay kit was obtained from Cytoskeleton. Histamine assay kit was purchased from Immuno-Biological Laboratories.

Male ICR mice were purchased from Vital River Laboratory Animal Technology, Co., Ltd. (Beijing, China). Mice were at 8–10 weeks of age and randomly divided into experimental groups on a body weight-stratified basis. Animals were kept under specific-pathogen-free conditions.

Human umbilical vein endothelial cells (HUVECs) were kindly provided by Dr. Song (Academy of Military Sciences, Beijing, China), who purchased them from American Type Culture Collection (ATCC, Manassas, VA, United States). Cells were used between passages 10–15 and were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37 °C in a 5% CO₂ incubator (HERAcell, Heraeus, Hanau, Germany).

The recommended clinical dose of SHLI is 60 mg/kg. In the present study, mice were given a dose of 600 mg/kg, calculated as per the equation for the conversion of animal doses (FDA, 2005).

During the sensitization phase, mice received intraperitoneal injection (i.p.) of a mixture containing equal volume of SHLI and AHG (SHLI/AHG) (300 mg/kg), normal saline/AHG (NS/AHG) or OVA (20 mg/kg) on days 1, 3, and 5.

Vascular leakage was assessed on day 19. Briefly, immunized mice were challenged by intravenous slow injection (i.v.) (approximately 2 min) of a mixture containing equal volume of SHLI and Evans blue (EB) (SHLI/EB) at 600 mg/kg or normal saline/EB (NS/EB). The naive mice received a single dose (i.v.) of SHLI/EB (600 mg/kg) or NS/EB. Thirty minutes after drug/EB administration, SHLI-induced changes on vascular permeability were analyzed by measuring the plasma protein leakage. EB extravasation in each ear was given a score of 1 to 6 based on the blue staining area (Han et al., 2016). The ears of each mouse were then preserved in formamide for EB extraction. The amount of EB extravasation was determined at an absorption wavelength of 610 nm using a microplate reader (Thermo Scientific Varioskan Flash, Thermo Scientific, United States).

Passive cutaneous anaphylaxis (PCA) test was performed on day 19. Immunization sera were collected and 50 μL of serially diluted sera (1, 1/2, 1/4, and 1/8) was intradermally injected into the dorsal skin of recipient mice. Two hours later, the recipient mice were challenged by administration (i.v.) of SHLI/EB (600 mg/kg), NS/EB, or OVA/EB (40 mg/kg). After 30 min, the mice were observed for EB extravasation into the subcutaneous tissue. When the blue spot in the dorsal region had a diameter of 5 mm or more, the PCA test result was considered positive.

Naive mice were treated (i.v.) with a single dose of NS/EB or SHLI/EB at 150, 300, or 600 mg/kg. Vascular leakage was assessed as described in the previous section. Punch biopsies of the left ears (8 mm diameter) were weighted and complete left ears were preserved in formamide for EB extraction. The right ears, lungs, and intestines were preserved in 10% neutral-buffered formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E). Slides were viewed using an Olympus BX53 microscope (Olympus, Tokyo, Japan) and images were obtained by Cell Sens Standard v 1.9 digital imaging software (Olympus).

In order to confirm the role of the RhoA/ROCK signaling pathway in SHLI-induced vascular leakage, mice were injected (i.p.) with 30 mg/kg fasudil hydrochloride once daily for three consecutive days. Thirty minutes after the last dose of fasudil, 600 mg/kg SHLI/EB was administrated (i.v.). In parallel groups, mice were treated (i.p.) with normal saline for 3 days and then administered (i.v.) 600 mg/kg SHLI/EB or NS/EB. Assessment of vascular leakage and histopathological evaluation were performed as described above.

Mice were injected (i.v.) with normal saline or 600 mg/kg SHLI. Five minutes after drug administration, blood samples were withdrawn and centrifuged at 3000 × g at 4°C for 15 min. Histamine concentration in the plasma was analyzed using the mouse histamine assay kit according to the manufacturer’s instructions.

Endothelial permeability was assessed using FITC-dextran flux measurements. HUVECs were grown in a transwell insert chamber (0.4 μm, Millicell, Merck Millipore, Cork, Ireland) until a confluent endothelial monolayer was formed. The cells were stimulated with SHLI (0.05, 0.1, or 0.15 mg/mL) for 1 h. Another group of cells were pretreated with 10 μM fasudil hydrochloride for 30 min and then incubated with 0.15 mg/mL SHLI for 1 h. The concentrations of SHLI used for stimulation corresponded to its equivalent concentration in the blood. These in vitro experimental concentrations of SHLI were derived from values converted from the in vivo blood concentrations where its major components were detected (Zeng et al., 2010; Zhou et al., 2013). In addition, the values were lower than the 5% maximal inhibitory concentration (IC₅). For the assessment of endothelial monolayer permeability, 400 μL DMEM containing FITC-dextran (1 mg/mL) was loaded in the upper part of the chamber. FITC-dextran that diffused through the cells was collected from the lower part of the chamber after 10, 20, 30, 40, 50, and 60 min, respectively. The fluorescence in the lower chamber was determined using a microplate reader (Thermo Scientific Varioskan Flash). Fluorescent apparent permeability coefficient (P_app, cm/s) of the samples was calculated (Wang L.Y. et al., 2011).

Rhodamine-phalloidin staining was used for visualizing F-actin. Endothelial monolayers or HUVECs were cultured on a 48-well plate and incubated with SHLI (0.05, 0.1, or 0.15 mg/mL) for 1 h, or pretreated with 10 μM fasudil hydrochloride for 30 min and then stimulated with 0.15 mg/mL SHLI for 1 h. Cells were then stained with 5 μg/mL rhodamine-phalloidin for 1 h at 37°C. Cells were visualized and images were obtained using Olympus IX71 fluorescent microscope (Olympus) and DP2-BSW v 2.2 software (Olympus).

In vivo, animals were administered (i.v.) 600 mg/kg SHLI, or pretreated with 30 mg/kg fasudil hydrochloride (i.p.) and then injected (i.v.) with SHLI as described above. Ears, lungs, and intestines were collected from single SHLI-treated animals at 15, 30, or 60 min after SHLI administration and from fasudil-pretreated mice at 30 min after SHLI injection. Tissues were homogenized in lysis buffer and lysates were centrifuged at 12,000 × g at 4°C for 10 min. Supernatants were stored at -80°C for western blot assay.

In vitro, HUVECs were incubated with 0.15 mg/mL SHLI for 15, 30, or 60 min, or pretreated with 10 μM fasudil hydrochloride for 30 min and then stimulated with 0.15 mg/mL SHLI for 30 min. Cells were collected and lysed at the scheduled time points. The supernatants of lysates were acquired as mentioned above.

Activity of RhoA was assessed by a pull down assay according to the manufacturer’s instructions. Briefly, GTP-RhoA was purified from the supernatants of lysates from the ears, lungs, and intestines of mice or HUVECs with the Rho binding domain region of rhotekin protein bound to glutathione-sepharose beads. The GTP-RhoA thus obtained was analyzed by western blot assay.

Equal amounts of protein (50 μg) were loaded and separated on sodium dodecyl sulfate polyacrylamide gel (SDS-PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes. The membranes were incubated with respective primary antibodies of anti-p-MLC2 (1:500), anti-MLC2 (1:500), anti-p-MYPT1 (1:500), anti-MYPT1 (1:500), and anti-RhoA (1:500) at 4°C over-night. After washing the membranes, they were incubated with the corresponding secondary antibodies at room temperature for 2 h. Protein bands were visualized using enhanced chemiluminescence detection substrate.

Data are expressed as mean (M) ± standard error of mean (SEM). Quantitative data were analyzed using the one-way analysis of variance (ANOVA) method. The score of vascular leakage of the ear was analyzed using the Rank-test. Statistical analyses were performed with SPSS 16.0 software. p < 0.05 was considered statistically significant.

Evans blue binds plasma albumin forming a firm albumin-EB complex and is widely employed as a marker to detect vascular leakage. In the present study, the extent of vascular permeability was evaluated by assessing EB extravasation in the ears of mice. In the vehicle control groups, none of the mice exhibited visible blue staining, indicating the inability of EB to induce vascular leakage on its own. However, significantly enhanced EB extravasation was observed in the ears of animals treated (i.v.) with SHLI/EB. This reaction was initiated approximately 10–15 min after administration (i.v.) of SHLI/EB and peaked by 25 min. Interestingly, EB extravasation was similar in SHLI-immunized and unimmunized mice (Figures 1A–C). Prior sensitization with SHLI had no effect on this vascular hyperpermeability reaction. These results suggest the role of non-immune mechanisms in eliciting vascular leakage caused by SHLI.

In order to confirm the absence of the involvement of immune mechanisms in SHLI-induced vascular leakage, IgE-mediated PCA test was performed. Significant blue spots were observed in all OVA-treated animals (the positive control group). However, mice administered with SHLI showed no dermal response (Table 1) indicating that SHLI could not elicit IgE production in mice. These results confirm that SHLI-induced vascular leakage was a non-immune-mediated pseudo-allergic reaction.

To understand better the role of SHLI-induced pseudo-allergic reactions, dose-dependent effect of SHLI/EB was analyzed in vivo in mice. Thirty minutes after a single injection of SHLI/EB, significant dose-dependent vascular hyperpermeability was observed (Figure 2A). The staining area and the amount of EB extravasation closely correlated with the dose of SHLI/EB administered (Figures 2B,C). SHLI also caused an increase in the ear weight of mice, which was about 11.2 and 18.1% in mice administered with 300 and 600 mg/kg SHLI, respectively, compared to the normal saline group (Figure 2D). In addition, SHLI triggered augmented histamine release in mice (Figure 2E), a feature generally seen in clinical pseudo-allergic reactions (Ring et al., 1999, 2014; McNeil et al., 2015).

As SHLI-induced ADRs are typically seen in the skin, gastrointestinal tract, and respiratory system, we examined the pathological alterations in the ears, lungs, and intestines of SHLI-treated mice. Consistent with increased vascular leakage, apparent dose-dependent ear thickening, edema, and neutrophil infiltration were observed in the ears (Figures 2F–I). In a dose-dependent manner, SHLI induced the swelling of pneumocytes and looseness of cytoplasm resulting in interstitial edema and alveolar wall-widening (Figures 2J–M). Microscopic examination of intestines revealed exudates and edema (Figures 2N–Q).

Vascular permeability is closely related to endothelial barrier function. Hence, we examined the effect of SHLI on HUVECs. The role of SHLI in endothelial permeability was analyzed by Transwell assay. In addition, cells were stained with rhodamine-phalloidin to visualize F-actin. Stimulation with SHLI for 1 h significantly enhanced the diffusion of FITC-dextran through a confluent endothelial monolayer in a concentration-dependent manner. The permeability coefficient of FITC-dextran was significantly increased in SHLI-treated groups than in the control group (Figure 3A). Consistent with the endothelial monolayer permeability assay, SHLI also evoked reorganization of actin cytoskeleton. Compared to the basal state represented by few thin stress fibers occasionally present at the cell periphery, stimulation with SHLI induced reorganization and assembly of F-actin to form thicker stress fibers that were diffused across the cells (Figures 3B–I).

Endothelial barrier dysfunction accompanied by formation and distribution of stress fibers is associated with the activation of RhoA (van Nieuw Amerongen et al., 1998, 2000). In order to study the role of the RhoA/ROCK signaling pathway in SHLI-induced endothelial barrier disruption, the protein levels of GTP-RhoA, p-MYPT1, and p-MLC2 in HUVECs stimulated with SHLI were analyzed. SHLI activated the RhoA/ROCK signaling pathway in a time-dependent manner. The expression of GTP-RhoA was up-regulated within 15 min and decreased by 60 min of SHLI stimulation (Figure 3J). Correspondingly, expression of p-MYPT1 and p-MLC2 were significantly enhanced by 15 min and peaked at 15–30 min after SHLI treatment (Figures 3K,L).

In order to confirm the involvement of the RhoA/ROCK signaling pathway in SHLI-induced endothelial changes, we additionally addressed the effect of fasudil, a specific inhibitor of ROCK, on SHLI-mediated breakdown of endothelial function. Preincubation of HUVECs with fasudil hydrochloride significantly decreased the over expression of p-MYPT1 and p-MLC2 elicited by SHLI (Figures 4A,B). In addition, SHLI-induced rearrangement of actin cytoskeleton and increased formation and distribution of stress fibers were inhibited by fasudil (Figures 4C–H). Fasudil hydrochloride significantly decreased the enhanced permeability coefficient induced by SHLI, thereby manifesting a protective effect as was evident in the fasudil pretreatment group compared to the single SHLI-treated group (Figure 4I).

To determine the role of the RhoA/ROCK signaling pathway in SHLI-induced pseudo-allergic reactions, the protein levels of GTP-RhoA, p-MYPT1, and p-MLC2 in the ears, lungs, and intestines of mice administered with SHLI were analyzed. Our results suggested activation of the RhoA/ROCK signaling pathway by SHLI as the expression of GTP-RhoA, p-MYPT1, and p-MLC2 in the ears, lungs, and intestines were significantly elevated after 15 min of SHLI administration and mostly remained consistently high for 15–60 min post-administration (Figures 5A–I). The time course of these augmented proteins correlated with the manifestation of vascular leakage in mice treated with SHLI, indicating a possible relation between the two processes.

To confirm the above findings, the effect of ROCK inhibitor in abrogating SHLI-induced pseudo-allergic reactions was analyzed. Protein assay results indicated that fasudil hydrochloride inhibited the enhanced expression of p-MYPT1 and p-MLC2 in the ears, lungs, and intestines of mice induced by SHLI administration (Figures 6A–F). Accordingly, mice treated with fasudil prior to administration of SHLI were obviously resistant to vascular dysfunction evoked by SHLI (Figure 6G). In addition, enhanced EB extravasation seen in the ears of mice and increase in the ear weight caused by SHLI were significantly alleviated by fasudil hydrochloride (Figures 6H–J). The histopathological examination also confirmed that SHLI-induced edema and exudates in the ears, lungs, and intestines were improved by the inhibitor. Remission of pathological alterations in the ear, alveolar wall, and intestinal villa were evidently observed in the fasudil pretreatment group compared to the single SHLI-treated group (Figures 6K–S).