Disruption of gamma-glutamyl leukotrienase results in disruption of leukotriene D(4) synthesis in vivo and attenuation of the acute inflammatory response

Abstract

To study the function of gamma-glutamyl leukotrienase (GGL), a newly identified member of the gamma-glutamyl transpeptidase (GGT) family, we generated null mutations in GGL (GGL(tm1)) and in both GGL and GGT (GGL(tm1)-GGT(tm1)) by a serial targeting strategy using embryonic stem cells. Mice homozygous for GGL(tm1) show no obvious phenotypic changes. Mice deficient in both GGT and GGL have a phenotype similar to the GGT-deficient mice, but approximately 70% of these mice die before 4 weeks of age, at least 2 months earlier than mice deficient only in GGT. These double-mutant mice are unable to cleave leukotriene C(4) (LTC(4)) to LTD(4), indicating that this conversion is completely dependent on the two enzymes, and in some organs (spleen and uterus) deletion of GGL alone abolished more than 90% of this activity. In an experimental model of peritonitis, GGL alone is responsible for the generation of peritoneal LTD(4). Further, during the development of peritonitis, GGL-deficient mice show an attenuation in neutrophil recruitment but not of plasma protein influx. These findings demonstrate an important role for GGL in the inflammatory response and suggest that LTC(4) and LTD(4) have distinctly different functions in the inflammatory process.

FIG. 1

Generation of ES cells carrying GGT^tml and GGL^tml mutations at both cis and trans locations. (A) An ES clone heterozygous for GGT^tml was used for gene targeting at the GGL locus. In one of the GGT alleles, the first noncoding exon (I) and coding exons 1 and 2 had been replaced by PGK-hprt cassette (23). Note that the two genes are only ∼3 kb apart. On the map, only the first exon of GGL is depicted. B, BamHI; X, XhoI; H, HindIII. (B) A targeting construct was designed such that the first exon of GGL was replaced by a PGKneo selective marker cassette; the MC1tk cassette was used for negative selection. (C and D) Predicted mutant alleles after homologous recombination. When the targeted deletion of GGL (GGL^tml) occurs on the same chromosome as GGT^tml, the two mutations are in cis (C); when GGT^tml and GGT^tml are on different chromosomes, they are in trans (D). (E) Southern blot analysis of ES cells using an external probe for the GGL gene (a 0.25-kb PflMI/SacI fragment). The probe identifies an 8.2-kb BamHI band (wild type) as well as a 4.2-kb BamHI band (mutant). ES cell clones positive for both bands may carry GGT-GGL double mutations (either in cis or in trans). Y, GGL locus targeted; N, GGL locus not targeted. (F) Southern blot analysis to distinguish cis and trans double mutations. A probe isolated from the 3′ hprt sequence was used to hybridize XhoI-digested ES cell DNA. A 31-kb band represents the allele carrying only GGT^tml (GGL^tml is on the other allele; thus, the two mutations are in trans [T]); the 14-kb band carries both alleles (the two mutations are in cis [C]).

FIG. 2

Survival of GGT^tml/tml-GGL^tml/tml mice (n = 70) compared to that of wild-type mice (n = 73).

FIG. 3

Analysis of endogenous cysteinyl LT in lavage fluid from zymosan A-induced peritonitis by HPLC. Three hours after zymosan A injection, cysteinyl LT in the supernatants of the lavage fluid were isolated. Prostaglandin B₁ (PGB₁) (Sigma) was used to correct for recovery of the LT. Peaks of PGB₁, LTC₄, and LTE₄ were authenticated by comparing them with retention times of commercial standards (Cayman Chemical Co.). Peak UI is an unidentified nonspecific peak.

FIG. 4

Effect of GGL deficiency on plasma protein extravasation (A) and neutrophil infiltration (B) in a model of experimental peritonitis. GGL-null and wild-type mice were injected intraperitoneally with 1 mg of zymosan A. At 0, 1, 2, 3, 4, 6, and 24 h, mice were euthanatized and peritoneal lavage was performed with 4 ml of PBS. To evaluate plasma extravasation, the total protein concentration in the supernatant of the lavage fluid was measured (10); for quantification of neutrophil accumulation, MPO activity was determined in the cell pellets of the lavage fluid. Data shown are means ± standard errors of the mean. n = 3 to 8; ∗, P < 0.05; ∗∗, P < 0.01.