Innate immune responses in peptidoglycan recognition protein L-deficient mice

Abstract

Peptidoglycan recognition proteins (PGRPs) constitute a family of innate immune recognition molecules. In Drosophila, distinct PGRPs bind to peptidoglycans on gram-positive or gram-negative bacteria and provide essential signals upstream of the Toll and Imd pathways required for immunity against infection. Four PGRPs, PGRP-L, -S, -Ialpha, and -Ibeta, are expressed from three genes in mammals. In this paper, we provide direct evidence that the longest family member, PGRP-L, is a secreted serum protein with the capacity to multimerize. Using gene targeting to create PGRP-L-deficient mice, we demonstrate little contribution by PGRP-L to systemic challenge using gram-negative bacteria (Escherichia coli, slightly less susceptible), Gram-positive bacteria (Staphylococcus aureus), or yeast (Candida albicans). Peritoneal macrophages from PGRP-L-deficient mice produced decreased amounts of the inflammatory cytokines interleukin 6 and tumor necrosis factor alpha when stimulated with E. coli or lipopolysaccharide, but comparable amounts when stimulated with S. aureus, C. albicans, or their cell wall components. Additionally, these cells produced similar amounts of cytokines when challenged with gram-positive or -negative peptidoglycans. In contrast to its critical role in immunity in flies, PGRP-L is largely dispensable for mammalian immunity against bacteria and fungi.

FIG. 1.

PGRP-L is a secreted serum protein. (A) HEK 293 cells were transfected with PGRP-L-myc (lanes 3, 4, 7, and 8) or PGRP-S-myc (lanes 1, 2, 5, and 6). Brefeldin A treatment (lanes 2, 4, 6, and 8) was used for 12 h to block protein secretion. Total cell lysates and culture supernatants were blotted with anti-Myc antibodies. (B) Drosophila Schneider cells were stably transfected with pAC5.1PGRP-L (lane L) or pAC5.1PGRP-S (lane S) or were mock transfected (lane C). Total cell lysates and culture supernatants were blotted with anti-V5 antibodies. (C) Mouse serum or PGRP-L-containing Schneider cell culture supernatants (lane L) were subject to immunoprecipitation with rat anti-PGRP-L antibodies. Immunoprecipitates were blotted with rabbit anti-PGRP-L. (D) HEK 293 cells were cotransfected with PGRP-L-myc and PGRP-L-V5 or PGRP-S-myc and PGRP-L-V5 or were mock transfected. Culture supernatants were incubated with anti-Myc antibodies. Immunoprecipitates were blotted with anti-V5 antibodies.

FIG. 2.

Generation of PGRP-L-deficient mouse. (A) The PGRP-L gene locus is depicted as a single line, with exons represented as filled boxes. Relevant restriction enzyme sites are shown. B, BamHI; H3, HindIII; RI, EcoRI. The organization of the targeting construct and the arrangement of the recombined PGRP-L allele are also shown. The dotted box indicates the GFP cassette that was engineered in-frame with the PGRP-L ATG start codon. The drug selection neomycin cassette (Neo) is depicted as a striped box, with two flanking loxP sites depicted as filled triangles. (B) Genotyping with Southern blot analysis. Tail DNA was subjected to agarose gel electrophoresis. The presence of the mutant allele (7.5 kb) and the wild-type allele (9.6 kb) was determined by Southern blotting, using a probe upstream of the homologous recombined sequence. (C) RT-PCR assay to determine PGRP-L expression. Liver RNA from wild-type, heterozygous, and homozygous gene-targeted mice was extracted and reverse transcribed into cDNA. PGRP-L expression was determined by PCR. The expression of hypoxanthine phosphoribosyltransferase was used as a control. (D) Serum from wild-type, heterozygous, and homozygous gene-targeted mice was immunoprecipitated using rat anti-PGRP-L antibodies. The immunoprecipitates were blotted with rabbit anti-PGRP-L antibodies.

FIG. 3.

E. coli and LPS challenges of PGRP-L-deficient mice. (A) Ten wild-type or PGRP-L-deficient animals were injected intraperitoneally with 2 million (left panel) or 5 million (right panel) live E. coli bacteria and were monitored for 5 days. Survival is plotted against time in days. (B) Three wild-type or PGRP-L-deficient animals were injected intraperitoneally with 2 million E. coli bacteria. Serum IL-6 and TNF-α levels were determined by ELISA 90 min after injection. (C) Ten wild-type or PGRP-L-deficient (KO) mice were injected intraperitoneally with 2 mg of LPS and were monitored for 10 days. Survival is plotted against time in days.

FIG. 4.

S. aureus and C. albicans challenges of PGRP-L-deficient mice. (A) Ten wild-type or PGRP-L-deficient mice were injected intraperitoneally with 10 million (left panel) or 1 billion (right panel) live S. aureus bacteria and were monitored for a 10- or 1-day period, respectively. Survival is plotted against time. (B) C. albicans (0.8 million viable organisms) was injected intravenously into 10 wild-type or PGRP-L-deficient mice. Survival was monitored over a 7-day period.

FIG. 5.

Cytokine production by thioglycolate-elicited peritoneal macrophages from PGRP-L-deficient mice. (A) 2 × 10⁵ peritoneal macrophages from wild-type (open bars) or PGRP-L-deficient (striped bars) animals were stimulated in the presence of 50 U of IFN-γ/ml with the various TLR-2 and TLR-4 ligands shown. TNF-α in supernatants was determined by ELISA after 24 h. (B) IL-6 levels in supernatants from peritoneal macrophages prepared as for panel A. (C) Peritoneal macrophages (2 × 10⁵) from wild-type (open bars) or PGRP-L-deficient (striped bars) animals were stimulated in the presence of 50 U of IFN-γ/ml with 10 μg of S. aureus PGN, lysozyme-digested S. aureus PGN, or B. subtilis PGN/ml. TNF-α levels in supernatants were determined by ELISA after 24 h. (D) IL-6 levels in supernatants from peritoneal macrophages prepared as for panel C.