Targeted Gene Disruption Reveals the Role of Cysteinyl Leukotriene 1 Receptor in the Enhanced Vascular Permeability of Mice Undergoing Acute Inflammatory Responses*

The cysteinyl leukotrienes (cysLTs), leukotriene (LT) C4, LTD4, and LTE4, are proinflammatory lipid mediators generated in the mouse by hematopoietic cells such as macrophages and mast cells. There are two mouse receptors for the cysLTs, CysLT1 receptor (CysLT1R) and CysLT2R, which are 38% homologous and are located on mouse chromosomes X and 14, respectively. To clarify the different roles of the CysLT1R and CysLT2R in inflammatory responses in vivo, we generated CysLT1R-deficient mice by targeted gene disruption. These mice developed normally and were fertile. In an intracellular calcium mobilization assay with fura-2 acetoxymethyl ester, peritoneal macrophages from wild-type littermates, which express both CysLT1R and CysLT2R, responded substantially to 1 × 10−6 m LTD4 and slightly to 1 × 10−6 m LTC4, whereas the macrophages from CysLT1R-deficient mice did not respond to either LTD4 or LTC4. Plasma protein extravasation, but not neutrophil infiltration, was significantly reduced in CysLT1R-deficient mice subjected to zymosan A-induced peritoneal inflammation. Plasma protein extravasation was also significantly diminished in CysLT1R-deficient mice undergoing IgE-mediated passive cutaneous anaphylaxis as compared with the wild-type mice. Thus, the cysLTs generated in vivo by either monocytes/macrophages or mast cells utilize CysLT1R for the response of the microvasculature in acute inflammation.

The cysteinyl leukotrienes (cysLTs), 1 leukotriene (LT) C 4 , LTD 4 , and LTE 4 , are proinflammatory lipid mediators (1,2) that have been implicated in allergic and asthmatic inflammatory responses on the basis of the efficacy of specific intervention with their biosynthetic inhibitors (3) or their receptor blockers (4). LTC 4 synthase (LTC 4 S) is the major enzyme responsible for the generation of LTC 4 , the parent compound of the cysLTs (5-7). LTC 4 S conjugates reduced glutathione to an epoxide metabolite of arachidonic acid, LTA 4 , generated by 5-lipoxygenase in the presence of the 5-lipoxygenase-activating protein (8,9). After carrier-mediated export (10), the sequential cleavage of glutamic acid and glycine from the glutathione moiety of LTC 4 yields LTD 4 and LTE 4 (11,12), respectively. LTC 4 S is expressed in various types of hematopoietic cells such as mast cells, basophils, eosinophils, and monocytes/macrophages. Neutrophils, which lack LTC 4 S but express LTA 4 hydrolase (13), make a dihydroxy leukotriene, LTB 4 , which has potent chemotactic activity via the LTB 4 receptors, BLT1 and BLT2 (14,15).
Two types of human receptors for cysLTs, designated cysteinyl leukotriene 1 receptor (CysLT 1 R) and CysLT 2 R, belonging to the seven-transmembrane, G protein-coupled receptor family, were recently cloned and shown to be 38% homologous (16,17). The rank order of affinities of the cysLTs for the CysLT 1 R and CysLT 2 R defined with transfected cells is LTD 4 Ͼ Ͼ LTC 4 Ͼ LTE 4 Ͼ Ͼ LTB 4 and LTD 4 ϭ LTC 4 Ͼ Ͼ LTE 4 Ͼ Ͼ LTB 4 , respectively. The genes for human CysLT 1 R and CysLT 2 R map to chromosome Xq13-q21 and 13q14, respectively. CysLT 1 R is expressed on airway smooth muscle, alveolar macrophages, peripheral blood monocytes, eosinophils (16,18), and endothelial cells (19). CysLT 2 R is expressed on alveolar macrophages, airway smooth muscle, cardiac Purkinje cells, adrenal medulla cells, peripheral blood leukocytes, and brain cells (17). We and others (20 -22) have reported that mouse CysLT 1 R has two isoforms of cDNA resulting from alternative splicing and that each can function as a receptor for LTD 4 in transfected cells with a ligand preference similar to that of human CysLT 1 R. Mouse CysLT 2 R was recently cloned and exhibited a ligand profile of LTC 4 ϭ LTD 4 Ͼ Ͼ LTE 4 (23,24). Mouse CysLT 1 R and CysLT 2 R are 38.9% homologous in amino acid sequence, and their gene loci are mapped to chromosomes X and 14, respectively.
We have previously shown that the increased vascular permeability in zymosan A-induced, monocyte/macrophagedependent peritoneal inflammation and in IgE-mediated, mast cell-dependent passive cutaneous anaphylaxis is diminished by ϳ50% in LTC 4 S null mice as compared with their wild-type littermates (7). Recently, Shi et al. (25) demonstrated that ␥-glutamyl leukotrienase null mice subjected to zymosan A-induced peritonitis had diminished neutrophil infiltration but intact plasma protein extravasation. In as much as the conversion of LTC 4 to LTD 4 was completely abolished in the peritoneal cavity of the ␥-glutamyl leukotrienase null mice, their findings suggest different roles for LTC 4 and LTD 4 through the same or different receptors. To clarify the role of CysLT 1 R relative to CysLT 2 R in vivo, we generated CysLT 1 R gene-disrupted mice by homologous recombination and examined plasma protein extravasation and neutrophil infiltration in zymosan A-induced peritoneal inflammation and plasma protein extravasation in IgE-mediated passive cutaneous anaphylaxis.

EXPERIMENTAL PROCEDURES
Generation of CysLT 1 R Gene-disrupted Mice-A 7-kb mouse genomic DNA fragment containing exons II-IV of the CysLT 1 R gene (20) was subcloned into a pBluescriptII vector (Stratagene, La Jolla, CA). A neo gene cassette from pMC1Neo (Stratagene) was inserted to replace 278 nucleotides of the coding region of the mouse CysLT 1 R gene, and the herpes simplex virus thymidine kinase (TK) gene was inserted at the 5Ј-end of the gene. The resultant targeting vector was linearized and electroporated into a C57BL/6 mouse embryonic stem cell line, ES-MK (26). The embryonic stem cells were selected with G418 (200 g/ml; Invitrogen) and ganciclovir (2 M), and homologous recombination was confirmed by Southern blot analysis of ScaI-digested genomic DNA from each embryonic stem cell clone with a ϳ400-bp 3Ј-fragment as a probe located outside the targeting vector (Fig. 1A). The embryonic stem cell clones, verified by Southern blot analysis, were microinjected into blastocysts from BALB/c mice, and chimeric males were obtained. Chimeras were bred to C57BL/6 females, and offspring were genotyped by Southern blot analysis of the tail DNA. Heterozygous females were bred to C57BL/6 males to obtain CysLT 1 R(Ϫ) C57BL/6 males. These CysLT 1 R(Ϫ) or (ϩ) C57BL/6 males were bred to CysLT 1 R(ϩ/Ϫ) C57BL/6 females to obtain CysLT 1 R(Ϫ/Ϫ) and CysLT 1 R(ϩ/ϩ) C57BL/6 females. All experiments except for passive cutaneous anaphylaxis were performed with the F 2 and F 3 generations of CysLT 1 R(Ϫ) and CysLT 1 R(ϩ) males. Passive cutaneous anaphylaxis was performed with the F 3 generation of CysLT 1 R(Ϫ) and CysLT 1 R(ϩ) males and F 3 generation of CysLT 1 R(Ϫ/Ϫ) and CysLT 1 R(ϩ/ϩ) females. All animal studies were approved by the Animal Care and Use Committee of the Dana-Farber Cancer Institute.
Northern Blot Analysis-Total RNA from the lung tissue of CysLT 1 R(Ϫ) and CysLT 1 R(ϩ) mice was isolated with Tri-Reagent (Sigma). A 20-g sample of the total RNA was resolved by electrophoresis on a formaldehyde-denatured gel and transferred to a nylon membrane (Pall Corp., Ann Arbor, MI) with 20 ϫ SSC for 24 h. Hybridizations with 32 P-labeled mouse CysLT 1 R and glyceraldehyde-3-phosphate dehydrogenase cDNAs were performed as described (20).
Isolation of Peritoneal Macrophages and Calcium Mobilization Assay-Resident peritoneal macrophages were isolated as described by Qiu et al. (27). Briefly, peritoneal lavage was performed with 10 ml of ice-cold phosphate-buffered saline (PBS). After the lavage fluid was centrifuged at 500 ϫ g for 5 min, the cells were cultured in Petri dishes in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin for 3 h at 37°C in a humidified atmosphere with 5% CO 2 . The nonadherent cells were removed by extensive washing. The adherent cells were collected by dispersion with 0.53 mM EDTA in Hanks' balanced salt solution, re-suspended in Dulbecco's modified Eagle's medium, seeded at 1 ϫ 10 6 cells onto a glass coverslip (13-mm diameter) in 24-well plates, and cultured for 18 h at 37°C in a humidified atmosphere with 5% CO 2 . The cells were washed twice with Hanks' balanced salt solution containing 1 mM CaCl 2 , 1 mM MgCl 2 , and 0.1% bovine serum albumin (HBSA 2ϩ ) and incubated with 2.5 M fura-2 acetoxymethyl ester (Molecular Probes, Eugene, OR) in the presence of 2.5 mM probenecid for 30 min at 37°C. After the cells were labeled, the coverslips were washed and placed in a diagonal position in a standard 1-cm square methacrylate cuvette containing 2 ml of HBSA 2ϩ . The cuvette was fitted with a rubber O-ring to position the coverslip just above a magnetic stirring bar. The cells were stimulated with LTB 4 , LTC 4 , LTD 4 , LTE 4 (Cayman Chemical, Ann Arbor, MI) or ATP (Sigma). Fluorescence output was measured with excitation at 340 and 380 nm in a fluorescence spectrophotometer (Hitachi F-4500, Japan), and the relative ratio of fluorescence emitted at 510 nm was recorded. Assays of the cellular response to LTC 4 were performed in the presence of 50 mM serine-borate to inhibit the conversion of LTC 4 to LTD 4 (28).
Zymosan A-induced Peritoneal Inflammation-Each mouse received an intravenous injection of 0.5% Evans blue dye (10 ml of dye solution/kg of body weight) in PBS immediately before the intraperitoneal injection of 1 ml of zymosan A suspension (1 mg/ml in PBS; Sigma). Mice were euthanized by CO 2 before and 30, 60, 120, and 240 min after the injection of zymosan A and underwent peritoneal lavage with 4 ml of cold PBS. Each peritoneal lavage fluid was divided equally into two tubes, 2 ml for Evans blue dye extravasation and cell counts and 2 ml for the myeloperoxidase (MPO) assay. Cells were sedimented from the lavage fluid by centrifugation at 500 ϫ g for 5 min, and Evans blue dye extravasation was assessed by light spectrophotometry of the supernatants at 610 nm.
The cell pellets were suspended in 100 l of PBS, and the cells were counted. Cells (2 ϫ 10 4 ) were cytospun onto a glass slide, stained with Diff-Quik (Dade Behring, Deerfield, IL), and analyzed for cell types.
The total neutrophils in the lavage fluid were calculated from the percentage of the cell type and the total cell count.
The cell pellets from 2 ml of the lavage fluid were suspended in 200 l of 50 mM potassium phosphate buffer (pH 6.0) with 0.5% hexadecyltrimethylammonium bromide (Sigma), frozen and thawed three times, and sonicated at level 3 for 1 min at 4°C with a Branson sonifier (Branson Ultrasonics Corp., Danbury, CT). The supernatants of the cell extracts were assayed for MPO as described (29 -31). Briefly, 10 l of diluted samples were added to a 96-well plate. The reaction was initiated by the addition of 190 l of assay buffer containing 0.167 mg/ml of o-dianosidine (Sigma) and 0.0005% hydrogen peroxide. The rate of change of absorbance at 405 nm was monitored in kinetic mode, and V max was calculated by a plate reader (Molecular Device, Sunnyvale, CA). Levels of MPO were determined from the calibration curve with human neutrophil MPO (Sigma) as a standard.
The LTC 4 S(Ϫ/Ϫ) and LTC 4 S(ϩ/ϩ) male mice (7) used for the MPO assay and cell counts were from the intercrossing of LTC 4 S(ϩ/Ϫ) males and females that had been backcrossed to the C57BL/6 strain at the N 5 generation.
Passive Cutaneous Anaphylaxis-Mice received intradermal injections of 25 ng of mouse monoclonal anti-dinitrophenyl (DNP) IgE in 25 l of saline in the right ear and 25 l of saline only in the left ear. After 20 h, mice were injected intravenously with 100 g of DNP-human serum albumin and 1% Evans blue dye in 100 l of PBS. The mice were euthanized 15 and 30 min after the intravenous injection, and a 6-mmdiameter disc of tissue was obtained from the center of each ear. Each disc was incubated in 200 l of formamide at 55°C for 48 h. Extravasation of Evans blue dye was quantitated by spectrophotometric analysis of the formamide extracts at 610 nm, and vascular permeability enhancement was calculated as the net difference between the sensitized and control ears.
Statistical Analysis-The results of the experiments were expressed as means Ϯ S.E. Student's t test was used for the statistical analysis in cases in which the variance was homogeneous, and Welch's test was used when the variance was heterogeneous. A value of p Ͻ 0.05 was considered significant.

RESULTS AND DISCUSSION
Generation of CysLT 1 R Gene-disrupted Mice-The targeting vector was designed so that the neo gene insertion interrupts the coding region that is common to the long and short isoforms of the mouse CysLT 1 R gene (Ref. 20; Fig. 1A). Out of 31 embryonic stem cell clones that survived in the presence of G418 and ganciclovir, two were identified as targeted clones by Southern blot analysis for the 4.8-kilobase (kb) band. The embryonic stem cell clones were microinjected into BALB/c-derived blastocysts. Two male chimeric mice (Ͼ70% chimerism by coat color) were obtained from each clone and bred to C57BL/6 females. Since the CysLT 1 R gene is on the X chromosome, heterozygotes obtained at F 1 generation were female. The heterozygous females were bred to C57BL/6 males to obtain CysLT 1 R(Ϫ) males. These CysLT 1 R(Ϫ) males were further bred to CysLT 1 R(ϩ/Ϫ) females. Southern blot analysis of ScaIdigested DNA from the progeny demonstrated a 4.8-kb band for the disrupted gene and a 4.0-kb band for the wild-type gene; it also revealed that the ratio was 1:1 for CysLT 1 R(Ϫ) and CysLT 1 R(ϩ) males and for CysLT 1 R(Ϫ/Ϫ) and CysLT 1 R(ϩ/Ϫ) females, respectively, as illustrated for one litter (Fig. 1B). The CysLT 1 R(Ϫ) males and CysLT 1 R(Ϫ/Ϫ) females developed normally and were fertile.
Since mouse CysLT 1 R mRNA is most abundantly expressed in the lung (20), we examined whether homologous recombination by the targeting vector disrupted expression of the CysLT 1 R mRNA in the lung. By Northern blot analysis with total RNAs, CysLT 1 R mRNA was not detected in CysLT 1 R(Ϫ) mice, whereas mRNAs with sizes of 1.4 and 4.0 kb were detected in the lung of CysLT 1 R(ϩ) mice (Fig. 1C). This result suggests that the transcript from the recombinant allele is unstable due to displacement of 278 bp of the coding region by the neo gene.
To seek functional evidence of the effect of the CysLT 1 R gene disruption, we utilized cysLT-evoked intracellular calcium mo-bilization in peritoneal macrophages, which expressed both CysLT 1 R and CysLT 2 R mRNA by reverse transcriptase-PCR (24). The macrophages from wild-type mice showed a rapid intracellular calcium flux in response to stimulation with 1 ϫ 10 Ϫ6 M LTD 4 ( Fig. 2A), a lesser signal in response to 1 ϫ 10 Ϫ6 M LTC 4 (Fig. 2B) or LTE 4 (data not shown), and no response to LTB 4 (data not shown). The preference for LTD 4 is consistent with the ligand profile obtained with Chinese hamster ovary cells stably expressing either the long or the short isoform of the mouse CysLT 1 R cDNA (20). In contrast, the macrophages from the CysLT 1 R(Ϫ) mice did not show an intracellular calcium flux in response to stimulation with either 1 ϫ 10 Ϫ6 M LTD 4 (Fig. 2C) or LTC 4 (Fig. 2D). That macrophages from the CysLT 1 R(Ϫ) mice showed a prominent intracellular calcium flux to ATP indicates that the signal transduction mechanism for intracellular calcium flux is intact. These results indicate that CysLT 1 R is the major receptor for cysLTs in peritoneal macrophages.
That LTD 4 can stimulate rat alveolar macrophages in vitro to increase the mRNA levels of inflammatory cytokines, such as macrophage inflammatory protein (MIP)-1␣ and tumor necrosis factor (TNF)-␣, has suggested a priming function of the cysLTs (32). Mouse alveolar (33) and peritoneal (34) macro-phages and in vitro cultured mouse (35) and human (36) mast cells are immune effectors that can generate LTC 4 . More recently, CysLT 1 R expression has been recognized on human alveolar macrophages (16) and human culture-derived mast cells (37) by immunodetection. Moreover, LTC 4 and LTD 4 can stimulate interleukin (IL)-4-primed human mast cells to transcribe and release cytokines, such as IL-5, TNF-␣, and MIP-1␤. The CysLT 1 R antagonist, MK-571, not only blocks this induction of proinflammatory cytokines by exogenous cysLTs, but also partially inhibits their expression followed by Fc⑀RI-mediated activation, revealing that CysLT 1 R mediates both innate and adaptive immune responses (38). The close developmental relationship of mast cells and monocytes (39,40) suggests that CysLT 1 R may have a similar effector function for macrophages. Furthermore, CysLT 1 R expression can be upregulated by IL-4 and IL-13 in human peripheral blood monocytes (41) and by IL-5 in eosinophil-differentiated HL-60 cells (42). These findings imply that CysLT 1 R may contribute to the cellular phase of a Th2 cell-driven inflammatory response by augmenting production of inflammatory cytokines by immune effector cells.
Role of CysLT 1 R in Zymosan A-induced Peritoneal Inflammation-We examined the role of CysLT 1 R in the in vivo monocyte/macrophage-dependent inflammatory response to zymosan A-induced peritoneal inflammation (30,31). Plasma protein extravasation assessed 30 -240 min after zymosan A injection was significantly suppressed in CysLT 1 R(Ϫ) mice at all time points as compared with the wild-type mice (Fig. 3A). This result was essentially identical to that obtained with LTC 4 S null mice, which lack the capacity to generate cysLTs (7).
We also examined neutrophil infiltration, which is generally apparent at 120 min and peaks at 240 min after the injection of zymosan A (25,30,43). As assessed at these time points with either neutrophil counts or MPO activity, there was no significant difference in neutrophil infiltration between CysLT 1 R(Ϫ) and (ϩ) mice (Fig. 3B, left). The findings were similar for the LTC 4 S(Ϫ/Ϫ) mice (Fig. 3B, right), which do not generate cys-LTs but do provide LTB 4 (7). Because of a study showing that plasma protein extravasation in response to zymosan A-in- duced intraperitoneal inflammation is intact while neutrophil accumulation is attenuated in ␥-glutamyl leukotrienase null mice (25), our results could imply that the LTC 4 -CysLT 1 R signal is sufficient to increase vascular permeability but not for neutrophil-directed function. However, our additional finding that intraperitoneal neutrophil accumulation is intact in LTC 4 S null mice 120 and 240 min after zymosan A injection suggests that cysLTs are not critical to this response in our strains. Our results indicate that a cysLT-CysLT 1 R-initiated stimulus plays a major role in the early phase of plasma protein extravasation, but not in the neutrophil infiltration that follows in a later phase of zymosan A-induced peritoneal inflammation.
Role of CysLT 1 R in Mast Cell-dependent Passive Cutaneous Anaphylaxis-Because we previously demonstrated that the edema induced in the ears of LTC 4 S null mice by passive cutaneous anaphylaxis is reduced by 50% as compared with their wild-type littermates (7), we next determined whether the CysLT 1 R is critical to the alteration of vascular permeability in passive cutaneous anaphylaxis by using CysLT 1 R-deficient mice. Preliminary experiments revealed that a change of ear thickness after the antigen challenge is more difficult to detect in C57BL/6 mice than in BALB/c or FVB mice, and therefore we used Evans blue dye extravasation to detect the plasma protein leakage induced by passive cutaneous anaphylaxis (44 -46). Evans blue dye extravasation in the ear of CysLT 1 R-deficient mice was reduced by ϳ50% at 15 min and by ϳ80% at 30 min after antigen challenge as compared with their wild-type littermates (Fig. 4). These results indicate that the increased vascular permeability mediated by cysLTs in passive cutaneous anaphylaxis significantly involves their action through the CysLT 1 R.
We produced CysLT 1 R-deficient mice by using C57BL/6 strain-derived embryonic stem cells because the X-linked location of the gene allowed generation of strain purity for F 1 heterozygous females. Another advantage of using the C57BL/6 strain for phenotypic analyses of the gene-disrupted mice could be the recent finding that in this strain the expression levels of CysLT 1 R and CysLT 2 R are higher than in the 129 strain (24). We then demonstrated that CysLT 1 R is the major functional CysLT receptor on peritoneal macrophages by the intracellular calcium mobilization assay. Finally, our in vivo findings with the CysLT 1 R gene-disrupted mice reveal a major role for the cysLTs-CysLT 1 R signal in innate and adaptive immune responses of vascular permeability enhancement during zymosan A-induced peritoneal inflammation and IgE-mediated, mast cell-dependent local anaphylaxis.