View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

The LDL -Related LRP6 Mediates Internalization and Lethality of Anthrax Toxin

Wensheng Wei,1 Quan Lu,1 G. Jilani Chaudry,3,4 Stephen H. Leppla,3 and Stanley N. Cohen1,2,* 1 Department of Genetics, Stanford University School of Medicine, Stanford, CA 94305, USA 2 Department of Medicine, Stanford University School of Medicine, Stanford, CA 94305, USA 3 Microbial Pathogenesis Section, Division of Intramural Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892, USA 4 Present address: Department of Biology, University of Texas at San Antonio, San Antonio, TX 78249, USA *Contact: [email protected] DOI 10.1016/j.cell.2005.12.045

SUMMARY During infection by B. anthracis, PA binds to receptors on the cell surface and is cleaved by the protease furin, Toxins produced by Bacillus anthracis and releasing a 20 kDa fragment and enabling the residual other microbial pathogens require functions of receptor-bound C-terminal 63 kDa PA peptide fragment host cell to yield toxic effects. Here we (PA63) to form heptamers which can interact with LF or show that low density receptor- EF. After entering target cells, PA/LF and PA/EF com- related protein 6 (LRP6), previously known to plexes translocate across the endosomal membrane into be a coreceptor for the , the cytosol where they dissociate, allowing LF and EF to exert their toxic effects (Leppla, 2000). LF is a Zn2+ prote- is required for anthrax toxin lethality in mamma- ase that cleaves the amino terminus of nearly all MAPK lian cells. Downregulation of LRP6 or coexpres- kinases (Duesbery et al., 1998; Vitale et al., 1998). EF is sion of a truncated LRP6 dominant-negative a calmodulin-dependent adenylate cyclase that elevates peptide inhibited cellular uptake of complexes intracellular levels of cAMP, producing profound clinical containing the protective antigen (PA) carrier of edema and impairing the immune response to infection anthrax toxin moieties and protected targeted (O’Brien et al., 1985; Hoover et al., 1994). cells from death, as did antibodies against epi- Two distinct cell surface act as receptors for topes in the LRP6 extracellular domain. Fluores- PA: ATR (anthrax toxin receptor) (Bradley et al., 2001; cence microscopy and biochemical analyses Liu and Leppla, 2003), which is encoded by alternatively showed that LRP6 enables toxin internalization spliced transcripts of the tumor endothelial marker 8 by interacting at the cell surface with PA re- (TEM8) (Carson-Walter et al., 2001), and CMG2, which has 60% identity with ATR and is encoded by cap- ceptors TEM8/ATR and/or CMG2 to form a multi- illary morphogenesis gene 2 (Scobie et al., 2003). Low component complex that enters cells upon PA density lipoprotein receptor-related protein 6 (LRP6) is binding. Our results, which reveal a previously a member of a gene family encoding low-density lipopro- unsuspected biological role for LRP6, identify tein receptors (LDLRs) that reside on the cell surface and LRP6 as a potential target for countermeasures participate in endocytosis of their ligands (Stefansson against anthrax toxin lethality. et al., 1995; Howell and Herz, 2001; Yu et al., 2001). As a key component of a complex that mediates the trans- duction of signals from secreted Wnt proteins to b , INTRODUCTION LRP6 functions in the regulation of embryogenesis and cell proliferation (e.g., Pinson et al., 2000; Tamai et al., Anthrax, a lethal disease of humans and other animals 2000; Semenov et al., 2001). During experiments that caused by the gram-positive spore-forming eubacterium, used expressed sequence tags (ESTs) to inactivate chro- Bacillus anthracis (Leppla, 2000; Bhatnagar and Batra, mosomal genes in human cell populations (Lu et al., 2004), 2001; Collier and Young, 2003), has been the focus of together with a phenotypic screen for cellular genes re- both biological interest and bioterrorism concerns. The quired for anthrax toxicity, we discovered that LRP6 inter- virulence of B. anthracis infection is produced by separate acts with the anthrax toxin receptors TEM8/ATR and complexes formed by interaction of a carrier protein, pro- CGM2 to promote entry into the cytoplasm of PA/receptor tective antigen (PA, 83 kDa) with lethal factor (LF, 90 kDa) complexes and consequently that LRP6 deficiency miti- or edema factor (EF, 89 kDa) (Singh et al., 1989) moieties. gates the effects of anthrax toxin. Our studies identify

Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. 1141 Figure 1. Evidence that LRP6 Deficiency Mitigates Toxicity to PA + FP59 (A) Cytotoxicity of the naı¨ve M2182tTA and clone ATR43 to PA/ FP59. The MTT assay was performed as described in Experimental Procedures.The values shown represent the mean and standard deviation, as defined by error bars in this and other figures. (B) Effect of tetracycline on resistance of ATR43 to PA/FP59. Cells were seeded at a concentration of 5 103 cells/well (96-well plate) with or without tetracycline (Tet, 1 mg/ml) and treated with PA (30 ng/ml or 3.7 1010 M) plus FP59 (50 ng/ml) for 2 days before being assayed by MTT. The values shown represent the mean and standard deviation (bars).

1142 Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. LRP6 as a potential target for the prophylaxis and treat- that corresponds to a 228 nt image clone (image clone ment of anthrax cytotoxicity. 285207, accession number N66273) (Hillier et al., 1996) annotated in GenBank as an intron located between exons RESULTS 21 and 22 of the low-density lipoprotein receptor-related protein 6 (lrp6) gene. This cDNA had been cloned in pLEST Isolation and Characterization of Cells 30 to the Tet-regulated lentiviral promoter and in antisense Showing Decreased Toxin Susceptibility orientation relative to the promoter. We designed a phenotypic screen aimed at discovering cellular genes that specifically affect the actions of PA in Decreased Expression of LRP6 Results the pathway to anthrax toxicity. In these experiments, in Toxin Resistance we used a transactivated tetracycline (Tet)-controlled pro- Western blot analysis of ATR43 cell extracts (Figure 1C) in- moter (Tet-off system) (Gossen and Bujard, 1992; Gossen dicated decreased intensity of bands migrating at the pre- et al., 1995) on the lentiviral vector pLEST to activate anti- viously identified positions of LRP6 monomer and dimer sense transcription 40,000 human ESTs (Lu et al., 2004) (Liu et al., 2003a), as detected with murine monoclonal an- introduced into a population of M2182 human prostate tibody directed against an epitope present in both LRP6 carcinoma cells. We then isolated clones resistant to kill- and the related protein, LRP5. We confirmed the identity ing by a mixture of PA and FP59, a fusion of the PA binding of these bands as LRP6 monomers and dimers by electro- domain of LF to the ADP ribosylation domain of Pseudo- phoresis of lysates containing an N-terminally-tagged monas aeruginosa exotoxin A. As PA enables the entry Flag-LRP6 protein in SDS-PAGE in the presence of the re- of FP59 into cells and exotoxin-A mediated lethality (Arora ducing agent b mercaptoethanol (b ME), followed by im- and Leppla, 1992), we reasoned that cells made defective munoblotting with anti-Flag antibody (Figure 1D); the two in PA internalization by antisense expression of ESTs highly reproducible faint bands migrating more slowly would show decreased killing by PA + FP59. than the Flag-tagged LRP6 dimer (Figure 1C) may repre- Screening of a cell library having an estimated complex- sent multimeric forms of LRP6. The biological effects of re- ity of 2 105 independent insertions of the pLEST con- duction of LRP6 protein in ATR43 were evident also from struct in of M2182 cells yielded 73 colonies the observed impairment of the cellular response of this resistant to PA + FP59, nine of which showed downregu- clone to Wnt, as indicated by sharply decreased activity lation of resistance by tetracycline. ATR43, a clone that of a b catenin-regulated promoter on the Super8XTOP- survived a PA concentration >25-fold the concentration flash reporter plasmid (Veeman et al., 2003), which was that was lethal in the parental cell line when combined introduced into this cell clone (Figure 1E). with 50 ng/ml of FP59 (Figure 1A), was selected for further While the above experiments argue strongly for the role study. Growth of ATR43 cells for 48 hr in the presence of of LRP6 in the phenotype observed for ATR43 cells, as- Tet (1 mg/ml) decreased their ability to survive exposure signment of the EST sequence that mediates tetracy- to the toxin while having little effect on growth of the pa- cline-dependent regulation of toxin resistance in ATR43 rental line (Figure 1B). cells to a region of the lrp6 gene that has been annotated Analysis of the PCR-amplified EST present in ATR43 in GenBank as an intron prompted additional studies to fur- revealed a single chromosomally integrated sequence ther establish the causal role of LRP6 deficiency in toxin

(C) Immunoblot analysis of lysates from naı¨ve M2182tTA and mutant clone ATR43 cells using murine monoclonal antibody directed against human LRP5/6. a tubulin bands were measured in Western blots as internal controls. In SDS-PAGE gels, LRP6 appeared in both monomer (*) and dimer (**) forms in the presence of the reducing agent, b mercaptoethanol (b ME). (D) Confirmation of the molecular nature of both monomer (*) and dimer (**) bands of LRP6 proteins. N-terminal-tagged Flag-LRP6 was transfected into 293T cell, and the expressing lysates from transiently transfected cells electrophoresed under SDS-PAGE conditions in the presence of b ME were probed in immunoblots using anti-Flag monoclonal antibody. (E) Effect of LRP6 deficiency on cellular response to Wnt activation. The data shown are the summed results of three independent experiments. The values shown represent the mean and standard deviation (bars). siRNA-sc refers to a scrambled sequence siRNA control. (F) Effect of LRP6 deficiency induced in naı¨ve cells. The EST identified in the ATR43 mutant was reintroduced into pLEST in the antisense orientation. Pooled populations of infected M2182tTA clones in 6-well plates were treated with PA/FP59 and photographed by microscopy at a magnification of 100 . M2182/pLEST-LIB refers to EST library in M2182 cells serving as a control. (G) Effect of lrp6-specific siRNA on transcript abundance. lrp6 transcripts were measured by quantitative RT-PCR and normalized to b actin mRNA. Filled-in bars indicate transcript abundance compared with wild-type M2182. Each data point represents an average of three repeats. The values shown represent the mean and standard deviation (bars). (H and I) Immunoblot assays of lysates of M2182 cells expressing siRNAs directed against lrp6 transcripts. Antibodies used for detection were rabbit polyclonal antibody H300 (H) against human LRP6 protein and murine monoclonal antibody against human LRP5/6 (I). Anti-a-tubulin antibody was used to calibrate loading of lysates on gels. Both monomer (*) and dimer (**) forms of LRP6 protein bands are shown. Antibody specificity was con- firmed by blotting the same membrane using antibody that had been preabsorbed with an excess of the peptide used as the immunogen (I, middle). (J and K) Cytotoxicity of PA + FP59 in M2182 cells treated with siRNAs (J) and in M2182/siRNA3 cells expressing LRP6 protein (K). Cell viability was assessed as described in Figure 1A. Two individual M2182/siRNA3 clones expressing the mutant LRP6m3 construct were subjected to the PA + FP59 toxin assay (K). A clone containing the expression vector alone (pcDNA3.1) was included as a control for effects of the extended procedure per se on siRNA-modulated interference with toxicity.

Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. 1143 Figure 2. Effect of LRP6 on PA Binding and Internalization (A) Effect of LRP6 deficiency on mouse macrophage sensitivity to PA + LF. pLENTI-SUPER constructs expressing siRNAs directed against murine lrp6 transcripts were introduced into RAW264.7 cells by transfection, and the cells acquiring the vector were selected by blasticidin and pooled for cytotoxicity assays. Controls were RAW264.7 cells expressing human lrp6 siRNA1, which shares no homology with the mouse lrp6 sequence, and the naı¨ve RAW264.7 cells not subjected to transfection. Cells at 1 103 cells per well in 96-well plates were treated with serially diluted PA and 500 ng/ml of LF for 2 days. Cell viability was assessed using MTT assay as described in Figure 1A. This experiment was repeated five times with separate trans- fection assays, and similar quantitative differences in experimental and control samples were observed. (B and C) Western blot analysis of bound (B) and internalized (C) PA in different cells was conducted following the protocol as described in Exper- imental Procedures. PA83 is the full-length form of PA (83 kDa), and the oligomer represents the heptamer of PA63. a tubulin abundance was mea- sured as an internal control.

1144 Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. resistance. First, a plasmid (pATR43) containing cDNA of LF) to kill macrophages, which are a natural target for LF. image clone 285207 that was newly inserted into the pLEST Three siRNAs corresponding to sequences of the murine vector in antisense orientation with respect to the Tet-reg- lrp6 gene were inserted in the pLENTI-SUPER vector ulated promoter was introduced into naı¨ve M2182tTA cells. and expressed stably in the murine macrophage cell line As was observed for ATR43, these cells showed decreased RAW264.7. Controls were nontransfected RAW264.7 cells sensitivity to PA + FP59 (Figure 1F), confirming the role of and cells expressing siRNA1 directed against a human the EST sequence in toxin resistance. The effects of inacti- lrp6 sequence absent in the mouse gene. As shown in vating lrp6 on the sensitivity of M2182 cells to PA + FP59 Figure 2A, MTT assay indicated increased survival in were further confirmed in cell clones that stably produced each of the three mouse-sequence-specific lrp6 siRNA 21–23 nt siRNAs complementary to three different exonic pools of clones exposed to PA + LF. sequences of lrp6 mRNA. Downregulation of lrp6 tran- Western blot analysis of lysates from M2182 cells made scripts by two of these siRNAs, siRNA1 and siRNA3, was resistant to toxin by either siRNA or antisense RNA against observed by quantitative RT-PCR (Figure 1G) along with lrp6 transcripts showed a decrease in 83 kDa PA mono- (1) reduction of the steady-state level of LRP6 protein as in- mer bound to the cell surface (Figure 2B) and reduction dicated by the summed densities of LRP6 monomer and of an SDS-resistant oligomeric form of PA that character- multimer bands (Figures 1H and 1I); (2) an impaired cellular istically survives proteolytic treatment when cell cultures response to Wnt activation (Figure 1E); and (3) reduced exposed to toxin complexes at 4ºC are shifted to 37ºC and toxin susceptibility (Figure 1J). Preabsorption of the pep- the toxin complexes are internalized (Liu et al., 2003a) tide employed as immunogen resulted in disappearance (Figure 2C). Additionally, deconvolution fluorescence mi- of the Western blot signals (Figure 1I, middle panel), al- croscopy using Alexa 488-labeled PA indicated that the though the ratio of LRP6 monomer and dimer varied in dif- abundance of PA both at the cell surface (Figure 2D) and ferent experiments and for different LRP6 antibodies. in the cytoplasm (Figure 2E) was reduced in all LRP6- That the observed siRNA effects on toxin-induced le- deficient clones examined, consistent with our biochemi- thality resulted from inhibition of LRP6 function was further cal evidence of decreased PA binding and internalization. confirmed by concurrent overexpression of full-length lrp6 Internalized PA colocalized with the early endosomal cDNA in two independent toxin-resistant clones (Figure 1K; marker EEA1 (data not shown), in agreement with evi- M2182/siRNA3 carrying pLRP6m3-4 or pLRP6m3-8 ver- dence that surface bound PA complexes are translocated sus M2182/siRNA3 carrying pcDNA3.1 vector alone). In into early endosomes after internalization (Leppla, 1995). these experiments, we created silent mutations at seven sites in the 21-nucleotide sequence targeted by the siRNA The Intracellular Domain of LRP6 Is Essential to prevent siRNA3 inhibition of the overexpressed lrp6 for PA-Mediated Internalization and Toxin Lethality cDNA. Using a fixed concentration of FP59 (50 ng/ml), The data presented above suggest that deficiency of the calculated values of EC50 (effective concentration of LRP6 decreases both binding and internalization of PA. PA required to kill 50% of cells tested) were increased 4.3- An LRP6 mutant protein (i.e., LRP6DC) that lacks most to 5.1-fold (calculated values for M2182, M2182/siRNA3_ of the cytoplasmic domain (Figure 2F) and has been re- pLRP6m3-4, M2182/siRNA3_pLRP6m3-8, and M2182/ ported to have dominant-negative effects on Wnt signal- 11 10 siRNA3_pcDNA3.1 are 3 10 M, 5.4 10 M, 4.5 ing mediated by the coexisting wild-type LRP6 protein 10 9 10 M, and 2.3 10 M, respectively). (Tamai et al., 2000) and an additional truncated LRP6 mu- tant protein that we expressed from pcDNA3.1 were LRP6 Deficiency Results in Impaired Binding tested for their ability to block PA-mediated cytotoxicity and Internalization of PA of FP59 in M2182 cells. Cell populations expressing each The known cell-surface location and previously elucidated LRP6DC mutant protein were treated with PA (50 ng/ml) + biological roles of LRP6 suggested that LRP6 deficiency FP59 (50 ng/ml) for 2 days, followed by extended growth might alter toxin lethality by impairing PA functions. To on toxin-free medium. Whereas no survivors were ob- test this hypothesis, we examined the effects of lrp6-spe- served among naı¨ve M2182 cells or cells transfected with cific siRNAs on the ability of native anthrax toxin (i.e., PA + pcDNA3.1, colonies were seen within several days in cells

(D) For PA binding assay, cells were incubated with Fluor 488-labeled PA for 1 hr at 4ºC before being examined by deconvolution microscopy following the protocol described in Experimental Procedures. Scale bar = 2.8 m. (E) For PA internalization assays, cells were incubated with Fluor 488-labeled PA (1 mg/ml) for 45 min at 37ºC followed by deconvolution microscopy. Scale bar = 2.8 m. (F) Schematic structure of dominant-negative mutant LRP6DC and wild-type LRP6 proteins. LRP6DC was generated by deleting residues from 1440 onward (Tamai et al., 2000), as described in Experimental Procedures. (G) Effect of LRP6DC production on cellular response to Wnt activation. Luciferase reporter gene assay was carried out as in Figure 1E. The data shown are the averaged results of three independent experiments. The values shown represent the mean and standard deviation (bars). (H) Cytotoxicity assay of M2182 expressing LRP6DC, pLRP6DC-1, and pLRP6DC-8. Cell viability was assessed as described in Figure 1A. (I and J) Western blot analysis of bound (I) and internalized (J) PA in LRP6DC mutants. PA binding and internalization assays were carried out as de- scribed in Experimental Procedures using anti-PA monoclonal antibodies. a tubulin was used as an internal control.

Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. 1145 Figure 3. Specificity of LRP6 Effect on Cell Susceptibility to Anthrax Toxin (A) Effect of Axin production on cellular response to Wnt activation. Cells in 6-well plates (40%–70% confluence) were transiently transfected with the following three DNAs: pcDNA3.1-AxinMyc (2 mg/well) or pcDNA3.1 (2 mg/well); Super8XTOPflash (1 mg/well); and pcDNA-lacZ plasmid (0.1 mg/well). Luciferase reporter gene assay was carried out as in Figure 1E. The data shown are the averaged re- sults of three independent experiments. The values shown represent the mean and stan- dard deviation (bars). (B) Cytotoxicity assay of M2182-expressing Axin. Cells transfected with the Axin expression construct as in Figure 3A were subjected to the PA/FP59 toxin assay. The values shown represent the mean and standard deviation (bars). (C and D) Assay of ligand-induced EGF recep- tor (EGFR) degradation. (C) EGFR degradation in M2182 cells and its LRP6-deficient mutants, ATR43, and M2182/siRNA3. (D) Effects of overexpression of VPS4, VPS4 (E228Q), and TSG101 (M95A) on EGF-induced EGFR degra- dation. M2182 cells were transfected in 6-well plates with pcDNA-EGFR (200 ng/well) along with one of the following: VPS4, VPS4 (E228Q), or TSG101 (M95A) (1 mg/well for each of the three). Assays were performed as described in Experimental Procedures. (E) Cytotoxicity assay of M2182 expressing VPS4, VPS4 (E228Q), or TSG101 (M95A). Cells transfected with different expression con- structs, as in Figure 3D, were subjected to the PA/FP59 toxin assay. The values shown represent the mean and standard deviation (bars).

expressing the pLRP6DC proteins (data not shown). Ran- Disruption of Wnt Signaling or of Generalized domly picked clones from transfected cells containing Mechanisms of Endocytic Trafficking the LRP6DC expression constructs were expanded to Does Not Impair Toxin Lethality generate the M2182/pLRP6DC-1 and M2182/pLRP6DC- Is disruption of Wnt signaling a mechanism underlying the 8 cell lines; assay of b catenin in these lines showed resistance to PA-mediated toxin lethality we observed in sharply reduced Wnt signaling (Figure 2G), confirming cells defective in LRP6 function? To address this question, the dominant-negative effects of the LRP6DC proteins. we examined the effects of Axin, which regulates Wnt Both clonal lines showed a dramatic and highly reproduc- signaling downstream of entry of the Wnt/receptor com- ible decrease in killing by PA + FP59 (Figure 2H). More- plex into cells (Zeng et al., 1997; Itoh et al., 1998; Willert over, whereas binding of the 83 kDa PA monomer to the et al., 1999), on the resistance of M2182 cells to PA + cell surface was not detectably affected by overexpres- FP59. Whereas adventitious expression of Axin prevented sion of LRP6DC mutant proteins (Figure 2I), PA internaliza- induction of b catenin following addition of Wnt protein to tion, as monitored by the presence of PA oligomer after cell cultures (Figure 3A), susceptibility to PA/FP59 toxin shift of the cell culture to 37ºC, was impaired (Figure 2J). was unaltered by Axin expression (Figure 3B), indicating These results argue strongly that LRP6 plays an essential that interference with Wnt signaling per se does not result role in accomplishing the internalization of PA and con- in toxin resistance. sequently in conferring cellular susceptibility to anthrax Whether mitigation of anthrax toxin lethality by LRP6 de- toxin. Moreover, the observed dominant-negative effects ficiency is associated with more generalized effects on en- of the LRP6DC on PA-mediated cell death imply that the docytic trafficking was addressed firstly by determining if intracellular domain of LRP6, unlike the intracellular do- LRP6-deficient cells show normal internalization and en- main of the PA receptor, ATR/TEM8 (Liu and Leppla, docytic trafficking of a prototypical cell-surface receptor, 2003), is essential for toxin lethality. the receptor (EGFR), and secondly

1146 Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. Figure 4. Binding of LRP6 Extracellular Domain to TEM8 and CMG2 (A) 293T cells were transfected with LRP6N-Myc. Cell medium containing LRP6N-Myc protein was incubated with PA83 on ice for 30 min before being mixed with agarose-conjugated anti-c-Myc antibody or G protein agarose conjugated with PA rabbit polyclonal antibody. After incubation for an ad- ditional 3 hr on a rotator in a cold room, the agarose beads were washed and subjected to immunoblot analysis. (B) 293T cells were transfected with Flag-LRP6, TEM8-HA, or both. Cell lysates were treated by addition of anti-HA-conjugated agarose beads or anti- Flag-conjugated affinity gel. The precipitated proteins were then subjected to immunoblot analysis probed with anti-Flag antibody conjugated with horseradish peroxidase (HRP) (top) or anti-HA-HRP (bottom). (C) 293T cells were transfected with LRP6N-Myc, TEM8-HA, or both. Cell lysates plus medium were treated by addition of anti-HA-conjugated or anti- c-Myc conjugated agarose beads. The precipitated proteins were then subjected to immunoblot analysis probed with anti-c-Myc-HRP. (D) 293T cells were transfected with LRP6N-Myc, CMG2-EGFP, or both. Cell lysates plus medium were treated with either anti-GFP-conjugated aga- rose beads or anti-c-Myc-conjugated agarose beads. The precipitated proteins were then subjected to immunoblot analysis probed with anti-c-Myc- HRP (top) or anti-GFP-HRP (bottom). (E) 293T cells were transfected with LRP6N-Myc, Myc-LRP6DN, TEM8-HA, or a combination of two of them. Cell lysates combined with medium were treated with anti-HA-conjugated agarose beads before being subjected to immunoblot analysis probed with anti-c-Myc-HRP. (F) 293T cells were transfected with LRP6N-Myc, TEM8-HA, or both. Cell lysates combined with medium were treated with anti-HA-conjugated agarose beads in the presence or absence of PA (0.5 mg/ml) before being subjected to immunoblot analysis probed with anti-c-Myc-HRP. Chemi- luminescence of Western blot bands was quantitated by using the VersaDoc Imaging System (Model 1000, Bio-Rad) and its software (Quantity One, Version 4.5.1). by investigating whether cells having broad defects in en- but unchanged lethality to PA + FP59 (Figure 3E). Thus, docytic trafficking exhibit resistance to toxin lethality. We LRP6 deficiency does not result in a generalized defect in observed that epidermal growth factor-induced internali- endocytic trafficking, and defective endocytic trafficking zation and degradation of EGFR was similar in parental per se is not sufficient to mitigate toxin effects. M2182 cells, in clone ATR43, and in cells made toxin resis- tant by siRNA directed against LRP6 (Figure 3C). Con- LRP6 Protein Interaction with PA Receptors TEM8 versely, cells that overexpress the mutant TSG101 or and CMG2 Is Required for Toxin Internalization VPS4 proteins and consequently are broadly defective in Multiple coimmunoprecipitation (co-IP) experiments endocytic trafficking (Bishop and Woodman, 2001; Strack showed interaction between LRP6 and both of the previ- et al., 2003; Lu et al., 2003) showed the expected abnor- ously identified cellular receptors for PA (i.e., TEM8 mality of EGFR internalization/degradation (Figure 3D) and CMG2) but no direct association of LRP6 with PA

Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. 1147 Figure 5. LRP6 Effects on Internalization and Localization of PA and Its Receptors (A and B) Deconvolution microscopy images (three independent image fields for each) showing the colocalization of Flag-LRP6 (red) and PA oligo- mers (green). M2182 cells were first transfected with Flag-LRP6 two days prior to fluorescence microscopy. Cells were incubated without PA (A) or with Fluor-488-conjugated PA (1 mg/ml) at 37ºC for 45 min (B). Flag-LRP6 protein labeled by Fluor-594 (red) is shown in the left panels (A and B). The PA oligomer signal (green) is shown in the middle panels. The right panels show the merged images, including DAPI nuclear staining (blue). Scale bar (A and B) = 2.8 m. (C) Deconvolution microscopy images showing colocalization of Fluor-594-labeled TEM8-HA (red) with Fluor-488-conjugated PA (1 mg/ml) (green) after cells were incubated with PA (1 mg/ml) for 45 min at 37ºC (scale bar = 2.8 m). The top and bottom panels show two independent fields. The right panels show the merged images, including DAPI nuclear staining (blue). (D) Deconvolution microscopy images showing the colocalization of Flag-LRP6 (red) and CMG2-EGFP (enhanced green fluorescent protein) after cells were incubated with PA (1 mg/ml) for 45 min at 37ºC. Scale bar = 2.8 m. (E and F) M2182 cells transfected with TEM8-HA or CMG2-EGFP were first incubated with or without PA (1 mg/ml) at 37ºC for 45 min before decon- volution microscopy. (E) Microscopy images showing aggregates (white arrows) of internalized TEM8-HA (red) in M2182 (top panels) or M2182/siRNA1 cells (bottom panels) in the absence (left panels) or presence (right panels) of PA. Scale bar = 5.6 m.

1148 Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. (Figure 4A). In these experiments, the C-terminal Myc- negative LRP6DC prevented the occurrence of cytoplas- tagged extracellular domain of LRP6 (LRP6N-Myc) was mic aggregates of Alexa Fluor-594-stained TEM8-HA harvested from 293T cell growth medium and incubated (Figure 5E) and EGFP-labeled CMG2 (Figure 5F), which with PA or with lysates containing HA-tagged TEM8 normally were observed after incubation of cells with PA (TEM8-HA) or enhanced green fluorescent protein-tagged for 45 min at 37ºC(Figure 5E, top panels)—providing fur- (EGFP-tagged) CMG2 (CMG2-EGFP); 293T lysates ex- ther evidence that the PA-mediated internalization of pressing N-terminal Flag-tagged full-length LRP6 (Flag- TEM8 and CMG2 is dependent on LRP6. Collectively, LRP6) were also examined by co-IP assay. LRP6 was our immunoprecipitation and fluorescence microscopy detected in precipitates of TEM8-HA (Figure 4B, top), and results argue strongly that LRP6 forms a complex with TEM8 was observed in precipitates of Flag-LRP6 (Fig- PA receptors that accomplishes the translocation of PA ure 4B, bottom). Similar co-IP experiments carried out us- and the bound toxin moieties into the cytoplasm. ing TEM8-HA and LRP6N-Myc, which contains only the extracellular domain of LRP6 (i.e., LRP6N), indicated that LRP6 Antibodies Protect Cells this domain is sufficient for interaction with TEM8 (Fig- from Anthrax Cytotoxicity ure 4C). In additional experiments, we observed LRP6N We generated rabbit polyclonal antibodies against two in precipitates of CMG2-EGFP (Figure 4D, top) and found synthetic peptides that correspond to different segments CMG2 in precipitates containing LRP6N-Myc (Figure 4D, of the extracellular domain of LRP6 and tested these anti- bottom). However, we observed no interaction between bodies (N204, directed against amino acids 204–223, and either TEM8 or CMG2 and an LRP6-truncated protein N251, directed against amino acids 515–534) (Figure 6A) lacking the extracellular domain (Figure 4E). As the precip- for their ability to protect RAW264.7 macrophages from itated bands from co-IP experiments that detected the the lethal effects of anthrax toxin (PA + LF). Dosage- interaction of LRP6 and PA receptors were faint in com- dependent increased survival was observed for cells parison with bands in the positive control lanes (Figures treated with N251, whereas antibody N204 provided no 4B–4D), we suspect that efficient interaction between detectable protection from toxin killing (Figure 6B). Signif- LRP6 and PA receptors may involve accessory protein(s) icantly, protection by antibody to N251 was partially re- that were not overexpressed in these experiments and versed by the addition of peptide N251 but not by N204 which thus may be present in limiting amounts. Interest- (Figure 6C). H300, a commercially available rabbit poly- ingly, while interaction binding of the LRP6 extracellular clonal antibody generated against a LRP6 human protein domain to TEM8 did not require PA, PA expression en- fragment (amino acids 1314–1613) that includes a se- hanced this interaction (Figure 4F); we have not observed quence of the cytoplasmic domain as well as segments enhancement of the LRP6-CMG2 interaction by PA (data of the extracellular and transmembrane domains (Fig- not shown). ure 6A), also provided dosage-dependent protection As LRP6 initiates Wnt signaling when it enters cells to- against killing by PA + LF (Figure 6D). gether with other members of a Wnt-based complex (Mao et al., 2001), we hypothesized that LRP6 may func- tion similarly to promote internalization of PA and its re- DISCUSSION ceptors; deconvolution fluorescence microscopy experi- ments provided evidence of just such a role. As has Role of LRP6 in Anthrax Toxicity been observed previously (Cong et al., 2004), in the ab- Using antisense RNA generated intracellularly against a li- sence of PA, deconvolution microscopy showed that brary of ESTs to inactivate chromosomal genes globally in Alexa Fluor-594-stained (red) Flag-labeled LRP6 was populations of human cells plus an experimental strategy present mainly in the cytoplasm as well as on the cell that selected cell clones resistant to the actions of cyto- surface (Figure 5A). Addition of PA altered the location of toxins, we identified genes whose functional inactivation internalized red fluorescent LRP6, resulting in its polar ag- is causally related to toxin resistance. This approach un- gregation within the cytoplasm (Figure 5B). Furthermore, covered lrp6, a member of a gene family encoding low- green fluorescence-labeled PA colocalized with cytoplas- density lipoprotein (LDL) receptor-related proteins (LRPs) mic LRP6 (Figure 5B) and also with internalized TEM8 that reside on the cell surface and participate in the deliv- (Figure 5C). We were not able routinely to overexpress ery of a variety of ligands into the cytoplasm (Orth et al., LRP6 concurrently with either TEM8 or CMG2 in the 1992; Kounnas et al., 1995; Stefansson et al., 1995; Ho- same cells (M2182); however, in rare instances when con- ward et al., 1996; Li et al., 2001). The results reported current overexpression of both Flag-LRP6 and CMG2- here indicate that the LRP6 protein, which previously EGFP was achieved, colocalization of LRP6 and CMG2 has been known principally as a mediator of the Wnt sig- was observed in cells treated with PA (e.g., Figure 5D). naling pathway, has a hitherto unsuspected role in the de- Expression of siRNA directed against LRP6 or dominant- livery of anthrax toxins into cells.

(F) Microscopy images showing aggregates (white arrows in four images of top right panels) and localization of CMG2-EGFP (green) in M2182 (top panels) or M2182/pLRP6DC-1 (bottom panels) without (left panels) or with (right panels) PA. Four independent image fields are used for each panel. Scale bar = 11 m.

Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. 1149 Figure 6. Effect of LRP6 Antibodies on Cytotoxicity of PA Plus LF (A) Peptides used as immunogens to produce the antibodies. H300 is a rabbit polyclonal antibody raised against C-terminal of human LRP6 (1314– 1613). N251 and N204 are rabbit polyclonal antibodies raised against small peptides targeting different segments of LRP6 extracellular domain. (B and C) Effects of LRP6 N251 and N204 on killing of RAW264.7 cell by PA/LF. Cell viability and the specificity of antibody protection (C) were as- sessed as described in Experimental Procedures. The antibody concentration is the relative concentration of bleed (1 = 4.88E+7 dilution). The values shown represent the mean and standard deviation (bars). (D) Effect of LRP6 commercially available antibody H300 and rabbit polyclonal antibody raised against Cyclin E (M-20) (both from Santa Cruz Bio- technology, Inc) as a control. The methods used are the same as above (Figure 6B). The values shown represent the mean and standard deviation (bars).

The role of LRP6 in PA-mediated toxigenicity was biochemical evidence and deconvolution immunofluores- confirmed by multiple lines of evidence: (1) regulation cence microscopy results showing reduced internalization of the toxin-resistance phenotype in clone ATR43 by tetra- of PA in ATR43 cells and in cells expressing siRNAs cycline (Figure 1B); (2) reproduction of the toxin-resis- directed against lrp6 mRNA (Figure 2); and (7) our demon- tance phenotype by antisense expression of the lrp6 stration using both fluorescence microscopy and immu- EST in naı¨ve cells (Figure 1F); (3) reproduction of the toxin- noprecipitation approaches that the LRP6 protein in- resistance phenotype by siRNA corresponding to se- teracts with the PA receptors, TEM8 and CMG2, and quences in the lrp6 open reading frame (ORF) and partial controls PA-triggered internalization of the receptors (Fig- reversal of this effect by overexpression of LRP6 protein ures 4 and 5). containing silent mutations that make it nonsusceptible to inhibition by the siRNA (Figures 1J and 1K); (4) creation Biological Functions of LRP6 and Related Proteins of toxin resistance by expression of a dominant-negative LDL receptors and LDL receptor-related proteins (LRPs) truncated LRP6 mutant protein lacking the cytoplasmic are prototypes of macromolecules that interact with and domain of the protein (Figure 2H); (5) protection of cells deliver ligands into cells via endocytosis (Schneider and from anthrax toxicity by antibodies directed against Nimpf, 2003). LRP6 differs from other members of its fam- epitopes in the LRP6 extracellular domain (Figure 6); (6) ily in that it contains a unique pattern of repeats of the

1150 Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. Figure 7. Model for LRP6-Promoted Internalization of Multicomponent Complex Containing TEM8/ATR and/or CMG2 LRP6 is proposed to act as a coreceptor for PA through its binding to TEM8/ATR and/or CMG2. PA addition triggers the internalization of this com- plex, which delivers PA and associated toxin molecules into cells. epidermal growth factor receptor (EGFR) and LDL recep- development (Gong et al., 2001), and its presence does tor (LDLR) sequences in the extracellular domain (Brown not mitigate either the embryonic lethality of LRP6 null mu- et al., 1998). A notable characteristic of LRP proteins is tations (Pinson et al., 2000) or the effects of LRP6 defi- their multifunctional nature. For example, LRP1 functions ciency on anthrax toxicity (the studies reported here). as a receptor for apoE-rich remnant , for lipo- The EST library fragment (i.e., image clone 285207) protein lipase, for a 2-macroglobulin/protease com- identified in ATR43 cells corresponds to a sequence anno- plexes, for plasminogen activator/inhibitor complexes, tated as being present in an intron of human lrp6 RNA. for the active protease tissue-type plasminogen activator, While the occurrence of an apparently intronic fragment and for the surface binding domain of exotoxin A from in the widely used EST library may indicate contamination Pseudomonas aeruginosa (Willnow and Herz, 1994). Our of the library by genomic DNA, interference with LRP6 finding that LRP6 acts as a coreceptor to promote inter- function by antisense RNA directed against this image nalization of the PA component of anthrax toxin in addition clone sequence suggests that it may in fact be exonic. to its previously known role in Wnt signaling indicates that Analysis using the Human BLAT Search program (http:// LRP family proteins act more broadly in mediating the genome.ucsc.edu/cgi-bin/hgBlat) indicated that the EST pathogenic effects of microbial toxins. Possibly related that constitutes image clone 285207 corresponds also to our findings are results of DNA membrane array-based to a sequence of the noncoding strand of an intron of a experiments showing that exposure of macrophages to Bcl-2-like gene (Bcl2L14). the anthrax lethal toxin perturbs transcription of a variety of genes regulated by GSK-3b—including some impli- Mechanistic Model for LRP6 Effects cated in Wnt signaling (Tucker et al., 2003). on Toxin Lethality Interestingly, the possibility that LRP1 and LRP6 can The data reported here suggest the following model for the interact functionally is suggested by recent evidence (Zil- role of LRP6 in anthrax toxicity (Figure 7). Independently of berberg et al., 2004) that LRP1 can sequester human Fz1 the presence of PA, LRP6 protein at the cell surface is ca- to disrupt the receptor/coreceptor complex formation and pable of interacting with two proteins previously identified repress canonical Wnt signaling, raising the possibility as PA receptors: TEM8/ATR and CMG2 (Bradley et al., that LRP6 may also modulate exotoxin A internalization. 2001; Scobie et al., 2003). Our data suggest that LRP6 de- LRP5, another member of the LRP family that has 71% ficiency reduces PA binding to the cell surface (Figure 2B) amino acid sequence identity to human LRP6, has also and also indicate that PA addition enhances binding of been implicated in Wnt signaling (Tamai et al., 2000); how- LRP6 to the PA receptors. Upon addition of PA, a complex ever, LRP5 has distinct effects on bone accrual and eye containing PA, LRP6, and the PA receptors TEM8 or

Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. 1151 CMG2 is formed and then internalized together with the M2182 cell line and 3 104 /ml for RAW264.7 cell line. Various concen- associated toxin moieties, LF or EF. This model is consis- trations of PA combined with a fixed concentration of FP59 (50 ng/ml) tent with earlier evidence that neither TEM8 nor CMG2 is (for M2182 line) or LF (500 ng/ml) (for RAW264.7 line) were added to wells, and cells were incubated for 48 hr at 37ºC. Cell viability was mea- sufficient to carry toxin into the cells. The essential role sured by adding 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium of LRP6 in enabling entry of receptor bound anthrax toxin bromide (MTT) (Sigma) to a final concentration of 1 mg/ml for 1 hr at into cells parallels its previously demonstrated role in Wnt 37ºC. The supernatant was carefully removed before the addition of pathway signaling (Mao et al., 2002). In that case, the 50 ml of solubilization solution (10% SDS in 0.01 N HCl) and subse- transmembrane protein Kremen2 forms a ternary complex quently of 200 ml of PBS to each well. Spectrophotometer readings at with the Wnt antagonist DKK-1 and LRP6, triggering 570 nm were determined using a GENios Multi-Detection Reader (TECAN, Phenix Research Products). Cell viability was normalized to LRP6-mediated endocytosis of all three components wells lacking PA. Each data point shown in figures for MTT assays rep- and consequent modulation of Wnt/b-catenin signaling. resents the averaging of results from eight wells. Whether other cellular proteins are included in the PA/re- ceptor/coreceptor complex and the stoichiometry of com- Preparation of LRP6-Specific Short Interfering RNAs plex formation currently is not known. Three human-specific lrp6 sequences (tgctacgattgtagttgga, cgaatc Our observation of perinuclear localization of adventi- gaattgaggtgtc, tgtctcgaggtaaatcaat) and three murine-specific lrp6 tiously expressed Flag-tagged LRP6 in M2182 cells (Fig- sequences (gagaatgcaacgattgtag, gtgctttgtttaattgatc, atacacatctggtt gagat) were selected. The siRNA duplexes were cloned into the ure 5) is consistent with recent independent evidence pLENTI-SUPER lentiviral vector (derived from pSUPER) (Brummel- that LRP6 can exist intracellularly in the perinuclear region kamp et al., 2002). siRNA constructs were then introduced into cells as well as at the plasma membrane of 293T cells (Cong by infection. Stable clones of cells carrying siRNAs were selected us- et al., 2004). Addition of PA increased the abundance of ing blasticidin (10 mg/ml) (Sigma). The scrambled siRNA control was cytoplasmic LRP6 and altered its location, producing po- a product from Ambion (Austin, TX). lar aggregates that colocalize with PA (Figure 5B). The Assay for Luciferase Reporter Gene Activity and EGF ability of antibodies against LRP6 extracellular domains Receptor Degradation to protect cells from anthrax toxin killing (Figure 6) argues These assays were performed as reported in (Veeman et al., 2003; Lu that at least some of the LRP6 protein that colocalizes with et al., 2003) and described in detail in Supplemental Data. PA in the cytoplasm is derived from the cell surface. The ring-shaped heptamer structure generated by spontane- Biochemical Assay of PA Binding and Internalization ous oligomerization of receptor bound PA (i.e., the pre- Methods were as described by Liu et al. (2003a, 2003b), except that pore) (Milne et al., 1994) is responsible for carrying EF or the final concentration of PA and its incubation time were 0.05 mg/ml for 2 hr and 1 mg/ml for 45 min, respectively. LF into early endosomes, where PA heptamers form an aqueous transmembrane channel that mediates trans- Western Blotting and Coimmunoprecipitation location of EF and LF to the cytosol (Blaustein et al., Western blotting was performed essentially as described by Harlow 1989). Whether LRP6 and the PA receptors are involved and Lane (1999) using precast 10% or 3%–8% gels (Criterion XT Pre- mechanistically in the formation of the prepore structure cast Gel, Bio-Rad) for electrophoresis and standard conditions. The has not been determined. soluble cell proteins were quantitated using the BCA protein assay kit (Pierce Biotechnology, Inc.). For coimmunoprecipitation assays, Our discovery of the role of LRP6 in anthrax toxicity and 293T cells, 48 hr after transfection, were lysed in NP40 buffer contain- the demonstration that antibodies directed against the ex- ing a protease inhibitor cocktail (Complete Mini, Roche Diagnostics, tracellular domain of LRP6 can protect cells grown in cul- Inc.). The lysates were centrifuged at 16,000 g, and the soluble frac- ture against killing by anthrax toxin suggest that the immu- tion was immunoprecipitated with the desired antibody-conjugated nological targeting of LRP6 may prove useful in protecting agarose beads (Santa Cruz Biotechnology, Inc.). The cell medium against the effects of accumulated toxin during the late was included in LRP6N-Myc immunoprecipitation reactions because stages of anthrax disease when antibacterial methods LRP6N-Myc is secreted into the supernatant (Tamai et al., 2000) after being expressed. normally are no longer of therapeutic value. Antibody Reagents EXPERIMENTAL PROCEDURES Polyclonal antibody against PA was generated as described in Liu et al. (2003b), and monoclonal antibody against PA was ordered from Novus Cell culture, construction and screening of cDNA libraries, and con- Biologicals (NB 600-407). Rabbit polyclonal antibody against human struction and cloning of expression vectors have carried out using LRP6 (H300) was purchased from Santa Cruz Biotechnology, Inc., variations of methods reported previously (Lu et al., 2004); these are and monoclonal anti-LRP5/6 antibody that was generated against described in detail in Supplemental Data. a synthetic peptide derived from an amino acid sequence (DTGT DRIEVTR) of the second YWTD repeat of LRP6 was obtained from Toxin Treatment and Cell Viability Assay A&G Pharmaceuticals (Baltimore, MD) and BioVision (Mountain View, Proteins produced as described previously (Leppla, 1988) include wild- CA). Rabbit polyclonal antibodies N204 & N251 were generated for type PA, FP59, and LF. Cytotoxicity assays for M2182 were performed us by Proteintech Group, Inc. (Chicago, IL) using synthetic peptides by treating cells at a concentration of 5 104 cells/well (in 6-well plates) derived from amino acid sequences (NFIHKSNLDGTNRQAVVKGS with PA (30 ng/ml) plus FP59 (50 ng/ml) for 2 days. Cells were cultured in for N204 and EVMNTDGTGRRVLVEDKIPH for N251). Agarose-conju- toxin-free medium for 3 days after toxin challenge before being stained gated monoclonal antibodies of c-Myc (9E10), HA (F-7), and GFP (B-2) with crystal violet for microscope imaging and photography. For MTT and horse radish peroxidase (HRP)-conjugated monoclonal antibodies assay, cells were seeded in 96-well plates (100 ml/well) 1 day prior to against c-Myc (9E10), HA (F-7), and GFP (B-2) were all obtained from toxin treatment; the cell concentration used was 5 104 /ml for Santa Cruz Biotechnology, Inc. Anti-Flag (M2) monoclonal antibodies

1152 Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. against the HRP-conjugate and affinity gels used for co-IP experi- Blaustein, R.O., Koehler, T.M., Collier, R.J., and Finkelstein, A. (1989). ments were obtained from Sigma, Inc. Anthrax toxin: channel-forming activity of protective antigen in planar phospholipid bilayers. Proc. Natl. Acad. Sci. USA 86, 2209–2213. Immunofluorescence Microscopy Bradley, K.A., Mogridge, J., Mourez, M., Collier, R.J., and Young, J.A. PA protein was labeled with Alexa Fluor 488 using the protein labeling (2001). Identification of the cellular receptor for anthrax toxin. Nature kit (A-10235, Molecular Probes, Inc.). The potency of PA after labeling 414, 225–229. was fully retained as indicated by the MTT assay (data not shown). The Brown, S.D., Twells, R.C., Hey, P.J., Cox, R.D., Levy, E.R., Soderman, concentrations of PA-Alexa 488 used in analysis of PA binding and A.R., Metzker, M.L., Caskey, C.T., Todd, J.A., and Hess, J.F. (1998). processing were 0.6 mg/ml and 1 mg/ml, respectively. Immunostaining Isolation and characterization of LRP6, a novel member of the low den- was performed according to a standard protocol (Harlow and Lane, sity lipoprotein receptor gene family. Biochem. Biophys. Res. Com- 1999). The primary antibody anti-Flag (M2) (murine monoclonal, mun. 248, 879–888. Sigma, Inc.) was used at a dilution of 1:200. Alexa-Fluor 594 goat Brummelkamp, T.R., Bernards, R., and Agami, R. (2002). A system for anti-mouse IgG (Molecular Probes) was used as the secondary anti- stable expression of short interfering RNAs in mammalian cells. Sci- body at a dilution of 1:400. The preconjugated antibody anti-HA-Alexa ence 296, 550–553. Fluor-594 (murine monoclonal IgG1, 16B12, Molecular Probes) was also used. The cover slips were mounted onto slides with a ProLong Carson-Walter, E.B., Watkins, D.N., Nanda, A., Vogelstein, B., Kinzler, Gold antifade reagent containing DAPI solution (Molecular Probes), K.W., and St Croix, B. (2001). Cell surface tumor endothelial markers and cells were examined under a deconvolution microscope (Applied are conserved in mice and humans. Cancer Res. 61, 6649–6655. Precision, Issaquah, WA). Collier, R.J., and Young, J.A. (2003). Anthrax toxin. Annu. Rev. Cell Dev. Biol. 19, 45–70. Antibody Protection Assay Cong, F., Schweizer, L., and Varmus, H. (2004). Wnt signals across the 3 104/ml of RAW264.7 cells were first seeded in 96-well plates plasma membrane to activate the beta-catenin pathway by forming (100 ml/well) the day prior to the assay. The cells were then incubated oligomers containing its receptors, and LRP. Development at 37ºC for 48 hr with 3 ng/ml of PA, 500 ng/ml of LF, and serial dilutions 131, 5103–5115. of antibodies. To test the specificity, the antibody was preabsorbed Duesbery, N.S., Webb, C.P., Leppla, S.H., Gordon, V.M., Klimpel, with the N204 or the N251 peptide prior to assay. The antibody was K.R., Copeland, T.D., Ahn, N.G., Oskarsson, M.K., Fukasawa, K., incubated with peptide in the ratio of 2 mg peptide /1 ml final bleed Paull, K.D., and Vande Woude, G.F. (1998). Proteolytic inactivation on a rotator at 4ºC overnight, followed by a centrifugation at 14,000 of MAP-kinase-kinase by anthrax lethal factor. Science 280, 734–737. RPM for 10 min to remove the precipitate. Toxin sensitivity was deter- Gong, Y., Slee, R.B., Fukai, N., Rawadi, G., Roman-Roman, S., Regi- mined by the cell viability as described above. nato, A.M., Wang, H., Cundy, T., Glorieux, F.H., Lev, D., et al. (2001). LDL receptor-related protein 5 (LRP5) affects bone accrual and eye Supplemental Data development. Cell 107, 513–523. Supplemental Data include Experimental Procedures and References Gossen, M., and Bujard, H. (1992). Tight control of in and can be found with this article online at http://www.cell.com/cgi/ mammalian cells by tetracycline-responsive promoters. Proc. Natl. content/full/124/6/1141/DC1/. Acad. Sci. USA 89, 5547–5551. Gossen, M., Freundlieb, S., Bender, G., Muller, G., Hillen, W., and ACKNOWLEDGMENTS Bujard, H. (1995). Transcriptional activation by tetracyclines in mam- malian cells. Science 268, 1766–1769. We thank Dr. Joy Ware for providing us with the human prostate can- Harlow, E., and Lane, D. (1999). Using Antibodies: A Laboratory Man- cer cell line M2182, Dr. Xi He for providing LRP6DC and pLRP6N-Myc ual (Plainview, NY: Cold Spring Harbor Lab Press). constructs, Dr. Stuart Aaronson for pFlag-LRP6 and Myc-LRP6DN, Dr. John Young for CMG2-EGFP, Dr. Randall Moon for the Super8- Hillier, L.D., Lennon, G., Becker, M., Bonaldo, M.F., Chiapelli, B., Chis- XTOPflash reporter plasmid, Dr. Laurie E. Ailles for pcDNA3.1-Axin- soe, S., Dietrich, N., DuBuque, T., Favello, A., Gish, W., et al. (1996). Myc, and Yuchin Teng for technical support. These investigations Generation and analysis of 280,000 human expressed sequence were supported by California Breast Cancer Research Program grants tags. Genome Res. 6, 807–828. to W.W. and Q.L. and by a grant from the Defense Advanced Research Hoover, D.L., Friedlander, A.M., Rogers, L.C., Yoon, I.K., Warren, R.L., Projects Agency (DARPA) to S.N.C. and Cross, A.S. (1994). Anthrax edema toxin differentially regulates lipopolysaccharide-induced monocyte production of tumor necrosis Received: June 7, 2005 factor alpha and interleukin-6 by increasing intracellular cyclic AMP. Revised: November 18, 2005 Infect. Immun. 62, 4432–4439. Accepted: December 20, 2005 Howard, G.C., Misra, U.K., DeCamp, D.L., and Pizzo, S.V. (1996). Al- Published: March 23, 2006 tered interaction of Cis-dichlorodiammineplatinum(II)–modified alpha 2-macroglobulin (alpha 2M) with the low density lipoprotein recep- tor-related protein/alpha 2M receptor but not the alpha 2M signaling REFERENCES receptor. J. Clin. Invest. 97, 1193–1203. Howell, B.W., and Herz, J. (2001). The LDL receptor gene family: sig- Arora, N., Klimpel, K.R., Singh, Y., and Leppla, S.H. (1992). Fusions of naling functions during development. Curr. Opin. Neurobiol. 11, 74–81. anthrax toxin lethal factor to the ADP-ribosylation domain of Pseudo- Itoh, K., Krupnik, V.E., and Sokol, S.Y. (1998). Axis determination in monas exotoxin A are potent cytotoxins which are translocated to Xenopus involves biochemical interactions of axin, glycogen synthase the cytosol of mammalian cells. J. Biol. Chem. 267, 15542–15548. kinase 3 and beta-catenin. Curr. Biol. 8, 591–594. Bhatnagar, R., and Batra, S. (2001). Anthrax toxin. Crit. Rev. Microbiol. Kounnas, M.Z., Moir, R.D., Rebeck, G.W., Bush, A.I., Argraves, W.S., 27, 167–200. Tanzi, R.E., Hyman, B.T., and Strickland, D.K. (1995). LDL receptor- Bishop, N., and Woodman, P. (2001). TSG101/mammalian VPS23 and related protein, a multifunctional ApoE receptor, binds secreted mammalian VPS28 interact directly and are recruited to VPS4-induced beta-amyloid precursor protein and mediates its degradation. Cell endosomes. J. Biol. Chem. 276, 11735–11742. 82, 331–340.

Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc. 1153 Leppla, S.H. (1988). Production and purification of anthrax toxin. Scobie, H.M., Rainey, G.J., Bradley, K.A., and Young, J.A. (2003). Hu- Methods Enzymol. 165, 103–116. man capillary morphogenesis protein 2 functions as an anthrax toxin Leppla, S.H. (1995). Bacterial Toxins and Virulence Factors in Disease. receptor. Proc. Natl. Acad. Sci. USA 100, 5170–5174. In Handbook of Natural Toxins, B.I.J. Moss, M. Vaughan, and A. Tu, Semenov, M.V., Tamai, K., Brott, B.K., Kuhl, M., Sokol, S., and He, X. eds. (New York: Marcel Dekker, Inc.), pp. 543–572. (2001). Head inducer -1 is a ligand for Wnt coreceptor LRP6. Curr. Biol. 11, 951–961. Leppla, S.H. (2000). Anthrax Toxin. In Bacterial Protein Toxins, K. Aktories, and I. Just, eds. (Berlin: Springer), pp. 445–472. Singh, Y., Chaudhary, V.K., and Leppla, S.H. (1989). A deleted variant of Bacillus anthracis protective antigen is non-toxic and blocks anthrax Li, Y., Lu, W., Marzolo, M.P., and Bu, G. (2001). Differential functions of toxin action in vivo. J. Biol. Chem. 264, 19103–19107. members of the low density lipoprotein receptor family suggested by their distinct endocytosis rates. J. Biol. Chem. 276, 18000–18006. Stefansson, S., Chappell, D.A., Argraves, K.M., Strickland, D.K., and Argraves, W.S. (1995). Glycoprotein 330/low density lipoprotein recep- Liu, G., Bafico, A., Harris, V.K., and Aaronson, S.A. (2003a). A novel tor-related protein-2 mediates endocytosis of low density lipoproteins mechanism for Wnt activation of canonical signaling through the via interaction with B100. J. Biol. Chem. 270, 19417– LRP6 receptor. Mol. Cell. Biol. 23, 5825–5835. 19421. Liu, S., and Leppla, S.H. (2003). Cell surface tumor endothelium Strack, B., Calistri, A., Craig, S., Popova, E., and Gottlinger, H.G. marker 8 cytoplasmic tail-independent anthrax toxin binding, proteo- (2003). AIP1/ALIX is a binding partner for HIV-1 p6 and EIAV p9 func- lytic processing, oligomer formation, and internalization. J. Biol. tioning in virus budding. Cell 114, 689–699. Chem. 278, 5227–5234. Tamai, K., Semenov, M., Kato, Y., Spokony, R., Liu, C., Katsuyama, Y., Liu, S., Schubert, R.L., Bugge, T.H., and Leppla, S.H. (2003b). Anthrax Hess, F., Saint-Jeannet, J.P., and He, X. (2000). LDL-receptor-related toxin: structures, functions and tumour targeting. Expert Opin. Biol. proteins in Wnt signal transduction. Nature 407, 530–535. Ther. 3, 843–853. Tucker, A.E., Salles, I.I., Voth, D.E., Ortiz-Leduc, W., Wang, H., Lu, Q., Hope, L.W., Brasch, M., Reinhard, C., and Cohen, S.N. (2003). Dozmorov, I., Centola, M., and Ballard, J.D. (2003). Decreased glyco- TSG101 interaction with HRS mediates endosomal trafficking and re- gen synthase kinase 3-beta levels and related physiological changes in ceptor down-regulation. Proc. Natl. Acad. Sci. USA 100, 7626–7631. Bacillus anthracis lethal toxin-treated macrophages. Cell. Microbiol. 5, Lu, Q., Wei, W., Kowalski, P.E., Chang, A.C., and Cohen, S.N. (2004). 523–532. EST-based genome-wide gene inactivation identifies ARAP3 as a host Veeman, M.T., Slusarski, D.C., Kaykas, A., Louie, S.H., and Moon, R.T. protein affecting cellular susceptibility to anthrax toxin. Proc. Natl. (2003). Zebrafish prickle, a modulator of noncanonical Wnt/Fz signal- Acad. Sci. USA 101, 17246–17251. ing, regulates gastrulation movements. Curr. Biol. 13, 680–685. Mao, B., Wu, W., Li, Y., Hoppe, D., Stannek, P., Glinka, A., and Niehrs, Vitale, G., Pellizzari, R., Recchi, C., Napolitani, G., Mock, M., and C. (2001). LDL-receptor-related protein 6 is a receptor for Dickkopf Montecucco, C. (1998). Anthrax lethal factor cleaves the N-terminus proteins. Nature 411, 321–325. of MAPKKs and induces tyrosine/threonine phosphorylation of MAPKs in cultured macrophages. Biochem. Biophys. Res. Commun. 248, Mao, B., Wu, W., Davidson, G., Marhold, J., Li, M., Mechler, B.M., De- 706–711. lius, H., Hoppe, D., Stannek, P., Walter, C., et al. (2002). Kremen pro- teins are Dickkopf receptors that regulate Wnt/beta-catenin signalling. Willert, K., Logan, C.Y., Arora, A., Fish, M., and Nusse, R. (1999). A Dro- Nature 417, 664–667. sophila Axin homolog, Daxin, inhibits Wnt signaling. Development 126, 4165–4173. Milne, J.C., Furlong, D., Hanna, P.C., Wall, J.S., and Collier, R.J. (1994). Anthrax protective antigen forms oligomers during intoxication Willnow, T.E., and Herz, J. (1994). Genetic deficiency in low density of mammalian cells. J. Biol. Chem. 269, 20607–20612. lipoprotein receptor-related protein confers cellular resistance to Pseudomonas exotoxin A. Evidence that this protein is required for O’Brien, J., Friedlander, A., Dreier, T., Ezzell, J., and Leppla, S. (1985). uptake and degradation of multiple ligands. J. Cell Sci. 107, 719–726. Effects of anthrax toxin components on human neutrophils. Infect. Im- Yu, K.C., Chen, W., and Cooper, A.D. (2001). LDL receptor-related pro- mun. 47, 306–310. tein mediates cell-surface clustering and hepatic sequestration of chy- Orth, K., Madison, E.L., Gething, M.J., Sambrook, J.F., and Herz, J. lomicron remnants in LDLR-deficient mice. J. Clin. Invest. 107, 1387– (1992). Complexes of tissue-type plasminogen activator and its serpin 1394. inhibitor plasminogen-activator inhibitor type 1 are internalized by Zeng, L., Fagotto, F., Zhang, T., Hsu, W., Vasicek, T.J., Perry, W.L., means of the low density lipoprotein receptor-related protein/alpha 3rd, Lee, J.J., Tilghman, S.M., Gumbiner, B.M., and Costantini, F. 2-macroglobulin receptor. Proc. Natl. Acad. Sci. USA 89, 7422–7426. (1997). The mouse Fused locus encodes Axin, an inhibitor of the Wnt Pinson, K.I., Brennan, J., Monkley, S., Avery, B.J., and Skarnes, W.C. signaling pathway that regulates embryonic axis formation. Cell 90, (2000). An LDL-receptor-related protein mediates Wnt signalling in 181–192. mice. Nature 407, 535–538. Zilberberg, A., Yaniv, A., and Gazit, A. (2004). The low-density lipopro- Schneider, W.J., and Nimpf, J. (2003). LDL receptor relatives at the tein receptor-1, LRP1, interacts with the human frizzled-1 (HFz1) and crossroad of endocytosis and signaling. Cell. Mol. Life Sci. 60, 892– downregulates the canonical Wnt signaling pathway. J. Biol. Chem. 903. 279, 17535–17542.

1154 Cell 124, 1141–1154, March 24, 2006 ª2006 Elsevier Inc.