If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
* This work was supported by grants from the Swedish Research Council (to D. H. and H. S.), the Wallenberg Consortium North (to D. H. and H. S.), and the European Union Quality of Life and Management of Living Resources Program (to H. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The peptidoglycan recognition protein PGRP-LC is a major activator of the imd/Relish pathway in the Drosophila immune response. Three transcripts are generated by alternative splicing of the complex PGRP-LC gene. The encoded transmembrane proteins share an identical intracellular part, but each has a separate extracellular PGRP-domain: x, y, or a. Here we show that two of these isoforms play unique roles in the response to different microorganisms. Using RNA interference in Drosophila mbn-2 cells, we found that PGRP-LCx is the only isoform required to mediate signals from Gram-positive bacteria and purified bacterial peptidoglycan. By contrast, the recognition of Gram-negative bacteria and bacterial lipopolysaccharide requires both PGRP-LCa and LCx. The third isoform, LCy, is expressed at lower levels and may be partially redundant. Two additional PGRP domains in the gene cluster, z and w, are both included in a single transcript of a separate gene, PGRP-LF. Suppression of this transcript does not block the response to any of the microorganisms tested.
Insects and mammals use similar mechanisms and molecular pathways to recognize and eliminate invading microorganisms. The best studied example is the humoral immune response of the fruit fly Drosophila melanogaster, where two major immune pathways signal the presence of microbes and mediate the production of antimicrobial peptides, the imd-Relish pathway and the Toll/Dif pathway (
) was screened using Hybond-NX membranes (Amersham Biosciences) and radioactive probes against the PGRP domains x, y, w, and z. We screened between 3 and 11 × 105 plaques with the different probes, corresponding to 0.6–2.2 times the estimated complexity of the library. The probes were made as described for Northern blot detection. Positive pBK-CMV phagemid clones derived from the ZAP Express vector (Stratagene) were obtained by in vivo excision using the ExAssist helper phage (Stratagene). EST clones of PGRP-LF were obtained from the Berkeley Drosophila Genome Project/Howard Hughes Medical Institute EST Project. The cDNA clones were sequenced using the Dyenamic ET terminator sequencing kit (Amersham Biosciences) and 17- to 18-base-long primers from Cybergene AB (Huddinge, Sweden).
RT-PCR—Total RNA from induced flies, induced mbn-2 cells, and untreated mbn-2 cells was used for the superscript one-step RT-PCR with platinum taq (Invitrogen) reactions with one common 5′ PGRP-LC primer and one y-specific 3′ primer. The reaction was run at 50 °C 30 min and 94 °C 2 min, followed by 30 cycles (94 °C 15 s, 50 °C 30 s, 72 °C 1 min), and ended by 72 °C 7 min. The product was cloned into the pCR2.1Topo vector using Topo TA cloning (Invitogen) and sequenced.
Cells, Microorganisms, and Cell Wall Isolates—The mbn-2 cell line (
) was grown at 25 °C in Schneider's medium with 10% fetal calf serum. The bacteria Micrococcus luteus Ml11, Bacillus megaterium Bm11, and Enterobacter cloacae β12 were originally obtained from Hans G. Boman (see Ref.
and references therein). The Erwinia carotovora carotovora SSC3193 wild type strain was obtained from Kenneth Söderhäll (Uppsala University, Uppsala, Sweden), and Escherichia coli O55:B5 was received from the CCUG (Culture Collection, University of Göteborg, Göteborg, Sweden). The fungus Dipodascopsis uninucleata var. wickerhamii was obtained from the Fungal Biodiversity Center (Utrecht, The Netherlands).
Insoluble peptidoglycan was prepared from B. megaterium Bm11, as described in Ref.
. LPS from E. coli O55:B5 was obtained from Fluka and taken up in distilled water. Before use, the stock solution was pretreated for 1 h at 62 °C. Peptidoglycan contaminants were removed by incubating LPS preparations (2 mg/ml) with PGRP-SC1B (20 μg/ml) at 25 °C for 12 h. The activity of the enzyme preparation was confirmed on purified peptidoglycan as described in Ref.
. In a separate experiment it was shown that LPS (2 mg/ml) did not inhibit the PGRP-SC1B degradation of peptidoglycan (data not shown). To make a fungal cell fragment preparation, 5 ml of D. uninucleata cell pellet was washed and resuspended in 50 ml of sterile Ringer's solution. The cells were fragmented in an ice-chilled Bead Beater chamber in 45 1-min runs, with 50 ml of 0.5-mm glass beads in a total volume of 100 ml.
RNAi—We amplified the PGRP regions of interest from cDNA or genomic DNA by PCR with primers containing the T7 sequence. The fragments were subcloned into pCR2.1Topo by using the Topo TA Cloning kit (Invitrogen). As templates for in vitro transcription (Ribomax large scale RNA production system T7 kit, Invitrogen), we used 10 μg of linear plasmid DNA containing the PGRP sequences, flanked by a T7-promoter on each side. The dsRNA for the different genes and domains corresponded to the following parts and lengths: PGRP-LC, 861 bp from the common exons 2 and 3; LCa domain, 523 bp; LCx domain, 354 bp; LCy domain, 584 bp; and PGRP-LF, 738 bp covering both domains. For RNAi, we plated 5 ml of mbn-2 cell suspension, 1 million cells/ml, in Schneider's medium with 10% fetal calf serum. The cells were kept at 25 °C and transfected 1 day later with 10 μg of dsRNA. Three days after transfection, the mbn-2 cells were induced with washed live bacteria, homogenized fungal cells, insoluble peptidoglycan, LPS, or sterile Ringer's solution as control for 6 h (2 h for LPS). The pellets of bacterial overnight cultures were washed and resuspended 1:100 in sterile Ringer's solution, and 15 μl were used per plate. The quantity of fungal cell fragments used for induction corresponds to 1.5 μl of cell pellet per plate. The final concentration of peptidoglycan and LPS was 1 μg/ml cell culture. The mbn-2 cells were harvested on ice after 6 h, centrifuged for 10 min at 500 × g, and total RNA was extracted.
RNA Preparation, Northern Blot, and Hybridization—Total RNA was prepared with TRIzol (Invitrogen) and dissolved in 50 μl RNase free water. For Northern blots, 5 μl (∼15 μg) of total RNA per lane was run on a 1% agarose gel containing formaldehyde. Hybridization was performed under high stringency conditions (50% formamide, 42 °C). Inserts from the cDNA clones pAttA for Attacin A, k-7 for Cecropin A, and BSJM108 for Diptericin (
) were cut out with EcoRI and used as templates for probes. The probes were 32P-labeled with the Rediprime II kit from Amersham Biosciences. Radioactivity was monitored using a PhosphorImager (Amersham Biosciences).
RESULTS AND DISCUSSION
Alternative Transcripts in the Complex PGRP-LC Locus— The PGRP-LC locus contains five PGRP homology domains (
). To identify transcripts that express the neighboring y, z and w PGRP domains, we screened a hemocyte-enriched cDNA library. We found 8 clones out of 450,000 screened plaques for PGRP-LCx, but only one for LCy out of 1.1 million. This single y clone corresponds to an aberrantly terminated PGRP-LCx transcript, in which the y domain was part of the 3′-untranslated region. We therefore resorted to RT-PCR and could amplify a functional PGRP-LCy-specific band with RNA from mbn-2 cells and immunostimulated adult flies. This PGRP-LCy transcript is completely analogous to PGRP-LCx and uses the same splice donor site in the third exon, 57 bp upstream of the one used for the PGRP-LCa transcript. As a result, the linker between the PGRP domain and the predicted transmembrane region is 19 amino acid residues shorter in PGRP-LCx and LCy than in LCa.
Eighteen out of 300,000 clones were positive for the w domain. At least 12, possibly all, of the w-containing clones also hybridized to the z probe, suggesting that both domains are included in a single transcript. This was confirmed by sequencing two of the cDNA clones and is also consistent with one 5′ EST sequence and one fully sequenced transcript from the genome project. It does not overlap with the PGRP-LC transcripts and therefore defines a unique PGRP gene, PGRP-LF. The encoded 369-amino acid PGRP-LF protein includes the z domain in exons 2 and 3 and the w domain in exon 4 (Fig. 1A). Membrane topology prediction suggests that the extracellular PGRP domains are immediately preceded by a transmembrane region and a very short intracellular domain of 23 residues (Fig. 1B).
Alternative Splice Forms of PGRP-LC with Different Specificity—The ability of the PGRP-LC gene to express three alternative PGRP domains, x, y, and a, suggests that the PGRP-LC isoforms display different recognition capabilities to various microbial patterns. To investigate the functions of the three PGRP-LC isoforms and of PGRP-LF, we used transcript-specific RNAi in mbn-2 cells. Control experiments show that double-stranded RNA directed against the unique exons suppress the corresponding transcript levels by at least 90%, without affecting the other alternative transcripts (Fig. 2, C–F). We then tested the cells for their ability to respond to the Gram-positive bacteria M. luteus and B. megaterium, the Gram-negative bacteria E. cloacae, E. carotovora, and E. coli, and one fungus, D. uninucleata. We monitored the immune response by Northern blot, using probes for the antimicrobial Diptericin, Attacin A, and Cecropin A genes, which are regulated by the imd/Relish pathway in this cell line. Our results support the suggestion that the PGRP-LC isoforms display distinct recognition abilities. PGRP-LCx is absolutely required for induction by all tested bacteria, and suppression of this transcript has the same effect as suppression of all PGRP-LC transcripts with dsRNA from the common exons 2 and 3 (Fig. 2, A and B). By contrast, the removal of PGRP-LCa specifically blocks the inducibility of the imd/Relish pathway by Gram-negative bacteria, without affecting the induction by Gram-positive bacteria. This clearly demonstrates a functional difference between PGRP-LCa and LCx.
A Specific Role of PGRP-LCa in the Response to LPS—The specific requirement of PGRP-LCa for induction by Gram-negative bacteria strongly suggested that this isoform could be involved in the response to LPS. Since LPS preparations are likely to be contaminated with peptidoglycan, we pretreated E. coli LPS with purified PGRP-SC1B, an amidase that degrades peptidoglycan and removes its immunostimulatory effect on Drosophila cells (
). Fig. 3 shows that a peptidoglycan-free LPS preparation induces the mbn-2 cells in a PGRP-LCa-dependent way, while crude LPS signals via PGRP-LCx in a PGRP-LCa-independent manner. This changed pattern confirms the effectiveness of the peptidoglycan removal. The chemical structure of peptidoglycan is identical in E. coli and B. megaterium (
), and induction by peptidoglycan from the latter species does not require PGRP-LCa (Fig. 3). The PGRP-LCa-dependent induction by LPS can therefore not be due to contaminating peptidoglycan, as suggested by Leulier et al. (
). It must be caused by some other factor in the preparation, most likely LPS itself. Thus, we conclude that whereas PGRP-LCx is the only isoform needed for induction by peptidoglycan, the response to LPS or Gram-negative bacteria requires both PGRP-LCa and LCx. This suggests that LPS is indeed the major inducing factor in the Gram-negative bacterial envelope.
Quantification of the Northern blot shows that stimulation by peptidoglycan-free LPS is about 7-fold weaker than that of living Gram-negative bacteria. This might be related to how the LPS molecules are presented in the micelles of the purified sample. The colloidal state of LPS is probably important and we obtain more reproducible induction when the stock solution has been heat-treated.
Unlike PGRP-LCa and PGRP-LCx, dsRNA against PGRP-LCy and PGRP-LF did not suppress the immune response (Fig. 2B). If anything, the induction is even slightly increased by dsRNA against PGRP-LF, although this could be an indirect effect. Trying to find functions for PGRP-LCy and PGRP-LF, we expanded our experiments by using the yeast-like fungus D. uninucleata, which is highly immunostimulatory in adult flies, in a Relish-dependent manner (
). However, neither intact cells (not shown) nor cell fragments of this microorganism induce any response in the mbn-2 cells (Fig. 2B), most likely due to missing factors in the cell culture system.
In conclusion, our experiments demonstrate that alternative splicing of PGRP-LC contributes to the capacity of the system to respond to both LPS and peptidoglycan. It is likely that PGRP-LCx acts as a direct pattern recognition molecule for peptidoglycans,
but we cannot exclude the possibility that other components are also involved. The simultaneous requirement of two splice forms for the response to LPS suggests that the PGRPs may act as heterodimers or perhaps as higher multimers. This would further add to the flexibility of the system.
We thank Sophia Ekengren for advice on LPS pretreatment and Svenja Stöven, Tobias Hainzl, Michael Williams, Karin Johansson, István Andó, and Magda-Lena Wiklund for helpful discussions.