Involvement of Histamine and RhoA/ROCK in Penicillin Immediate Hypersensitivity Reactions

The mechanism of penicillin immediate hypersensitivity reactions has not been completely elucidated. These reactions are generally considered to be mediated by IgE, but penicillin-specific IgE could not be detected in most cases. This study demonstrated that penicillin was able to cause vascular hyperpermeability in a mouse model mimicking clinical symptoms of penicillin immediate hypersensitivity reactions. The first exposure to penicillin also induced immediate edema and exudative reactions in ears and lungs of mice in a dose-dependent manner. Vasodilation was noted in microvessels in ears. These reactions were unlikely to be immune-mediated reactions, because no penicillin-specific IgE was produced. Furthermore, penicillin treatment directly elicited rapid histamine release. Penicillin also led to F-actin reorganization in human umbilical vein endothelial cells and increased the permeability of the endothelial monolayer. Activation of the RhoA/ROCK signaling pathway was observed in ears and lungs of mice and in endothelial cells after treatment with penicillin. Both an anti-histamine agent and a ROCK inhibitor attenuated penicillin immediate hypersensitivity reactions in mice. This study presents a novel mechanism of penicillin immediate hypersensitivity reactions and suggests a potential preventive approach against these reactions.

The mechanism of penicillin immediate hypersensitivity reactions has not been completely elucidated. This study is the first to investigate whether penicillin treatment might induce IgE-independent vascular hyperpermeability in a mouse model which mimics the typical clinical symptoms of immediate allergic reactions to penicillin. We demonstrate that histamine release and the RhoA/ROCK signaling pathway play important roles in penicillin NAHRs, and the results suggest that an anti-histamine agent and a ROCK inhibitor might be useful for the prevention and treatment of some penicillin hypersensitivity reactions.

Penicillin Induces Immediate Non-allergic Hypersensitivity Reactions in Mice.
To examine the extent of vascular leakage induced by penicillin, we used Evans blue (EB) as a marker of plasma protein extravasation, because it binds plasma albumin with high affinity and forms an albumin-EB complex. This property has been widely applied to assess vascular leakage and the extent of inflammatory extravasation 30,31 . We injected penicillin intravenously with EB and determined vascular leakage by observing EB leakage in the ear. As a control, EB alone did not cause visible vascular leakage in mice, as indicated by EB extravasation. Dose-dependent vascular leakage was observed in ears after penicillin-sensitized mice received a challenge treatment of penicillin (1.25, 5, or 20 kU/mouse) in combination with EB. EB extravasation occurred at approximately 7-10 minutes and reached a peak at approximately 20 minutes after penicillin treatment. Interestingly, marked dose-dependent EB extravasation was also observed in unsensitized mice receiving a single dose of penicillin/EB at the same dose levels as those for the penicillin challenge in sensitized mice. Prior sensitization did not enhance the extent of penicillin-induced EB extravasation compared with that in the penicillin single-dose group (Fig. 1a,b,d,e). The microvascular leakage of EB noted in penicillin-treated mice was similar to that caused by compound 48/80, a mast cell degranulation inducer (Fig. 1c,f,g).
We further ruled out the role of penicillin-specific IgE in vascular leakage. The passive cutaneous anaphylaxis (PCA) test is a standard method of measuring allergen-specific IgE antibody levels in allergy mouse models 32 . As a positive control, ovalbumin (OVA) induced a 100% positive response, whereas no visible reaction was observed in penicillin-treated groups (Table 1). This result indicated that penicillin-specific IgE was not responsible for the vascular hyperpermeability in the mice treated with penicillin.
To additionally investigate the IgE-independent reactions induced by penicillin in mice, ear and lung weight data were obtained at 30 minutes after penicillin treatment, and histological observations were performed to evaluate whether penicillin causes ear edema and pulmonary exudation. A single dose of penicillin at 250 or 500 kU/kg significantly enhanced ear weight by 20% and 35% compared with that in the control group (Fig. 1h). The increase in ear weight was dose dependent. The microscopic examination confirmed that the increase in ear weight in the penicillin-treated groups was due to edema, which was characterized by ear thickening, inflammatory cell infiltration, and microvascular dilation ( Fig. 1j-m). Moreover, a single dose of penicillin at 250 or 500 kU/kg also resulted in pulmonary inflammation. The differences in weight between wet and dry lungs (lung weight reduction) increased by 12% and 17%, respectively, compared with that of the control group (Fig. 1i). The differences in weight were also dose dependent. The histological changes, including congestion, edema with alveolar septal broadening, and perivascular and peribronchial monocyte infiltration were observed in the lungs of mice in the penicillin-treated groups ( Fig. 1n-q).
We next directly observed the local vascular changes in ears. Treatment with 500 kU/kg penicillin significantly dilated microvessels. Vascular dilation, edema and leakage were noted ever since 5 minutes and got severe gradually within 30 minutes after penicillin treatment ( Fig. 1r-u). The vessel diameter change of penicillin group was approximately 1.4-fold, 2-fold, and 2.5-fold compared with normal saline group at 5 minutes, 15 minutes and 30 minutes after treatment (Fig. 1v).
Histamine Release Contributes to Penicillin NAHRs. We tested whether penicillin directly elicits histamine release by performing in vitro and in vivo experiments. In the in vitro experiment, histamine release in rat basophilic leukemia-2H3 (RBL-2H3) cells markedly increased in a concentration-dependent manner after the cells were treated with penicillin (20, 40, or 60 kU/mL) for 30 minutes (Fig. 2a). In the in vivo experiment, the histamine concentration in plasma was notably higher in mice treated with a single dose of 500 kU/kg penicillin, with a 33% higher concentration than that in the control group (Fig. 2b). Moreover, pretreatment with the anti-histamine drug loratadine significantly alleviated the vascular leakage, ear swelling and pulmonary inflammation induced by a single dose of 500 kU/kg penicillin. Penicillin-induced vascular hyperpermeability, increased ear weight, vasodilation and edema were ameliorated in the loratadine pretreatment group (Fig. 2c-m). Microscopic examination showed that penicillin-induced edema, congestion and inflammatory cell infiltration in the ears and lungs were attenuated by pretreatment with loratadine ( Fig. 2n-s).
Penicillin Enhances Endothelial Monolayer Permeability through Activation of the RhoA/ROCK Signaling Pathway. Because penicillin-induced vascular leakage is correlated with vascular endothelial barrier dysfunction, we evaluated whether penicillin enhances endothelial permeability through an in vitro experiment. By assessing FITC-dextran diffusion through the confluent endothelial monolayer, we found that incubation with penicillin (1.25, 5, or 20 kU/mL) for 1 hour significantly increased the human umbilical vein endothelial cell (HUVEC) monolayer permeability (characterized by fluorescence apparent permeability coefficients, P app ) in a concentration-dependent manner (Fig. 3a). These results demonstrated that penicillin enhanced endothelial permeability, which in turn led to vascular hyperpermeability.
To determine whether penicillin induces alterations in the actin cytoskeleton, actin filaments (F-actin) were stained and observed. Rhodamine-phalloidin staining of confluent endothelial monolayer or un-confluent HUVECs both demonstrated that stimulation with penicillin (1.25, 5, or 20 kU/mL) for 1 hour resulted in noticeable F-actin accumulation and reorganization in HUVECs. F-actin intensity was increased in a concentration-dependent manner ( Fig. 3b-i). Stress fiber formation and assembly may dissociate cell connections and cause a separation of cells and looseness of the endothelial barrier, thereby contributing to an increase in endothelial permeability.
We further determined whether the RhoA/ROCK signaling pathway is involved in the penicillin-induced endothelial permeability increase and F-actin reorganization in HUVECs. The results indicated that incubation with 5 kU/mL penicillin activated the RhoA/ROCK signaling pathway in a time-dependent manner. The protein levels of GTP-bound RhoA (GTP-RhoA), ROCK1, phospho-myosin phosphatase targeting subunit 1 (p-MYPT1) and phospho-myosin light chain 2 (p-MLC2) in HUVECs were increased within 15 minutes of penicillin exposure and reached maximum levels after 30 minutes of penicillin treatment (Fig. 3j).

RhoA/ROCK Signaling Pathway Plays a Role in Penicillin NAHRs.
Because the RhoA/ROCK signaling pathway is involved in penicillin-induced endothelial hyperpermeability, we next verified the possible contribution of this pathway in penicillin-induced vascular leakage, which is closely associated with penicillin NAHRs. GTP-RhoA, ROCK1, p-MYPT1, and p-MLC2 expression levels were markedly increased in a time-dependent manner in the ears and lungs of mice treated with a single dose of 500 kU/kg penicillin. These proteins were expressed at peak levels 15-30 minutes after penicillin treatment and then gradually declined but did not reach baseline levels until 120 minutes after penicillin injection (Fig. 5a,b).
We further explored the effect of fasudil on penicillin NAHRs in mice. Pretreatment with fasudil hydrochloride significantly attenuated the penicillin-induced augmentation of ROCK1, p-MYPT1 and p-MLC2 protein levels in the ears and lungs of mice (Fig. 6a,b). Notably, when mice were pretreated with fasudil hydrochloride, the vascular leakage increase, ear swelling, and pulmonary inflammation evoked by a single dose of 500 kU/kg penicillin were markedly alleviated. Penicillin-induced vascular hyperpermeability, increased ear weight, vasodilation and edema were significantly prevented by pretreatment with fasudil hydrochloride (Fig. 6c-m). Moreover, histological observations confirmed that edema and inflammatory exudation in the ears and lungs were largely eliminated in the fasudil hydrochloride pretreatment group (Fig. 6n-s).

Discussion
Whether penicillin causes NAHRs in addition to immediate AHRs is unknown. In the current study, we demonstrated that the first exposure of mice to penicillin through intravenous injection rapidly provoked significant vascular hyperpermeability. Because typical symptoms of penicillin immediate hypersensitivity reactions, which include rash, urticaria, angioedema, bronchospasm, and anaphylaxis are related to the skin and respiratory systems, we chose the ear (skin) and lung as indicative organs to evaluate the underlying mechanism of penicillin hypersensitivity reactions. As a consequence of penicillin treatment, tissue edema and exudative inflammation appeared in the ear and lung. Prior sensitization with penicillin did not enhance the vascular hyperpermeability reaction. No abnormal signs were observed during the penicillin sensitization phase. There was no evidence of penicillin-specific IgE elicited by penicillin sensitization, as determined by PCA tests. Additionally, penicillin directly stimulated histamine release, and loratadine markedly attenuated penicillin immediate reactions in mice. Thus, penicillin-induced vascular hyperpermeability in mice might be NAHRs. FDA guideline on immunotoxicology evaluations indicates that when signs of anaphylaxis are observed in animal studies, follow-up studies should be performed. However, no effective and validated animal experimental models for NAHRs have been available 33 . This model is the first animal model of penicillin NAHRs, which mimics aspects of the clinical manifestations of penicillin immediate allergic reactions.
As a low molecular weight drug, penicillin is incapable of producing an immunogenic response. Instead, penicillin or metabolites could act as haptens binding to carrier proteins to induce a specific antibody response 34 . Previous animal studies have shown that cutaneous anaphylactic reactions can be provoked by using penicillin/ metabolite-protein conjugates in sensitized animals [35][36][37] . In this study, because we sought to determine whether penicillin induces vascular hyperpermeability independent of IgE, mice were treated with penicillin without protein conjugates. According to the results of the current study, no specific IgE was produced in the mice treated with penicillin that presented induced vascular hyperpermeability.
The dosing instructions for commercially available penicillin injections indicate that penicillin is clinically used at doses varying from 2 × 10 6 units/day to 2 × 10 7 units/day for intravenous infusion. The doses of penicillin used in mice (125, 250, and 500 kU/kg) in this study were within the clinical dose range according to the conversion of animal doses to human equivalent doses 38 . Apparently, as indicated in this study, penicillin-induced vascular hyperpermeability in mice occurred in a dose-dependent manner. It could be reasonably speculated that some of the penicillin reactions are elicited by high doses.
Agents directly activate mast cells and basophils in an IgE-independent manner, and elicit the release of histamine and other inflammatory mediators, is one of the primary reasons for the occurrence of NAHRs 15,17 . Histamine release is reported to be responsible for NAHRs caused by opioid drugs 39 , iodinated contrast material 40 , bradykinin 41 , vancomycin 42 , and polymyxins 43 . In this study, penicillin induced histamine release in both in vitro and in vivo experiments, and anti-histamine drug diminished penicillin-induced NAHRs. Therefore, our results indicate that penicillin directly elicits histamine release, which may be related to penicillin-induced NAHRs.
Clinical manifestations, including rash, urticaria and angioedema, which occur most frequently in patients with immediate penicillin hypersensitivity, are the consequences of vascular hyperpermeability. Vascular permeability is mainly determined by properties of the vascular wall and hemodynamic factors. Endothelial barrier function is one of the main factors influencing vascular permeability 44 . Endothelial permeability is partly mediated by paracellular mechanisms, which are controlled by the dynamic opening and closing of endothelial cell-cell junctions 21,45 . Endothelial junctions have been proposed to bind to cytoskeletal and signaling proteins via associated proteins and to play crucial roles in maintaining vascular integrity and barrier function 45,46 . The dissociation of vascular endothelial cell-cell junctions, which is usually caused by actin cytoskeleton reorganization, widens the intercellular space, increases vascular permeability, and facilitates fluid and plasma protein movement 21,47 . RhoA, one member of the Rho GTPase family, is a key regulatory molecule for organization of the actin Scientific RepoRts | 6:33192 | DOI: 10.1038/srep33192 cytoskeleton and integrity of the intercellular junctions 48 . RhoA induces the formation of actin stress fibers at least in part by increasing the phosphorylation of MLC, which is mediated by its downstream effector ROCK 49,50 . When RhoA is in an activated GTP-bound form, it binds ROCK, thereby enhancing ROCK's kinase activity 51 .   ROCK directly phosphorylates MLC. In the same time, ROCK also inactivates myosin light chain phosphatase (MLCP) by phosphorylating the myosin-binding subunit (MYPT1), which is considered as a hallmark of ROCK activity 44,52 . Both functions of ROCK would result in phosphorylation of MLC, which in turn lead to stress fiber formation 18 . Our results indicate that penicillin promotes actin stress fiber formation and assembly in HUVECs and increases the permeability of the endothelial monolayer. This effect leads to fluid leakage from the circulatory system into the interstitial space, thus resulting in interstitial edema and exudative inflammation, as observed in the ears and lungs of mice treated with a single dose of penicillin. Penicillin-induced exudation in the ears and lungs of mice and HUVEC monolayer hyperpermeability were accompanied by increases in GTP-RhoA, ROCK1, p-MYPT1 and p-MLC2 levels. Pretreatment with the ROCK inhibitor fasudil hydrochloride significantly blocked penicillin-induced endothelial cell stress fiber reorganization and attenuated endothelial permeability in the confluent HUVEC monolayer. Consistently with these results, penicillin-induced vascular hyperpermeability in mice was also alleviated by fasudil hydrochloride. Thus, this study strongly suggests that the RhoA/ROCK signaling pathway plays an important role in penicillin NAHRs. Meanwhile, vascular dilation is also one factor affecting vascular permeability. Vasodilation elevates blood flow and increases the intravascular pressure, which results in hyperpermeability of vascular barrier 53 . It has been reported that the vascular hyperpermeability induced by many inflammatory agents such as histamine 54 and bradykinin 55 are related with vasodilation. The present study reveals that penicillin provokes microvascular dilation immediately after the first exposure, which is associated with penicillin-induced vascular leakage and also responsible for penicillin NAHRs.
Approximately 10% of the worldwide population report a history of penicillin allergy, although only a few of these patients have been confirmed to be truly allergic 1-3 . Currently, the widely accepted main reasons for the high misunderstanding of penicillin allergy include the lack of previous clinical diagnoses for patients labeled as allergic, the dependence of the "allergy" symptoms on the illness or interactions between the illness and penicillin, and the attenuation of penicillin-specific IgE over time 3 . However, on the basis of the results of this study, we consider that penicillin causes NAHRs in addition to well-known IgE-mediated AHRs. This underlying mechanism may also contribute to the low reproducibility of penicillin allergies. When patients' immediate reactions are NAHRs instead of AHRs, the penicillin-specific IgE could not be detected in further diagnosis of penicillin allergy. In addition, this mechanism may explain some cases in which diagnostic tests are negative despite the presence of immediate reactions to penicillin, which are mainly manifested by urticaria, pruritus, and angioedema 56 . According to our research, NAHRs rapidly occur after the first penicillin exposure in a dose-dependent manner, thus underscoring the necessity of optimizing penicillin administration schemes and dosages. When the penicillin dosage is controlled, it is possible that some patients who have previously exhibited penicillin hypersensitivity would be able to tolerate penicillin and safely use it again. We suggest that the mechanisms of penicillin NAHRs include direct histamine release and RhoA/ROCK signaling pathway activation. Thus, administration of anti-histamine agents or inhibitors of the RhoA/ROCK signaling pathway may be useful in preventive approaches against penicillin NAHRs. Furthermore, this animal model of penicillin NAHRs should be a useful tool for further exploring the mechanisms of drug hypersensitivity reactions. This study reveals a novel mechanism of penicillin hypersensitivity reactions and provides a potential path for the clinical use of penicillin in a subgroup of patients with penicillin hypersensitivity. Animals. Specific-pathogen-free male ICR mice were received at 8-10 weeks of age and randomized on a body weight-stratified basis.

Methods
Cell Culture. HUVECs were kindly provided by Dr. Song (Academy of Military Medical Sciences, Beijing, China). RBL-2H3 cells were obtained from the cell bank of Shanghai Science Academy. HUVECs and RBL-2H3 cells were respectively cultured in Dulbecco's modified Eagle medium containing 10% fetal bovine serum (FBS) and minimum essential medium containing 10% FBS at 37 °C with 5% CO 2 . Assessment of Vascular Leakage in Mice. In the sensitization experiment, mice were immunized by intraperitoneal injection (ip) of 0.4 mL of a mixture consisting of an equal volume of normal saline/aluminum hydroxide gel (AHG) (Thermo Fisher Scientific, Rockford, IL, USA) or penicillin (Huabei Pharmaceutical, Shijiazhuang, Hebei, CHN)/AHG (2.5, 10, or 40 kU/animal) on days 1, 3, 5 and 19, and challenged by intravenous slow injection (iv) (approximately 2 minutes) of 0.4 mL of a mixture consisting of equal volume of normal saline/EB, or penicillin/EB (1.25, 5, or 20 kU/animal) on day 33. In the unsensitization experiment, the naive mice received (iv) a single dose of normal saline/EB, compound 48/80 (Sigma-Aldrich, Louis, MO, USA)/EB (0.075 mg/animal), or penicillin/EB at the same doses as those of challenge treatment in sensitized mice. Vascular leakage was assessed by evaluating blue staining in the ear 30 minutes after drug/EB administration. Each ear was given a score of 1 to 6 according to the blue area in the ear, where "1" denoted no visible blue area, and "2 to 6" represented visible blue areas of < 12.5%, 12.5-25%, 25-50%, 50-75%, and > 75%, respectively. The ears of each mouse were preserved in 2 mL of formamide for EB extraction.
Passive Cutaneous Anaphylaxis Test. Mice were immunized (ip) by penicillin/AHG or normal saline/ AHG on days 1, 3, 5 and 19 as described above or immunized (ip) by 0.4 mL OVA (Sigma-Aldrich, Louis, MO, USA) (0.8 mg/animal) at the same dose schedule as that of the penicillin groups. Immunization sera were collected on day 33. Fifty microliters of serially diluted sera (1, 1/2, 1/4, and 1/8) was intradermally injected into the dorsal skin of depilated recipient mice. After two hours, the recipient mice were challenged by administration (iv) of 0.4 mL of normal saline/EB, OVA/EB (0.4 mg/animal) or penicillin/EB (1.25, 5, or 20 kU/animal). The PCA test results were considered positive when the blue dorsal region had a diameter of 5 mm or more 30 minutes after the challenge injection.

Evaluation of Ear and Lung Pathological Changes.
Mice were treated (iv) with penicillin (125, 250, or 500 kU/kg) or normal saline and euthanized thirty minutes after dosing. Punch biopsies of the left ears (5 mm diameter) and the left lobes of the lungs (wet lung) were weighed immediately, and then the left lobes of the lungs were dried at 80 °C for 24 hours and weighed again (dry lung). The differences in weight between wet and dry lungs were calculated. The right ears and the rest of the lungs were preserved in formalin for histopathological evaluation.
Vessel in vivo Microscopy. Mice were anesthetized and treated (iv) with 500 kU/kg penicillin or normal saline. The ear vessels were monitored at pre-dose, and at 5, 15, and 30 minutes after dosing. The change in vessel diameter (%) was calculated according to equation (1), where D t and D 0 represent the vessel diameter at t minutes after dosing and at pre-dose, respectively.
(1) t 0 Histamine Assay. RBL-2H3 cells were stimulated with penicillin (20, 40, or 60 kU/mL) for 30 minutes. The concentrations of penicillin used here were lower than the IC 5 . The supernatants were used as extracellular samples (EC), and the cell lysates prepared by Triton 100-PBS were used as intracellular samples (IC). For histamine assays, 100 μ L EC or IC samples were mixed with 50 μ L 0.4 mol/L NaOH and 10 μ L of 0.1% o-phthalaldehyde methanol solution. After 10 minutes of incubation, 50 μ L of 0.5 mol/L HCl was added to stop the reaction 57 . The fluorescence was measured at an excitation wavelength of 360 nm and an emission wavelength of 450 nm using a microplate reader. The ratio of histamine release was calculated according to equation (2) where EC represents EC fluorescence, and IC represents IC fluorescence.
Mice were treated (iv) with 500 kU/kg penicillin or normal saline. Blood was collected 5 minutes after penicillin or normal saline injection and centrifuged to obtain plasma. The histamine concentrations were determined by enzyme-linked immunosorbent assays (Immuno-Biological Laboratories, MN, USA) according to the manufacturer's instructions. Briefly, the samples were acylated for 1 hour, added to the histamine microtiter plate, and incubated with histamine antiserum for 18 hours at 4 °C. After the plate was washed, it was incubated with enzyme conjugate for 1 hour and reacted with substrate for 20 minutes at room temperature. The reaction was stopped by addition of stop solution to each well. The absorbance was measured using a microplate reader at 450 nm.
Actin Cytoskeleton Observation. HUVECs were seeded at a cell density of 2 × 10 4 cells/well or 1 × 10 5 cells/well in a 48-well plate. After cultured for 2 days, cells or confluent monolayer were stimulated with penicillin (1.25, 5, or 20 kU/mL) for 1 hour or pretreated with 10 μ M fasudil hydrochloride (Chase Sun Pharmaceutical, Tianjin, CHN) for 30 minutes 58 and then treated with 20 kU/mL penicillin for 1 hour. The concentrations of penicillin used were lower than the IC 5 . Cells were fixed with formalin for 10 minutes, blocked in normal saline containing 1% BSA for 20 minutes, and stained with 5 μ g/mL rhodamine-phalloidin (Sigma-Aldrich, Louis, MO, USA) for 1 hour. Fluorescence images were captured by using a fluorescence microscope.
Transendothelial Permeability Assay. HUVECs were plated in a transwell insert chambers (0.4 μ m pore size, Millicell, Merck Millipore, Cork, IRL) at a density of 4 × 10 4 cells/chamber and formed confluent monolayer in 4 days. Then cells were incubated with penicillin (1.25, 5, or 20 kU/mL) for 1 hour or pretreated with 10 μ M fasudil hydrochloride for 30 minutes and then treated with 20 kU/mL penicillin for 1 hour. Medium containing FITC-dextran (MW 4 × 10 4 kDa, Sigma-Aldrich, Louis, MO, USA) (1 mg/mL) was loaded in the upper compartment of the chamber. FITC-dextran that diffused through the endothelial monolayer to the lower compartment was collected after 10, 20, 30, 40, 50, and 60 minutes 59 . The concentration of FITC-dextran in the lower compartment was measured at an excitation wavelength of 485 nm and an emission wavelength of 510 nm by using a microplate reader. The fluorescence P app (cm/s) was calculated according to equation (3) where [A], t, A, V and [L] represent the cumulative concentration of fluorescence in the receiver compartment (mg/mL), time (s), the area of membrane surface (cm 2 ), and the donor concentration of fluorescence (mg/mL), respectively 60 .
Effect of Fasudil Hydrochloride and Loratadine on Penicillin NAHRs. Mice were pretreated with